Archaeological and Ethnographic Perspectives On Food, Politics, and Power Edited by Michael Diet/er and Brian Hayden

ARCHAEOLOGICAL AND
ETHNOGRAPHIC PERSPECTIVES
ON FOOD, POLITICS, AND POWER
Edited by Michael Diet/er and Brian Hayden
SMITHSONIAN INSTITUTION PRESS
WASHINGTON AND LONDON

S M I T H S O N I A N S E R I E S I N A R C H A E O LO G I C AL I N Q U I R Y
Bruce D. Smith and Robert McC. Adams, Series Editors
The Smithsonian Series in Archaeological Inquiry presents original case studies that ad
dress important general research problems and demonstrate the values of particular the
oretical and/ or methodological approaches. Titles include well-focused edited collections
as well as works by individual authors. The series is open to all subject areas, geographi
cal regions, and theoretical modes.
Advisory Board
Linda Cordell, University of Colorado
Kent V. Flannery; University of Michigan
George C. Frison, University of Wyoming
Roger C. Green, University of Auckland
David Hurst Thomas, American Museum of Natural History
John E. Yellen, National Science Foundation
© 2001 by the Smithsonian Institution
All rights reserved
COPY EDITOR: Peter Donovan

PRODUCTION EDITOR: Dukejobns
DESIGNER: Amber Frid-Jimenez
Library of Congress Cataloging-in-Publication Data

Feasts : archaeological and ethnographic perspectives on food, politics, and power I edited by
Michael Dietler, Brian Hayden.
p. cm.
Includes bibliographical references and index.
ISBN 1-56098-861-4 (alk. paper) - ISBN 1-56098-840-1 (paper : alk. paper)
1. Festivals--Congresses. Fasts and feasts--Congresses. I. Dietler, Michael.
II. Hayden, Brian.
GT3930.F4 2001
394.26-<lcZl 00-061932
British Library Cataloguing-in-Publication Data available
Manufactured in the United States of America
08 07 06 05 04 03 02 01 5 4 3 2 1
8 The paper used in this publication meets the minimum requirements of the American
National Standard for Information Sciences--Permanence of Paper for Printed Library
Materials ANSI 239.48-1984.
For permission to reproduce illustrations appearing in this book, please correspond directly
with the owners of the works, as listed in the individual captions. The Smithsonian Institution
Press does not retain reproduction rights for these illustrations individually, or maintain a file
of addresses for photo sources.
Contents
List of Illustrations vii

List ofTables ix
Contributors xi
1
Digesting the Feast-Good to Eat, Good to Dri n k, Good
toThi n k: An Introduction 1
Michael Diet/er and Brian Hayden
PA RT 1 : E T H N O G R A P H I C P E R S P E C T I V E S
2
Fabulous Feasts: A Prolegomenon to the Importance of Feasting 23
Brian Hayden
3
Theorizing the Feast: Rituals of Consumption, Com mensal Pol itics,
and Power in African Contexts 65
Michael Diet/er
4
Of Feasti ng and Value: Enga Feasts in a Historica l Perspective
(Pa pua New Guinea) 115
Polly Wiessner
5
Akha Feasti ng: An Ethnoa rchaeological Perspective 1 44
Michael J. Clarke
6
Polynesian Feasting in Ethnohistoric, Ethnographic, and Archaeological Con
texts: A Comparison ofThree Societies 1 68
Patrick II. Kirch
7
Feasti ng for Prosperity: A Study of Southern
Northwest Coast Feasting 1 85
James R. Perodie
Contents
8
The Big Drin k: Feast a n d Forum in the Upper Amazon 21s
Warren R . DeBoer
9
Feasts a n d Labor Mobi lization: Dissecting a Funda mental
Economic Practice 240
Michael Diet/er and Ingrid Herbich
PA RT 2 : A R C H A E O LO G I C A L P E R S P E C T I V E S
10
The Evolution of Ritua l Feasting System s in Prehispanic
Phi l i ppine Chiefdoms 267
Laura Lee Junker
11
Feasting a n d the Emergence of Platform Mound Ceremonialism
i n Eastern North America 31 1
Vernon James Knight
12
A Case of Ritua l Feasting at the Cahokia Site 334
Lucretia S. Kelly
13
Feasting on the Peri phery:The Production of Ritua l Feasti ng and V i l l age
Festiva l s at the Ceren S ite, El Salvador 368
Linda A. Brown
14
Feasting in the Ancient Nea r East 391
Denise Schmandt-Besserat
15
Ga rbage a n d the Modern American Feast 404
Douglas C. Wilson and William L. Rathje
Index 423
vi
Illustrations
2.1. Schema of feast types 38

2.2. Oversized food-preparation vessels in the household
of a Yao village headman 48
2.3. Multiple outside hearths for a Hmong funeral 50
2.4. Multiple outside hearths for a Ta Oi marriage feast in Vietnam 51
2.5. Closeup of the food-preparation area in Figure 2.4 52
2.6. Vietnamese lineage shrine used for lineage feasts 54
2.7. Water-buffalo horns from feasts displayed on a house 56
2.8. Pig mandibles on display in an Akha administrator's house 56
3.1. Luo communal beer drinking pot and beer fermentation pot 97
3.2. How beer is consumed from thago through a long vine-stem straw 98
3.3. An iconic representation of a Luo feast 99
3.4. A Luo woman engaged in drying sinoho 100
4.1. Post-sweet potato networks of exchange, warfare, and ritual 122
4.2. Number of pigs contributed by big-men in secular vs. sacred feasting 130
4.3. Schematic drawing of a Kepele cult site 134
5.1. The number of pigs that each clan in Sam Soong village owns 150
5.2. Elder lineage women and peer guests during an Akha marriage feast 152
5.3. Lineage elders, a Ritual Reciter, and the younger lineage host
at a curing feast 154
5.4. Large cooking pots and woks 157
5.5. The multimodal size distribution of woks and pots in an Akha village 159
6.1. The main ceremonial plaza on Hivaoa, Marquesas Islands 176
7.1. A potlatch at Alert Bay1prior to 1914 186
8.1. Schema contrasting male-associated core with female-associated
periphery 217
8.2. Schema contrasting friends, affines, and enemies 219
8.3. Conibo-Shipibo vessel forms 224
8.4. Artifacts associated with the ani shreati 226
8.5. Pottery vessels associated with the ani shreati 227
8.6. Schema for marriage and sex 231
9.1. Schematic representation of Collective Work Events (CWEs) 242
9.2. Iron hoe blade from the precolonial era 250
9.3. Schematic links between subsistence production, marriage,
and iron production in Samia 252
9.4. Schematic representation of the role of work feasts in a multi-centric
economy 253
10.1. Tenth- to nineteenth-century Philippine chiefdoms 269
vii
Illustrations
10.Z. Faunal remains in household middens of elite and non-elite residences

at Tanjay Z87
10.3. Typical Chinese porcelain "serving" assemblages found at Philippine sites Z90
10.4. Percentages of porcelain forms by weight in excavations
of Tanjay and Cebu Z91
10.5. Comparison of ceramic assemblages in elite vs. non-elite
habitation zones Z9Z
10.6. Characteristic "serving assemblages" from Philippine sites Z94
10.7. Schema for relationships of expanded ritual feasting
in Philippine chiefdoms 300
11.1. Truncated Mounds in the Eastern U.S., 100 B.c.-700 A.O. 314
11.Z. McKeithen Mound A, plan of features 316
ll.3. Walling Platform Mound, plan of features 317
11.4. Cold Springs, plan of features 318
11.5. Cold Springs, cross-sections of large postholes 319
11.6. Garden Creek Mound No. Z, plan of features 3ZO
11.7. Timucuan method of smoking meat on a scaffold,
sixteenth-century Florida 3Z6
lZ.l. The American Bottom region 335
lZ.Z. Late prehistoric chronology for the American Bottom 336
lZ.3. Site plan of central Cahokia 338
lZ.4. Portions of deer belonging to the high, mid, and low Food Utility Index (FUI)
categories 341
lZ.5. A profile of one of the sub-Mound 51 excavation units 343
lZ.6. Change in percentage of NISP of deer at Cahokia 345
lZ.7. Summary of faunal materials from sub-Mound 51 pit 346
lZ.8. Zone DZ birds from sub-Mound 51 348
lZ.9. Zone DZ fish from sub-Mound 51 350
13.1. Map of western El Salvador 369
13.Z. Site map for Ceren, El Salvador 371
13.3. Plan view of Structure 10 at Ceren, El Salvador 373
13.4. Artist's reconstruction of Structure 10 374
13.5. Household 1, Structure 10, and Structure lZ showing positions
of metates 376
13.6. Women's work group using household space to grind corn
for a ritual feast 379
14.1. Standard of Ur 393
14.Z. Perforated plaque from Nippur 395
14.3. Perforated plaque from Khafaje 395
14.4. Perforated plaque from Khafaje 396
14.5. Seal impressions from Ur 396
14.6. Seal impression from unknown site 397
15.1. A Marriott hotel advertisement for a restaurant feast 418
viii
Tables
2.1. Archaeological Signatures of Feasts 40

4.1. The Social Components of Feasting and Corresponding
Political Strategies 118
4.2. Chronological Scheme of Events Discussed in Text 120
4.3. The Six Stages of the Aeatee Cult of Central Enga and Its Relation
to the Tee Cycle 135
5.1. Comparison of Rich and Poor Families 148
5.2. Akha Feast Types 153
6.1. Key Contrasts between Tikopia, Marquesas, and Hawaiian Case Studies 170
6.2. Principal Kinds of Feasts in Tikopia 171
6.3. Principal Kinds of Feasts in the Marquesas 174
7.1. Marriage Wealth Exchanges between Two High-Ranking Kwakiutl
Families 208
8.1. Classification of the Conibo-Shipibo Ceramic Assemblages 225
8.2. Distribution of Beer and Nonbeer Vessels 228
8.3. Synopsis of Schematic Polarities: The Conibo-Shipibo Point of View 231
11.1. Truncated Mounds in the Eastern United States, 100 B.C.-700 A.O. 315
11.2. Associations of Certain Woodland Platform Mounds 322
15.1. Artifacts Related to a Thanksgiving Feast in Sample C45-120291 409
15.2. Artifacts Related to a Presidents' Day Feast in Sample C7-022394 412
15.3. Artifacts Related to a Midweek Party in Sample A7-031194 414
ix·
Contributors
Linda A. Brown, Department of Anthropology; University of Colorado, Boulder

Michael ]. Clarke, Department of Archaeology; Simon Fraser University
Warren R. DeBoer, Department of Anthropology; Queens College, City University
of New York
Michael Dietler, Department of Anthropology; University of Chicago
Brian Hayden, Department of Archaeology; Simon Fraser University
Ingrid Herbich, Department of Anthropology; University of Chicago
Laura Lee Junker, Department of Anthropology; Western Michigan University
Lucretia S. Kelly; Department of Anthropology; Washington University; St. Louis
Patrick V. Kirch, Department of Anthropology; Unive ;sity of California, Berkeley
VernonJames Knight, Department of Anthropology; University of Alabama
James R. Perodie, Department of Archaeology; Simon Fraser University
William L. Rathje, Department of Anthropology; University of Arizona
Denise Schmandt-Besserat, Center for Middle Eastern Studies, University of Texas
at Austin
Polly Wiessner, Department of Anthropology; University of Utah
Douglas C. Wilson, National Park Service, Vancouver, Washington
xi
1
1
DIGESTING THE FEAST: GOOD TO EAT,
GOOD TO DRINK, GOOD TO THINK
A N I N T RO D U CT I O N
Michael Diet/er and Brian Hayden
With apologies to Levi-Strauss for yet another usurping of his famous dictum, a
central argument of this volume is surely that feasts are "good to think." Indeed,
these chapters collectively indicate the appropriateness of this statement in two
crucial senses. In the first place, they convincingly demonstrate that thinking
about feasts can, and should, provide an important point of departure for under
standing culture and social life in both past and present societies. In particular, the
time is long overdue for feasts to be taken seriously by archaeologists as a signifi
cant-perhaps a central-social practice. Secondly, however, these chapters also
show that transforming feasts from something considered as epiphenomena!
trivia worthy of little more than bemused speculation into a subject of recog
nized importance and analytical utility is by no means a straightforward proposi-
Michael Dietler and Brian Hayden
tion. We need to think seriously and critically about what feasts are, how they op
erate, and how we can detect and interpret them. Otherwise, they risk becoming
one more ill-digested archaeological interpretive fad.
One of the main stimuli for undertaking this book, and for convening the sym
posium that generated it, was that both of the editors came to the conclusion
over a decade ago that feasts are an extremely significant aspect of social life on a
worldwide scale, and that understanding them is crucial for apprehending and
comprehending many social and cultural processes in ancient societies. However,
we arrived at this mutual conclusion by quite different routes and from very dif
ferent theoretical dispositions.
As will be evident from our respective chapters and prior publications on the
subject, Hayden's perspective (see Chapter 2) is firmly grounded in a cultural
ecology orientation. Furthermore, he was originally led to explore this line of re
search largely by his interest in the evolutionary implications of feasting in the
transformation of sociopolitical structures in hunter-gatherer societies of the Pa
cific Northwest and Paleolithic Europe. Dietler's perspective (see Chapter 3), on
the other hand, grew out of the domain of social theory often referred to by the
ecologically inclined as "culturalist" -that is, a perspective that takes culture seri
ously as a historical agent as well as a historical product-although, because of its
determinist implications, this label is one that few scholars would actually apply
to themselves. Hence, Dietler's perspective actually has a strong social orienta
tion, with roots in practice theory and political economy (in a nutshell, there is a
central concern with the intricate relationship between culture and power).
Moreover, he was induced to investigate feasting as a result of archaeological and
ethnographic research among agrarian societies in Europe and Africa.
This very fact of arriving at a conjuncture of interest and understanding from
such strikingly different directions strengthened our convictions that, despite our
continuing dialogue of disagreement over a number of issues (see below, and
Chapters 2 and 3), we had stumbled upon something of genuine and general
significance.
One of the more disconcerting conclusions that we also agreed upon, how
ever, was that feasting was both severely undertheorized in the existing literature
and lacked systematic empirical documentation of a kind that would be useful
for archaeological interpretation. This is not to claim that feasting has been com
pletely ignored in the earlier archaeological and anthropological literature, or
that significant insights have not been forthcoming (e.g., see Friedman 1984;
Friedman and Rowlands 1978). However, these earlier works generally included
analyses of feasts as peripheral observations: they did not generate a systematic
theoretical exploration of the subject and they failed to engender a sustained ar
chaeological pursuit of their implications. Hence, we sought to convene a group
2
D I G E STI N G T H E F E A ST
of other scholars who have recently begun to pursue similar issues in a wide va
riety of both ethnographic and archaeological contexts in order to forge a better,
if still provisional, understanding of the nature of the feast as a social institution
and cultural practice. As the chapters in this volume attest, these scholars also ex
hibit a healthy diversity of theoretical orientations and ideas about how to un
derstand feasts. Consequently, this brief introductory essay is less an attempt to
force these chapters into a strained synthetic summary than it is a selective high
lighting of a few key themes that indicate the current state of research and sug
gest future directions.
C AT E G O R I E S A N D D E F I N I T I O N S
Perhaps the most obvious question that needs to be addressed at this point is,
''Are we all talking about the same thing when we use the word feast?" Obviously,
we are not really talking about a "thing" at all, but about a category used to de
scribe collectively a diverse set of cultural practices. Hence, equally important to
ask is the related question, "Is this a useful analytical category?" In other words, it
is clear that we all use the rubric feast to cover a wide range of cultural practices.
Are we agreed on the criteria by which we selectively lump these practices to
gether and exclude others, and does the resulting category have enough coher
ence and specificity to be useful? We believe the answer to both questions is
"yes," with a few stipulations.
Each paper provides at least a general working definition of feasts. These defi
nitions show some variability in both the details (such as the requisite social dis
tance of the participants and size of the gathering) and the ways they are
phrased; and some are more explicit than others. However, common to all of
these is the idea that feasts are events essentially constituted by the communal
consumption of food and/ or drink. Most authors are also explicit in differentiat
ing such food-consumption events from both everyday domestic meals and from
the simple exchange of food without communal consumption. These are impor
tant distinctions to maintain if the category is to have analytical utility. To the ex
tent that one begins to conflate feasts with the general exchange of food or with
the other kinds of transactions for which feasts may serve as a context, one pre
cludes both understanding feasts as a specific social practice and understanding
the important semiotic and functional relationships between feasts and these
other kinds of practices.
Dietler (Chapter 3) also argues that it is crucial to recognize and understand
feasting as a particular form of ritual activity. This is what distinguishes feasts
from daily meals, gives them their peculiar power in articulating social relations
and action, and makes them analytically approachable by building upon an exist
ing body of theoretical work. The dramaturgical effects usually associated with
3
feasting (e.g. , singing, dancing, inebriation, oratorical displays) underline the rit
ual nature of these events. They help to create the experience of "condensed
meaning" (Cohen 1979) and render feasts such ideal stages for other important
social transactions, such as prestations of valuables, making alliances, and, as
Wiessner (Chapter 4) demonstrates, the construction of value.
In contrast, Hayden (Chapter 2) uses a somewhat broader definition, simply
stipulating that any unusual occasion accompanied by an unusual shared meal
should be considered a feast. Moreover, in his view, ritual aspects may or may not
be identifiable archaeologically and may not even be overtly manifest in some
ethnographic cases. Other authors propose yet other minor variations for the
definition of feasts. However, despite these differences, there is still a very large
degree of agreement on identifying many kinds of shared meals as feasts when
confronted with empirical cases in ethnographic contexts.
Given these stipulations, it seems clear that "feasts" is a category that has suffi
cient specificity to be analytically powerful yet brings within its scope a signifi
cant range of important practices around the world and through time. Hence
"feasts" is a productive category: It is "good to think. "
Within the domain of practices that we designate as feasts, there are many
possible ways to categorize the range of differences and similarities. This fact ex
plains the considerable diversity of classificatory schemes brought to bear on the
subject by the authors in this volume. While some readers may find the lack of a
uniform classification troubling or disappointing, we would suggest that this di
versity need not worry us and is, in fact, a good thing-especially at this stage of
theoretical development. Classifications are, after all, simply tools of analysis.
Different ones will be appropriate for different purposes. The criteria for select
ing them should be simply the effectiveness they demonstrate in achieving the
goals for which they were designed, the interest of those goals, and, of course,
the logical consistency of the classifications. As research progresses in this rela
tively novel field, our various ways of characterizing and understanding feasts
will undoubtedly improve. But this does not mean that we will move toward the
development of a single typology of feasts. In our view that would be counter
productive: in the social sciences, classificatory reification is generally the enemy
of understanding.
Hayden (Chapter 2) has expressed the hope that eventually we may develop an
archaeological classification of feasts based on material remains. Others (e.g. , Diet
ler, Chapter 3) are less sanguine about this prospect, believing that the culturally
constituted nature of feasts mandates that archaeological interpretation will al
ways require the construction of a richly textured and culturally specific contex
tual argument grounded in a theoretical understanding of the complexities of
feasting in comparative ethnographic perspective-in other words, a version of
4
D I G E S T I N G T H E F E AST
what Hayden calls "triangulation." The difficulties of distinguishing on a priori

grounds between, for example, the permutations of ritual practices that are used
to mark feasts off from everyday meals, and those that are used to mark off social
classes in "diacritical" feasts (see Dietler, Chapter 3) suggest that uniform typolo
gies of material signatures may not help us much beyond the mere identification
of the existence of feasting.
Whatever the eventual resolution of this discussion, what is clear is that mov
ing beyond the basic identification of presence toward understanding what kinds
of feasts were being mounted, what kinds of ritual work were being undertaken,
and what the social ramifications were, requires complex and nuanced forms of
recursive argumentation. However, even if a handy interpretive formula may ul
timately be unattainable for archaeologists (as well as for cultural anthropologists
and historians), this does not mean that we cannot develop a much better under
standing of the material dimension of feasting. Indeed, at present, the lack of
precise information on this issue is one of the biggest impediments to the ar
chaeological investigation of feasting and the evaluation of the relative plausibil
ity of different archaeological interpretations. For the most part, archaeologists
do not yet know precisely what to look for. The exemplary archaeological studies
by Junker (Chapter 10), Knight (Chapter n), Kelly (Chapter 12), Brown (Chapter
13), and Schrnandt-Besserat (Chapter 14) in this volume point the way toward
profitable strategies for interrogating the archaeological record for this kind of
information (see also, for example, Blitz 1993; Clark and Blake 1994; Dietler 1990,
1996, 1999b; Gardeisen 1999; Moore 1989; Morris 1979; Murray 1995). But much
more primary ethnographic research focusing specifically on feasting, of the kind
undertaken by Clarke (Chapter 5), Hayden (Chapter 2), and Dietler and Herbich
(Chapters 3 and 9), is urgently needed. Moreover, this work should be undertaken
by scholars already familiar with the problems of detecting feasts in archaeologi
cal contexts. Again, it is the synergistic dialectical conversation between these
domains that will propel the development of theory and method for both.
As a final note on classification, a brief word about the classificatory structure
of this volume is also in order. Part One assembles considerations of feasting
based upon ethnographic and historical data, either through comparative analy
sis or individual case studies. These chapters use these rich sources of informa
tion about social life to develop a fuller understanding of feasting practices and
they discuss the material implications of the analyses they undertake. Several of
these chapters are explicitly concerned with developing new theoretical ap
proaches to feasting, and all offer at least implicit theoretical insights. Moreover,
all have as a major goal assessing and aiding the possibilities for archaeological in
terpretation of feasts, although, given the limitations of space, they do not actu
ally involve the application of such insights to specific archaeological cases.
5
Thus, Hayden (Chapter 2) examines feasting from a cross-cultural and ecolog

ical perspective toward the development of a general framework for the archaeo
logical detection and interpretation of feasts. Dietler (Chapter 3) uses the Luo
and other African cases to argue for the inherently political role of feasts and to
develop a general theoretical analysis of the complex relationship between feasts,
commensality, and power. Wiessner (Chapter 4) provides a rich ethnohistorical
analysis of the role of feasts in religious cults and strategies of New Guinea big
men for promoting their self-interests and transforming conceptions of value.
Clarke (Chapter 5) shows how feasts are used by the Akha tribes of Thailand to
establish socioeconomic safety networks. Kirch (Chapter 6) uses a comparison of
three Polynesian cases to demonstrate how the magnitude of, and facilities asso
ciated with, feasting increase with increasing size and complexity of the polities.
Perodie (Chapter 7) documents the range of feasting on the Northwest Coast of
North America and illustrates what advantages accrued to hosts. DeBoer (Chap
ter 8) provides an important example of Amazonian feasting with its unique girl's
initiation features and competitive male boasting, and reflects upon long-term
historical transformations and continuities. Finally, Dietler and Herbich (Chapter
9) use a comparative examination of the political-economic dynamics of work
feasts among the Samia of Africa and other societies to propose a general theo
retical model for the crucial role of feasts in labor mobilization and exploitation.
Part Two contains chapters that move in the opposite interpretive direction.
That is, they attempt to grapple with the detection of feasting in the material
record and to then make plausible inferences about the social life and culture of
the people who were producing and participating in those feasts. These creatively
ingenious studies illustrate both the difficulties and potential of this domain of
inquiry, as well as the necessity of a recursive dialectical conversation between
ethnographically derived theory and archaeological data. At first glance, it may
appear curious that Wilson and Rathje's (Chapter 15) study of garbage in present
day Tucson should be classed together with these latter chapters; but, in fact, the
logic of this interpretive exercise is thoroughly archaeological. The samples of
refuse were obtained by means other than excavation, but these material items
are then used to reason inferentially back to an interpretation of the practices
that produced them based upon an understanding of the general social context
(but without direct observation of those specific practices). Hence, they provide
a good heuristic test of the possibilities for understanding feasting through mate
rial remains in a social and cultural context with which many readers will have
some familiarity:
Other chapters in this section offer a range of more obviously archaeological
case studies.Junker (Chapter 10) provides a regional analysis of the economic and
political role of feasts in the very dynamic cultural environment of the Philip-
6
D I G E STI N G T H E F E AST
pines during the second millennium A.O. Knight (Chapter n) explores the impor
tant role of feasting in the creation and use of the first platform mounds among
the Woodland cultures of the southeastern United States during the first millen
nium A.O. , while Kelly (Chapter 12) documents large-scale feasting at the largest
Mississippian sites during the early stages of the succeeding millennium. Brown
(Chapter 13) presents an analysis of new evidence for the use of special feasting
structures for lineage feasts in non-elite Maya contexts around A.O. 600. Finally,
moving to the ancient Near East, Schmandt-Besserat (Chapter 14) explores the
,
use of feasts by Sumerian elites to collect surpluses.
It is worth emphasizing that, despite the impressive diversity of geographical
and temporal contexts represented by the chapters in this section, the intention
was clearly not to provide a comprehensive global coverage of feasting in prehis
tory: Hence, the absence of studies from certain regions should by no means be
taken as an indication that the editors judged that feasting was not an important
feature of the ancient social landscape in those areas or that the potential for the
archaeological analysis of feasts is less good in those contexts. Far from it. Eu
rope, for example, is not covered in this volume even though feasting is now in
creasingly recognized to have been a practice of major importance implicated in
various kinds of social change throughout the region and despite the fact that
one of the editors has previously undertaken several studies of feasting in differ
ent European archaeological contexts (Dietler 1990, 1996, l999a, l999b; see also
Murray 1995; Schmitt Pantel 1992; Sherratt 1991). The intention of Part Two was
simply to provide a selected set of fresh cases demonstrating cutting-edge possi
bilities for the identification and interpretation of feasts in a diverse, rather than
exhaustive, range of archaeological contexts.
F E A S T S A N D F E AST I N G
Another important question to pose is whether one legitimately can discuss feast
ing in archaeological contexts without actually being able to identify specific
feasts. In other wbrds, can one detect the traces of a practice, or process, without
necessarily being able to identify its constituent events? We would strongly sug
gest that one can. Indeed, archaeologists should have no inherent methodologi
cal objection to this kind of procedure. We do it all the time for other processes,
such as trade or agriculture, for which specific events are usually very difficult, if
not impossible, to identify. We feel quite confident in assuming the existence of
trade, for example, on the basis of general regional patterns reflecting the results
of its operation, without ever being able to point to a particular place where ex
changes actually took place. We would suggest that the same is true of feasting.
The requirements are that we have a good theoretical understanding of the social
roles of feasting and their permutations, and that we know what the relevant
7
diagnostic criteria are and how to look for them in regional archaeological data.
Archaeologists know quite well how to sniff out trade and characterize its quan
titative and qualitative dimensions through the use of, for example, petrographic,
chemical, and stylistic analysis of ceramics found in contexts of consumption.
But they are generally not yet familiar with the ways that one can detect and
characterize feasting as a process. Hence, they are less alert to the possibilities and
more skeptical of the endeavor unless one can produce a Pompeiiesque example
of a feast.
Junker's (Chapter rn) analysis of feasting in the Philippines is of great interest
precisely because it is an excellent example of how the analysis of feasting as a
process can be convincingly and profitably done through sensitive examination
of things like regional patterns of ceramic import consumption. A similar case
has been made for different regional patterns of feasting in Iron Age Europe
(Dietler 1990, 1996) with the subsequent identification of an actual feast event
(Dietler l999b; Gardeisen 1999). What such studies demand is a careful contex
tual analysis of patterns of consumption. That is, not merely looking at the dis
tribution of ceramic types or wares over the landscape, but undertaking an
analysis of the quantitative, context-specific distribution of specific forms and
the patterns of association where they are found. It involves asking, for exam
ple, whether certain kinds of tableware are found in all or only some graves or
settlements in a region, in all or only some domestic or ceremonial contexts on
those settlements, associated with what other kinds and sizes of ceramic cook
ing vessels and in what relative quantities in each of those contexts, associated
with what kinds of faunal remains, and so on. It also requires looking at trans
formations of these patterns over time. Such analysis can be very revealing not
only of the existence of feasting, but, more importantly, of its forms and his
torical significance. Of course, as Schmandt-Besserat (Chapter 14) has nicely
demonstrated with her Near Eastern study, the possibilities for understanding
regional feasting practices are greatly enhanced by the presence of contempo
rary pictorial and textual representations (see also, for example, Arsenault 1992;
Joffe 1998; Schmitt Pantel 1992). But a great deal is possible even in the absence
of these latter kinds of data.
Hayden (Chapter 2) provides a summary of the many kinds of material evi
dence that can potentially be used for dealing with feasting archaeologically, but
virtually all the contributions to this volume address this issue in one way or an
other. One of the interesting features to emerge from this discussion is that feast
ing actually has an advantage over trade as a subject of archaeological
investigation in that it offers much better possibilities for being able to identify at
least some specific feasting events. This is true, in part, because feasting activities
by their very nature produce copious amounts of distinctive refuse at the loca-
8
tions where they occur, and feasting locations are often associated with notable
ritual structures. In contrast, specific trade activity areas are more difficult to
identify; and the act of exchange produces little if any distinctive refuse. Ironi
cally, because feasts often provide the context for exchange events, the advantages
of detecting feasting in the archaeological record may even help us to better un
derstand trade. Methods for accomplishing this are still being worked out, but
the contributions by Brown (Chapter 13), Kelly (Chapter 12), and Knight (Chapter
II) all offer exemplary case studies showing how the presence of feast events can
be teased out of various permutations of spatial, faunal, and artifactual data. As
these chapters also ably demonstrate, the mere identification of the existence of
feasts at individual archaeological sites is not the ultimate goal of such research.
Rather, we want to know what they mean in terms of the societies that produced
them-and this requires situating them within broader regional patterns.
Part of the potential visibility of feasts derives from the fact that, to reiterate
once again, they are ritual (or, as Hayden would qualify it, "ritualized") events.
This means that they are commonly a central element of life-crisis ceremonies
such as initiations, weddings, and burials, of which at least the latter have a good
chance of being preserved as single-event archaeological sites. More broadly, it
means that the same kinds of diacritica (e.g., qualitative and quantitative differ
ences in food consumed, spatial segregation, architectonic elaboration for dra
maturgical effect, etc.) that may be used, often in combination, to symbolically
mark feasts off from daily meals as ritual events may also make them stand out in
the archaeological record-if we know how to look for them. That, of course,
will require that we become sensitized to their existence and importance, and
that we develop the theoretical competence to deal with their operation in a
more sophisticated manner. This task is a large and complex one, but a task in
which we can already see considerable productive strides being made.
One of the encouraging signs to emerge out of the studies in this volume is
the frequent association of feasting with spatial differentiation or architectonic
elaboration that may be readily recognizable. This is certainly the case in
Brown's (Chapter 13) study of non-elite feasting at Cereo in El Salvador, Kelly's
(Chapter 12) study of Cahokia, Knight's (Chapter II) analysis of platform
mounds, two out of three of Kirch's (Chapter 6) Polynesian cases, and the pot
latches reported by Perodie (Chapter 7), to name a few prominent examples. It is
important to emphasize that not only were these structures specially con
structed sites for feasting, but they were most probably constructed through
feasting. That is, the more marked the architectonic elaboration, the more such
features represent the congealed labor of work feasts and are, in effect, an ad
vertisement of the feasts that went into their construction (see Dietler and Her
bich, Chapter 9). In other cases (e.g., Clarke's Akha example, Kirch's Tikopia
9
example, the Luo feasts in the homestead reported by Dietler, Chapter 3), where
there is less architectonic elaboration or smaller feasts are held in or near do
mestic contexts, the detection of feasting sites may be considerably more diffi
cult. However, under extraordinary conditions it may still be possible to
recognize feasts in domestic contexts through analysis of permutations of fau
nal and artifactual remains (e.g., Clarke, Chapter 5; Dietler l999b; Gardeisen
1999). Evidence for the presence of alcohol may be another usefully widespread
diagnostic sign of feasts. In most small-scale societies, and particularly with pre
distillation forms of alcohol, drinking is not a part of daily meals-it is some
thing reserved for, and indexical of, feasts (see Dietler 1990 and Chapter 3). As
Wilson and Rathje's (Chapter 15) study shows, even in present-day Tucson there
is a strong association between alcohol and feasts.
F EASTS A N D G E N D E R
One of the topics that we would suggest still needs much more explicit treatment
and fuller elaboration is the gender relations that underlie, and are reproduced
and transformed through, feasts. Of course, these are by no means uniform. But
one can already begin to discern a few significant tendencies that require further
research and theoretical discussion.
In the first place, feasting practices almost always act to mark and naturalize
gender categories (see Dietler, Chapter 3). That is, even in societies with a strong
egalitarian ethos, feasts serve to define and inculcate social categories-and gen
der categories are among the most common distinctions marked by these rituals.
Such marking occurs through a wide variety of symbolic diacritica that may be
combined in different permutations. These may include: (1) spatial segregation or
positioning while eating (i.e., differences indicated by men and women eating in
different locations or seated in standardized configurations: alternate seating, op
posite sides of a room, inner and outer circles, etc.), (2) temporal distinctions
(i.e., differences indicated by order of serving), (3) qualitative distinctions (i.e., dif
ferences in the nature of the food, drink or serving vessels offered to men and
,
women), (4) quantitative distinctions (i.e., differences in the amounts of food or

drink served to women and men), or (5) behavioral expectations (i.e., differences
in the ways women and men are expected to act during and after feasting: for ex
ample, who is permitted to act drunk, who may talk while eating, who may reach
for food, who retires from the meal first, etc.). Where social classes exist, these di
acritical patterns may vary between classes such that gender is marked in differ
ent ways within each class-a situation suggested, for example, in Kirch's
(Chapter 6) discussion of Hawaiian royal feasting.
In addition to this aspect, however, gender considerations must also enter the
10
analysis of feasts because feasting frequently involves a gendered asymmetry in

terms of labor and benefits. That is, very often female labor largely supports a
system of feasting in which men are the primary beneficiaries in the political
arena. Female labor is frequently of primary importance in the domain of agri
cultural production and the raising of household domestic animals (such as fowl
and swine), although the relative gendered contribution in this domain is highly
variable (Boserup 1970; Guyer 1988). Far more common, however, is a dominant
female contribution to the crucial culinary and serving labor that transforms raw
food ingredients into feasts. These labor inputs are one of the main reasons why
there is such a strong linkage between polygyny and male political power (see
Dietler, Chapter 3; Dietler and Herbich, Chapter 9). In brief, cases where women
provide the agricultural, culinary, and serving labor for male political activities
are quite common (Bohannan and Bohannan 1968 provides a typical example).
However, cases of the inverse pattern (where men consistently provide the agri
cultural, culinary, and serving labor that underwrites feasts formally hosted by
women) may exist, but they are extremely rare.
These features need not be interpreted as a universal pattern of exploitation,
however. In some cases there is a more balanced, or even male-dominated, pat
tern of labor in the production of feasts (especially in the butchering and cooking
of animals), if not in the production of daily meals (e.g., the Akha case described
by Clarke, Chapter 5). Moreover, women may share in the status and political
benefits from their labor by being members of an influential household or line
age (in matrilineal contexts). They may also derive considerable categorical and
individual status from their central role in the furnishing of hospitality (e.g., see
Gero 1992; March 1998). And, in many societies, women do host their own work
feasts and other feast events, although usually on a smaller scale than men. Fi
nally, in the recent past, the common traditional female monopolization of cook
ing ·and brewing duties has frequently given them opportunities for entering the
monetized market economy and gaining a source of income (e.g., through beer
sales) that has enabled them to acquire considerable independence and intrafa
milial power under changing socioeconomic conditions (e.g., see Colson and
Scudder 1988; Netting 1964).
Hence, feasts are intimately implicated in the representation, reproduction,
and transformation of gender identity and in the gendered division of feasting
labor and benefits, although in complex and variable ways. Archaeologists can ill
afford to ignore these features if we want to use feasting as a way of perceiving
and understanding the development of sociopolitical relations in ancient soci
eties. However, we urgently need more careful attention to such matters in
ethnographic studies of feasting and the development of ethnological theory on
feasts.
11
E C O L O G I C A L M AT E R I A L I S M V E R S U S C U LT U R E A N D P O W E R
There is an interesting ontological tension underlying this volume that also mer
its a few comments. Indeed, a major part of the interest of the book is the dia
logue it presents between radically different perspectives on the same set of
practices. Although the most explicitly articulated contrast in ontological posi
tions is that between the two editors alluded to earlier, the implications of this
debate are apparent throughout the volume. Hayden (Chapter 2) makes a force
ful case for an ecologically grounded materialist consideration of feasting, and
this approach is closely echoed in several other contributions (especially Perodie,
Chapter 7, and Clarke, Chapter 5). Others find this perspective less compelling
and approach the subject from quite different theoretical positions (see especially
Dietler, Chapter 3; Wiessner, Chapter 4; Kirch, Chapter 6; DeBoer, Chapter 8;
Dietler and Herbich, Chapter 9). These perspectives need not necessarily be en
tirely in conflict with each other: Wiessner (Chapter 4) and Junker (Chapter rn),
for example, discuss the ecological constraints on the historical development of
feasting patterns within interpretive frameworks that are not ecologically deter
ministic and that are quite sensitive to the importance of culture and historical
contingency. Moreover, a central concern with political power and the prominent
role attributed to it in all the chapters provide a very significant basis for common
ground. One can reach agreement on the forms and importance of political
power while ultimately viewing it as either a major type of ecological behavior
or as a culturally defined field of social action. It is for this reason that Hayden de
scribes his approach as "political ecology."
However, a chapter claiming to fulfill an introductory role would be sorely re
miss to the extent that it failed to reveal and explore a few of the divergences in
basic ontological premises that characterize the volume. Clearly, neither of the
editors undertakes this analysis from a disinterested position. Hence, caveat lector.
Hayden (Chapter 2) and Perodie (Chapter 7) insist that a practice as ubiquitous
and enormously "expensive" as feasting must have some "practical benefits,"
with an understanding of practicality rooted in the perspective of Marvin Harris.
They acknowledge. that idiosyncratic values motivate some people to use their
resources and power in nonrational, non-self-interested, nonpredictable fash
ions. However, they argue that in aggregate, people do tend to make decisions
based on their own self-interests and the information or choices that are avail
able. In this respect, they act in ecqlogically (and economically) rational terms.
Idiosyncratic variations do occur but rarely are accepted, supported, or perpetu
ated by the communities at large for any length of time. In archaeological evolu
tionary terms, these idiosyncrasies become 'background noise" for the basic
trends that form the archaeological record. In this outlook, political power plays
12
a major role in promoting and defending individual or factional self-interests�

which is why feasting becomes so important in traditional societies that can af
ford to produce the necessary surpluses.
Other perspectives, including perhaps most explicitly that of Dietler (Chapter
3), strongly disagree with aspects of this ecological/materialist vision, although
not all from precisely the same theoretical basis. Without wishing to refight the
"Culture and Practical Reason" battle (see Sahlins 1976, 1995), Dietler feels it is im
portant to signal the quite widespread anthropological position, in which his and
several other contributions to this volume are rooted, which holds that "practi
cality" is not a universal principle of bottom-line materialism, but is a culturally
constructed concept. To the extent that Hayden and Perodie mean by practical
benefits something like the proposition that feasting involves a strong element of
"strategic," self-interested political action, whether consciously acknowledged or
euphemized in the shared "sincere fictions" (Bourdieu 1990) that make possible
the reproduction of the system, it is probable that all of the authors would con
cede some common ground. What then becomes the crucial question is the cul
turally specific definition of appropriate goals and strategic paths. Obviously,
relentless material accumulation is not a universal goal of self-interested action:
in many cases it will lead to scorn, ostracism, witchcraft accusations, and an early
death. In such cultural contexts, the skilled self-interested actor will play by a
quite different set of rules and toward a very different end than the capitalist rob
ber baron. As the various chapters indicate, each in its own way, the interest of
the feast is not simply that it enables the accumulation of wealth or material
goods, but that it is a remarkably supple ritual practice that allows the strategic
reciprocal conversion of economic and symbolic capital toward a wide variety of
culturally appropriate political goals. This is what accounts for the striking ubiq
uity and durability of the feast as an institutionalized practice in the face of dra
matic social transformations, as illustrated in particularly remarkable fashion by
DeBoer's (Chapter 8) Shipibo-Conibo example.
There is an additional element of the "practical benefits" argument that moves
the debate beyond common ground. This is the principle expressed by Hayden
(Chapter 2) that, because of its costliness and ubiquity, the feast must be a form of
"adaptive behavior" -that it must have practical benefits for reproduction and in
dividual survival. Here a number of the authors would politely disagree. In
Dietler's view, this functionalist logic would necessitate drawing the same con
clusion about American professional football games and rock concerts. It is hard
to see the significance of either of these for the survival of American society at
large, and both are clearly deleterious to the long-term life of the performers.
Similarly, feasts can easily escalate into conditions that are dangerous both for the
13
health of the society and for individuals involved. The Hawaiian royal pattern is a
classic case in point (Kirch, Chapter 6; Sahlins 1992), as were Enga pig feasts
(Wiessner, Chapter 4), the feasting customs among the Masa of Africa (Garine
1996), and Luo funeral feasts, which often impoverish families (Dietler, Chapter
3). Hence, from this point of view, the value of a concept such as "adaptive be
havior" in explaining cultural practices seems highly questionable and the
demonstration of a causal connection seems illusory at best.
From Hayden's perspective, on the other hand, the adaptiveness of rock con
certs is not to be found so much in the present cultural context, but in the early
genetic evolutionary roots of the human species where ecstatic rituals included
rhythms and music that enhanced emotional bonds between individuals and
groups critical for surviving severe periods of stress (Hayden 1987, l993:1wff).
Emotional aggressivity toward groups identified as enemies, whether in sports or
real conflicts, may have a similar origin and may still be part of our innate emo
tional heritage (Hayden 1993:175-176). In more immediate terms, it is clear that, ir
respective of the physical effects these events may have on the players involved, the
players clearly feel that they are benefiting handsomely, which their multimillion
dollar contracts amply confirm. Hayden argues that benefits of comparable scale
ought to exist for the principal players in feasts that escalate in costs far beyond
the norm, such as the royal Hawaiian and Enga feasts. However, all-out competi
tive feasting undoubtedly entails both winners and losers who risk everything in
their economic battles; and as in competitive businesses, those who are losers are
typically ruined and crushed so that it is not surprising to find families impover
ished by feasting where feasting becomes competitive.
Much of this disagreement turns around different views of culture. From
Dietler's perspective, feasts, like many other important cultural practices, have
little direct significance for "survival" (except occasionally in the negative sense
when practices sometimes have deleterious unintended consequences) and the
concept of "adaptation" is powerless to explain either the generation of the myr
iad forms that feasts take or their social significance. On the other hand, he would
claim that feasts have a great deal to do with politics and power, and that much of
what Hayden calls "adaptation" is simply social competition of highly variable,
unpredictable, contradictory, and largely unverifiable adaptive significance for so
cieties, social groups, and individuals, respectively Alternatively, for Hayden and
others, politics and the creation of differential power or social safety nets through
feasting play critical roles in economic success, reproduction, and survival.
Within Dietler's perspective, Hayden and Perodie's dismissal of "prestige" and
"status" as inconsequential psychological phenomena is also highly contentious.
For Dietler, these words describe crucial aspects of the kind of symbolic capital
that is a necessary condition for becoming an influential member of society
14
They are the preconditions for developing the moral authority to influence
group decisions, exert leadership, and wield power-or to resist the power of
others. They are the essential elements of the possibility of political action. In
deed, the case can be made that, even in late-capitalist America, they are what the
accumulation of economic capital is ultimately all about. From Hayden's per
spective, on the other hand, "prestige" and "status" in traditional societies are
simply euphemisms for economic success and political power.
Finally, from Dietler's ontological position, it is also important to envisage cul
ture not as something that is destroyed by confrontation with "external realities,"
or that withers away or can be abandoned in the face of opportunities for self
advancement (c£ Hayden, Chapter 2)-a kind of optional, external decorative
fa<;:ade covering a universal bedrock core of materialist rationality. Nor is it some
thing inherited from the past as a static bundle of traits. Rather, it is a way of per
ceiving and thinking about the world, and of solving the problems of daily life
through the application of distinctive categorical and analogical understandings.
Hence, culture is not an alternative to a universal "practicality," but rather the
very way that practicality is constituted.
Ultimately, Hayden feels (somewhat more optimistically than Dietler) that
some of these contentious issues may be only a matter of difference in emphasis
and that more common ground may be recognized in the future. Both of us
argue that some of the agendas behind the hosting of feasts are unvoiced but
driven by political or other self-interested considerations. Moreover, neither ap
proach sees cultural norms as imposing such stringent controls on human behav
ior that no innovation or change ever takes place. On the other hand, neither
approach goes so far as to claim that cultural traditions do not impose some con
straints on the behavioral and conceptual options that people must choose from.
In Hayden's view, the emphasis, however, is more on people's penchant to recog
nize the existing constraints and to use cultural concepts as well as technology
simply as tools to achieve their own self-interested goals. If suitable concepts, val
ues, or other cultural tools are not available, highly motivated individuals typi
cally set about trying to create them, as documented in Wiessner's Enga data
(Chapter 4; see also Wiessner and Tumu 1998). Whether they succeed or not de
pends upon the economic costs, the effectiveness of the introduced elements,
and the competing self-interests of other community members. That is where se
lection comes in. While there may be idiosyncratic variations among people and
even communities in the short term, self-interest in the long run and for most
people is ultimately characterized by basic ecological imperatives of survival, re
production, and health.
In Dietler's view, people's very concepts of self-interest are constituted not by
universal ecological imperatives, but through logics of action that are defined by
15
both specific cultural context and the social situation of actors and which are, at
the same time, inculcated and continually transformed through the practices of
everyday life and ritual. Hayden acknowledges this factor, but argues that what
he views as the distortion or redirection of self-interest through cultural values
applies primarily to situations where the practical impacts are not extreme. He
believes that, as the consequences of cultural values become increasingly detri
mental to individual self-interests, people must eventually refuse to accept values
that authorities promote, even under extreme threats of retribution. It is difficult
to explain revolutions otherwise. In Dietler's view, this conclusion, with its classic
"false consciousness" vision of culture as something that can be equated with
values promoted by authorities, underlines again the fundamentally different
conceptions of culture held by the editors (for a fuller discussion of this perspec
tive on the relationship between ideology, hegemony, and culture, see Dietler
1999a) .
It is important that the reader have a clear sense of the differences outlined
above. This is not simply an arcane bone of friendly contention between the edi
tors, but a crucial issue that is manifested throughout the book in often quite sub
tle ways. We are not, of course, suggesting that all of the chapters can be lined up
on either side of a binary "great divide" defined by the terms of this debate, and
we have, in fact, hesitated to speak for other authors in specifying their diverse,
and often implicit, ontological premises. What we are suggesting is both that
such fundamental differences are important to consider in reading the ensuing
chapters and, perhaps even more importantly, that the issue offeasts is not the prod
uct of a particular theoretical camp, but can be approached profitably from a variety
of quite different theoretical orientations. In fact, what is surprising is not that
the various authors of this book have contrasting ontological positions, but that,
given those positions, we are able to agree on so much. That we all see feasts as
an extremely important cultural practice with characteristic social and political
roles, despite our respectful divergence on fundamental theoretical matters, is,
we believe, a strong endorsement of the viability of pursuing the exploration of
feasting.
F EASTS A N D SOCIAL C H A N G E
One of the main reasons for exploring the subject of feasts was the suspicion that
they have been intimately involved in processes of social change. That is, it was
suspected that they were not simply epiphenomenal reflections of changes in cul
ture and society, but central arenas of social action that have had a profound im
pact on the course of historical transformations. Indeed, Hayden (1990, Chapter
2) has even suggested that the origins of agriculture may be tied to the produc
tion demands generated in political feasting contests, while Dietler (1990, 1996)
16
D I G E S T I N G T H E F E AST
has shown how feasts have served as arenas for the articulation and entanglement
of colonial encounters and the transformation of tastes, value, and relations of
power. Many of the papers in this volume have contributed other novel and com
pelling arguments for the role of feasts in social change.
Wiessner's (Chapter 4) rich historical study, for example, shows how feasts
among the Enga of New Guinea have acted as ritual theaters for the cultural con
struction and transformation of value (see also Wiessner and Tumu 1998). The
dramaturgically charged presentation of the pig by big-men resulted in its incor
poration into feasting and exchange networks in ways that had major conse
quences for the historical transformation of politics and the ecology of the
region. Similarly, Junker's (Chapter rn) analysis shows how the importation of ce
ramics for feasts played into the competitive rivalries between chiefs in the highly
unstable political landscape of the prehispanic Philippines and how feasting be
came a socially transformative practice through its constant realignment of al
liances and patron-client networks. Dietler (Chapter 3) uses the case of the Luo of
Kenya to show how feasts can serve as mechanisms for the transformation of in
formal power into institutionalized formal political roles, and how dependent the
maintenance of authority is on the practice of feasting. Finally, Dietler and Her
bich (Chapter 9) show how the work feast can become a mechanism of labor ex
ploitation that can result in spiraling asymmetries in economic and symbolic
capital and in the emergence of social stratification in egalitarian societies.
A N I N V I TAT I O N TO T H E F E A S T
These chapters offer a compelling collective demonstration that feasts are indeed
"good to think" in many ways. Moreover, they go a long way toward showing ar
chaeologists how to think about feasts. Clearly, much empirical and theoretical
work remains to be done in order to improve our ability to deal with the inter
pretive subtleties of feasting in archaeological contexts. But there is now a solid
foundation upon which to build. It is no longer possible to ignore the significant
role that feasts have played in the social, political, and economic domains of life
around the world and throughout history: Archaeologists need to be aware of
this and to develop the skills necessary to seek and interpret evidence of feasts.
But our challenge at this stage is also to not let this new awareness dissipate
into yet another ill-digested vogue or an oversimplified mechanistic model of
feasts to meet all occasions. The detection of some form of feasting in the ar
chaeological record is, we would strongly emphasize, but the first of many hur
dles that must be overcome before archaeologists will be able to reap the rich
insights into the past that we think the in-depth study of feasting has to offer. In
vocations of feasting are clearly becoming increasingly popular in the recent
archaeological literature, but sometimes this appears to amount to little more
17
than signaling the presence of feasts, as if their significance were uniform and
self-explanatory. However, we emphatically reiterate that, without an adequately
theorized and contextualized analysis of feasting, the mere documentation of
the existence of such practices will not yield the kind of understanding of prehis
toric societies and their social, political, and economic dynamics for which this
domain of activity holds such heuristic promise. It is crucial to identify the spe
cific nature of prehistoric feasts in particular cases and to explain how and why
they operated in specific socioeconomic contexts that can be inferred from ar
chaeological remains or other data. Hence, although the task is complex, we
need to further develop the exploration of feasts as a powerful, versatile, and sub
tle analytical tool capable of providing a window of entry into the diverse array
of forms of political and economic action and social relations in ancient societies.
The collective effort represented in this volume provides a solid basis for opti
mism and further work in this quest.
REFERENCES
Arsenault, D.
1992 Pratiques alimentaires rituelles clans la societe mochica: Le contexte du fes
tin. Recherches Ammndiennes au Quebec 22:45-64.
Blitz, ].
1993 Big Pots for Big Shots: Feasting and Storage in a Mississippian Community.
American Antiquity 58:80--96.
Bohannan, P., and L. Bohannan
1968 Tiv Economy. Evanston: Northwestern University Press.
Boserup, E.
1970 Women's Role in Economic Development. London: Allen and Unwin.
Bourdieu, P.
1990 The Logic of Practice. Stanford: Stanford University Press.
Clark, ]. E., and M. Blake
1994 The Power of Prestige: Competitive Generosity and the Emergence of
Ranked Societies in Lowland Mesoamerica. In Factional Competition and Polit
ical Development in the New World, edited by E. Brumfield and]. Fox, pp. 17-30.
Cambridge: Cambridge University Press.
Cohen, A.
1979 Political Symbolism. Annual Review of Anthropology 8:87-n3.
Colson, E., and T. Scudder
1988 For Prayer and Profit: The Ritual, Economic, and Social Importance of Beer in
Gwembe District, Zambia, 1950-1982. Stanford: Stanford University Press.
Dietler, M.
1990 Driven by Drink: The Role of Drinking in the Political Economy and the
Case of Early Iron Age France. journal of Anthropological Archaeology
9:352-406.
18
D I G E STI N G T H E F E A ST
1996 Feasts and Commensal Politics in the Political Economy: Food, Power, and
Status in Prehistoric Europe. In Food and the Status Quest: An Interdisciplinary
Perspective, edited by P. Wiessner and W. Schiefenhavel, pp. 87-125. Oxford:
Berghahn Books.
l999a Rituals of Commensality and the Politics of State Formation in the
"Princely" Societies of Early Iron Age Europe. In Les princes de la protohistoire
et !'emergence de l'etat, edited by P. Ruby; pp. l3s-152. Naples: Cahiers du Cen
tre Jean Berard, Institut Fran�ais de Naples 17-Collection de l' Ecole Fran
�aise de Rome 252.
l999b Reflections on Lattois Society during the 4th Century B.C. In Lattara 12 :
Recherches sur le quatrieme siecle avant notre ere a Lattes, edited by M. Py; pp.
663-680. Lattes: Association pour la Recherche Archeologique en Languedoc
Oriental.
Friedman, J.
1984 Tribes, States, and Transformations. In Marxist Analyses and Social Anthropol
ogy, edited by M. Bloch, pp. 161-202. London: Tavistock.
Friedman)., and M. J. Rowlands
1978 Notes toward an Epigenetic Model of the Evolution of "Civilisation." In The
Evolution of Social Systems, edited by J. Friedman and M. J. Rowlands, pp.
201-278. London: Duck.worth.
Gardeisen, A.
1999 Decoupe et consommation de viande au debut du Ne siecle avant notre ere:
Quelques elements de boucherie gauloise. In Lattara 12: Recherches sur le qua
tribne sifcle avant notre ere a Lattes, edited by M. Py; pp. 569--588. Lattes: Asso
ciation pour la Recherche Archeologique en Languedoc Oriental.
Garine, I. de
1996 Food and the Status Quest in Five African Cultures. In Food and the Status
Quest: An Interdisciplinary Perspective, edited by P. Wiessner and W. Schiefen
h5vel, pp. 193-218. Oxford: Berghahn Books.
Gero, ). M.
1992 Feasts and Females: Gender Ideology and Political Meals in the Andes. Nor
wegian Archaeological Review 25:15-30.
Guyer, ).
1988 The Multiplication of Labor: Historical Methods in the Study of Gender and
Agricultural Change in Modern Africa. Current Anthropology 29:247-272.
Hayden, B.
1987 Alliances and Ritual Ecstasy: Human Responses to Resource Stress. Journal
for the Scientific Study of Religion 26:81---9 1 .
1990 Nimrods, Piscators, Pluckers, and Planters: The Emergence of Food Produc
tion. Journal of Anthropological Archaeology 9:31-69.
1993 Archaeology: The Science of Once and Future Things. New York: W. H. Freeman.
Joffe, A.
1998 Alcohol and Social Complexity in Ancient Western Asia. Current Anthropology
39:297-322.
19
March, K. S.
1998 Hospitality, Women, and the Efficacy of Beer. In Food and Gender: Identity and
Power, edited by C. M. Counihan and S. L. Kaplan, pp. 45-80. Amsterdam:
Harwood Academic Publishers.
Moore, ]. D.
1989 Pre-Hispanic Beer in Coastal Peru: Technology and Social Context of Prehis
toric Production. American Anthropologist 91:682-695.
Morris, C .
1979 Maize Beer in the Economics, Politics, and Religion of the Inca Empire. In
Fermented Foods in Nutrition, edited by C . Gastineau, W Darby, and T. Turner,
pp. 21-34. New York: Academic Press.
Murray; M.
1995 Viereckschanzen and Feasting: Socio-Political Ritual in Iron-Age Europe.
journal of European Archaeology 3:125-15r.
Netting, R.
1964 Beer as a Locus of Value among the West African Kofyar. American Anthropol
ogist 66:375-384.
Sahlins, M.
1976 Culture and Practical Reason. Chicago: University of Chicago Press.
1992 The Economics of Develop-man in the Pacific. Res 2r:r2-25.
1995 How "Natives " Think: About Captain Cook, for Example. Chicago: University of
Chicago Press.
Schmitt Pantel, P.
1992 La Cite au banquet: histoire des repas publics dans les cites grecques. Collection de
l' E cole Franc;:aise de Rome, 157· Paris: De Boccard.
Sherratt, A.
1991 Sacred and Profane Substances: The Ritual Use of Narcotics in Later Prehis
toric Europe. In Sacred and Profane: Proceedings of a Conference on Archaeology,
Ritual and Religion, edited by P. Garwood, D. Jennings, R. Skeates, and ].
Toms, pp. 50-64. Oxford: Oxford University Committee for Archaeology, no.
32.
Wiessner, P., and A. Tumu
1998 Historical Vines: Enga Networks of Exchange, Ritual, and Warfare in Papua New
Guinea. Washington, D.C . : Smithsonian Institution Press.
20
Pa rt 1
ETH N O G RA P H I C
P E R S P E CT I V E S
2
FA B U LOUS FEASTS
A P R O L E G O M E N O N TO T H E I M P O RTA N C E O F F E A S T I N G
Brian Hayden
Everyt h i n g that is n ot g iven is lost
Hindu proverb
I am always a bit uncomfortable and self-conscious when people ask me what I

am studying in my research in far-off lands. I anticipate the knowing smiles and
mirthful expressions when I say that I am studying .feasting. I can see that most
people think that here is someone who has found a way to use taxpayer dollars to
achieve personal bliss. Well, it is a gluttonous, thankless task that most serious
scholars have avoided, perhaps for fear of ridicule. But someone has to do it!
Feasting behavior has been largely ignored by archaeologists since the incep
tion of the discipline, and by anthropologists for the last two decades. The fact
that there is almost no body - of archaeological interpretation involving feasts is
probably in part due to the limited theoretical attention devoted to the topic in
general anthropology. While there are many descriptive accounts of feasts in ear-
23
Brian Hayden
lier ethnographies, few anthropologists address the theoretical importance of

feasting in any specific areas (the Northwest Coast potlatch constituting a notable
exception-see Suttles 1968; Piddocke 1965; Harris 1971:248, 394; Ruyle 1973) . Al
most no attempt has been made to examine feasting from a cross-cultural or eco
logical perspective (exceptions include Mauss 1924; Rappaport 1968; Suttles 1968;
Dalton 1977). Perhaps occidental researchers have been biased i n their views o f
the importance o f feasting, attributing such behavior t o a sybaritic self-indulgent
aspect of human nature that is unworthy of serious attention. Perhaps archaeol
ogists have simply written the study of feasting off as a frivolous type of psycho
logical self-gratification that pleasure-loving individuals engage in, but which is
not particularly important for understanding adaptive behavior, economics, so
cial structures, or cultural change, especially given the assumed difficulty of de
tecting feasting in the sparse material remains of the archaeological record.
Whether these or other reasons account for the relative absence of studies on
feasting in recent decades, we intend to change the perspective of the discipline.
� sting behavior in traditional societies is not simply self-indulgent, social, or
gustatory gratification on the part of pleasure-seeking individuals. Rather, as the
following studies demonstrate, feasting is emerging as one of the most powerful
cross-cultural explanatory concepts for understanding an entire range of cultural
processes and dynamics ranging from the generation and transformation of sur
pluses, to the emerg�nce of social and poli�-ine q-;;ili
ties, to the creatiop _2_f
prestige technologies including specialized domesticated foods, and to the under-
1,
writing of elites in complex societie in the past, archaeologists have studied
:J
such things as prestige technology; regional exchange, domestication, and many
other material domains; but archaeologists have neglected the study of one of
the most critical causal phenomena capable of tying changes in all these domains
together: namely; feasting.
There are some good a priori reasons for considering feasting as being an im
portant adaptive behavior for human beings. Cultural ecology maintains that be
haviors that are (r) widespread, (2) persistent over time, and (3) expensive in terms
of time, resources, and / or energy should have definite adaptive values. By adap
tive value, I mean behavior that has some practical benefits for reproduction and
survival rather than the psychological self-gratification, ego-grooming, pride,
prestige, or status benefits that are usually assumed to explain feasting behavior.
Although these psychological benefits certainly are part and parcel of feasting
and may explain some of the individual motivations for hosting or attending
feasts, such psychological factors do not adequately explain the magnitude of
many feasts or why others besides the hosts agree to contribute time, energy; and
surpluses to feasts.
From an ecological point of view, it is clear that feasting is extremely wide-
24
A P R O L E G O M E N O N TO T H E I M P O RTA N C E O F F E ASTI N G
spread, if not universal; that it is extremely persistent, probably dating back to the
Upper Paleolithic (Conkey 1980), if not before; and that in many instances it re
quires years of preparation and surplus accumulation, extending even into fu
ture, debt-ridden years due to the deficit financing of feasts. Clearly, there is
something of substantial importance transpiring in these cases, as emic state
ments about the importance of feasts so frequently proclaim. What are these
weighty matters that archaeological and anthropological inquiries have been so
silent about until now?
For the vanguard of archaeologists and anthropologists that has taken up the
quest to find the underlying significance of feasting, the challenge is daunting.
Yet there are many clues provided by some of the excellent descriptive ethnogra
phies written in the 1970s and in earlier years. There are also early written records
and pictorial representations of feasts (as demonstrated in Chapter 14) . There are
important principles of inference provided by animal ecologists concerning the
observation of behavior, its context, consequences, and inferred adaptive signifi
cance. Moreover, there are ongoing possibilities of making new observations
about feasting in societies that still use feasting to structure their economic, so
cial, and political worlds. In addition, we can ask individuals in traditional soci
eties why they host or support feasts. We hope to employ all of these avenues in
trying to understand the role and significance of feasting behavior. We will mon
itor the specific interactions, the magnitude of the event, the cost and effort ex
pended by hosts, the outcomes, and the material signatures of these events. We
will ask probing questions of those most involved in feasts. We are confident that
important conclusions will result.
At the outset, I suggest that it is useful to make several important distinctions
in analyzing feasts. Form versus symbolical content is one distinction. By form, I
refer to the overall nature of the behavior that we are trying to explain, whether
large-scale feasts, the construction of massive architecture, or the manufacture of
costly items that are meant to create spectator reactions of awe. By content, I refer
to the specific symbolical meaning emically attributed to a specific behavior or to
the_creation of a particular obj ect. Such content could be the particular meaning
· attributed to the presentation of a particular kind of food, pretexts for holding
feasts (e.g. , to please an ancestor), the meaning of spiral decorations, or any num
ber of other symbolical meanings. Previous studies of feasting have often be
come bogged down in the myriad, culture-specific, indigenous symbols of
feasting behavior. Perhaps this is why there have been few attempts at cross
cultural or ecological understanding of feasting behavior. I suggest that our first
task is to understand the reason for the emergence of a particular behavioral
form, like competitive feasting or the building of megaliths. If we can resolve
these issues, then it may be of interest to go on and attempt to disentangle the
25
Brian Hayden
problems related to symbolical meanings. Many of the articles in this volume, in

fact, focus primarily on questions of form and function rather than symbolic
content of feasting.
Five key chapters directly address the basic issues of the underlying functions
of feasts. The analysis by Michael Clarke on the creation of political and social
safety nets via feasting among the Akha (Chapter 5) is an important ethnoarchae
ological foundation study of feasting. Equally important are Laura Junker's
analysis of how feasting was used in Philippine chiefdoms to create political and
military alliances to control lucrative trade (Chapter rn), Michael Dietler's analy
sis of African feasting in order to centralize political and economic power (Chap
ter 3), and Michael Dietler and Ingrid Herbich's documentation of how work
feasts can increase socioeconomic inequalities (Chapter 9). Polly Wiessner
(Chapter 4) adds an important dimension by looking at sacred feasts in New
Guinea meant to promote solidarity and support for secular feasts where self
interest is pursued in a much less restrained fashion. Then in a refreshing reex
amination of the potlatch, James Perodie also examines the diverse underlying
motives for feasting on the Northwest Co ast (Chapter 7) .
.
ECOLOGY A N D F EASTI N G
To broaden the context for understanding feasting behavior from an ecological
viewpoint, I propose that it is worth considering feasting as one component, al
beit a major component, of what I like to refer to as "social technology." Social
technology can be defined as the creation and maintenance of social relation
ships that are predicated on securing access to resources, labor, or security (see
Keesing 1975:122 for a general discussion how social factors are related to ecol
ogy). Other facets of social technology certainly include many aspects of kinship
(real and fictive), ritual (Hayden 1987), gift giving, and language. Indeed, since all
these behaviors can have the same ultimate goal, such as establishing subsistence
or defensive alliances (e.g., Wiessner 1982; Dalton 1977), it is far from coincidental
that they all tend to occur and be used together in the same contexts, although
the relative emphasis may vary from one instance to another. Development of
such an elaborate social technology (rather than material technology, communi
cation, or intellectual abilities) is perhaps the most distinctive aspect of human
nature that sets us apart from the rest of the animal kingdom. Feasts, unlike kin
ship and language, have clear archaeological consequences, which I will enumer
ate shortly. Cultural and behavioral ecologists have been remarkably resistant to
examining the adaptive value of feasting and other social technological behav
iors, choosing instead to examine optimal foraging choices of resources and
monitoring cost benefits of resource exploitation. Perhaps they have adhered too
closely to the ecological model established by mainstream biological ecologists.
26
Most ecologists tend to emphasize the fact that human beings are animals like
many others in the world and that human behavior, too, should conform to eco
logical models. They are probably correct in thinking that differences in language
ability and intelligence by themselves do not necessarily change the fundamental
nature of ecological adaptations. Yet, there is a critical difference between the be
havior of other animals and human behavior in more complex cultural systems
that does create a vast gulf between humans and animals. This difference has an
extraordinarily profound effect on how people use resources and on which eco
logical models can be applied to human behavior. This difference is simply: the
ability of humans to transform food surpluses into other kinds of usefal or desirable
goods and services_. Because of the distinctive dynamics of human behavior in this
domain, I refer to its study as "political ecology."
Other animals and insects, such as squirrels and bees, may harvest and store
food surpluses; however, no other animals transform those surpluses into usable
nonfood items or services that have real consequences for survival (Hayden
1994,1998). For other animals, the value of food resources is ultimately limited by
individual or group metabolic needs. In contrast, humans invest surplus food and
labor, with a remarkable ability to expand consumption constraints. Feasting and
gift giving are probably the principal means for transforming surpluses in order
to improve chances of survival and reproduction. Feasting is, above all, con
cerned with surpluses, their production, use, transformation, control, and distri
bution. We use feasts to display our success whereas other animals use antlers,
plumage, calls, colors, or manes (Zahavi and Zahavi 1997) . In fact, numerous an
ecdotal accounts, as well as several more systematic studies (e.g. , Blanton and
Taylor 1995; Izikowitz .1951 :341, 354; Stanish 1994; Cowgill 1996; Blackburn
1976:242; Friedman 1975; Friedman and Rowlands 1977=208-214-see also Wiess
ner' s Chapter 4) indicate that the drive to achieve advantages throughfeasting is prob
ably the single most important impetus behind the intensified production of
surpluses beyond household needs for survival. The transformation of surpluses
is a unique human ability, but one not without rudimentary origins in other pri
mates, as exemplified by food sharing for favors. Transformation of surpluses cre
ated an entirely new ecological dynamic that has been consuming world resources at
a geometrically increasing rate since the advent of transegalitarian feasting and
prestige technologies (Hayden 1998) .
N E W A P P R O A C H E S TO F E A S T I N G B E H AV I O R
D EF I N ITI O N S
Given this new approach to the study of feasting, what specific questions need to
be asked? At the outset, we need an operational definition of feasting. There are
also a number of important issues that need to be resolved. Perhaps the most
27
Brian Hayden
difficult issue is how to meaningfully categorize the extensive variety of feasts

recorded ethnographically. We should determine which of these types may have
had practical benefits and what those benefits are likely to have been. Under
standing how different types of feasts are related to social, economic, and politi
cal dynamics of traditional communities constitutes yet another maj or
theoretical undertaking. Finally, understanding how different types of feasts can
be recognized archaeologically is an area that is critical for prehistorians but
largely unexplored. These are the major themes of this book. I would suggest
that if we are ever to get beyond the detailed, myopic view of many social an
thropologists who see the trees well enough but who have little idea about the
forest, it is essential to take a cross-cultural and comparative perspective .
To begin with definitions, I would like to propose that a feast be defined as any
sharing between two or more people of specialfoods (i. e . , foods not generally served at
daily meals) in a meal for a special purpose or occasion. As Dietler (1996: 89) has
pointed out, there is usually a highly ritualized component to feasting as well;
however, it is conceivable that in some cases the ritualized content may be fairly
minimal (as at some dinner parties in contemporary societies). Daily family
meals can also be highly ritualized (including prayers or the saying of grace) but
would not constitute feasts . The definition that I use above explicitly excludes gift
exchanges without meals; it excludes food offerings to spirits on the part of indi
viduals or groups who do not themselves eat or drink; it excludes communion
style consumption in churches; it excludes sacrifices in which there is no
consumption of food or special drink; and it excludes group meals shared for me
chanical convenience, as in cafeteria-style meals that are part of workshops, ritu
als, or other events . Other slightly different definitions of feasts are presented by
Michael Dietler, Polly Wiessner, Michael Clarke, Laura Junker, and Linda Brown
(Chapters (3, 4, 5, ro, and 13). However, in archaeological operational terms, the
use of foods in a nondomestic pattern is a recurring central feature (Chapters 5,
10-13 by Clarke, Junker, Knight, Kelly, and Brown) .
The term adaptive value simply indicates behavior that generates some practi
cal benefit for survival, reproduction, health, or standard of living.
PURPOSES
Probably the most contentious issue in the study of feasts is why feasts are held. Do
hosts really expect to gain practical benefits from every feast that they hold? If they
deny such motives, or are not conscious of them, is it possible to still impute an
adaptive value to such feasting behavior? Have the original or barely conscious
practical purposes been lost sight of amidst the fog of ideology and symbolism that
is usually generated by the aggrandizer-promoters of feasts? Have religion and rit
ual obscured the real relationships of power vested in the elites, as Marx argued, or
become the shared "sincere fictions" of entire communities, as Bourdieu (1990)
28
suggests? To take an example from Western culture, how would most people ex
plain the lavish dinners and exchanging of presents at Christmas? Many, if not most
people would explain these feasts in symbolical (content) religious terms, as cele
brations of the birth of Christ. And this is precisely the kind of explanation that is
most common in the ethnographic descriptions of feasting (as funerals necessary
for the safe j ourney of the dead's soul, etc.). Yet, few social scientists today would
hesitate to attribute a more basic, underlying, explanation to Christmas feasts as es
sentially highly ritualized events that are meant to create social solidarity and coop
erative bonds between family members and close social affiliates.
In order to reach such conclusions, it is not necessary to depend on emic ratio
nales, pretexts, or ideology to explain why specific types of feasting occur. How
ever, it would be foolish to entirely ignore what practitioners, especially
insightful practitioners, have to say about these matters as well. Questions about
the symbolical content of feasts and about the formal functions of feasts repre
sent very different levels of inquiry. In my fieldwork, I have always found that
people in traditional societies can easily articulate answers concerning symbolic
content, but rarely seem to comprehend questions about formal aspects of feasts
or other behavior such as why funeral feasts are so large and expensive. Thus, a
judicious combination of empirical observation and emic commentary is proba
bly the optimal avenue of inquiry. Empirical observation of behavior, contexts,
materials, and outcomes has been the standard for interpreting animal behavior
and much human behavior for well over a century. It can also be successfully used
in the study of feasting especially if we ever hope to deal with formal questions
of function. Although archaeologists must rely heavily on ethnologists and eth
noarchaeologists to understand the basic nature of feasting behavior (as well as
all other basic forms of human behavior), once its material correlates have been
established, archaeologists are in an important position to delineate the evolution
of such behavior, and understand the conditions under which it emerged and di
versified into specific types of feasts. The studies in this volume by Junker, Di
etler, Clarke, and Perodie constitute ample evidence of the utility of this formal
ethnographic approach.
P RACT I C A L B E N E F I TS
Given our present state o f knowledge about feasting, what are the most obvious
or important practical benefits that can be proposed to account for the substan
tial time and resources expended in some feasts? There are nine basic types of
practical benefits that I would suggest occur most commonly. Feasting can:
r. mobilize labor;
2. create cooperative relationships within groups or conversely, exclude dif
ferent groups;
29
Brian Hayden
3. create cooperative alliances between social groups (including political sup

port between households);
4. invest surpluses and generate profits;
5. attract desirable mates, labor, allies, or wealth exchanges by advertising the
success of the group;
6. create political power (control over resources and labor) through the cre-
ation of a network of reciprocal debts;
7. extract surplus produce from the general populace for elite use;
8. solicit favors;
9. compensate for transgressions.
Except for work feasts, penalty feasts, and solicitation feasts, all other benefits
of feasting revolve around the creation or maintenance of important social relation
ships. These relationships may be important for different reasons (defense, mar
riage, wealth accumulation, and so on-Dalton 1977:202-204), but establishing
desirable social relationships constitutes the bottom line for many feasts. This is
an extremely important feature of some feasts that has far-reaching implications
for the transformation of technology and culture. Notably, in order to create a fa
vorable disposition among guests, hosts generally try to demonstrate the special
ness and importance of the guests by presenting them with special foods, drinks,
gifts, or ritual displays as indications of how much the hosts value their guests.
The more important the relationship is, the more effort, time, and resources
should be involved in giving an impressive feast. In potlatch-type alliance feasts,
Dalton (1977:207) stresses that feasts are made as lavish as possible to demonstrate
the wealth and power of the hosts, and therefore the desirability of hosts as allies.
Trying to impress one's guests, for whatever reason, means obtaining and prepar
ing labor-intensive foods, drinks , serving vessels, prestige items, and ritual
items-an extremely dynamic and powerful engine for the generation of cultural
and technological change (Hayden 1998).
The above benefits of feasting, are, of course, not meant to exclude competi
tive attempts to diminish rivals. However, I view such competitive feasting as sub
sumed under strategies to obtain political power, including displays of success
(which may only make sense in competitive contexts) . It should also be noted
that there is probably a major difference between the practical benefits (listed
above) to the hosts as a group versus the practical benefits to the promoter and
organizer of feasts who, in the most active cases, is usually a Triple A personality
(aggressive, aggrandizive, and accumulative) out to maximize his own self
interest, wealth, and political power (see Hayden 1995). Not all lineage heads who
are responsible for organizing lineage feasts and maintaining alliances advanta
geous for their lineage promote their own self-interest over their lineage's inter-
30
est. But demographic characteristics ensure that some lineage heads, clan heads,
and community heads will be aggressive aggrandizers, and they will be the most
notable ones with the most impact on cultural change. These aggressive aggran
dizers promote the benefits of feasting to the community or their support group
and try to get as many people to produce as much surplus as possible. The ag
grandizers then assume as much control over the use of surpluses for their own
benefit as possible. These aggrandizers attempt to set the agendas, they manipu
late war, peace, and alliances, and they create cults and secret societies to further
their own self-centered goals (Chapters 4 and 7; also Hayden 1995) .
While many indigenous informants and social anthropologists emphasize psy
chological reasons that people give feasts, I would like to argue that concepts such
as "status" and "prestige," as usually used in Western society, really hold little ex
planatory value on their own. If they are to be entertained as causal motivations
for holding large costly feasts, they should be scrutinized carefully I have three ob
j ections to these concepts. First, they posit some inherent (and in many accounts,
economically self-destructive) drive for psychological approval from others. Such
desires for approval, or at least acceptance, certainly exist in many people, but the
variability and distribution of their intensity has never been documented, nor has
the emergence of these emotions been convincingly tied to any cultural or biolog
ical evolutionary theory that I am aware of. Thus, one wonders what their selective
advantage might have been, and why such traits should have become widespread
in human populations as posited by proponents of these views.
Secondly, admitting that some people do have strong psychological penchants
for seeking approval from others, the magnitude of expenses involved in some
feasts (with supposedly no other benefit than achieving approval from others)
seems out of line with the ego-gratification received. Indeed, in some societies
such as the Akha, hosts are unusually obsequious and display little satisfaction or
pride as noted by Clarke (Chapter 5). Nor does it seem realistic that the desire for
approval would generally outweigh desires for personal gain or turn most rational
beings into economic lemmings. Moreover, it is highly unlikely that the prestige or
status resulting from giving gifts or feasts would be long-lived, just as gift giving,
by itself, in contemporary society is a tenuous means of establishing influence or
prestige. Compulsive gift-givers are just as likely to be ridiculed as fools behind
their backs, as they are to be revered for their gifts. In order for gift giving to have
longer lasting effects, it must be backed up by other features, such as the creation
of agreed upon or de facto contractual debts-an aspect of traditional feasting
that Laura Junker strongly emphasizes in her chapter, as do Mauss (1924), Dalton
(1977:205, 207), Gosden (1989), Lightfoot and Feinman (1982:66), and others.
Thirdly; given the variability in human genetic and personal developmental
histories, it is certainly possible that there will always be a few people who feel
31
Brian Hayden
compelled to accumulate large amounts of wealth in order to be able to dispense

it simply in the hopes of acquiring honor or prestige-especially perhaps under
conditions of cultural collapse (e.g., see Dalton 1977:207 on changes in the pot
latch). The laws of chance dictate that such idiosyncratic individuals will emerge
from time to time. However, I contend that such personality types are rare and
that the vast majority of individuals in all cultures have, and always will follow
their own practical self-interests in the long run. It is doubtful that trying to es
tablish cultural traditions and institutions that are costly to individuals solely on
the basis of a purported widespread desire for honor, status, or prestige would be
successful or long-lasting without some real practical benefits to back up psycho
logical benefits or other ideological pretexts (e.g. , pacifying the ancestors, or
courting the favor of the gods).
Thus, it seems unlikely that Western concepts of status, prestige, honor, or ap
proval from others by themselves play a dominant role in the institutionalized
hosting of feasts involving large numbers of people and major expenditures of
time, energy, and resources. However, it must be emphasized that all of these are
Western sociocultural analytical concepts. English terms may only be the closest
approximations to indigenous terms with significantly different connotations. So
cial anthropologists have often taken these translations at face value. However, I
am convinced that a careful reading of the texts and the implications of status or
prestige clearly indicate that traditional usage of these terms carries a very differ
ent meaning from the Western translations. In brief, my impression is that while
the Western terms status and prestige carry connotations of psychological gratifi
cation stemming from the approval of others, equivalent terms in transegalitar
ian societies generally carry a different set of connotations relating more to
economic success, political success and power, reliability in honoring debts, and
the ability to organize people for a variety of purposes. Citing Voss (1987=131),
Laura Junker makes the point explicitly: in the Philippines, the prestige came
from the creation of social debts through feasting. Dalton (1977:207) also clearly
views the lavishness of feasts as statements of wealth and power and success.
Similarly, Friedman and Rowlands point out that the surpluses used to host feasts
result in prestige, but a "prestige whose cultural content is very different than in
our own society" (1977:207). The status gained is "the 'social value' attributed to
the ability . . . to produce such wealth." In Chapter 4, Polly Wiessner provides
additional observations along these lines. She suggests that prestige accrued to
those who brought benefits to the group; however this "prestige" was really a li
cense for aggrandizers to produce wealth, marry many women, and retain ser
vants. In summary, I would argue that a more appropriate translation of status
and prestige in contemporary terms might be "credit rating."
The antithesis of the status that ambitious people seek, especially in transegal-
32
itarian societies, is embodied by people who are systematically reviled as "lazy;"

"moochers," "dead skin," or "rubbish men" (see Chapter 7; also Yan 1996:103).
The vilification of such people is a recurring theme, if not an obsession, of elites
in many transegalitarian cultures. The consequences of being considered a
moocher or rubbish man are that one's credit rating is low or nonexistent for
loans, reciprocal gifts, reciprocal feasts, and other events because such people are
unreliable and unproductive . In many cases, these people cannot marry because
of these liabilities. They are the people who do not support aggrandizers' strate
gies for generating surpluses. The worst fear of aggrandizers is to have to default
on feasting and other debts. This precipitates their fall to the status of rubbish
man, a person to whom no one will lend wealth or only at exorbitant rates, a per
son who has essentially declared bankruptcy. They have no "status."
These are only my impressions gleaned from the ethnographies and others'
analyses in passing. While a full exegesis of "status" and "prestige" in transegal
itarian societies would be extremely enlightening, I am not prepared to under
take such a task here. Nevertheless, it is perhaps worth noting that a strong
argument can probably be made for this new meaning of status (i. e . , as success
recognition) as having an adaptive value. Evolutionarily, it should be adaptive to
imitate individuals who display clearly successful behavior, as it should also be
adaptive to reproduce with such individuals. In fact, the biological world is re
plete with examples of species where individuals compete with each other to
display their superior biological success, sometimes even to their own detri
ment, as in the case of elaborate moose or elk antlers (Beardsley 1993; Zahavi
and Zahavi 1997). Displays of success are vitally important for attracting desir
able mates and supporters or coworkers, both in the biological realm and in the
cultural realm. Feasts constitute one of the most prominent venues for display
ing success among humans, and "status" is the verbal symbolic expression of
that success.
The standard response of more sociocultural anthropologists at this juncture
in the dialogue is to invoke the power of cultural traditions, norms, values, and
beliefs in curtailing people's own practical self-interests and promoting the give
away of wealth for more etheric goals such as recognition and approval by oth
ers. While cultural norms may urge some people to take on costly social roles
without providing practical benefits, and while this may occur in some cases over
the short term, my experience in tribal villages indicates that even in the most
"tradition-bound" cultures such as the Maya, the Akha, and the Western Desert
Aborigines, when self-interest no longer accords with cultural traditions, then
the cultural traditions wither and die, provided other options are viable. As
Lewis-Williams notes: "Cosmological, social, religious and iconographic frame
works are not immutable givens. They are reproduced through complex
33
Brian Hayden
processes and forms of representations and are therefore always open to negotia
tion" (1997: 827-emphasis added) .
Similarly Blitz (1993a:23) argues that individuals do not merely react to the so
cial idiom of ritual but continually create, alter, reinterpret, or manipulate its in
formation content with dramatic results. Indeed, if this were not the case, we
would be in dire straits in trying to explain how and why cultures change. We
would be thrown back to relying on diffusion or similar explanations. In fact,
Marxism, Cultural Materialism, Post-Processualism, Ecology, Darwinism, and
Processualism all share a commitment to viewing individual actors as the locus
of change and as rationally pursuing their own self-interests as a fundamental
tenet of understanding the workings of culture.
There is no more dramatic demonstration of the relatively weak hold that cul
tural traditions have over people's behavior than the occurrence of noncon
formists and revolutionaries. In the tribal cultures where I have worked, I have
always been impressed at the large percentage of individuals who simply did not
buy into the traditional value system and who would be generally classified as ag
nostic or atheistic in contemporary terms. For example, in the very traditional
Maya Highland villages where I worked, about ro percent of the household heads
interviewed would have to be classified as a-religious or agnostic, while another
ro-20 percent of households had changed their beliefs and affiliation from tradi
tional community cults to more charismatic forms of Christianity. Izikowitz
(I95I:J2I) makes similar observations for Southeast Asian tribal societies. It seems
that in all cultures, there are probably significant proportions of independent
thinkers who evaluate situations on their own merits irrespective of handed
down traditional wisdom or promulgated ideologies. It is the individual who is
the locus of cultural selection (Harris 1979:61). Cultural values or ideas do not fet
ter individual actions to the point of acting against important self-interests-real
consequences from other community members do. It is the freethinkers who pi
oneer creative new thoughts and new behaviors, and who serve as models for
possible new cultural, social, political, and economic configurations. If this were
not the case, there would be no unions or revolutions, and little cultural change.
To illustrate the relative roles of ideology versus practical benefits, it is useful
to consider the parable of the shopping cart. Before 1970, supermarket storeown
ers appealed tirelessly to the traditional cultural values of customers (neatness,
courtesy, reciprocity, etc.) to return their shopping carts to the store rather than
leaving them in parking lots. However, appeal to these supposedly strong cul
tural norms had little effect on people's actual behavior until store owners began
adding a practical consequence to the return of shopping carts-the investment
of 25 cents to obtain a cart, redeemable upon the return of the cart to its proper
storage location. The change in behavior was dramatic. Even stronger cultural
34
A P R O L E G O M E N O N TO T H E I M P O RTA N C E O F F EASTI N G
values were involved in trying to prevent people from polluting and littering, all
to little avail until substantial fines and bottle / can deposits were establish�d. The
moral of these parables is, as Marvin Harris (1979:59, 270-277) long ago argued,
that while cultural values certainly do exist and are used to back up or justify ac
tual behavioral choices, practical consequences play a much more dominant role
in behavioral choices.
In terms of feasting, one of the most powerful enforcing criteria is the accept
ance of a contractual debt when one accepts an invitation to a feast intended to
create social bonds or reciprocal obligations. Debt relationships also prolong and
maintain the "status" associated with gift giving because the "superior" status is
active as long as the gift has not been repaid. And it usually takes one or more
years for many gifts to be repaid. Moreover, overdue unpaid debts entail socio
economic rupture between social groups (Dalton 1977:205, 207), engendering ad
ditional weighty consequences. But gifts and debts generally constitute
immediate advantages. The ulterior and adaptive advantages have been noted
above (alliances, attraction of mates and labor, investments, profits, performing
work, and political control through debts). It is certainly necessary to acknowl
edge the full range of idiosyncratic behaviors, but it is also necessary to carefully
differentiate these from the behavior of the maj ority of people, upon which
major traditions and institutions are based, including feasting.
A number of the chapters in this volume carefully document the dynamics,
benefits, and strategies of feasting using ethnographic and historical data. Laura
Junker provides a broad overview of the role of feasts in acquiring political and
economic power in Southeast Asia, and Michael Clarke focuses on the use of
feasts to establish economic and social security networks in tribal Thailand. Polly
Wiessner examines how big-men manipulate group solidarity and values
through feasts for their own benefit in New Guinea. Both James Perodie and
Michael Dietler document the broad range of practical benefits that feasts pro
vided on the historic Northwest Coast (Perodie) and sub-Saharan Africa
(Dietler). In an innovative analysis of Sumerian data, Denise Schmandt-Besserat
also demonstrates how feasting was used by early state elites to underwrite early
Mesopotamian state economies.
VA R I E T I E S O F F E A S T S
Having discussed the raison d'etre of feasts, it is now possible to deal with the
question of how to classify feasts, for not all feasts are the same. Indeed, there is a
wide range of different typologies of feasts. It is possible to classify feasts by:
1. symbolical content;
2. inferred functions (types of practical benefits);
35
Brian Hayden
3. size;
4. goals of creating social bonds vs. achievement of more immediate, lim-
ited objectives;
5. the use of prestige materials or other archaeological indicators;
6. participating, or core, social units;
7. horizontal vs. vertical social relationships between guest and host;
8. the kind of reciprocity involved;
9. the degree of obligation (social necessity vs. self.initiated hosting);
rn. seasonal or calendrical occurrences vs. life or economic conditions.
Yan (1996:44ff) discusses classification in regard to the parallel phenomenon

of gift giving. The traditional ethnographic approach has been to classify feasts
according to their emically stated purposes, based on the specific symbolical
pretexts for holding feasts: harvest feasts, ancestral feasts, marriage feasts, fu
neral feasts, puberty feasts, and others. For the understanding of feasting as a
behavioral form (in contrast to the symbolical content of feasting), emic classi
fications are clearly the least useful type of classification, although some do
tend to have specific functions and can be classed within one or more specific
functional types of feast (e.g. , community harvest feasts as solidarity and/ or
promotional feasts).
All of the other approaches have much more utility for archaeologists, and in
fact, many of these other classificatory approaches are used by authors in this
volume. Michael Clarke uses the range of social groups involved, James Perodie
uses reciprocity expectations, and Michael Dietler uses a political function classi
fication. Polly Wiessner uses a secular versus sacred distinction with differing pat
terns of refuse and special structures. My own preference is to adopt a functional
approach as an intermediary step to generating an archaeological classification
based on material remains. Unfortunately, because of the complexity of feasting
behavior, there are seldom feasts with pure functions, and in some cases the same
emic pretext for a feast (e.g. , a funeral feast) may have quite different functions in
different households (e.g. , in a poor versus a rich household). In order to deal
with such distinctions, it is either necessary to add a great deal more detail to our
descriptions and analyses of feasts or to create more subcategories of functional
or traditional emic categories of feasts. However, I am hopeful that this kind of
confusion will largely disappear when we pass to archaeological classifications of
feasts. The endeavor to create such a useful archaeological typology is still in its
earliest stages and is one of the principal reasons why we have assembled this
volume.
Of the ten approaches to classification just listed, the following are probably
the most common or have the most potential for use by prehistorians.
36
A P R O L E G O M E N O N TO T H E I M P O RTA N C E O F F E A STI N G
1 . Emic: Ernie feast types are theoretically almost limitless, since the kinds of
pretexts that might be used for holding feasts are only constrained by
human creativity. People are constantly trying out new pretexts for hold
ing feasts or giving socially binding gifts (Yan 1996:60 ). However, it seems
that people everywhere find some pretexts for holding feasts much more
compelling than others. Thus, there are some extremely common recur
rent emic types. These include feasts for marriages, funerals, children's
maturation events, plantings, harvests, work, status, new houses, making
war and making peace, rituals, ancestor worship, and noteworthy celestial
or seasonal events. In Chapter 3, Dietler provides a discussion of why such
emotionally, theatrically, and ritually charged contexts are especially appro
priate for feasting events.
2. Functional: In terms of practical purposes or social functions, there is a
fairly narrow range of important proximate benefits likely to be derived
from hosting feasts. These benefits can be grouped into two main divi
sions: creating cooperation, alliance, or social distinctions on the one hand
versus economic benefits on the other hand (Fig. 2.1). Feasts in the first di
vision include those meant to create cooperative in-group relationships or
distinctiveness between groups (Michael Dietler's diacritical feasts); feasts
to create alliances between groups; feasts to create political support; and
feasts to attract desirable mates and labor via advertising. The division of
feasts having to do with proximate economic benefits includes feasts held
to accomplish some task, investment feasts (including high-cost matura
tion events), competitive feasts, solicitation feasts, and feasts given in lieu
of punishment. It may be possible that another basic category is common
in which emergency situations (due to climate, disease, or other catastro
phes) make people willing to surrender surpluses to charismatic leaders
who promise relief if a large enough feast or ritual can be held. It probably
makes most sense to view these situations as being manipulated purely for
the benefit of those organizing the feasts who try to expropriate the sur
pluses of others. They therefore might be termed "exploitative" or
"calamity" feasts. I have argued that despot aggrandizers often intention
ally create disputes between villages precisely in order to create a climate
of crisis and impending calamity that the despots manipulate for their own
advantage, largely via the need for alliance and compensation feasts (Hay-
. den 1995). Although some of the details of Michael Dietler's terms and
functional classification of feasts (Chapter 3) may differ from the ones I use,
we use many of the same basic distinctions.
I feel that it is important to attempt to distinguish between different
functional types of feasts because the consequences for the material cul-
37
Brian Hayden
SOLIDARITY SOLICITATION
FEASTS FEASTS &
(Within groups) P U N I S H M ENT
FEASTS
RECIPROCAL COM PETITIVE

FEASTS FEASTS
(Profits from loans
&
(Between groups)
interest)
FEASTS TO
POLITICAL SUPPORT ACQU I R E
FEASTS POLITICAL
(To obtain political POSITIONS
supporters) (As a formal
criterion for political
advancement)
PROMOTIONAL
FEASTS MATURATION
(To display success & FEASTS
attract labor or (Investment of
economic support) surpluses in
children)
WORK PARTY
FEASTS
TRI BUTE
FEASTS
Figure 2.1. A schematic diagram of the most important divisions in

the types of feasts discussed in this chapter.
ture expressions should be quite different and because the consequences for
community dynamics and impacts on technological changes should be very differ
ent. For instance, sohdarity feasts, no matter what their size, should entail
minimal departures from standard daily foods or material items, whereas
competitive and promotional feasts should represent major departures
with consequent pressures to develop and change both food and material
technologies.
3. Size: Feasts vary enormously in size from the minimum of a two-person
(dyadic) sohcitation or friendship (sohdarity) dinner to intercommunity
events involving hundreds or thousands of people. Clearly the smaller
types of feasts will be difficult to identify archaeologically unless the re
mains of special foods or special serving/ preparation vessels are present.
Even then, it may be difficult to determine whether such remains are from
small special meals or whether they were part of larger, more lavish spe-
38
cial feasts. However, one might expect some differences to occur with in
creasing sizes of feasts. In general, perhaps the more people involved, the
more specialized and numerous food items, food preparation facilities
(hearths, roasting pits, kitchens), food preparation and serving vessels, and
architectural structures will be. Thus, irrespective of function or emic pur
pose, one useful way of classifying feasts for archaeological purposes may
be by sheer scale. While this approach enables us to make some sense of
the archaeological data, I would argue that it might obscure some impor
tant potential distinctions that might be made with a bit more diligent
enquiry.
4. Social bonds vs. limited goals: Many limited-goal feasts (Yan 1996) probably
leave minimally distinctive material remains. Work feasts, solicitation
feasts, punishment feasts, crisis feasts, and similar events are either rela
tively small, and/ or are unlikely to involve particularly special foods or
vessels. In these cases, we need to develop better criterion for distinguish
ing such feasts from normal eating. However, Michael Dietler (personal
communication) has observed some work feasts where lavish food and
prestige wares were used in order to attract more and better-quality labor.
Thus, limited-goal types of feasts may be either minimally distinctive from
normal meals except for size and perhaps location, or they may be hard to
distinguish from other feast types. There may be some useful distinctions
for archaeologists in this approach to classification, but the issue needs to
be explored in greater depth and more empirical data are sorely needed.
5. Social units: One way of adding meaning to size classifications of feasts is
by examining the social units involved in feasts, as suggested by Michael
Clarke. Social units range from individual families, to lineages, to clans, to
communities, and extend to regions. While there are some distinctions
that can be made at each level, the higher levels tend to include lower lev
els (e.g. , lineages include individual families), and thus, each level of social
unit tends to be associated with a specific size range (number of partici
pants) of feasting. This is a productive way of analyzing feasts for archaeo
logical purposes, although within each social group it may still be
desirable to try to' distinguish feasts held for different purposes, especially
solidarity versus promotional or competitive feasts.
6. Material-based (archaeological) classifications: In my estimation, it would be
ideal if we could develop an archaeological classification of feasts that ex
actly matched ethnographically known or inferred functions of feasts.
However, at this juncture it seems highly unlikely that such a high degree
of specificity will soon be attainable using archaeological remains. Instead,
the archaeologically attainable is more likely to involve various combina-
39
Brian Hayden
tions of material indicators, creating new distinctions between feasting

events that will be significant for understanding how past cultures func
tioned and changed. Initially, it is probably most useful simply to list the
kinds of material remains that can be used to identify feasting events from
the past (see Table 2.1). Various values and combinations of these same
TABLE 2.1
Archaeologica l Signatures of Feasts
Food Rare or labor-intensive plant or animal species (especially

condiments, spices, and domestic animals)
Special "recreational" foods (e.g., tobacco, opium, cannabis, and
alcohol)
Quantity of food items
Evidence of waste of food items (e.g., deposition of articulated
joints, unprocessed bone)
Preparation vessels Unusual types (e.g., for beer-making, chili-grinding, perhaps
initial appearance of cooking pots)
Unusual large sizes
Unusual numbers
Serving vessels Unusual quality or materials (e.g. , first occurrence of pottery or
highly decorated or specially finished pottery, large gourds,
stone bowls)
Unusual size of serving vessels
Unusual numbers of serving vessels
Food-preparation facilities Unusual size of facilities (e.g., large roasting pits or hearths)
Unusual number of facilities (e.g., several hearths in a row)
Unusual location or construction of facilities
Special food-disposal features Bone dumps
Special refuse fires containing feasting items
Feasting middens
Feasting facilities Special structures (temporary vs. permanent) for highest-ranking
guests and hosts, or for large numbers of people
Special display facilities, scaffolds, poles, or other features
Special locations Mortuary or remote locations that are clearly not habitation sites
(e.g., in front of Megalithic tombs, at henge monuments,
inside caves)
Loci associated with nuclear households, residential corporate
households, large feasting middens or central community
spaces
Associated prestige items Presence or absence, and relative abundance of prestige items
typically used in different types of feasts (e.g. , ritual display
items, feathers, shell jewelry)
The destruction of wealth or prestige items (via intentional
breakage or burial)
40
TABLE 2.1 continued
Ritualized items of etiquette Smoking or other narcotic paraphernalia

Ritualized vessels for consumption of alcohol, chocolate, kava, or
other prestige drinks
Paraphernalia for public rituals Dance masks or costume elements
Existence of aggrandizers Wealthy burials; social or site hierarchies; large residences with
high storage per capita
Recordkeeping devices The presence or absence and frequency of tally sticks, counting
tokens, or symbolic pictograms
Pictorial and written records of
feasts
Food-storage facilities Stables, storage pits, granaries
Resource characteristics Abundance, intensified exploitation, invulnerability to
overexploitation
factors should be useful in trying to distinguish different types of feasting

events, no matter which classification approach one adopts.
Undoubtedly, more types of observations can be added to the list in Table 2.1.
However, even at this early stage of theory development, archaeologists probably
have more variables to work with than might be initially assumed in their quest
to deal with past feasting behavior. Many chapters in this volume focus on a num
ber of specific material indicators of feasts. Warren DeBoer discusses the unusual
size and decoration of feasting vessels in the Amazon as does Laura Junker for
the prehispanic Philippines, Linda Brown for the Maya, and Michael Clarke for
the Hill Tribes of Thailand. DeBoer also provides a good, albeit somewhat un
usual example of specialized ritual paraphernalia used in overall feast contexts, as
does Linda Brown for prehistoric rural Maya feasting. Almost all chapters deal
with unusual foods consumed in feasts. The dominant role of animals in feasting
is emphasized by Knight for early Moundbuilders, Lucretia Kelly at Cahokia,
Junker for the early Philippine chiefdoms, and Clarke for contemporary Thai Hill
Tribes. Large game such as deer, and large domestic animals feature most promi
nently in all these contexts, leading to interesting models involving the role of
feasting in the domestication of animals. Brown makes similar observations for
the Maya but is also able to document the use of special plant foods such as
achiote. Similarly, the special nature of feasting refuse is documented in many of
these chapters (e.g. , Kelly, Knight, Wiessner), and is the special focus of Wilson
and Rathje's analysis of contemporary industrial garbage.
A number of papers focus on the emergence of specialized structures related
to feasting activities. Linda Brown identifies a rural Maya feasting and ritual
41
Brian Hayden
structure, possibly related to lineage ancestor worship. James Knight argues that
the early platform mounds of the southeastern United States constitute special
ized feasting structures with scaffolds, while Patrick Kirch documents the spec
tacular elaborations of special feasting structures and food-preparation facilities
in increasingly complex levels of Polynesian chiefdoms. In the case of the south
eastern platform mounds and the simpler Polynesian structures, these may be
the first examples of specialized public architecture that occur in the regional cul
ture sequences, and I think it is worth asking whether the initial appearance of
specialized public structures in many more, or perhaps all, regional cultural se
quences may be related to feasting activities. Finally, in an impressive tour de
force centered on Sumerian remains, Denise Schmandt-Besserat reminds us that
there is a great deal to be gleaned from pictorial representations and early written
texts, not only concerning the existence of feasting, but its dynamics as well.
In order to determine how these various types of observations might fit to
gether iii a provisional archaeological classification, it would probably be useful
to see what material patterning can be gleaned from a brief overview of feasting
in several different types of cultures. I will consider this topic next.
T H E E VO L U T I O N O F F E A S T I N G : G E N E R A L I Z E D
H U N T E R / G AT H E R E R S
There is little information available on feasting among generalized hunter I gath
erers. Whether this is because it is absent or rare as Polly Wiessner suggests (1996,
personal communication), or because it may have a different character, or be"
cause it has simply not been reported in the ethnographic literature is unclear.
Sharp (1994) argues that all Chipewyan meals are solidarity-enhancing events by
their very nature and that no special meals are used for those purposes. On the
other hand, among other groups there are a few indications that large game kills
and ritual gatherings were sometimes accompanied by festive meals (Sandal 1966;
Richardson and Ianzelo 1974). Whether these were special meals or simply logis
tically convenient ways of organizing meals for a gathering of people is unclear.
In the existing literature, there are few indications that most ritual meals had any
special connotation or involved any special foods or preparations, although large
quantities of meat with high fat content seem to have been highly valued by
many groups.
There are also occasional statements that special cuts of meat were, or could
be, reserved for men's consumption at special ritual locations (Woodburn 1966).
This is perhaps the closest approximation that exists to the feasts documented for
more complex types of societies. I am unaware of feasting accounts involving in
terband visiting and alliance brokering, or the kinds of meals that may have been
served at such events. Nevertheless, is difficult to imagine that no special gastro-
42
nomic hospitality would b e displayed for potential or continuing allies in other

bands, even if only on a dyadic partnership basis. Such alliances could have been
sought for many reasons, including defense, refuges in times of starvation, mar
riage, and rituals (Wiessner 1977, 1982).
Even if it is demonstrated that some kinds of alliance and solidarity feasts do
occur among generalized hunter I gatherers, they may differ in character from
solidarity and alliance feasts in more complex societies. This would be because of
three important factors:
1) Feasting as we are most familiar with it among more complex groups is
largely predicated on the accumulation, storage, and use of surpluses. Most gen
eralized hunter I gatherers do not accumulate or store surpluses, and they have
few if any mechanisms for dealing with, or transforming surpluses (Wiessner
1982, 1996). Hunter/ gatherers certainly do invite their allies to partake of unusu
ally desirable and abundant resource occurrences when hosts have far more food
in their territory than they can use (Flood 1980). But in these instances, all fami
lies have a glut of the same abundant foods, and there is really no reason for any
sharing between families.
2) In addition, there are few ways in which generalized hunter I gatherers can
create special foods by intensifying labor to procure or process foods. There is lit
tle need for developing more intensified ways of processing foods (collecting and
grinding cereal grains may be one exception) . Moreover, the specialized technol
ogy needed for more intensified preparations would only be cumbersome for
highly nomadic groups. Similarly; it is difficult to predict when most specialty (and
even routinely used) species might be found or procured because almost by defini
tion these will be rare or difficult to get, and all such food procurement is on a
probabilistic or encounter basis. AB Richard Lee (1979:243) documented, even the
!Kung only kill about 28 medium or large game animals per year per band. The
yearly average per hunter is only o.6. Moreover, st9rage is not practiced. Thus, it
would be difficult to plan a feast for any particular time and be confident of actu
ally being able to procure or serve specialty foods such as big game meat. Trying
to return such food gifts in any reciprocal fashion would similarly become almost
impossible because of lack of predictability and storage capabilities.
3) There is also an extremely strong egalitarian and sharing ethic that charac
terizes most, if not all, generalized hunter I gatherers and which is clearly adap
tive for survival (Winterhalder 1996; Wiessner 1996). Given such a strong ethic,
the idea of someone "giving" a feast seems at odds with the rest of the dominant
and adaptive cultural values. The sharing of food is simply expected, if not de
manded, in all social, ritual, and other contexts. There are certainly "host" groups
and "visitor" groups, however these distinctions may not go beyond the control
over access to resources or the active versus passive roles that each group might
43
Brian Hayden
play in rituals. The idea of a "host" who gives a feast based on his stockpiling of
foods or gifts, and his control over family and other labor, simply seems to be an
tithetical to the rest of generalized hunter I gatherer existence. These factors may
account for the relatively ordinary, perhaps even mechanical, sharing of food be
tween hunter I gatherer social units.
In addition to these factors, the best-documented example of specialty foods
being consumed (the hind leg of large animals consumed at men's ritual sites
Woodburn 1966) seems to represent consumption at remote and obscure ritual
locations, removed from normal living sites. It is questionable as to whether such
locations would be very visible archaeologically, whether they would be recorded
in any archaeological surveys, or whether they could meaningfully be distin
guished from transit encampments on the basis of their artifactual or feature
contents.
Thus, if feasting did exist at the generalized hunter I gatherer level, it is not cer
tain. that the foods involved were significantly different from daily seasonal fare,
that there would be any clear host or guest relationship, or that the remains could
be identified archaeologically. While feasting certainly plays a pivotal role in es
tablishing alliances among more complex cultures, it may be that for generalized
hunter/ gatherers, the forging and maintenance of alliances were dealt with by
alternative social techniques such as marriage, the negotiating of kinship rela
tionships and classes, gift giving, and participation in ritual cults (Hayden 1987). A
great deal more basic research is required dealing with generalized hunter/gath
erers before a clear picture of feasting at this important base level emerges.
T R A N S E G A L I TA R I A N H U N T E R / G AT H E R E R S A N D
H O RT I C U LT U R A L I S T S
With the emergence of transegalitarian societies (those between chiefdoms and
true egalitarian societies), the full range of feasting that has been previously dis
cussed becomes established. A range of other developments characterizes transe
galitarian societies. These developments include the production of reliable
surpluses, storage of food and valuables, private ownership of resources and prod
ucts, the transformation of surpluses into prestige items, economically based com
petition, and the establishment of contractual debts. It is probably these features
that also make feasting a viable means of transforming surplus food into other de
sirable currencies such as establishing alliances in order to reduce the risk of starva
tion, attack, and other problems of existence. In fact, there appears to be a maj or
shift in the types of risk-reducing strategies used by generalized hunter / gatherers
versus transegalitarian hunter I gatherers. Generalized hunter I gatherers place by
far the greatest emphasis on sharing and alliance formation via kinship and rituals
(Wiessner 1982; Hayden 1987). Complex hunter/ gatherers, in contrast, put much
44
A P R O L E G O M E N O N TO T H E I M P O RTA N C E O F F EASTI N G
more emphasis on food storage, raiding, wealth accumulation, and the creation of
alliances via the consumption or giving away of economic surpluses in feasting. All
transegalitarian societies use these new capacities for surplus production, storage,
,
wealth accumulation, and economically based competition to create many types of
feasts. Some of the most interesting types are alliance feasts between families, line
ages, clans, or communities.
Another type of transegalitarian feast is the promotional feast that advertises
the economic, social, and political success of a social group and the desirability of
becoming affiliated with such groups either via marriage, wealth exchanges, po
litical support, or by adoption. Typically, the construction of a new house, the in
stallation of major new wealth items, the death of an administrator and
installation of a new administrator, and marriages or maturation events, are all
viewed as prime opportunities to display success.
Profit and investment feasts probably also emerge where the transegalitarian
surplus base is ample enough. These include work feasts, reciprocal food and
wealth exchange feasts, high-expense maturation feasts, and competitive feasts
where there is a contractual debt to return the amount received plus a substantial
increment or interest. Similarly, manipulative, or calamity, feasts probably only
occur in transegalitarian and more complex societies.
Another key component of the new feasting types that emerge with transegal
itarian societies is the presence of specialized witnesses or recordkeepers. Their
function is to ensure that none of the contractual debts are forgotten or neg
lected, and to ensure that the full value of elevated individuals is a matter of pub
lic record. Because the amounts in these transactions can be staggering at times,
we frequently find examples of simple recording devices or counters being em
ployed, whether in the form of notched sticks, engraved bones or stones, knotted
cords, or clay tokens (Schmandt-Besserat 1986; Wilson and Towne 1978; Lewis
1969:214). In fact, given these ethnographic occurrences and the transegalitarian
nature of European Upper Paleolithic art, it is worth wondering if some of the
numerous cases of "periodic notation" on bone and stone such as those reported
by Marshack (1997) might not be feasting and debt tallies. Typically, too, the
counting systems of ethnographic transegalitarian societies reach into the hun
dreds and thousands, far beyond the l-20 numeration limit of most generalized
hunter I gatherers.
It is worth emphasizing again, that feasts are simply one of a number of social
strategies used to achieve specific goals. It should come as no surprise to find that
the different techniques and strategies used to achieve these goals are often inter
mixed in order to ensure that the desired effect is actually achieved. Thus, the
reaffirmation of kinship ties, the giving of gifts (the return of which is obligatory
if social rupture and hostilities are to be avoided-Dalton 1977:205, 207), and
45
Brian Hayden
especially ritual all tend to play important subsidiary roles in feasts, while feasting
often plays an important supporting role in events where kinship or ritual is the
central focus.
C H I E F D O M S A N D E A R LY S TAT E S
1here are undoubtedly additional permutations and developments of feasting
events that take place with the emergence of chiefdom and state-level societies.
Chiefdoms, especially, are probably much more variable in both organizational and
feasting characteristics than I had initially assumed (Hayden 1995:64). On the basis
of Laura Junker's data, for instance, it seems that an important distinction should
be made between the sociopolitical dynamics of territory-based versus labor-based
chiefdoms. Other distinctions, especially those related to the po�ulation size of
chiefdoms, are also undoubtedly important. However, the complexities of the issue
(not the least of which is how to distinguish between complex transegalitarian so
cieties and chiefdoms-see Miller and Boxberger 1994) require more detailed con
sideration than is possible here. Nevertheless, it is clear that major changes in
feasting behavior take place with the transition from transegalitarian to stratified
societies. We may not be able to identify all these changes or provide coherent ex
planations for them at this point, but it does seem that changes in the scale and
function of feasts broach new dimensions in many chiefdoms and early states such
as discussed by Kelly, Kirch, Junker, Dietler, and Schmandt-Besserat. In particular
the pretext of feasting seems to be manipulated by chiefs and elites in early states in
order to collect surpluses from the populace (alluded to by Pat Kirch [1984:263] and
dealt with in more detail by Laura Junker, Michael Dietler, and Denise Schmandt
Besserat in this volume and by Urry [1993] for . New Zealand). 1he feasts held at
Mesoamerican ballcourts and documented by Fox (1996), may well have served a
similar function. 1his change appears to coincide with the holding of feasts on pre
viously unheard-of scales of size, generating enormous amounts of specialized
feasting refuse. 1he spectacular 65,000-cubic-meter Bronze Age feasting midden at
East Chisenbury near Stonehenge is probably only the first of this type of deposit
to be recognized (McOmish 1996). 1his is a prime candidate for remains of chiefly
level tribute feasts. 1he feasting refuse that fills ditches surrounding British cause
wayed enclosures probably represents earlier versions of tribute feasts among
smaller chiefdoms. For now, however, I will limit my discussion to transegalitarian
feasting, with which I am most familiar.
A R C H A E O LO G I C A L C O N S T E L L AT I O N S O F F E A S T I N G T R A I T S
Here, I would like to examine some broad distinctions in feasting that might be
identifiable in · the archaeological record. Several basic questions need to be ad
dressed. First, is it possible to identify feasting events in the archaeological
46
record? Considering the criteria in Table 2.1, the answer is clearly, yes. Second, is it
possible to identify the level of involvement of specific households in feasting?
The answer is yes. Third, is it possible to distinguish meaningful types of feasts?
Again the answer is yes.
We can affirm without hesitation that it is certainly possible to identify at least
some feasts in the archaeological record. There are many instances of feasting
documented in prehistory, especially in the European Neolithic where abundant
food remains of specialized nature have been found in front of, or in, megalithic
tombs and associated with causewayed enclosures or other ring-ditch monu
ments (Bradley 1984; Hedges 1984). There are many other examples documented
in this volume. The earliest clear evidence in world prehistory is documented at
Hallan Cerni (Rosenberg and Davis 1992) .
In terms of cultural evolution, it is probable that feasting was taking place at
some Upper Paleolithic transegalitarian, faunally rich and diversified aggregation
sites such as Altamira, Cueto de la Mina (Conkey 1980), or Enlene (personal ob
servation). If so, it should come as no surprise that good examples of feasting
also occur in relatively affluent Epipaleolithic transegalitarian sites such as Hallan
Cerni (Rosenberg and Davis 1992) and some Natufian sites (Byrd 1989:78-80).
However, in most of these cases, feasting has been documented because of its
unusually large scale and sometimes special location. Large-scale occurrences are
not particularly difficult to identify as feasts. The real question of importance is
whether the existence of smaller-scale feasts can be detected, and whether differ
ent kinds of feasts at several scales of size can be distinguished. There is also the
possibility that some feasting loci were kept meticulously clean as important sa
cred areas, as with kivas. Identifying special structures related to feasting when
there are no food remains may pose considerable challenges. However, initially,
we will concentrate on situations with clearer feasting indications.
Intermediate size feasting (with about 10---s o people) rarely seems to leave the
, kinds of features, structural remains, or abundant refuse that facilitates the iden
tification of precise feasting locations (unless intermediate-size feasts are held at
special locations such as megalithic tombs) . Thus, let us turn to the second ques
tion: whether households engaged in such feasting can be identified, for it is in
the household assemblage that the existence of intermediate-size feasts should
be the most apparent archaeologically. Work by Michael Clarke (Chapter 2) and
John Blitz (1993a, 1993b) clearly indicate that it is possible to identify households
involved in feasting (Fig. 2.2). It is probably even possible to identify the relative
frequency and size of feasts on the basis of the proportionate number of serving
and food-preparation vessels as well as their absolute size. The larger the prepa
ration and serving vessels, the larger should be the number of regular feast
participants.
47
Brian Hayden
Figure 2.2. Storage of oversized food-preparation vessels and extra serving vessels used
for feasting in the household of a Yao village headman. The occurrence of large sizes and
high frequencies of serving or preparation vessels clearly identifies this household as in
volved in major feasting activities.
On the basis of ethnographic observations among the Maya and the Akha, it
appears that, in general, only households that regularly host feasts of a certain size
acquire vessels and containers appropriate for that size feast. For infrequent feasts
of unusual size, the food-preparation and serving needs of particular households
are generally met by borrowing the required materials. Of course, care must be
taken to differentiate food-preparation and serving vessels from storage vessels,
but this should not present a major problem in most cases. In some cases, the fine
quality or decoration of serving vessels makes it abundantly clear that certain
wares were primarily used in feasting contexts, for example the Peten Poly
chromes of the Maya, the Kamares ware of Crete (Day and Wilson 1998). In
other cases, the contexts and distributions of certain wares, for example the Ne
olithic Grooved ware of the British Isles (Sherratt 1991:55) or European beakers
indicate feasting functions. In any event, let us assume for now that the identifi
cation of feasting vessels can be reliably carried out for specific archaeological
48
households or communities and that the ceramic tradition is relatively devel

oped. Where there are both daily ceramics and feasting ceramics for large gather
ings, we would expect to find histograms of the sizes of individual types of
preparation and serving vessels to exhibit some degree of bimodality or even tri
modality (e.g., Clarke's data in Chapter 5, Fig. 5.5; also Blitz l993b:fig. 4). The dif
ference in size between the modes (as well as the amplitude of the modes) should
reflect the relative size of the feasting groups and the frequency or intensity of
feasting for the household, corporate residence, community, or other sampling
unit. Households that did not host significant numbers of intermediate or high
level feasts would be expected to have unimodal size distributions of vessel types.
However, it must be kept in mind that in early or initial ceramic assemblages, all
ceramic vessels may have been used for feast preparation and serving (Hayden
l995b; Clark and Gosser 1995) . Such situations might be characterized by uni
modal vessel-size distributions as well.
Where domesticated animals are used for feasting, the absolute and propor
tionate frequencies of various sized and aged animals represented in the refuse
associated with a particular household can also be used to gauge the frequency
and size of various kinds of feasts, either feasts hosted by the household (perhaps
where cranial and low-utility bones predominate) or feasts at which the house
hold members were guests (where bone remains from high-utility cuts predomi
nate, although reciprocal feasting would probably lead to a mixture of high- and
low-utility bones) . Theoretically, the larger the maximum-sized animal, the larger
the feast should be. This relationship seems to be generally backed up by empiri
cal observations (e.g. , Michael Clarke's data). The degree of bone reduction ver
sus waste is probably also related to the size of the feasts, with much less waste
occurring in small or intermediate feasts. At individual households and corporate
residences, the occurrence of outside hearths or roasting pits may also be indica
tive of the hosting of unusually large feasts (Figs. 2.3-2.5) . Small and intermedi
ate-sized feasts are generally prepared over normal household hearths. By far the
most common, if not exclusive use of domestic animals in transegalitarian soci
eties is for feasting (Keswani 1994; Blanton and Taylor 1995; Hayden 1990 ). On the
basis of ethnographic observations among the Akha by Michael Clarke, both
food vessels and food wastes occurring in the toft zones of individual households
appear to reflect real frequencies and sizes of feasting involvement. Thus, even
without the occurrence of special facilities or obvious feasting middens, it should
be possible to identify feasting involvement at the household level using a num
ber of indicators.
The third question, that of determining the purpose of feasts, is the least ex
plored, but perhaps the most interesting. At the outset, we may iterate the possi
bility that generalized hunter I gatherers may not host any feasts in the usual
49
Brian Hayden
Figure 2.3. For large feasts such as this Hmong funeral, multiple
outside hearths are the typical solution to preparing large-scale
meals. One hearth not being used occurs in the foreground; two
others in use occur behind it. Note also the very large size of
food-preparation and serving vessels on the hearths, on the
ground, and on the table.
sense, or that if they do, their feasts probably most closely resemble what I would
call solidarity feasts. In many (but perhaps not all) solidarity feasts, there seems to
be minimal attempt at expending large amounts of resources or consuming spe
cial foods; there is little ostentatious display, and all participants often make equal
contributions. Thus, these feasts should be minimally differentiated from daily
meals, and may only differ from them in terms of size. These characteristics
seem consistent with what we currently know about generalized hunter/ gath
erer communal meals.
At the transegalitarian level, there are several heuristic principles that can be
proposed for differentiating feasting functions, all of which require far closer
scrutiny with empirical data, refinement, and reevaluation. We can begin, first,
50
Figure 2.4. The multiple outside hearths made for a Ta Oi marriage feast in Vietnam.
Note the location only a few meters from the rear of the longhouse where about 1 00
guests were served.
by postulating that solidarity feasts will often be the least materially distinctive as
just discussed.
Second, there is probably a relationship between the quantity and quality and
cost of prestige / display items on the one hand, and the degree of promotion and
competition involved in the feasts on the other hand, whether at the interlineage
level or at the intercommunity level. Among transegalitarian societies, there fre
quently seems to be only a loose fit between household feasting activity and
household wealth or political power in a community. As Michael Clarke argues in
his analysis of the Akha, this is because multifamily socioeconomic and political
alliances (usually based on kinship), rather than nuclear families, appear to be the
most important units of competition in most transegalitarian communities. By
themselves, single households simply appear to be too vulnerable to vicissitudes
of economic, social, and political life and stand little chance of winning any com
petitive struggles against multifamily alliances such as strong lineages . This is
clearly documented in Condominas (1977:86-87, 95, mo, 123, 139, 151, 156) where
51
Figure 2.5. A closeup of the food-preparation area in Figure 2.4 showing several hearth�
in use and the large food-preparation vessels.
poor or politically isolated families could be disenfranchised or even enslaved or

killed for minor transgressions or disputes with the more powerful. Thus, indi
vidual households generally seek alliances with the strongest lineage (or other)
network open to them, and it is in their interests to see that the lineage is
strengthened as much as possible. Therefore, they make substantial contribu
tions to all lineage marriage, funeral, house building, and other feasts, and they
participate in the many lineage solidarity feasts .
In terms of feasting, this results in the major promotional and competitive loci
occurring at lineage or community levels rather than at the nuclear household
level. The fact that individual households may exhibit elaborate feasting para
phernalia but not necessarily be the most wealthy or the most powerful house
hold in the lineage or clan or community (although this is often the case) can
distress some archaeologists. Yet, from an archaeological perspective, such con
cordance of material with economic, social, and political roles of individual
households is not of critical importance, since we know all too well that varying
strategies and individual idiosyncrasies create a great deal of variability at the
52
individual-household level (Hayden and Cannon 1984). What is of the utmost im
portance is the fact that such materials as the foods and paraphernalia associated
with different levels and types of feasting exist at all, and then that these items are
differentially distributed among the social units within a given community and
that the magnitude of these differences can be monitored. It is the existence of
the overall feasting system and the magnitude of differentiation within the sys
tem that is of the greatest archaeological importance, not necessarily the correct
identification of each household's status within the community or the system or
within a specific lineage. However, to obtain a precise estimate of such differenti
ation within a community, excavations and analysis of individual household as
semblages is essential.
A third potential principle for interpreting transegalitarian feasts is based on an
initial perusal of the literature indicating that the purposeful destruction of high
cost wealth items (versus personal use) items via breakage, burning, permanent
burial (in offerings or graves), or submersion in deep water, may only occur in the
context of competitive feasts. Thus, the occurrence of destroyed wealth, as ex
emplified by the burned Liangzu burial offerings in China (Xuanpei 1992), the
bronze wealth objects dropped in British rivers or bogs (Bradley 1990 ), the highly
valued Northwest Coast coppers dropped into the sea, and many other exam
ples, can probably be used to distinguish simple promotional and/ or alliance
feasting from the more elaborate competitive feasting forms. The presence, fre
quency, and relative cost of major wealth items may also be a measure of the de
gree of surpluses and competitiveness involved in feasting.
A possible fourth principle is related to specialized structures. Though large
feasts can be for purposes of community solidarity, lineage or community pro
motion, or community competition, it seems unlikely that special structures for
feast-related activities characterize this entire range or scale of functions. Rather,
it seems possible that special structures, especially if they are permanent rather
than erected only for single events, may reflect more institutionalized competi
tive or promotional (or perhaps even alliance) feasting for the highest-ranking ad
ministrators of large groups such as lineages, clans, or communities (Fig. 2.6).
Exactly how lineage feasting and ritual structures operated in this context is not
entirely clear from the ethnographies that I have seen, but I suspect that they are
constructed primarily as promotional adjuncts to lineage feasts and rituals. On
the other hand, the specialized Kepele cult structures that Wiessner documents
in Chapter 4 for New Guinea indicate that such relationships are probably com
plex and require much more investigation.
If these and similar principles can be established as reliable material guides for
understanding past feasting behavior, it may be possible to classify archaeological
53
Brian Hayden
Figure 2.6. A Vietnamese lineage shrine used for lineage feasts. Only high-ranking lineage
men use the interior of this structure for feasting and rituals. Women, children, and low
ranking lineage members eat their meals outside the structure.
feasting events into a number of broad categories. Four such categories might
consist of:
1 . M I N I M A L LY D I S T I N C T I V E F E A S T S
There are n o archaeological examples o f minimally distinctive feasts proposed in
this volume. However, ethnographically, at least at some levels, solidarity feasts
are minimally distinctive . Most generalized hunter/ gatherer feasts are probably
minimally distinctive as well. In these cases, only the size of the food-preparation
and serving materials may differ from daily meals, as well perhaps as some food
species of minimally higher value (chickens, ducks, rabbits, small pigs). There are
many small-scale (less than rn people) household and moderate-sized (about
10-50 people) lineage solidarity feasts of this type, as well as dyadic friendship
feasts, solicitation feasts, and possibly punishment feasts-for which data are
sparse . Given the single event and unforeseen nature of punishment feasts, as well
as the likely desire of the giver to minimize costs, little special food-preparation
or serving paraphernalia would be expected to be used for punishment feasts, ex
cept through borrowing. Even moderate-size lineage solidarity feasts probably
54
occur as a rule within normal residences, without special paraphernalia other

than perhaps slightly larger ordinary preparation and serving vessels.
Where surpluses are limited, alliance feasts between lineages, clans, or com
munities may also be minimally distinctive in that prestige items might rarely be
used for display or for serving, although animals of greater value could be killed
to accommodate the larger numbers expected in such circumstances. Another in
stance where animals of unusual value might be used as part of solidarity feasts
is the large-scale community meal in which all interested households make con
tributions to obtain an unusually large domestic animal such as a large pig or
bull. The animal can be killed at a central location, but the meat is often divided
up according to individual household financial contributions. In Southeast Asia,
household members take the meat back to their houses for cooking and con
sumption with no vill age-wide communal meal, although considerable visiting
often occurs. Animal horns or mandibles from such communal events are usually
saved and placed in prominent locations such as the village priest's or headman's
house in order to display the community's ability to sponsor such events (Figs. 2. 7
and 2.8). As far as I know, beyond this simple corporate display, prestige items are
not associated with such solidarity feasts.
Small- and medium-sized work feasts probably share most of the material
characteristics of solidarity feasts. As with community feasts, many hosts proba
bly seek to minimize costs and might not use prestige serving vessels to feed
workers, especially if they draw mainly on kinship ties to recruit workers. Like
some community solidarity feasts, work feasts sometimes take place in unusual
locations (forests or springs). On the other hand, Michael Dietler (personal com
munication) has observed that as the number or required skill of workers in
creases beyond the near kin group, greater material inducements must be
provided to attract the required number or expertise of workers. Therefore, un
usually high-quality foods and serving vessels may be used, and other types of
entertainment or treats may be provided. This level of work feast might better be
classed with promotional feasts in terms of material patterning.
2 . P R O M OT I O N A L / A L L I A N C E F E A S T S
Although individual families sponsor marriage, funeral, ancestral, and new
house feasts, these are generally lineage or clan affairs in which the success of the
social group at large is on display (see Michael Clarke's analysis of Akha feasting).
There are, of course, important lineage or clan solidarity aspects to these events
as well. However, to the extent that advertising group success (to attract mates
and desirable labor) is important in these feasts, and to the extent that surpluses
are available, prestige items for promotional display (including serving vessels)
should increase in frequency and cost. Large-scale food-preparation facilities,
55
Figure 2.7 (top). Water-buffalo horns from feasts are frequently displayed on the houses
of the sponsoring individuals. Here, an Akha village head has placed horns from commu
nity feasts on the walls of his house as a display of community prosperity. In other
groups, such as Torajan communities in Sulawesi, horns are displayed on houses to indi
cate household or lineage wealth. Figure 2.8 (bottom). Pig mandibles are also conserved
after smaller feasts and used to display household, lineage, and sometimes community
feasting activities, as in this Akha administrator's house.
preparation vessels, serving vessels, and perhaps even temporary architecture

typically characterize these feasts. Such feasts also tend to be typified by a variety
of associated ritual paraphernalia used to indicate that participants use and un
derstand the same feasting conventions and recognize the contractual nature of
debts incurred by accepting invitations to feasts. Such paraphernalia includes
items such as pipes and the ritual sharing of tobacco smoking, special vessels for
the consumption of alcoholic or other drinks like chocolate, kava or coffee (e.g.,
Dietler 1990), and other ritualized narcotic paraphernalia. Community alliance
and promotional feasts are probably indistinguishable from lineage- or clan
sponsored promotional feasts in terms of size and material signatures although
locations may shift from individual family or corporate residences to central
spaces or facilities of communities. Specialized structures may be associated with
some levels of this type of feasting. Large household-sponsored work feasts,
household or lineage investment feasts, (such as maturation feasts for children in
wealthy families), as well as diacritical feasts as discussed in Chapter 3 and Dietler
(1996), may be archaeologically indistinguishable from promotional feasts, except
perhaps for size. Several of the archaeological examples in this volume probably
represent promotional and alliance feasting, in particular, the Maya lineage feast
ing and ritual structure documented by Brown and the Woodland platform
mounds described by Knight. By further examining these archaeological cultures
in more detail, it may be possible to further refine this gross level of interpreta
tion in specific cases. The feasting documented at Cahokia by Kelly and in the
Philippines by Junker may also incorporate promotional and alliance feasting
components, however their scale and the inclusion of costly prestige wares asso
ciated with chiefs seem to indicate other feasting components such as tribute or
competitive feasting.
3 . CO M P ET I T I V E F E ASTS
These feasts have all the material characteristics o f promotional feasts but have
even more and more costly prestige items, more prestige serving vessels, some
times intentionally destroyed high-value wealth items, and possibly more elabo
rate, permanent, specialized structures for feasting-related activities . Junker's
Philippine feasts likely included competitive feasts, given the importance of Chi
nese porcelain wares (Chapter rn).
There is bound to be blurring between the above divisions, especially as sur
pluses become abundant, which may lead to the incorporation of displays of suc
cess in feasts that are fundamentally predicated on promoting solidarity within a
social group. In fact, it is rare to find a "pure" feast in terms of function. As noted,
even the remains from community solidarity feasts are often displayed in the
form of bull horns displayed in the community as a sign that the community was
57
Brian Hayden
successful enough to sponsor one or more large-scale feasts. However, there are
real foci in both the intended purpose of feasts as well as in the material pattern
ing associated with different types of feasts. I am convinced that ignoring these
functional foci will only impede our understanding of feasts and our ability to
make sense of the archaeological remains of feasts.
4. TR I B UTE F EASTS
These feasts probably characterize chiefdom and early state levels of organiza
tion. They may be far larger than any other type of feast. They are held at regu
lar, calendrical intervals and should be as inclusive as possible within a given
polity. It seems unlikely that wealth distribution would be common since the goal
of these feasts is to amass as much surplus as possible, and to sequester as large a
proportion of it as possible for elite use. These types of feasts probably generate
a large amount of food refuse due to the large-scale consumption of most sur
plus food by the populace in order to motivate them to support the elites and
tribute feasts. Tribute feasts are likely to be intimately tied to rituals honoring
polity deities and in many cases are associated with monumental structures and
spaces associated with those deities. Schmandt-Besserat makes a convincing case
for tribute feasts in Sumer (Chapter 14), and similar events may be represented by
Kelly's Cahokian feasts (Chapter 12), the large British Neolithic and Bronze Age
feasting middens previously mentioned (McOmish 1996), and large-scale feasting
remains at Incan sites (Morris 1988).
P R I N C I P L E S A N D CO N C LU S I O N S
The major points that I wish to make about feasting can be summarized as
follows:
1. The high cost of some feasts, their widespread occurrence, and their per
sistence over time indicate that feasts are probably adaptive in an evolu
tionary and cultural ecological sense.
2. Feasts are techniques for transforming surp luses into socially; economically;
and politically useful currencies that can be used to further individual and
group self-interest and survival. This appears to be one of the most unique
and distinctive capabilities that distinguish humans from the rest of the an
imal world.
3. In transegalitarian societies, clear adaptive advantages can be established
via feasts that give households and lineages advantages in warfare, acquir
ing mates, help in emergencies and catastrophes, and in political control
over resources and people. Feasting frequently operates by establishing
long-term social ties and debts although immediate benefits are often ob-
58
tained through work feasts. Feasts can also b e used for financial invest
ment, to manipulate people, and perhaps for other purposes.
4. Feasting is probably the main dynamic factor behind the development of
prestige technologies, especially the development of food production and
the domestication of plants and animals.
5. Feasting in transegalitarian and other complex types of societies is predi
cated on the production and use of economic surpluses. Thus, the pres
ence and magnitude of storage facilities should be related to the type and
intensity of feasting. The greater the surpluses, the greater should be the
cost and frequency of prestige items, prestige foods, prestige serving ves
sels, and possibly feasting architecture. Also, the greater the surpluses, the
greater the expected promotional and competitive nature of the feasts.
6. Feasting is probably not a prominent part of generalized hunter I gatherer
behavior.
7. The larger the feast and the larger the surpluses, the greater the waste of
food and goods is expected to be, culminating in the intentional destruc
tion of wealth in competitive feasts. This seems supported by both ethno
graphic studies and studies of modern material culture such as that of
Wilson and Rathje in this volume.
8. Feasting may also be strongly implicated in the emergence of the first spe
cial-function architectural structures in transegalitarian communities.
9. As in most archaeological inquiries, single lines of evidence for inferring
behavior or social organization are tenuous at best, given the idiosyn
crasies of people, history, preservation, sampling, excavation, processing,
and interpretation. However, it should be apparent from the foregoing and
the many contributions in this volume, that in studying feasting, we are
not restricted to a single line of evidence. Rather, there are many different
types of data, both theoretical and empirical, that we can use to document
and investigate the prehistoric occurrence of feasting (Table 2.1). I strongly
endorse the use of as many lines of evidence as can be marshaled in study
ing feasting. Sociologists call this approach "triangulation," and it is a good
metaphor. It is also essential to continuously reexamine and reassess the
logic and perhaps premature generalizations that we are developing in try
ing to grapple with this relatively complex but potentially fruitful topic.
In conclusion, it is worth reiterating that our theoretical and comparative un

derstanding of feasting is abysmal, but we are taking active measures to remedy
this situation. The chapters in this volume provide eloquent testimony of the ad
vances that are possible. Nevertheless, the corpus of existing empirical observa
tions on material culture related to feasting is depressing. Traditional social
59
Brian Hayden
anthropology and ethnography have not been of great service to archaeologists

in this respect. It seems that the future development and validation of accurate
and reliable principles related to feasting must be undertaken by ethnoarchaeolo
gists or sympathetic ethnographers. There is a great deal of literature on the
topic in ethnographies, but most of it has debilitating gaps that render it of min
imal value for archaeological purposes. For instance, despite the great wealth of
descriptions on New Guinea feasts, there are no published plans of feasting struc
tures or even any indication of the locations of feasting or food-preparation facil
ities such as roasting pits for feasts. Polly Wiessner provides some of the first
illustrations in this volume. The task for ethnoarchaeologists and ethnographers
is ei:iormous, but it is not too late to launch such an undertaking. The gathering
of minds represented in this volume is ample testimony to the interest and po
tential relevance that the future study of feasting holds for us all.
AC K N OW L E DG M E N TS
My sincerest gratitude is extended to Ralana Maneeprasert and Chantaboon Sutthi (for
mer Director) of the Tribal Research Institute in Chiang Mai, Thailand, as well as to Pro
fessor Chang Quoc Vuong, Director of the Vietnamese Culture and Ecology Program at
Hanoi National University. Their unselfish sharing of knowledge and help in visiting tradi
tional Hill Tribe communities was invaluable in opening my eyes to the realities of Hill
Tribe culture and feasting. Funding for these investigations was provided by the Social Sci
ence and Humanities Research Council of Canada.
REFERENCES
Beardsley, Tim
1993 Honest Advertising. Scientific American 268 (5): 24-27.
Blackburn, Thomas
1976 Ceremonial Integration and Social Interaction in Aboriginal California. In
Native Californians: A Theoretical Retrospective, edited by Lowell Bean and
Thomas Blackburn, pp. 225-243. Socorro, N.M.: Ballena Press.
Blanton, Richard, and Jody Taylor
1995 Patterns of Exchange and the Social Production of Pigs in Highland New
Guinea.journal of Archaeological Research 3:n3-45.
Blitz, John
l993a Big Pots for Big Shots: Feasting and Storage in a Mississippian Community.
American Antiquity 58:80-96.
l993b Ancient Chiefdoms of the Tombigbee. Tuscaloosa: University of Alabama Press.
Bourdieu, P.
Bradley, Richard
1990 The Passage of Arms. Cambridge: Cambridge University Press.
60
Byrd, Brian
1989 The Natufian Encampment. Jutland Archaeological Society Publication XXIII:r.
Denmark: Aarhus University Press.
Clark, John, and Denis Gosser
1995 Reinventing Mesoamerica's First Pottery. In The Emergence of Pottery, edited
by W Barnett and J. Hoopes, pp. 209-22r. Washington, D. C . : Smithsonian In
stitution Press.
Condominas, Georges
1977 We Have Eaten the Forest. New York: Hill and Wang.
Conkey; Margaret
1980 The Identification of Prehistoric Hunter/ Gatherer Aggregation Sites: The
Case of Altimira. Current Anthropology 21:609-(530.
Cowgill, George
1996 Population, Human Nature, Knowing Actors, and Explaining the Onset of
C omplexity. In Debating Complexity, edited by D. Meyer, P. Dawson, and
D. Hanna, pp. 16-22. Calgary: University of Calgary Archaeology
Association.
Dalton, George
1977 Aboriginal Economies in Stateless Societies. In Exchange Systems in Prehistory,
edited by T. Earle and J. Ericson, pp. 191-212. New York: Academic Press.
Dat, Peter M . , and David E. Wilson
1998 Consuming Power: Kamares Ware in Protopalatial Knossos. Antiquity
72:350-358 .
Dietler, Michael
9:352-406.
Status in Prehistoric Europe. In Food and the Status Quest, edited by P. Wiess
ner and W Schiefenhovel, pp. 87-125. Oxford: Berghahn Books.
Flood, Josephine
1980 The Moth Hunters: Aboriginal Prehistory of the Australian Alps. Canberra: Aus
tralian Institute of Aboriginal Studies.
Fox, ]. G.
1996 Paying with Power: Ballcourts and Political Ritual in Southern Mesoamerica.
Current Anthropology 3T483-509.
Friedman, J.
ogy, edited by M. Bloch, pp. 161-202. London: Malaby Press.
Friedman, J. , and J. Rowlands
1977 Notes toward an Epigenetic Model of the Evolution of "Civilization." In The
Evolution of Social Systems, edited by J. Friedman and M. J. Rowlands, pp.
201-276. London: Duckworth.
61
Brian Hayden
Gosden, Chris
1989 Debt, Production, and Prehistory. journal of Anthropological Archaeology
8:355-387.
Harris, Marvin
1971 Culture, Man, and Nature. New York: Crowell.
1979 Cultural Materialism. New York: Random House.
Hayden, Brian
1987 Alliances and Ritual Ecstasy: Human Responses to Resource Stress. journal
for the Scientific Study of Religion 26 (1): 81--91.
1994 Competition, Labor, and Complex Hunter-Gatherers. In Key Issues in Hunter
Gatherer Research., edited by Ernest Burch Jr., pp. 223-239. Oxford: Berg Pub
lishers.
1995a Pathways to Power: Principles for Creating Socioeconomic Inequalities. In
Foundations of Social Inequality, edited by T. Douglas Price and Gary Feinman,
pp. 15-85. New York: Plenum Press.
1995b The Emergence of Prestige Technologies and Pottery. In The Emergence of
Pottery, edited by William Barnett and J. Hoopes, pp. 257-265. Washington,
D.C . : Smithsonian Institution Press.
1998 Practical and Prestige Technologies: The Evolution of Material Systems.
journal of Archaeological Method and Theory 5=1-55.
Hedges, John
1984 Tomb of the Eagles. London: John Murray.
Izikowitz, Karl
1951 Lamet: Hill Peasants in French Indochina. Uppsala: Goteborg.
Keesing, Roger
1975 Kin Groups and Social Structure. New York: Holt, Rinehart and Winston.
Keswani, Priscilla
1994 The Social Context of Animal Husbandry in Early Agricultural Societies.
journal of Anthropological Archaeology 13:255-277.
Kirch, Patrick
1984 The Evolution of the Polynesian Chiefdoms. Cambridge: Cambridge University
Press.
Lee, Richard
1979 The !Kung San. Cambridge: Cambridge University Press.
Lewis, Paul
1969 Ethnographic Notes on the Ak:has of Burma. Human Relations Area Files,
New Haven.
Lewis-Williams, ]. D.
1997 Agency, Art, and Altered Consciousness. Antiquity 71: 810-830.
Lightfoot, Kent, and Gary Feinman
1982 Social Differentiation and Leadership Development in Early Pithouse Vil
lages in the Mogollon Region of the American Southwest. American Antiquity
47:64-85.
62
Marshack, Alexander
1997 Paleolithic Image Making and Symboling in Europe and the Middle East: A
Comparative Review. In Beyond Art: Pleistocene Image and Symbol, edited by M.
Conkey, 0. Soffer, D. Stratmann, and N. Jablonski, pp. 53-91. Memoirs of the
California Academy of Sciences, No. 23.
Mauss, Marcel
1954
[1924) The Gift. New York: Free Press.
McOrnish, David
1996 East Chisenbury: Ritual and Rubbish at the British Bronze Age-Iron Age
Transition. Antiquity 70:68-76.
Miller, Bruce, and Daniel Boxberger
1994 Creating Chiefdoms: The Puget Sound Case. Ethnohistory 41 (2): 267-294.
Morris, Craig
1988 A City Fit for an Inka. Archaeology 41:43-49.
Piddocke, Stuart
1965 The Potlatch System of the Southern Kwakiutl: A New Perspective. South
westernjournal of Anthropology 21:244-264.
Rappaport, Roy
1968 Pigsfor the Ancestors. New Haven: Yale University Press.
Richardson, B., and T. Ianzelo
1974 Cree Hunters of Mistassini (film). National Film Board, Ottawa.
Rosenberg, Michael, and Michael Davis
1992 Hallan Cerni Tepesi, an Early Acerarnic Neolithic Site in Eastern Anatolia.
Anatolica 18:1-18 .
Ruyle, Eugene
1973 Slavery, Surplus, and Stratification on the Northwest Coast: The Ethnoener
getics of an Incipient Stratification System. Current Anthropology 14:603-617.
Sandal, Roger
1966 Walbiri Ritual at Ngama (film). Canberra: Australian Institute of Aboriginal
Studies.
Schmandt-Besserat, Denise
1986 An Ancient Token System: The Precursor to Numerals and Writing. Archae
ology 39 ( 6): 32-39.
Sharp, Henry S.
1994 Inverted Sacrifice. In Circumpolar Religion and Ecology: An Anthropology of the
North, edited by Takashi Irirnoto and Takako Yamada, pp. 253-271. Tokyo:
University of Tokyo Press.
Sherratt, Andrew
1991 Sacrifice and Profane Substances: The Ritual Use of Narcotics in Later Ne
olithic Europe. In Sacred and Profane, edited by Paul Garwood, D. Jennings, R.
Skeates, and]. Toms, pp. 50-64. Oxford: Oxford University Committee for Ar
chaeology.
63
Brian Hayden
Stanish, Charles
1994 The Hydraulic Hypothesis Revisited: Lake Titicaca Basin Raised Fields in
Theoretical Perspective. Latin American Antiquity 5:312-332.
Suttles, Wayne
1968 Coping with Abundance. In Man the Hunter, edited by R. Lee and I. Devore,
pp. 56-68. Chicago: Aldine.
Urry, Katherine
1993 Te Hakari: Feasting in Maori Society and Its Archaeological Implications.
Master of Arts Thesis, Department of Anthropology; University of Auckland.
Voss, Joachim
1987 The Politics of Pork and Rituals of Rice: Redistributive Feasting and Com
modity Circulation in Northern Luzon, the Philippines. In Beyond the New
Economic Anthropology, edited by ]. Clammer, pp. 121-141. New York: St. Mar
tin's Press.
Wiessner, Polly
1977 Hxaro: A Regional System of Reciprocity for Reducing Risk among the Kung
San. Ph.D. Dissertation, Department of Anthropology; University of Michi
gan, Ann Arbor.
1982 Beyond Willow Smoke and Dogs' Tails. American Antiquity 47:171-178.
1996 Leveling the Hunter: Constraints on the Status Quest in Foraging Societies.
In Food and the Status Quest, edited by P. Wiessner and Wulf Schiefenhovel,
pp. 171-192. Oxford: Berghahn Books.
Wilson, N. , and A. Towne
1978 Nisenan. In Handbook of North American Indians, Vol. 8, California, edited by R.
Heizer, pp. 387--J97. Washington, D.C . : Sinithsonian Institution.
Winterhalder, Bruce
1996 Social Foraging and the Behavioral Ecology of Intragroup Resource Trans
fers. Evolutionary Anthropology 5 (2): 46-57.
Woodburn, James
1966 The Hadza (film) . London: London School of Econoinics.
Yan, Xuang
1996 The Flow of Gifts. Stanford: Stanford University Press.
Zahavi, Amotz, and Avishag Zahavi
1997 The Handicap Principle. Oxford: Oxford University Press.
64
3
THEORIZING THE FEAST
R I T U A L S O F C O N S U M P T I O N , C O M M E N S A L P OL I T I C S , A N D
P O W E R I N A F R I C A N C O N T E XTS
Michael Diet/er
"Feast" is an analytical rubric used to describe forms of ritual activity that involve
the communal consumption of food and drink. Rituals of this kind play many
important social, economic, and political roles in the lives of peoples around the
world. As the chapters in this volume attest, recognition of this fact has been
growing rapidly among archaeologists recently, along with the fertile insights
that feasts may offer in understanding social relations and processes in ancient so
cieties. I would suggest that one of the reasons that a focus on feasting is, in fact,
crucial to archaeology is that it constitutes part of a central domain of social ac
tion that has been largely absent from archaeological analysis to date, much to
our detriment. Discussions of the transformation of political systems, for exam
ple, have tended rather crudely to link broad evolutionary processes to general
65
Michael Dietler
structural typologies without considering the intervening kinds of social prac

tices by which people actually negotiate relationships, pursue economic and po
litical goals, compete for power, and reproduce and contest ideological
representations of social order and authority. Hence, there has been a general
failure to deal effectively with issues of agency and to understand the ways in
which practice transforms structure.
In my view, it is essential for archaeologists to come to grips with the arenas of
social action in which, and the sets of practices by which, the micropolitics of
daily life are played out. This is the only way we will move beyond mechanistic
typological reductionism in understanding historical transformations of various
relations of power and in addressing such perennial issues as the development of
social stratification and political centralization. For example, it is undoubtedly
important to nuance our understanding of complex political structures with tax
onomic distinctions, such as that between hierarchy and heterarchy raised by
Crumley (1987); but it is equally important to attempt to understand the practices
by which individuals create, maintain, and contest positions of power and au
thority within systems structured in these ways and, in the pursuit of their con
flicting interests, transform the structures of the systems themselves. Put in
simpler terms, we need to think seriously and realistically about political life as it
is lived and experienced if we are to fill our analytical categories with meaningful
content and advance beyond mechanistic structural correlations, vague pro
nouncements about overdetermined social processes, and sweeping evolutionist
teleologies.
F E A S T S , P O L I T I C S , A N D A R C H A E O LO G Y
There has been much written recently about the need to develop a practice
oriented approach in archaeology, but rather few coherent suggestions or effec
tive demonstrations of how this can be accomplished. This is one of the principal
attractions of a focus on feasts. Although as yet curiously underacknowledged,
the "commensal politics" of feasting is a domain of political action that is both
extremely important on a worldwide scale and potentially accessible to archaeo
logical analysis (Dietler 1990, 1996, 1999a; Hayden 1990, 1996). Indeed, I would
contend both that feasts are inherently political and that they constitute a funda
mental instrument and theater of political relations. In making this statement, let
me explicitly emphasize that I manifestly do not mean to make the naive reduc
tionist argument that feasts are only about power; nor do I mean that they are the
only significant domain of political action. Far from it. But they are commonly an
important arena for the representation and manipulation of political relations,
and it behooves us to explore critically this dimension of such a widespread cul
tural institution. However, before we are able to fully exploit this promising av-
66
R I T U A L S O F CO N S U M PT I O N , C O M M E N S A L P O L I T I C S , A N D P O W E R I N A F R I C A N C O N T E XTS
enue of analysis, we need not only a greater range of empirical information

about the diagnostic characteristics of feasts, but, most crucially, a more devel
oped theoretical understanding of the nature of feasts as a distinctive kind of rit
ual practice. Ultimately, it is only through the latter that we will be able to
comprehend and exploit the former. This is by no means a simple or straightfor
ward proposition: it requires detailed, careful, and subtle analytical exploration
and argumentation.
As noted above, I define feasts explicitly as a form of public ritual activity cen
tered around the communal consumption of food and drink. Let me immedi
ately anticipate a common misunderstanding of this definition by some
archaeologists and make clear that identifying feasts as ritual activity does not
mean that they are necessarily highly elaborate ceremonies. A ritual act can be as
simple as making the sign of the cross upon entering a church, pouring a few
drops of beer on the threshold of a house as a libation, or throwing a small wine
and cheese reception for a visiting anthropologist who has just presented a collo
quium lecture. Moreover, as the last example suggests, rituals need not necessar
ily be "sacred" in character (Moore and Myerhoff 1985) . The defining criterion of
rituals is that they are in some way symbolically differentiated from everyday ac
tivities in terms of forms of action or purpose: in Kertzer's (1988:9) phrase, they
are "action wrapped in a web of symbolism." More will be said about this later.
For the moment, let me simply assert that, as with other types of ritual, feasts
provide an arena for both the highly condensed symbolic representation and the
active manipulation of social relations. Moreover, as a particular form of ritual in
which food and drink constitute the medium of expression and commensal con
sumption constitutes the basic symbolic idiom, feasts have some distinctive prop
erties (which, again, will be discussed in more detail later). 1
I n earlier publications, I used comparative ethnographic data t o develop a the
oretical discussion of several major political dimensions of feasting ritual, with
distinctions based upon a consideration of the conjuncture of the different polit
ical roles played by feasts and the nature of their symbolic action. These different
modes of commensal politics were labeled "entrepreneurial," "patron-role," and
"diacritical" feasts, and I used different contexts in prehistoric Europe to illustrate
how the application of this perspective can aid archaeological understanding of
ancient societies (see Dietler 1990, 1996, 1998, 1999a, 1999b ). In this chapter I use a
variety of ethnographic evidence from African agrarian societies to emend and
further elucidate these theoretical constructs and explore their utility for archae
ological interpretation.2
I focus upon Africa for several reasons. The most obvious reason is that I have
firsthand experience of it from having spent several years conducting ethno
graphic research there.3 More important than mere personal familiarity, how-
67
Michael Dietler
ever, is the fact that Africa is the ethnographic terrain that really gave birth to po
litical anthropology as a field (Amselle 1998 :58; Moore 1994). Because so much re
search during the colonial era was pragmatically driven from an early date by
attempts to understand the operation of politics in both myriad stateless societies
and the large centralized kingdoms that were encountered (e.g. , Fortes and
Evans-Pritchard 1940; Middleton and Tait 1958), the African literature is unusually
rich in comparative observations on, and insights into, structures of power and the
operation of politics in daily life. Moreover, despite the obvious cautious source
criticism necessary in negotiating much of the earlier structural-functionalist
political work, Africanists have remained at the vanguard of political analysis and
the theoretical exploration of power. Hence, Africa does offer an especially
promising context for investigating the political dimension of feasting.
But there is more. Africa is also of interest because it has frequently been sin
gled out by scholars as presenting some intriguingly distinctive characteristics in
the realm of food and politics. This is, after all, the continent that was designated
by Goody (1982) as the prototypic land without "cuisine." His book Cooking, Cui
sine, and Class was largely dedicated to explaining why African societies, even the
highly stratified kingdoms, had not developed the kind of markedly differenti
ated culinary practices that characterize Europe and China. I hasten to add that
one should be wary of Goody's rather sweeping regional generalizations, but
they do point to some interesting theoretical issues that are important for under
standing the archaeological interpretation of feasting. Likewise, several scholars
have recently suggested both that the nature of power in Africa differs funda
mentally from that in "the West" (in that it is centered more around consump
tion than around "transformation," that is, the capacity to consume rather than
the ability to get people to do things: Schatzberg 1993:446), and that it is insepara
bly associated with metaphors of food and its consumption (see Bayart 1993;
Lentz 1998; Schatzberg 1993). Again, Lentz (1998) quite rightly cautions against
accepting such broad generalizations and reifications, pointing out the tremen
dous diversity of political practices, strategies, and moral philosophies of power
in Africa. But, whether one ultimately accepts these arguments for African excep
tionalism or not, what such examples indicate is that African societies do furnish
a challenging context in which to examine and refine theoretical constructs con
cerning feasting and politics derived from broader surveys of ethnographic data.
F EASTS A N D T H E P O L I T I C A L E CO N O M Y
Before undertaking a more detailed analysis of the Inicropolitical dimensions of
feasting ritual, or what may be labeled "commensal politics," I will begin with the
general observation that in Africa, as elsewhere, feasts serve a wide variety of im
portant structural roles in the broader political economy. They create and main-
68
R I T U A L S O F C O N S U M PT I O N , C O M M E N S A L P O L I T I C S , A N D P O W E R I N A F R I C A N C O N T E XTS
tain social relations that bind people together in various intersecting groups and
networks on a wide range of scales, from the local household cluster to the re
gional political community. For example, they are extremely important in estab
lishing sentiments of friendship, kinship, and community solidarity, as well as in
cementing bonds between affine groups and political links between leaders of
various kinds. In this sense, they may be seen to perform, at a variety of scales,
the classic integrative function of creating communitas, which was identified by
earlier functionalist analysts of ritual (see especially Turner 1969). Unfortunately,
the relatively limited considerations of feasting by archaeologists have tended not
to penetrate much beyond this level until quite recently. As later discussion will
demonstrate, a more productive political analysis of feasting must also explore
the complex contradictory processes and tensions simultaneously operating in
feasting ritual. However, it is important to acknowledge this integrative function
which, among other things, enables feasts to act frequently as the nodal contexts
that articulate regional exchange systems: commensal hospitality establishes rela
tionships between exchange partners, affines, or political leaders and provides the
social ambiance for the exchange of valuables, bridewealth, and other goods that
circulate through a region. Feasts may also provide the main context for the arbi
tration of disputes, the passing of legal judgments, and the public acting out of
sanctions (ridicule, mimicry, ostracism, etc.) that maintain social control within a
community. In the important religious sphere, feasts also serve to provide links to
the gods or ancestors that can also be used to define the structure of relations be
tween social groups or categories within a region or community. They also pro
vide a crucial mechanism for the process of labor mobilization that underlies the
political economy and they serve to articulate conversions between spheres of
exchange (see Chapter 9).
Examples of these features are ubiquitous in the anthropological and historical
literature on Africa. But, important as these features are, a proper analysis of
feasting must move beyond functional consideration of such general structural
roles to examine the dynamic nature of feasts as privileged ritual sites of political
and economic practice, to show in detail how and why they work, and to demon
strate how feasts are implicated in social change.
F EASTS AS RITUAL, RITUAL AS POLITICS

As noted earlier, feasts are defined as public ritual events of communal food and
drink consumption. This means that they differ in some way from daily con
sumption practices; but at the same time, the ritual symbolism of feasts is consti
tuted through a complex semiotic relationship to daily consumption patterns,
and both form part of a common semiotic field (see Douglas 1984; Elias 1978). To
adapt a concept from linguistic analysis, feasts may be viewed as the "marked"
69
· Michael Dietler
form to the "unmarked" meal. To illustrate this idea through a simple example,
the "communion" event of the Catholic mass may be seen as essentially a feast
involving the ritual distribution and consumption of bread and wine. The mean
ing of this consumption event both derives from and plays upon its original
meaning in the context of daily meals, but is, at the same time, dramatically
transformed by the symbolic framing devices that distinguish the mass as a the
ater of ritual action. Of course, quotidian meals are also, to a certain extent, "rit
ualized" events in that they are highly structured sequences of action that serve
to shape the "habitus" (Bourdieu 1990) of individuals (inculcating dispositions
guiding practice and naturalizing the social order) and their constituent elements
can be manipulated subtly to make political statements (Appadurai 1981) . But
they differ from more formal ritual "feast" events in being generally less con
sciously public performances. The ways in which feasts are symbolically marked
as distinct from daily practice are variable, and extremely important for archaeol
ogists to understand. More will be said about this later. For the moment, it is im
portant to recognize that this relationship between feasts and daily meals is
crucial both to understanding the symbolic significance of feasts and to our very
ability to identify feasting archaeologically.
At this point, it is necessary to expand the discussion slightly and set feasts in a
broader theoretical context by saying a few words about the emerging anthropo
logical understanding of the nature of ritual in general and its relationship to pol
itics and power. One consistently common feature of recent views in cultural
anthropology is a rej ection of assumptions that continue to underlie many ar
chaeological interpretations: that ritual is a straightforward reflection of social
and political structure and/ or an inconsequentially epiphenomena! aspect of the
"superstructure" of society. The older Durkheimian functionalist view of ritual
as an adaptive mechanism (a kind of all-purpose adhesive substance) for the
maintenance of social solidarity (or "system equilibrium," in the terminology of
one of the more archaeologically popular versions of functionalism) is also now
generally recognized to be a partial and flawed understanding. This is not to deny
or ignore that rituals frequently serve to create and reproduce a sense of commu
nitas (Turner 1969; van Gennep 1960). But anthropological understanding of the
symbolic work of ritual has moved well beyond this feature, and attention has
now turned to the historically instrumental role of ritual in creating, defining,
and transforming structures of power.
The relationship between ritual and politics is seen to be an intimate one: to
paraphrase one recent review of the subject, there is no ritual without politics
and no politics without ritual (Kelly and Kaplan 1990:141). However, this relation
ship is also a complex one that has generated an extensive, and often contentious,
literature in anthropology (cf. Apter 1992; Bell 1996; Bloch 1989; Cohen 1979; Co-
70
maroff and Comaroff 1991, 1993; Kelly and Kaplan 1990; Kertzer 1988; Tambiah
1985; Turner 1969) . In this latter vein, many scholars (e.g. , Bloch 1989) see ritual as
essentially a conservative authoritarian force that acts to mystify asymmetrical
relations of power, while others (e.g. , Apter 1992; Comaroff and Comaroff 1991,
1993; Kertzer 1988) view it as an important historical force for both the repro
duction and the transformation of relations of power. This latter, more fluid,
practice-oriented perspective approaches ritual as an instrument of both domina
tion and resistance, as an arena for the symbolic naturalization, mystification,
and contestation of authority. It should go without saying, but let me reiterate an
earlier caveat by noting that, in treating the political dimension of ritual, one is
not attempting to reduce rituals, such as feasts, to an activity that is only about
politics. There is clearly a lot more of considerable significance going on. But rit
uals and politics are inseparably linked in ways that are important to understand.
The effectiveness of ritual in this domain stems from several features. As Cohen
(1979) has noted, the most emotionally compelling and effective political symbols
are those that are not overtly political but rather tend to have an ambiguous 'bivo
cality" melding intense personal experience of existential identity issues with
broader structures of power. By "condensing meaning" in this way, ritual symbol
ism infuses social norms and categories with emotion (Turner 1967:29 ). This is one
reason that traumatic life-crisis events, such as death, so commonly serve as a
major ritual arena for the manipulation of political symbolism (Morris 1992). The
emotional power of rituals also stems from certain theatrical media and sensory
mechanisms commonly employed (in various combinations) in performance that
tend to frame ritual as symbolically pregnant action marked off from other kinds
of daily practice, thus focusing people's attention and rendering them receptive to
episodes of heightened emotional experience. These devices include such things
as music, dancing, rhythmic verse, role acting, evocative staging and costumes,
and intoxication. Dramaturgical techniques such as the creation of images
through contrast and the dialectical resolution of contradictions merge emotional
catharsis with important pedagogical functions. Symbolic references to the past
are commonly invoked to create an impression of seainless continuity, and highly
formalized, repetitive sequences of action serve to limit the perception of alterna
tives and to naturalize the projected order by linking it to the "natural" experience
of the passage of time ( c£ Bloch 1989; Dietler and Herbich 1993; Moore and Myer
hoff 1985; Tambiah 1985; Turner 1967).
Like all ritual, feasts provide a site and a medium for the highly condensed
symbolic representation of social relations. However, again as with other ritual,
they express idealized concepts: the way people believe relations exist, or should
exist, rather than how they are necessarily manifested in daily activity. Such rep
resentations may either camouflage, naturalize, or contest asymmetries of
71
Michael Dietler
power, and struggles over the control of representations and their interpretation
by differentially situated actors are an important site of historical change. How
ever, in addition to this idealized representation of the social order, rituals also
offer the potential for manipulation by individuals or groups attempting to alter
or make statements about their relative position within that social order as it is
perceived, presented, and contested. As such, feasts are subject to simultaneous
manipulation for both ideological and more immediately personal goals. In other
words, individuals can use feasts to compete against each other without ques
tioning a shared vision of the social order that the feast reproduces and natural
izes, or they can use feasts to simultaneously struggle for personal position and
promote contrasting visions of the proper structure of the social world.
Feasts are a particularly powerful form of ritual activity that also have the
pragmatic virtue of being potentially visible in the archaeological record. Be
cause of their inherent emotive and symbolic power, feasts are very often inti
mately embedded in rites de passage or life-crisis ceremonies, such as funerals; and
it is this feature that often renders them archaeologically detectable as distinct
events. Moreover, the culinary nature of feasts generally necessitates the use of
containers for both preparation and consumption. Very frequently, over the past
10,000 years of human history at least, a substantial portion of these containers
has tended to be made of ceramic or metal, which preserve extremely well in the
archaeological record even when broken. Detecting feasts in the Paleolithic is, of
course, considerably more difficult (see Dietler 1996; Marshall 1993; Perles 1996);
and the political dimensions of feasting are somewhat different among forager
societies (see Hayden, Chapter 2; Wiessner 1996) than among the agrarian soci
eties discussed here.
The previously asserted potency of feasts as a particular form of ritual activity
derives from the fact that food and drink serve as the media of expression and
commensal hospitality constitutes the syntax in the context of a ritual of con
sumption. Food and drink are highly charged symbolic media because they are
"embodied material culture" : that is, a special form of material culture produced
specifically for ingestion into the body. They are a basic and continual human
physiological need, which are also a form of "highly condensed social fact" (see
Appadurai 1981:494) embodying relations of production and exchange and link
ing the domestic and political economies in a highly personalized way. Moreover,
although eating and drinking are among the few biologically essential acts, they
are never simply biological acts. Rather, they are learned "techniques du corps"
(Mauss 1935)---culturally patterned techniques of bodily comportment that are
expressive in a fundamental way of identity and difference. Alcoholic beverages
frequently have a privileged role in the feasting context because they are essen
tially food with certain psychoactive properties resulting from an alternative
72
R I T U A L S O F C O N S U M PT I O N , C O M M E N S A L P O L I T I C S , A N D P O W E R I N A F R I C A N CO N T EXTS
means of preparation that tend to amplify their significance in the important dra
maturgical aspects of ritual (Dietler 1990) . Moreover, this property of fermenta
tion as a quasi-magical transformation of food into a substance that, in turn,
transforms human consciousness augments the symbolic value of alcohol in the
common liminal aspects of rituals.
Both food and drink are also a highly perishable form of good, the full politico
symbolic potential of which is realized in the drama of public-consumption
events that constitute a prime arena for the reciprocal conversion of what Bour
dieu (1990) metaphorically calls "symbolic capital" and economic capital. Public
distribution and consumption of a basic need derives added symbolic salience
from its demonstration of confidence and managerial skill in the realm of pro
duction. More importantly, however, consumption is played out in the extremely
powerful idiom of commensal hospitality. I believe this feature is crucial to un
derstanding the political dimensions of feasts, and it is for this reason that I have
chosen to emphasize what I have called "commensal politics."
Anigbo has asserted that "Commensality is not essentially about expressing
love or intimacy" (1996:101-102), because it is clear that even individuals aggres
sively opposed to each other may use commensality to define their relationship.
However, commensality is a powerfully expressive trope of intimacy that creates
and reproduces relationships capable of encompassing sustained aggressive com
petition by effectively euphemizing it in a symbolic practice that encourages col
lective misrecognition of the self-interested nature of the process. And as
Bourdieu has pointed out
In the work of reproducing established relations-feasts, ceremonies, exchange 'of

gifts, visits or courtesies, and, above all , marriages-which is no less vital to the exis
tence of the group than the reproduction of the economic bases of its existence, the
labour required to conceal the function of the exchanges is as important as the
labour needed to perform this function. (199o:n2)
Hence, one begins to glimpse the symbolic force at the heart of commensal ritu
als. Feasts act as a form of symbolic "metaproduction," constituting and euphe
4
mizing broader social relations in terms of the basic commensal unit.
Furthermore, commensal hospitality may be viewed as a specialized form of
gift exchange that establishes the same relations of reciprocal obligation between
host and guest as between donor and receiver in the exchange of other more
durable types of objects (Mauss 1966). The major difference is that food is de
stroyed in the act of commensal consumption at a feast; and, moreover, de
stroyed by ingesting it into the body. This is a literal "embodiment" or
"incorporation" of the gift and the social debt that it engenders. Aside from the
powerful symbolic dimension of this practice, it also results in the pragmatic fact
73
Michael Dietler
that, unlike durable valuables, the food consumed cannot be recirculated (or
"reinvested") in other gift-exchange relationships: food must be produced anew
through agricultural and culinary labor in order to fulfill reciprocal obligations.
A clarification should be raised here, however, because food can also be used for
nondestructive exchange in the same fashion as durable valuables. In contrast to
the prepared food consumed at feasts, this food may be either raw (e.g. , yams,
sacks of flour), processed (e.g. , cooked or smoked meat), or even live potential
food (e.g. , chickens, goats, cattle). In the case of live animals, in particular, the po
tential for long-term reinvestment is obvious; but even the more perishable forms
may be quickly redeployed to a certain extent in other local exchange networks or
in subsequent commensal hospitality. The exchange of food in this manner may
take place completely outside of a commensal-consumption context that one
would properly call a feast; or a feast may serve as the arena for such exchanges. In
the latter case, different kinds of foods may sometimes be used for the feast and
the exchange transaction. Although both of the two political uses of food de
scribed above (commensal consumption and nondestructive gift exchange) may
take place at feasts, it is important for the analytical purposes of this discussion
that the distinction between them not be obscured by subsuming them both
under the general termfeasting. They are not the same thing (Dietler 1996).
Commensal consumption (which, to reiterate, is here taken to be a definitive
attribute of feasts) places obvious limitations on the possibilities of the guest/ re
ceiver to redeploy the food (s)he has received in the fulfillment of reciprocal obli
gations of other exchange relationships: it removes goods permanently and
immediately from circulation. It is thus a more temporally restrictive use of food
in manipulating social relations than is the nondestructive exchange pattern that
may or may not accompany a feast. Because of the commensal aspect, it is also a
potentially even more subtle manipulation.
The critical point to retain is that commensal hospitality centering on food and
drink distribution and consumption is a practice, which, like the exchange of
gifts, serves to establish and reproduce social relations. This is why feasts are
often viewed as mechanisms of social solidarity that serve to establish a sense of
community. However, as Mauss (1966) long ago pointed out, these are relations
of reciprocal obligation that simultaneously serve to create and define differences
in status. The relationship of giver to receiver, or host to guest, translates into a
relationship of social superiority and inferiority unless and until the equivalent
can be returned. 'l\.s the Bemba [of Zambia] say, 'You have eaten namba (the
sticky gum from the munamba tree) and it sticks to your stomach' . . . i.e., you
have filled your stomach with food from some one and it puts you under a per
manent obligation to him" (Richards 1939: 135) . In this feature, the potential of
hospitality to be manipulated as a tool in defining social relations, lies the crux of
74
R I T U A L S O F CO N S U M PT I O N , C O M M E N S A L P O L I T I C S , A N D P O W E R I N A F R I C A N CO N T E XTS
commensal politics. The hospitality of feasting is, of course, only one of many
potential fields of political action that may be variably articulated. As will be
shown in more detail in the later discussion of the Luo case, feasting may be
strategically used by individuals either to complement or to compete against
forms of prestige and power derived from other domains of competition for
symbolic capital, such as warfare, magic, gift giving, public oratory, etc. (cf. Bour
dieu 1990; Lemonnier 1990; Modjeska 1982). However, the special attribute of
feasting is that, because of the intimate nature of the practice of sharing food and
the symbolic power of the trope of commensality, of all forms of gift prestation
it is perhaps the most effective at subtly euphemizing the self-interested nature of
the process and creating a shared "sincere fiction" (in Bourdieu's apt phrase) of
disinterested generosity.
M O D E S O F COM M E N SA L POLITICS
I will now turn to some selected African empirical contexts in order to further ex
plicate several previously defined theoretical constructs: specifically, three differ
ent modes of commensal politics, or general patterns in the ways that feasts
operate symbolically in serving as sites and instruments of politics (Dietler 1996) .
One can, of course, propose a variety of more or less useful classifications of
feasts based upon a range of criteria, such as scale of inclusion (household,
neighborhood, community, etc.), specific cultural context (funerary feasts, mar
riage feasts, initiation feasts, etc.), or manifest and latent social and econoinic
functions (religious feasts, labor feasts, community celebrations, etc.; see Hayden
1996 and various chapters in this volume for some alternative classifications).
However, the value of a classification is entirely relative to the problem it is in
tended to solve.
The distinctions outlined here are analytical constructs designed to further un
derstanding of the specific problem of the political dimensions of feasting ritual.
As will become clear in the discussion to follow, a concept such as the "empow
ering feast" crosscuts many of the other potential categories noted above because
it highlights the ways that certain political processes are operative in all these ap
parently different feasting contexts. Hence, I am not really proposing here a ty
pology of "kinds of feasts" that can be linked directly to, for example, certain
patterned deposits of archaeological material (insofar as that might be possible).
Rather, I am attempting a heuristic dissection of the politico-symbolic dimension
of feasting as an institution. The application of insights derived from this analysis
to the archaeological record must always rely upon complex contextual argu
ments that accommodate the specific cultural conditions of a given case (see
Dietler 1990, 1996, 1999a) .
The first of these three modes of commensal politics to be analyzed is directed
75
Michael Dietler
toward the acquisition or creation of social (and economic) power and the latter
two are directed toward the maintenance of existing inequalities in power rela
tions. The first two operate primarily through an emphasis on quantity, and the
last operates through an emphasis on style. The first two work through the idiom
of donor I receiver, superiority I subordination relations within an inclusive bind
ing exchange dyad, whereas the latter works through the idiom of diacritical ex
clusion in an insider I outsider relationship.
EM POWE R I N G F EASTS
The first of these feast patterns, which I call the "empowering feast," involves the
manipulation of commensal hospitality toward the acquisition and maintenance
of certain forms of symbolic capital, and sometimes economic capital as well.
The term covers a range of symbolic consumption practices that are instrumen
tal in negotiating social positioning. In previous publications (Dietler 1996, 1999a),
I have referred to this category as the "entrepreneurial feast," but subsequent dis
cussions have led me to believe that this term has the potential to create some
misunderstanding. It was intended simply as a convenient trope, but runs the risk
of being interpreted literally as a sort of crude neoclassical economic concept.
The change in terminology also, I believe, helps to underline the fact that I am
not attempting to distinguish a type of specialized feast involving openly aggres
sive competitive contests (as distinct from, for example, a "harmonious egalitar
ian" community celebration) . Rather, I use the more passive term "empowering"
as a way to indicate an effective political role of feasting events of various kinds
rather than necessarily an overt intention of the hosts. Although this role is
sometimes fully; or at least partially; recognized by the participants, much of the
effectiveness of this political mechanism derives from the fact that it often entails
a kind of collective misrecognition or euphemization of the self-interested na
ture of the practice. It involves what Bourdieu characterizes as a "sincere fiction
of disinterested exchange" (1990:112). Indeed, a major part of Bourdieu's argu
ment about habitus is that the skill and grace of the genuinely competent social
actor relies upon that actor being unaware of the principles that inform his or her
actions. Awareness arises in the context of mistakes, of alternative actions that
raise uncertainties about precisely how one should act. Although the limited role
of consciousness in social action is an aspect of Bourdieu's work that is perhaps
overstated and subject to some question and criticism, nevertheless, I believe that
he is correct in identifying the euphemization of self-interest as an important as
pect of ritualized forms of exchange, such as feasts.
Another preliminary preemptive disclaimer is necessary to clarify the fact that
I have previously referred to empowering feasts as a domain of inherent social
"competition" (Dietler 1990, 1996) . This word also has the potential to give rise to
76
R I T U A L S O F C O N S U M PT I O N , C O M M E N S A L P O L I T I C S , A N D P O W E R I N A F R I C A N C O N TEXTS
misunderstandings, particularly in traversing linguistic frontiers in which the cul

tural coding of the term differs.5 Hence, let me reiterate that in using the term
competitive I am manifestly not referring only to activities that involve an overt ag
onistic challenge to monopolize power, with resulting explicit "winners" and
"losers." I have something more subtle in mind than the ideology of free-market
capitalism or football! Nor am I referring only to feasts that involve an escalating
scale of ostentatious reciprocal hospitality (of the well known New Guinea 'big
man" type: see Lemonnier 1990, 1996) .
Rather, I mean that feasts are inherently political, but with an understanding of
power in the sense it has acquired in the wake of work by Bourdieu (1990), Fou
cault (e.g. , 1980), and others: as a relational phenomenon rather than as a limited
good. Hence, the symbolic capital realized through empowering feasting is an in
herently "competitive" phenomenon in that it describes conditions of relative
asymmetries in relationships between people, and, moreover, asymmetries that
must be renegotiated continually through symbolic practices. This "competi
tion" is not necessarily one that strives toward aggressive domination and relent
less accumulation of power: it is often simply one of maintaining status among
peers or of defining one's peers. Nor is it necessarily one that directs an explicit
challenge to particular individuals or groups: it often involves simply a positive af
firmation of the prestige of the host and his /her group that implicates others
only in a relative, indirect, general sense. There is clearly a significant difference
between, for example, maintaining friendly reciprocal obligations with one's
neighbors in hosting small beer parties and the agonistic attempts by New
Guinea big-men to crush their rivals with hospitality. There are generally cultur
ally specific behavioral sanctions and moral philosophies of legitimate power
that restrict the escalation of such commensal practices and assure that cases of
the latter extreme form are fairly unusual. But some degree of competition is in
volved in all these empowering feast contexts. Those who do not keep up fall be
hind. Such practices always affect the relative status and influence of participants
and the quality of relationships. In this sense, commensal politics is always com
petitive in its effects, even though the political implications may be subtle, lim
ited, and thoroughly euphemized.
Consequently; it must also be recognized that, for example, feasts conceived
sincerely by the participants as harmonious celebrations of community identity
and unity are simultaneously arenas for manipulation and the acquisition of pres
tige, social credit, and the various forms of influence, or informal power, that
symbolic capital entails. These are not mutually exclusive functions that require,
or even enable, one to assign a given feast event to one of two alternative cate
gories (e.g., solidarity vs. competitive). Rather, one must recognize the complex
political polysemy of feasts. They both unite and divide at the same time. They si-
77
Michael Dietler
multaneously define relationships and boundaries. This feature may well entail
certain structural contradictions of interest, but it does not necessarily result in
conflict, or even the perception of incongruity, in the course of practice.
Finally, let me also emphasize that, in treating the political dimension of things
such as religious feasts, I am manifestly not attempting to make a vulgar reduc
.
tionist argument of the bottom-line "practical reason" variety. I do not wish to
reduce the participants to unidimensional cynical manipulators and deny their
religious sincerity and the affective motivational force of religious belie£ Quite
the contrary. Rather, I believe this is an issue of audience: it must be remembered
that all rituals, including feasts, have simultaneous multiple audiences. Religious
feasts, for example, are clearly directed at communicating with gods, ancestors,
or spiritual forces: they are a sincere attempt to "bring them to the table," so to
speak. But they are simultaneously directed toward an audience of living hu
mans, and perhaps several groups or categories of living humans. Feasts are pol
ysemous, in terms of audience, motivation, and forms of empowerment.
Concentrating on an analysis of the political should not be interpreted as a denial
of the importance of other dimensions.
Symbolic capital translates into an ability to influence group decisions or ac
tions. · This influence derives from the relations created and reproduced in the
process of personal interaction. In the case of feasting, those are multiple rela
tions of reciprocal obligation and temporary sentiments of social asymmetry be
tween host and guests created through displays of hospitality. The "power"
derived from this sort of commensal politics may range from a subtle and tem
porary affirmation of elevated status (such as attitudes of gratitude or deference)
to demands for special rights and leading managerial roles in group decisions. In
societies without formal specialized political roles, hosting feasts is very often a
major means of acquiring and maintaining the respect and prestige necessary to
exercise leadership. It does not create the power to command, but it does imbue
individuals with the moral authority that is a necessary condition to exert per
suasive influence.
In societies where institutionalized political roles or formal status distinctions
exist, but without fixed hereditary rules for determining who may fill them, host
ing feasts is often the means by which individuals assume and hold these roles
and statuses. In all such cases, this kind of power is continually being renegoti
ated, sustained, and contested through commensality. This form of commensal
politics has been described by various anthropologists in many contexts across
Africa (11ot to mention the Pacific, Latin America, Asia, and the rest of the
world) . Among the Yoruba of Nigeria, for example, men move up the social hier
archy by taking titles. This is accomplished by displays of prestige in feasts fur
nished with large quantities of beer or palm-wine (Obayemi 1976). Among the
78
Dorze of Ethiopia, assumption of the title of balak'a and its elevated political sta
tus requires the hosting of feasts so lavish that there is even some reluctance to
undertake the initiation procedure (Halperin and Olmstead 1976). Similarly,
among the Koma of Cameroon, there is a formalized age-grade system that leads
to the possibilities for individuals to become high-ranking initiates and respected
makers of policy within the village as they progressively gain access to more se
cret religious knowledge with each step. Moving up through this system requires
the sponsorship of special feasts known as "cattle dances" that are held by a man
to honor his wife and are fueled with a great deal of millet and sorghum beer and
beef. These can be held by a man only six or seven times in a lifetime, and the
ability to hold such a feast is decided by fellow villagers who judge whether an in
dividual has acquired the necessary symbolic and economic capital for the rank
to which he aspires. There are, of course, many other feasting contexts for ac
quiring personal prestige that are not tied directly to the age-grade structure.
These include beer parties hosted for gatherings on market days, for work feasts,
and for various ritual activities (Garine 1996) .
In societies with an egalitarian political ethos, the self-interested manipulative
nature of the process may be concealed or euphemized by the fact that it is car
ried out through the socially valued and integrated institution of generous hospi
tality, and it may even be perceived by the participants as a leveling device.
However, this apparent leveling is, in a sense, merely the conversion of economic
capital into symbolic capital. In fact, feasts may be used as a form of what Firth
(1983) has called "indebtedness engineering" every bit as much as the prestation
of valuables. This is quite clear in the cases where feasting is recognized by the
participants to be openly aggressive, as with the escalating beer feasts between
exchange partners among the Mambila of Nigeria where the failure to return a
yet more copious feast results in jeering and ridicule (Rehfisch 1987) . But it can be
equally operative in cases where competitive manipulation is more subtly euphe
mized and where there is no escalation of prestation.
Commensal hospitality may be manipulated in the empowering feast pattern
for economic advantage as well as for political power, especially through the in
stitution of the "work feast" ; and this was particularly true of societies in the
past. As this institutionalized practice is more thoroughly analyzed elsewhere in
this volume (see Chapter 9), I will simply note here that the "work feast" is a
form of labor mobilization practice found throughout Africa (and indeed,
around the world) . It constitutes one pole in a continuum of labor mobilization
practices, here called "collective work events" (CWE), for which the other pole is
the "work exchange." The work feast is an event in which a group of people is
called together to work on a specific project for a day and the participants are
then treated to food and/ or drink, after which the host owns the proceeds of the
79
Michael Dietler
day's labor. Before the development and spread of the capitalist monetary econ
omy, such CWEs were virtually the only means (excluding slavery) by which a
group larger than the domestic unit could be mobilized for a project requiring a
larger communal effort. This is particularly true of societies without centralized
political authority, but even obligatory forms of labor (corvee) organized by chiefs
or kings operate within this idiom.
Work feasts are extremely important in the political economy because of the
context they provide for the acquisition and conversion of symbolic and eco
nomic capital. In the first place, as with all other types of feast, they provide an
opportunity to make public statements about prestige and acquire symbolic cap
ital. A lavish work feast augments the reputation of the host in the same way that
sponsoring a communal ritual does. However, it also provides a means of
harnessing the labor of others in order to acquire economic capital that can sub
sequently be converted to symbolic capital by several means. In effect, work
feasts act as a mechanism of indirect conversion in multi-centric economies that
can provide a potential catalyst for increasing inequality in social relations (see
Chapter 9).
S O C I O E C O N O M I C PA R A M E T E R S O F E M P O W E R I N G F E A S T S
The empowering feast pattern operates o n a variety o f scales and in numerous
contexts within a given society. It may extend from the private hosting of a pot of
beer among a small group of friends, to the hosting of trade partners from an
other community, to the sponsorship of maj or community life-crisis ceremonies
and religious festivals. Guests may include members of the local community or
people from other communities. The extent of the symbolic capital derived from
these activities varies according to the context, the lavishness of the hospitality
provided, and the range of guests convened. The host may be either an individual
household, a kinship unit, or an entire community. In the latter cases there are
usually certain individuals who act as managers and derive prestige from their
role in successfully organizing and executing feasts that represent the group to
outsiders; hence prestige accrues to both the hosting group as a whole and to cer
tain influential individuals who can mobilize group activities.
Although most households will engage in some form of this kind of feasting
behavior, hosting large-scale feasts requires considerable planning, time, and
labor (for both agricultural production and culinary preparation), as well as large
surplus stocks of food and/ or drink. The kinds of food and drink traditionally
available in most African agrarian societies (and most prehistoric societies) would
generally have had very limited storability, especially once prepared for con
sumption. This would necessitate, in most cases, a large labor force for final
preparation and serving just prior to the feast as well as command of a large
80
ready supply of agricultural produce. The institutional arrangements for mobi

lizing these large supplies of labor and food vary a great deal from society to so
ciety, but in all cases the organization and execution of a large feast requires the
host to be a good manager. It is usually advantageous for a household sponsoring
a feast to be able to provide a large portion, if not the bulk, of the labor and raw
materials from its own reserves, and a high incidence of polygyny among big
men and other types of informal leaders is often cited in this connection ( c£
Boserup 1970 :37; Geschire 1982; Friedman 1984; Lemonnier 1990) .
In some cases work feasts may also be employed to harness the labor of others
in differentially increasing the productive base of certain households (see Chap
ter 9) . In most cases of very large feasts, however, the host must mobilize addi
tional food and labor contributions through personal networks of social
obligation. These networks of support are established by adept building up of
symbolic capital over the years through various arenas of prestige competition
and various deployments of economic capital. Hence a large, lavish feast is not
just an isolated event. It is a moment of public ritual drama in a continuous
process of political manipulation that serves as an advertisement of the scale of
the support base that a social manager has been able to construct through vari
ous transactions, at the same time that it produces further symbolic capital.
It is important to underline the significant scale of the resources that are de
voted to this kind of commensal political activity in most societies, and especially
to note the resources devoted to the production of alcoholic beverages for such
purposes (see Dietler 1990 :361-362) . One frequently sees archaeological estimates
of "subsistence" food production requirements that both ignore the importance
of alcoholic beverages and do not take such crucial festive requirements for social
reproduction and politics into account. Yet, where attempts have been made to
measure such things in ethnographic contexts, the figures are consistently im
pressive. Haggblade (1992), for example, noted that households in Botswana con
sume 15-20 percent of all the grain produced in the form of sorghum beer, much
of it consumed in work feasts during the harvest time when many men remain
intoxicated most of the time. Similarly, Richards (1939: 80) estimated that an aver
age household among the Bemba of Zambia used about 400 pounds of millet per
year in brewing beer, out of a total production of about 2,400 pounds of grain
(i. e., about 17 percent); and for chiefs, who commonly drink beer every day as
part of their duties of hospitality (and may virtually subsist on it), the quantity is
much higher. Netting (1964) estimated that the Kofyar of Nigeria consume about
40 gallons (151 liters) of millet beer per person each year, while annual consump
tion estimates for the city of Ouagadougou in Burkina Faso ran to 236 liters of
traditional beer per person, with half the annual grain consumption for a family
being in the form of beer (Pallier 1972; Saul 1981) . Likewise, Garine (1996) noted
81
Michael Dietler
that among the Koma of Cameroon sorghum beer provides about a third of the
total calories consumed during the year. He further calculated the large invest
ments involved in hosting different kinds of feasts. For one age-grade ceremony,
one needs 70 pots of beer (490 liters made from about 100 kg of cereal), plus an
other 50 kg of sorghum flour for 24 porridge balls, plus a number of cattle (that
are worth up to $400 each); for a cattle dance, one needs 75 pots of beer, 20 por
ridge balls, and the most prestigious cattle; and for the funeral of a woman, one
needs 37 pots of beer (Garine 1996) . Similarly, Reh:fisch (1987) noted that, among
the Mambila of Nigeria, one beer feast in the competitive series he studied mobi
lized over 480 pots of beer (plus 47 chickens, 1 sheep, a dog, kola nuts, and to
bacco) to counter a previous feast in which 430 pots of beer (and 30 chickens) had
been offered. In Manga (a Mossi town of about 7,000 inhabitants in Burkina
Faso), memorial ceremonies called kuure are the occasions for the most lavish
beer feasts. In one week, five kuure were held in one ward, consuming 1,900 kg of
red sorghum made into beer (with seven cartloads of wood-1,400 kg-required
for brewing and cooking for one of these feasts alone); and, during a single dry
season, within the town as a whole, IO tons of sorghum were converted into beer
for these memorial feasts alone, with a total annual festive consumption esti
mated at 14 tons of grain brewed for beer (Saul 1981) . Finally, among the Luo of
Kenya, funerals are the occasions for the most lavish feasts mounted in this soci
ety. These events frequently result in the serious impoverishment of the hosting
family, and the Kenyan government has even attempted to intervene legally to
limit the scale of Luo funerals.
All of this represents a substantial investment of agricultural and culinary
labor in the essentially political activity of acquiring and maintaining symbolic
capital and creating and sustaining social relationships. Moreover, contrary to
some persisting archaeological conceptions of economically autonomous do
mestic units, it represents a substantial portion of domestic agricultural produc
tion that is regularly dedicated from the beginning to flowing outside the
household and being consumed by people in other domestic units. Hence, it is
clear that recognizing the importance of feasting for both social reproduction
and political action in agrarian societies should provoke a corollary recognition
of the scale of productive labor and resources necessarily devoted to these crucial
features of social life. Feasts are an instrumental force in the organization of pro
duction as well as in the structuring of social relations and power.
PAT R O N - R O L E F E A S T S
The second major mode of commensal politics that may b e distinguished I will
call the "patron-role feast." This involves the formalized use of commensal hos
pitality to symbolically reiterate and legitimize institutionalized relations of
82
R I T U A L S O F C O N S U M PT I O N , C O M M E N S A L P O L I T I C S , A N D P O W E R I N A F R I C A N C O N T EXTS
asymmetrical social power. This corresponds to a specific form of what has tradi
tionally been called "redistribution" in the literature of economic anthropology
(cf. Polanyi 1957; Sahlins 1972). The operative symbolic trope behind this form of
commensal politics is the same as for the previous mode: the relationship of re
ciprocal obligation engendered through hospitality. In this case, however, the ex
pectation of equal reciprocation is no longer maintained. Rather, the acceptance
of a continually unequal pattern of hospitality symbolically expresses the formal
ization of unequal relations of status and power and ideologically naturalizes it
through repetition of an event that induces sentiments of social debt. On the one
hand, those who are continually in the role of guests are symbolically acknowl
edging their acceptance of subordinate status vis--vis the continual host. On the
other hand, the role of continual and generous host for the community at large
comes to be seen as a duty incumbent upon the person who occupies a particular
elevated status position or formal political role. Institutionalization of authority
relies on this binding asymmetrical commensal link between unequal partners in
a patron/ client relationship.
This is the principle that lies behind the regular lavish hospitality expected of
chiefs and kings in almost all societies where they exist, and certainly those in
Africa. This sense of obligation for generosity in a commensal context is nicely
encapsulated in the Baganda definition of the essential qualities of a good chief:
"beer, meat and politeness" (Mair 1934:183) . Among the Nyoro, also of Uganda,
the king was expected to regularly hold great feasts and give gifts, and many of
his special names emphasize this expected generosity. A decline in the lavishness
of the feasts provided by the king was cause for complaints. Chiefs under the
king were also expected to follow this pattern on a more local level (Beattie 1960 ).
Similarly, among the Pondo of South Africa, Hunter noted that "Generosity is a
primary virtue and the mark of a chief" It was particularly important for the
chief to dispense generous hospitality, and "there was always much beer at the
great places." Indeed, the Pondo word for "chief," Inkosi, is also the word in
everyday usage for "thank you" (1961:387-388) . Dillon provides a more detailed
idea of the scale of such obligations among the Meta of Cameroon:
The foremost duty of a Jon [village chief] in the mind of any Meta person was to
feed his people. This was done most lavishly when he provided several grand feasts
at the time of his installation. Yet the Jon also entertained more modestly on a regu
lar basis. Each time that the villagers worked for him he was obligated to feed them
when they had finished their task, and he hosted the entire village whenever he held
an annual celebration involving dancing. Likewise, if the village went to war, . . . the
Jon . . . had to provide the returning warriors with an appropriate reception. But
even if no such activities had taken place within a year, the people sometimes still
expected the Jon to give them a feast simply because he was their leader.
83
Michael Dietler
Besides hosting the entire village on special occasions, the Jon frequently enter
tained individuals and small groups. He was expected to have wine ready for such
visitors at any time, as well as for the mikum si [senior village notables] when they
met on the village rest day. Moreover, if there was a market in his village, he held
court in a house just outside of it, providing palm wine for both the local notables
and important visitors . . . .
Since the Jon was continually receiving visitors-on week days, on village rest
days, on special occasions, and on market days--he was in an excellent position to
use the norms of hospitality as a political tool. He could honor and reward allies as
well as cultivating the nonaligned. At the same time, he gained prestige with the en
tire community by feeding it well. (Dillon 1990:129-130)
Similarly; among the Bemba of Zambia, Richards noted that the chief was re
sponsible for feeding all those who provided tribute work on his corvee projects,
courtiers, executive officials, visiting councilors, and others. She estimated that
during one nine-month period the main chief provided food and beer for at least
one day for 561 men and 324 women who provided labor and, among others,
about 40 tribal councilors with their wives and retinue at least twice (1939:147). As
she noted, the culinary labor for this is provided by the multiple wives of the
chief, under the direction of the senior wife who was necessarily a woman with
"a good deal of organizing ability; capable of supervising younger wives, arrang
ing for the endless grinding and brewing required in the capital, and the stirring
of huge pots of porridge to be served in enormous eating-baskets about eight
times the size of an ordinary icipe" (1939:148). As she further stated, "The whole
of this system of distributing food is of course necessary to the chief if he is to
make gardens and conduct tribal business through his councilors. But it is more
than this. The giving of food, as in most African tribes, is an absolutely essential
attribute of chieftainship, just as it is of authority in the village or household"
(1939:148). Correspondingly; the failure of a chief to provide food for his subj ects
considerably weakens his prestige. "The tradition of the generous king survives
as a standard against which the modern ruler is constantly measured, and meas
ured to his disadvantage" (Richards 1939:264).
It is important to emphasize that this kind of practice is not, as has sometimes
been posited in functionalist accounts, necessarily a systemically adaptive means
of providing balanced food security for a population. Rather, it is first and fore
most a politico-symbolic device for legitimizing status differences, and any nutri
tional benefits to the population at large are highly variable (see Friedman 1984;
Hayden and Gargett 1990; Pryor 1977). This political function is underlined by the
fact that challenges to chiefly authority can also be launched through feasting.
Anigbo (1996) provides an excellent example of such a challenge among the Igbo
of Nigeria in the form of a case in which two contestants for the chiefship fought
84
over who had the right (by virtue of lineage seniority) to convene an important
"feast of yams," which sets the date for eating new yams. This conflict culmi
nated in a dispute over who would host the centrally important omabe (mask
feast) : each candidate ended up holding this feast on a different day, with the sup
porters of each boycotting the feast of his rival.
Chiefs raise food supplies for this lavish public hospitality in a variety of ways
(e.g. , see Hunter 1961:384-389; Richards 1939; Schapera 1938). Often tribute in food
and drink furnishes an important part, with individuals obligated to provide the
chief with a portion of their own production. For example, Gutmann (1926:346)
noted that Chagga chiefs collected part of their tribute in the form of a portion of
the banana beer brewed by households. He states that the people were happy to
render this tribute because it enabled the chief to maintain a continuing open
feast at his residence, which they liked to attend, but also that the chief's hench
men were constantly checking to make sure that no household brewed without
paying the beer tribute.
The work feast (especially in the more obligatory corve form), directed toward
the extensive fields of the chief, is another common mechanism for mobilizing
food stocks for such purposes (see Chapter 9 ). Among the Bemba of Zambia, for
example, Richards (1939) noted that chiefs organize the largest labor groups
found in the country to work their own fields: she estimated, for example, that
275 men-days and 210 women-days per year were required for the gardens of one
smaller chief for cutting and clearing branches, respectively (1939 :388). Moreover,
chiefs are very often ostentatiously polygynous in comparison to their people,
providing a large pool of household labor; and they sometimes have attached
forms of dependent labor (such as, in the past, slaves) .
D I A C R I T I C A L F E A STS
The third major mode o f commensal politics, which I will call the "diacritical
feast," involves the use of differentiated cuisine and styles of consumption as a di
acritical symbolic device to naturalize and reify concepts of ranked differences in
the status of social orders or classes (cf. Elias 1978; Goody 1982; Bourdieu 1984) .
Although it serves a somewhat similar general function to the previous pattern
(i. e., the naturalization and objectification of inequality in social relations), it dif
fers from it in several important respects. In the first place, the basis of symbolic
force shifts from quantity to matters of style and taste. Moreover, the emphasis
shifts from an asymmetrical commensal bond between unequal partners to a
statement of exclusive and unequal commensal circles: obligations of reciprocal
hospitality are no longer the basis of status claims and power.
This is the distinction made by Goody (1982) when he differentiated between
''hieratic" and ''hierarchical" systems of stratification in his discussion of the ori-
85
Michael Dietler
gins and significance of cuisine. According to Goody, the development of such

diacritical culinary practices is often linked to the development of specialized
food preparers for the elite class (replacing wives in this role, who become com
mensal partners), and commensal exclusivity is often accompanied by class en
dogamy. The feasting patterns of the Hawaiian kingdoms described by Kirch
(Chapter 6) are a classic example of this, and of what I mean by the diacritical
feast mode . Although Goody's dichotomy may be an overly broad generaliza
tion, it is clear that the practice of diacritical feasting transforms elite feasts into
what Appadurai (1986:21) has called "tournaments of value," which serve both to
define elite status membership and to channel social competition within clearly
defined boundaries. Diacritical stylistic distinctions may be based upon the use of
rare, expensive, or exotic foods or food ingredients. Or they may be orchestrated
through the use of elaborate food-service vessels and implements or architecton
ically distinguished settings that serve to "frame" elite consumption as a distinc
tive practice even when the food itself is not distinctive . Or they may be based
upon differences in the complexity of the pattern of preparation and consump
tion of food and the specialized knowledge and taste (i. e . , "cultural capital" :
Bourdieu 1984) that proper consumption entails.
Because this type of feasting relies upon style and taste for its symbolic force, it
is subject to emulation by those aspiring to higher status. Such emulation consti
tutes an attempted elevation of status through representational means, which
may focus on either (or both) the mimetic development of styles of action (man
ners, tastes, etc.) or the use and consumption of objects (foods, service vessels,
etc) that are materialized signs of a particular social identity. This can result in the
gradual spread through a society of foods and food practices by what Appadurai
(1986) has described as a "turnstile effect." This happened in ancient Greece with
the expansion of the symposion (wine-drinking party) from its aristocratic origins
throughout urban society (Dentzer 1982; Murray 1990 ), and it was a common fea
ture in the development of European bourgeois manners and food culture (Bour
dieu 1984; Elias 1978). Junk.er (Chapter rn) offers another example among
Philippine chiefdoms.
Such emulation, and the resulting devaluation of diacritical significance, can
be thwarted only by the imposition of sumptuary laws that restrict consumption
within clear social boundaries or by the use of exotic foods and consumption
paraphernalia, access to which can be controlled through elevated expense or
limited networks of acquisition. In the absence of effective means of monopo
lization, the weakening of diacritical symbolic force caused by emulation may
provoke continual shifts in elite tastes as they react to the process of imitation.
These shifts need not be solely in the direction of increasing elaboration. In many
cases, this reaction may be toward ostentatiously simpler, rather than more elab-
86
R I T U A L S O F C O N S U M PT I O N , C O M M E N S A L P O L I T I C S , A N D P O W E R I N A F R I C A N C O N T EXTS
orate, cuisine and/ or consumption paraphernalia, depending upon the nature of

the emulation being reacted against. The fluctuating trajectories of such changes
depend upon both the nature of historical precedents and opportunities for
strategic shifts presented by invention and the incorporation of exotic elements.
Africa is an interesting case in the analysis of the diacritical feast pattern pre
cisely because, according to Goody (1982), one should not find it there. The
Pando of South Africa provide a good example of the kind of situation that
Goody took to be typical of African societies: "In spite of the fact that chiefs were
the wealthiest men in the country, chiefs always lived very much as their people,
and most still do. At the great place there is more beer and meat than elsewhere,
but otherwise there is no difference between the diet of a chief and that of com
moners" (Hunter 1961:388).
However, although it is true that African societies do not appear to have devel
oped highly elaborated diacritical cuisines to the same extent as the states of Eu
rope and Asia, they are not without diacritical food practices that serve to
symbolically demarcate kings, chiefs, and nobles. Often African royal or noble
culinary diacritica are expressed in the form of special food avoidances or privi
leged consumption of certain animals of ritual significance rather than through
consumption of specially elaborated cuisine. For example, among the Nyoro of
Uganda, the king was not allowed to eat certain kinds of common foods thought
to be of low status (e.g. , sweet potatoes, cassava, and certain vegetables), and the
men he appointed as "crown wearers" (i. e., great chiefs of high status and politi
cal authority) had to observe the same restrictions. Moreover, the king's cooks
were not allowed to have sexual intercourse just before or during their alternat
ing periods of service in the palace (Beattie 1960). Among the Meta of
Cameroon, the fan [village chief] has exclusive rights to receive, butcher, and dis
pose of certain prestigious and dangerous animals known as "noble game" (e.g.,
leopards and pythons) from which he was believed to acquire power. He would
also share specifically prescribed parts of these animals with the senior village no
tables (Dillon 1990:133-135, 153-157) . Among the Mamprusi of Ghana, the king ob
serves all the common food prohibitions of his subj ects, but in addition he does
not eat goat, black fish, or the flesh of a variety of other animals associated with
earth-shrines. Moreover, his diet is restricted to highly esteemed items (e.g. ,
guinea fowl and millet porridge) prepared separately for him by a junior wife
under the supervision of the senior wife (Brown 1975:158-159 ) Among the Bemba .
of Zambia, even when traveling, a chief cannot eat cooked food offered by his
subjects because "porridge cooked on 'impure' fire would endanger his life"
(Richards 1939:138). Rather, raw food materials are offered by subjects and these
are cooked by one of the chief's wives on a fire that she creates herself. Moreover,
"chiefs visiting each other will exchange uncooked food to be prepared by their
87
Michael Dietler
respective staffs, but only a chief's head wife could send a royal visitor dishes of
porridge and relish" (Richards 1939:138). Finally, among the Igala of Nigeria the
king is considered to be divine and is believed to not eat at all. In fact, the king al
ways eats in seclusion and his food and meals are referred to only in euphemisms
(Boston 1968:204-205).
Hence, diacritical culinary practices differentiating certain elevated kinds or
categories of people clearly do exist even in Africa. However, they are by no
means universal among African kingdoms and chiefdoms. Moreover, these ritu
als are sometimes so exclusively focused as to be effectively noncommensal to the
extent that they perhaps defy the definition of a feast, and they tend to mark cer
tain institutionalized political roles rather than social classes. Nevertheless, such
practices do perform the ritual work noted above of reifying asymmetrical rela
tions of power through the symbolic manipulation of food consumption in a pat
tern that emphasizes difference and separation of at least a small elite segment of
the soaety.
F EASTS A N D S O C I A L B O U N DA R I ES
To b e analytically useful, the concept of diacritical feasts requires some further
cautious clarification. This is because nearly all feasts actually serve in some ways
to define social boundaries while simultaneously creating a sense of community.
That is, nearly all feasts serve to mark, reify, and inculcate diacritical distinctions
between social groups, categories, and statuses while at the same time establish
ing relationships across the boundaries that they define. Gender categories and
age distinctions, for example, are very commonly signaled in what I have distin
guished as "empowering" feasting practices even among peoples with a strongly
egalitarian political ethos. For example, among the Luo of Kenya (who do not
have "diacritical feasts" in the sense defined above), categorical distinctions be
tween men and women, between elders and younger men, and between kinship
groups are signaled at feasts by spatial criteria (i. e., who sits where and with
whom), temporal distinctions (i. e., the order of serving), by different types of
drinking vessels and practices, and by different types of beer and food (see the
later discussion of the Luo case) . Similar kinds of practices, in culturally specific
manifestations, are ubiquitous in the African ethnographic record (e.g., see Carl
son 1990; Hunter 1961; Karp 1980; Peristiany 1939; Richards 1939; Sangree 1962) .
Such practices can also be a subtle but powerful means of marking the social
ranking of individuals in hierarchies of prestige and influence. As Dillon noted
for the Meta of Cameroon, for example, they "were very sensitive to subtle dis
criminations reflected in hospitality, and one man's attitude toward another
might be significantly altered, depending on whether he had made a special effort
to serve him or offered only lame excuses. People attending important gatherings
88
also noted how various guests were treated, depending on their statuses and rela
tions with the host" (1990:130) .
Similarly, social groups or networks of various kinds (affines, age grades, etc.)
are frequently marked by the same kinds of practices that are used to make other
insider versus stranger distinctions. Concepts of ethnicity; for example, very fre
quently involve beliefs (of variable accuracy) about distinctive food tastes and
culinary practices. The Luo love fish and know that this distinguishes them from
their Kisii neighbors to the east who eschew fish. They also believe that their own
revulsion at the idea of eating caterpillars sets them apart from other neighbors
to the north. Feasts can be a theater for the symbolic manipulation of such culi
nary distinctions in the expression of sentiments of inclusion and exclusion at
various levels.
Alas, the situation is yet more complicated for archaeologists looking for evi
dence of what are here defined as diacritical feasts because similar symbolic de
vices can be used to mark categories of events as well as categories of people.
Particular care must be taken not to mistake the kinds of practices that may be
used to differentiate feasts in general (as ritual events) from everyday informal
consumption in societies without diacritical feasts for those used to differentiate
social classes in societies having diacritical feasts. In many cases, this former dis
tinction (i.e., marking feasts as ritual events) is accomplished simply by differ
ences in the sheer quantity of food and drink proffered and consumed, or by a
change in the location and timing of consumption. However, the same types of
devices used as symbolic diacritica in marking social distinctions may be em
ployed to distinguish ritual from quotidian practice by serving as "framing de
vices" that act as cues establishing the ritual significance of events (see Miller
1985:181-183) . For example, feasts may be marked by special foods (e.g. , ones that
are expensive, rare, exotic, especially rich, particularly sweet, intoxicating, etc.).
Among the Luo, for example, beer is not something consumed with everyday
meals and beef is a food that is normally reserved exclusively for larger feasts (al
though accompanied by the standard range of other daily foods) . Alternatively,
special service vessels or other paraphernalia (including special forms of clothing
or other bodily adornment), or special architectural staging, may be employed
for this marking purpose. To use the Luo as an example again, they have a dis
tinctive paired set of very large beer pots (called thago and dakong'o) that are used
only at important feasts (Herbich and Dietler 1989, 1992) . Hence, among the Luo,
beer, beef, and certain kinds of ceramic vessels are all indexical markers of feasts
as ritual events. Finally, atypical complexity in recipes or in the structured order
of service and consumption may also be used to invoke such distinctions (see
Douglas 1984) .
Unfortunately, there is no handy, universal rule of thumb that will enable the
89
Michael Dietler
archaeologist to distinguish readily between "diacritical feasts" and these other

boundary-marking practices (i.e., those marking both boundaries between social
groups and categories and boundaries between ritual and quotidian contexts) . But
I believe this disentangling of symbolic logic is both possible and useful in many
instances. Each case will require a careful and critical evaluation of the contextual
and associational patterns of the evidence and a multistranded, thickly textured
interpretive argument in order to differentiate between "diacritical feasts" mark
ing social classes and the diacritical use of cuisine to mark other social categories
or to mark feasts as special ritual events. To take a highly simplified hypothetical
example: special types of ceramic tableware that are found only in funerary con
texts, but in all funerary contexts, are more likely representative of the latter (that
is, marking the ritual nature of an event); whereas those found exclusively in male
graves, but in all male graves probably imply both a ritual and categorical distinc
tion; and large bronze drinking vessels found only in a limited number of very
wealthy burials most likely indicate the operation of "diacritical feasts." But the
plausibility of such an interpretation will depend upon other evidence from settle
ment data as well (see Dietler 1996 for archaeological examples).
It is important to point out that a general increase in, for example, the com
plexity or elaborateness of the decoration of tablewares in comparison to cook
ing wares (or of ceramics in general in comparison to a previous habitation level
or archaeological period) is not necessarily an indication of the use of style in the
development of diacritical feasts. This may simply be related to an increasing
"complexification" of food-consumption patterns (in the sense of Douglas 1984)
through more marked symbolic emphasis on distinctions such as that between
ritual and quotidian dining practice. The diacritical feast pattern rests on an ex
clusive sumptuary use of style in food-consumption rituals by certain social
classes whatever the relative complexity of food patterns within the society as a
whole. More will be said about these issues later, but for the moment it is useful
to open a brief parenthetical consideration of one of the most common categor
ical distinctions defined through feasting.
F EASTS A N D G E N D E R
As noted earlier, gender is one cultural category of social identity that is nearly
everywhere marked, reified, and naturalized to some extent through feasting
practices. In fact, gender is one of the most common categorical distinctions
made through food/ drink-related practices in general, albeit in a wide variety of
culturally specific ways (Bacon 1976; Child, Barry, and Bacon 1965; Counihan and
Kaplan 1998; Dietler 1990; Gefou-Madianou 1992; Herbich 1991; McDonald 1994).
As the Luo example discussed below illustrates (cf Karp 1980; Ngok.wey 1987 for
other African examples), such categorical boundary marking at feasts may be
90
based upon various permutations of symbolic diacritica, including: (1) spatial dis
tinctions (that is, segregation or other structured differential positioning of men
and women while eating), (2) temporal distinctions (such as order of serving or
consumption), (3) qualitative distinctions (for example, in the kinds of food, drink,
or service vessels men and women are given or are allowed to consume), (4)
quantitative distinctions (in the relative amounts of food or drink served to men
and women), or (5) behavioral distinctions (that is, differences in expected bodily
comportment between women and men during and after feasting, including
such things as permissible signs of intoxication, talking while eating, reaching for
food, serving or being served, withdrawing from the meal first, and so on) .
An important feature to signal here is that, where diacritical feasting (in the
sense defined above) is in operation, these patterns of gender differentiation may
vary greatly between social classes. In other words, gender may be marked in
quite different ways within the feasting practices of each class. For example,
Goody (1982) noted a frequent pattern in which, with the development of endog
amous social classes marked by restricted commensal circles and diacritical culi
nary practices, one often notes a shift in the position of women of the elite class
from food servers and prepares to commensal partners (with a corresponding de
velopment of specialist food preparers and servers, who are sometimes male).
This does not imply any corresponding change in gendered practices in feasting
among the non-elite classes; and one can anticipate in such cases a marked differ
ence between the classes in, for example, the spatial and behavioral distinctions
by which gender is marked.
It is also important to reiterate that feasting practices, although marking bound
aries of gender identities in the ways noted above, simultaneously express rela
tionships of mutual dependence across those boundaries that, in turn, represent
and naturalize ideologies structuring larger societal relations of production and
authority. This leads to a more general point I wish to emphasize : that under
standing the gender relations that underlie, and are reproduced through, feasts is a
crucial part of the project of theoretical analysis that is necessary to make feasting
a productive focus of archaeological inquiry. That is because, in addition to the
various aspects of symbolic representation noted above, feasting frequently is sus
tained by a gendered asymmetry in terms of labor and benefits. Specifically; fe
male labor (producing and processing the agricultural supplies that are essential
for feasts) often largely supports a system of feasting in which men are the pri
mary beneficiaries in the political arena. This is one of the main reasons why there
is such a strong linkage between polygyny and male political power in Africa and
elsewhere (c£ Boserup 1970:37; Clark 1980; Friedman 1984; Geschire 1982; Lemon
nier 1990; Vmcent 1971; also see Dietler and Herbich, Chapter 9).
Female labor is often of major, or even primary; importance in agricultural
91
Michael Dietler
production, although the relative gendered contribution in this domain is by no

means uniform (Boserup 1970; Guyer 1988). However, even more common is a
dominant female contribution to the crucial culinary and serving labor that
transforms raw food ingredients into feasts (Friedl 1975; Goody 1982) . Moreover,
although cases such as the Luo (described below), in which women provide the
agricultural, culinary, and serving labor for male political activities are quite com
mon (e.g., see Bohannan and Bohannan 1968; Clark 1980), examples of the in
verse pattern (where men consistently provide the agricultural, culinary, and
serving labor that underwrites feasts formally hosted by women) are extremely
rare, if they exist at all.
At first glance, it may be tempting to interpret this fact as a systematic form of
labor exploitation, in line with Marx's observation that women probably consti
tuted the first exploited class (Meillassoux 1975:78) . However, the question of ex
ploitation frequently hinges upon a subtle contextual consideration of the
question posed by Clark for the Kikuyu: are women "controllers of resources or
themselves resources controlled by men?" (Clark 1980:367) . Although exploita
tion is frequently a justifiable analytical conclusion, this is by no means a pattern
that is universal or even generalizable in a simple way. For example, in some soci
eties there is typically a more balanced, or even male-dominated, pattern of labor
in the production of feasts (although this generally does not extend to the prepa
ration of daily meals). Moreover, women may share in the status and political
benefits from their labor by being members of an influential household or line
age (in matrilineal contexts) . Their labor (and male dependence upon it) may also
be overtly recognized and valued, and women may even derive considerable cat
egorical and individual status from their central role in the furnishing of hospi
tality or in maintaining commensal relations with the gods (e.g. , see Gero 1992;
March 1998). And, in many societies, women do host their own work feasts and
other feast events, although usually on a smaller scale than men. For example,
among the Tiv of Nigeria, women host smaller work feasts than men, but these
"underscore the prestige of important women. If a woman calls a big hoeing
party and supplies generous amounts of food and beer, she will be called 'impor
tant woman' (shagba kwase) for months afterwards" (Bohannan and Bohannan
1968:73). Finally, the common traditional female monopolization of cooking and
brewing responsibilities has, with the penetration of the monetized market econ
omy, frequently presented women with opportunities for gaining a source of in
come (e.g. , through beer sales), and this has sometimes enabled them to acquire
considerable economic independence and intrafamilial power under changing
socioeconomic conditions (e.g., see Colson and Scudder 1988; Netting 1964).
The relationship between feasts and gender is clearly a complex but analyti
cally rich and important one. Feasts are intimately implicated in the representa-
92
R I T U A L S O F C O N S U M PT I O N , CO M M E N S A L P O L I T I C S , A N D P O W E R I N A F R I C A N C O N T EXTS
tion, reproduction, and transformation of gender identity; as well as in the gen

dered structuring of relations of production and power in society. This means
both that feasting is an important and potentially productive avenue for under
standing gender relations and roles in archaeological contexts, and that gender
must be an essential consideration in any analysis of feasting.
R E L AT I N G T H E M O D E S O F C O M M E N S A L P O L I T I C S
Let us now return to the consideration of the different modes of commensal pol
itics outlined earlier, because it is also necessary to say a few words about the rela
tionships of these modes to each other. The first thing to emphasize is that they
should decidedly not be interpreted as evolutionary stages that can be correlated
with, for example, outmoded evolutionary typologies of political organization
(band, tribe, chiefdom, state, etc.). There is, to be sure, an obvious correlation to
some extent with increasing social stratification and complexity of structures of
political power (for example, diacritical feasts, as defined here, are generally a fea
ture encountered exclusively among state societies-but not all states will have di
acritical feasts). However, rather than describing a series of successive
evolutionary stages, these feasting modes should be viewed as constituting a pro
gressively expansive repertoire of forms of political action through feasting. One
form does not replace another; some forms simply expand the range of commen
sal politics in operation. It is true that there have been, and are, societies in which
only empowering feasts are operative: this is the most basic and fundamentally
ubiquitous mode of commensal politics. However, societies in which diacritical
feasts are found are also certain to have each of the two other forms as well. In
other words, where cuisine is used as a diacritical symbolic device separating
classes, the politics of commensality will still be used by individuals or groups
jockeying for relative status within those classes. Furthermore, kings, chiefs, and
others in patron positions will often simultaneously employ unequal commensal
hospitality in the patron-role pattern to legitimize institutionalized political au
thority roles. Likewise, both empowering and patron-role feasts are likely to be
operative where the latter type is found: the use of redistributive hospitality by in
stitutionalized patrons (e.g. , "chiefs") to maintain the authority vested in their
roles does not preclude the use of hospitality by others to define their relative sta
tuses below that of such patrons, or its use by chiefs of different areas to negotiate
and define their relative statuses vis-a-vis each other, or indeed its use to contest
chiefly authority. Whatever kings, chiefs, or elite classes are doing with their food,
common households will continue to hold feasts in their own way to establish
community and personal relationships, mobilize labor, and build symbolic capital.
Hence, the "festive landscape" in any given society will most likely be a palimpsest
of several different modes of commensal politics operating in different contexts.
93
Michael Dietler
A second point to bear in mind is that the distinctions between the three
modes of commensal politics are not precisely of the same order-and this
fact has important implications for the role of feats in social change. The dif
ference between empowering feasts and patron-role feasts is really one of estab
lishing a transitional division along a continuum of expectations. The symbolic
logic of both is quite similar: both operate by defining a single "consumption
community" within which asymmetries are expressed and naturalized to differ
ent degrees by the sharing of food. It is really the extent of institutionalized
acceptance, or expectation, of a continuing pattern of unreciprocated or unbal
anced hospitality that defines the difference. As the example in the following sec
tion will show, there is often a subtle distinction between the two, and it is not
difficult to imagine how the patron-role feast may crystallize out of certain forms
of empowering feasts. It is also important to recognize that tensions and conflict
may actually be created when groups approach such feasts with different under
standings of their political logic: for example, when the hosts view the feast in the
patron-role mode and the guests view it in the empowering mode. This is partic
ularly a risk with feasting across cultural boundaries, where, for example, hosts
and guests are members of different ethnic groups that do not share the same
cultural codes and behavioral expectations. But it can also be manipulated con
sciously by individuals or groups who are quite aware of the conventions but
who, for example, choose to challenge chiefly authority by refusing to acknowl
edge a patron-role feast as such and treating it instead in the competitive empow
ering mode. This form of "festive revolution" is, of course, one of the many
ways in which feasting can become a site of contestation and a dynamic agent in
political change.
In contrast to the other two modes, the diacritical feast manifests a symbolic
logic that differs in kind. It serves to reify asymmetries along lines of class or social
order by defining the boundaries of separate "consumption-communities." It also,
of course, serves to solidify identity within those consumption-communities
through food sharing and the cultivation of shared tastes. Again, it is important to
emphasize that all feasting rituals involve boundary-defining practices. Social cate
gories such as age and gender, for example, are very commonly marked in the ways
noted in previous sections; and it is important for archaeologists to be aware of the
operation of such diacritical devices. But these other distinctions are established
within commensal networks through variations in food-sharing practices. What
are here called "diacritical feasts" represent a special kind of boundary-defining
practice based upon commensal exclusion that I believe is sufficiently different and
heuristically valuable to merit distinguishing categorically. As prior studies of pre
historic European contexts have shown (see Dietler 1996, 1999a), it can be a pro
ductive category for archaeological analysis.
94
LUO F EASTS
In order t o further clarify some of the more abstract points made earlier, I will
briefly treat several aspects of feasting among the Luo of western Kenya in some
what more detail than the other examples raised in the discussion. The Luo are a
Nilotic-speaking people who inhabit a region of about 10,000 km2 surrounding
the Winam Gulf, in the northeast corner of Lake Victoria. They have a patrilineal
kinship system and live in homesteads scattered across the countryside, which
are occupied by polygynous extended families with a patrilocal postmarital resi
dence pattern (see Dietler and Herbich 1989, 1993; Evans Pritchard 1949; Herbich
1987; Herbich and Dietler 1992, 1993; Shipton 1989; Southall 1952) .
Agriculture provides the base of their diet, and this is carried out by women in
scattered sets of small plots in the vicinity of the homestead. Grain crops include
several varieties of sorghum, millet, and maize. Root crops, especially sweet po
tatoes and cassava, are also important, as are various kinds of beans, greens,
lentils, and wild leaves. In some areas, bananas are also grown. Protein sources in
clude milk fish (caught in the Gulf and traded widely throughout the region),
,
chickens, sheep, and goats. Beef is also highly prized, but cattle are an important
symbol of wealth and are usually slaughtered only for feasts. Aside from pur
chased fish and sporadic "target" buying and selling of grain at the local markets,
most households grow most of the food they eat. There is little reliance on food
stuffs imported from outside the region (aside from salt and a few luxuries, such
as tea, sugar, and tobacco).
With these basic ingredients, the Luo manage to maintain a relatively varied
repertoire of dishes, and there are regional and family preferences for recipes.
The main meals are constituted around a thick, bread-like porridge (called kuon)
made from boiled sorghum or maize flour. This is the symbolically central in
gredient of the diet, and a Luo who has not eaten kuon will say that (s)he has
not eaten. Various stew-like dishes made from vegetables, fish, or meat serve es
sentially as a condiment to kuon. Snacks and lesser meals consist of a maize and
bean mixture (nyoyo), a thin millet or maize porridge (nyuka), sweet potatoes
and sour milk and other such dishes. The main alcoholic beverage is beer
,
(kong'o) made from millet and / or maize, although a distilled alcohol known as
chang'aa has also become popular in recent decades. These alcoholic beverages
are not items consumed with daily meals; rather, they are essential components
of feasts.
As with many African societies, there is usually a seasonal period of hunger
just before the main harvest of the year, when grain supplies tend to run low and
must be stretched. Luo history is also marked by periodic episodes of maj or
famine caused by crop failures and cattle epidemics. These episodes are known
by name and they were important enough to constitute many of the main hinges
95
Michael Dietler
of collective memory (see Dietler and Herbich 1993), or what Shipton (1990 :375)
has aptly called "the hitching-posts of history."
Feasts are an important element of Luo life, and they play most of the various
roles attributed to the "empowering" mode in the earlier discussion. The largest
feasts, and indeed the largest gatherings in the society outside of markets, take
place at funerals. These events are held at the homestead of the deceased and are
marked by the provision of large quantities of beer and beef, along with the stan
dard kuon and other foods. They are accompanied by ritual dramaturgical prac
tices such as parading of cattle, dancing, singing, speeches, and the recitation of
praise songs that recount the accomplishments of both the deceased and the
speakers. They often last for several days, during which a large group of lineage
members, affines, and neighbors must be kept satisfied with copious amounts of
food and drink. The prestige of the deceased and his / her family are thought to
be reflected in the size of the gathering capable of being assembled and sustained
at the funeral feast and the lavishness of the hospitality provided. Influential men
have the most ostentatiously lavish funerals, but every Luo is concerned about
having an impressive funeral mounted for him/her. This concern is often voiced
by older widows as a major reason for joining religious groups, as these assure
their followers of a proper funeral. As noted earlier, the scale of hospitality at fu
nerals is often so great that at least temporary impoverishment of the family may
result, and the lavishness of such feasts among the Luo and other west Kenyan
peoples is the subject of frequent harangues by government ministers and mem
bers of other ethnic groups (e.g., Mburu 1978).
Less spectacular feasts are also held for marriages, harvest celebrations, collec
tive labor mobilization, the founding of a new homestead, and a host of other
things (such as ceremonies concerning the birth of twins). Small-scale gatherings
of elders or meetings between friends are also often marked by sharing a pot of
beer. In general, feasts are distinguished from daily meals by several features.
Most commonly, these include the consumption of beer (and/ or chang'aa) and
beef, which are not everyday foods. They are also sometimes marked by the loca
tion of consumption and the use of special containers.
In the territory of Alego, for example, homesteads have a special shaded area
known as siwanda that serves as the place where senior men at feasts gather to
gether to drink beer and eat.6 At feasts of some importance, these elders will con
sume unfiltered beer out of a special large pot called a thago (Fig. 3.1). A thago may
be larger than a meter in diameter and a meter tall, and it is supported by being
partly buried in the ground at the siwanda. The men sit around the pot in a circle
drinking from long straws (oseke) made of hollow vine stems with a woven filter
on the end (Fig. 3.2). The possession of a personal straw, which one carries to
96
Figure 3.I. Photograph of Luo co mmunal beer drinking pot (thago), on left, and beer fer
mentation pot (dakong'o), on right. Scale in cm. (Photo by M. Dietler and I. Herbich)
beer drinks in a special bamboo case, is a clear sign of senior male status. Another
large pot, called a dakong'o (Fig. 3.1), in which the beer has been brewed, always
stands near the thago, and beer is removed from it and mixed with hot water for
consumption from the thago. Younger men will generally drink and eat in the
clear area in front of the house (known as laru) . They will also usually drink fil
tered beer that is served in a pot called mbiru, which is much smaller than a thago.
They will consume their beer by dipping large cups made from half of a hol
lowed gourd (agwata; sometimes now tin cans or enamel mugs) into the mbiru.
Women may also consume some beer in this fashion (Herbich 1991) . Every
household will have at least one mbiru, but the same is not true of thago. These
latter are large, expensive pots that may be owned by only a few of the wealthier
homesteads in neighborhood. Other homesteads will have to borrow a thago
when they wish to organize an important feast. These are highly prized pots that
usually are known within a neighborhood by a specific name, and many exam
ples we found in homes were over 50 years old, some much older. Figure 3.3 offers
a splendid iconic representation of such a feast, and the marking of social cate
gories by these practices. It is a mural from an abandoned beer hall in a Luo mar
ket center (such beer halls were outlawed by the Kenyan government during the
1970s as part of a longstanding, and unsuccessful, struggle by the colonial and
97
Figure 3.2. Photograph showing how beer is consumed from thago through a long vine
stem straw (oseke) . (Photo by M. Dietler and I. Herbich)
postcolonial Kenyan governments to exert state control over alcohol: see Ambler
1991) .
The consumption of food is also done on a communal basis. A "loaf" of kuon
will be served on a basket plate and shared by several diners who will break off
morsels and dip them into a common ceramic bowl (tawo) of the stew/ sauce
(sometimes these serving containers are now replaced by imported enamel
dishes) . These serving containers are generally not different from those used in
everyday meals.
Figure 3.4 is a photograph that rather ironically encapsulates the gendered re
lations of production that underlie Luo feasting. It portrays a woman working at
98
R I T U A L S O F C O N S U M PT I O N , C O M M E N S A L P O L I T I C S , A N D P O W E R I N A F R I C A N CO N T E XTS
Figure 3.3. Photograph of a mural on the exterior of an abandoned beer hall showing an
iconic representation of a Luo feast. Note the elder men drinking from a large commu
nal pot (thago) through long straws (oseke) while younger men and women drink from
cups, and a woman acts in a serving capacity. (Photo by M. Dietler and L Herbich)
drying sinoho (a processed grain flour product resulting from an initial stage in
the chafne operatoire of the beer fermentation process: see Herbich 1991) on the
floor of another abandoned beer hall with a feasting mural in the background.
This is one of the many laborious steps necessary to produce, store, and process
a sufficient quantity of grain to mount a feast. Women are the agricultural and
culinary labor force that lies behind the production of all Luo feasts, although
they share in the ensuing prestige and other benefits only indirectly, as wives of
the generous host. Women grow the crops, process them, and do the cooking,
brewing, and serving. This is one of the reasons that, in this polygynous society,
having many wives is not only a sign of wealth, but is essential for being able to
mount large feasts. Acquiring wives requires wealth and is a gradual process be
cause one must give a large amount of bridewealth to the woman's family in the
form of cattle and, now, often money (formerly, iron hoes were also given) . How
ever, multiple wives considerably expand the possibilities for a homestead to offer
lavish hospitality, which, as further discussion will show, has important political
implications (see also Dietler and Herbich, Chapter 9).
The Luo do not have anything resembling the agonistic competitive feasting
of New Guinea big-men or the escalating Mambila (Nigeria) bear feasts (Reh-
99
Michael Dietler
Figure 3.4. Photograph of a Luo woman engaged in drying sinoho (one step in the labori
ous chaine operatoire of beer production) on the floor of an abandoned beer hall with a
mural showing an idealized representation of a Luo feast in the background. (Photo by
M. Dietler and I. Herbich)
fisch 1987) . Yet generosity in commensal hospitality is an essential practice in

maintaining a man's prestige and influence, and the funeral feast is a final dia
matic affirmation of status. Luo feasts also provide a prime context for demon
strations of oratory, which is a highly valued skill that also brings prestige. This
oratory includes forms of ritualized boasting in which men extol their own
achievements and denigrate those of their rivals. Specialized praise singers, who
accompany their songs on a form of lyre (nyatiti), may also be employed for this
purpose.
Obviously; the Luo do not have diacritical feasts in the sense defined earlier (at
least those in the countryside-the situation of the Luo who have moved to the
capital city of Nairobi is somewhat different). The question of patron-role feasts
1 00
R I T U A L S O F C O N S U M PT I O N , CO M M E N S A L P O L I T I C S , A N D P O W E R I N A F R I C A N C O N T E XTS
is a little more complicated, and the issue is worth discussing in some detail
because the Luo case highlights the fluid boundary between empowering and
patron-role feasts noted earlier. Although they now live with a system of "chiefs"
constructed by the British colonial government and continued by the postcolo
nial Kenyan state, the Luo have traditionally had a strongly egalitarian political
ethos and lacked centralized authority. They do, however, have an indigenous
term, ruoth, that is used to refer to modern chiefs. In the precolonial era this term
more likely meant something closer to "leader" or "man of influence" than to
the institutionalized political role it has come to signify. However, oral histories
indicate that the degree to which individuals in the past were able to transform
their informal influence into naturalized positions of authority and power varied
somewhat from region to region.
Whisson (1961) offers an interesting case study of this process in the territory
of Asembo that both illustrates the means available of concentrating power in
the precolonial era (including feasting) and the ramilications this had during the
imposition of the colonial administration and its structure of institutionalized
chiefs. 7 Traditional Luo political organization has been described as a classic case
of the segmentary lineage system (Evans-Pritchard 1949; Southall 1952). The
modern administrative boundaries within Luo territory, which were defined dur
ing the colonial era, effectively froze into static form what had previously been a
series of highly dynamic factional and territorial struggles between competing
subgroups organized according to lineage affiliation and military expediency.
Based on oral histories, Whisson describes competition for leadership in the con
text of such factional struggles during the immediately precolonial and early
colonial eras in Asembo, the territory of the Luo subgroup known as JoAsembo
-
along the north coast of the Winam Gulf
One of the main functions of precolonial leaders was the arbitration of dis
putes within the smallest local territorial unit, the gweng'. Becoming an influen
tial leader required the building of prestige and moral authority, and these
qualities were acquired from several possible sources. The most immediate crite
ria were genealogical position and the strength of the lineage: the most ge
nealogically senior member of the dominant lineage of the gweng' had
responsibilities to settle disputes within the gweng', and he met with other similar
leaders to attempt to resolve disputes between gwenge. 8 Disputes that could not
be settled peacefully were resolved by fission and migration, or by armed con
flict. The segmentary lineage ideology structuring patterns of alliance and oppo
sition in conflict created opportunities for leadership by members of strong
lineage segments at all the points of segmentation. However, this was also aug
mented by the creation of pragmatic alliances in which strong lineages would se-
101
Michael Dietler
cure the support of weaker "jodak" (tenant) lineage groups that had settled in
their territory after being forced out or fissioning elsewhere.
Hence, as Whisson (1961:7) pointed out, the main sources of power that an in
dividual could manipulate came from: (1) being a senior member of a powerful
lineage, (2) personal ability in warfare, and (3) the capacity to marshal a signifi
cant amount of support in the face of conflict. Skill in the use of magical power
(bilo) was particularly important in winning prestige in the sphere of warfare. ]o
bilo (magicians) were feared and respected for their powers of divination and
their ability to use killing magic on enemies. The ability to rally support de
pended upon the accumulation of wealth and prestige, and it is in this domain
that feasting played an important role. Wealth in this context would be reckoned
in terms of cattle and wives, both of which were essential for the production of
feasts. Acquiring large numbers of cattle was greatly aided by skill in raiding
(which was itself a source of prestige). These cattle were used for prized meat at
feasts, but also for the payment of bridewealth that was necessary to acquire
wives. A large number of wives greatly increased the capacity of the homestead
for agricultural and culinary labor, so that wealthy men were able create and use
food surpluses to host feasts for the lineage leaders who assembled to discuss po
litical and judicial matters. As Whisson noted, this wealth (in cattle, wives, and
crops) was used to entertain "the leaders of the clans and subclans forming the
nucleus of a council or court and meeting in the home of the richest or most re
spected man. This man became ruoth, the leader" (1961:7).
The strongest leaders would be able to draw upon all three of the mutually
enhancing sources of power noted above. But a skilledjabilo from a weaker line
age who had accumulated the cattle and wives to host lavish feasts could even
overcome a genealogical handicap by rallying the support of other lineages and
creating political alliances. The British colonial government attempted to
squeeze this fairly loose and fluid set of political relations into their preconceived
model of "chiefdoms" operating as a hierarchical administrative system. They
imposed a model of institutionalized central authority with formalized political
roles and rules of succession upon a much more dynamic and competitive set of
political practices sustained by cultural perceptions of authority that were far
more contingent.
· The process by which the British "identified" Luo chiefs and the manipulations
that wertt on among competing Luo men of influence seeking to be named
chiefs is a complex tale. What is important to retain for the purposes of this dis
cussion of feasts is that the colonial situation under which these new chiefs oper
ated created contradictions that sometimes undermined their authority. These
new chiefs were agents of the state, but their ability to perform the functions that
the state demanded of them depended upon maintaining the traditional forms of
1 02
R I T U A L S O F CO N S U M PT I O N , CO M M E N S A L P O L I T I C S , A N D P O W E R I N A F R I C A N C O N T EXTS
symbolic capital. However, the suppression of warfare and raiding eliminated

both a major former arena for the acquisition of prestige and an important
source of the cattle that produced the wives and feasts necessary to operate suc
cessfully in the other major arena of political action. Government pay was not
sufficient to compensate for this loss and the state took a dim view of augment
ing income through bribery. As in the case of an early Asembo chief named
Odindo, those who were unable to keep up the lavish hospitality that people ex
pected of a traditional ruoth sometimes fell from power in the face of continual
scheming by rivals from other lineages (Whisson 1961:11). Others were able to
survive by better adapting to their role as agents of the state by having their sons
sent to mission schools and gaining the skills of literacy that the government par
ticularly prized. It is noticeable, however, that successful chiefs today are still con
spicuously more polygynous than the rest of their people. For example, one chief
in our research area had 45 wives when we arrived and over 50 when we left three
years later.
CO N C LU S I O N
The Luo examples should serve to give a better sense of the experience of com
mensal politics that lies behind the more abstract theoretical discussion offered
earlier. In particular, one can begin to understand the way in which the hospital
ity used in empowering feasts to acquire and maintain symbolic capital can be
come transformed into the institutionalized expectation of the patron-role feast;
and one can see how a failure to meet those expectations can seriously weaken
credibility and undercut authority. One can also get a better sense of the way that
feasting either combines or competes with other sources of symbolic capital
(prowess in warfare, oratorical skill, powers in magic, genealogical pedigree, gift
giving, etc.) to establish prestige and influence. Feasting is by no means the only
arena of political action, but it is very frequently an extremely important, if not
crucial, one. The Luo case also illustrates the subtle ways in which social cate
gories and boundaries are symbolically marked by the ritual practices of feasts,
and why those operative among the empowering feasts of the Luo are quite dif
ferent in their symbolic logic from those described for the diacritical feast. Finally,
this case also highlights the often unremarked gendered relations of production
that support commensal politics. The division of labor and symbolic capital
along gender lines is certainly not always identical to, nor so starkly realized as,
that among the Luo; but this is always an important consideration in the theo
rization of feasting.
For archaeologists, the implications of the discussion presented in this chapter
are several and important. In the first place, it is clear that feasts commonly serve
a variety of crucial structural roles in articulating the political economy of a wide
1 03
Michael Dietler
array of societies. It is also clear that feasts are a prime arena and instrument of
political action by individuals and social groups pursuing economic and political
goals and competing for influence within their social worlds. However, the ways
in which feasts serve the acquisition and transformation of symbolic and eco
nomic capital are extremely complex, and archaeologists need a well-developed
theoretical understanding of the nature of feasting ritual if we are to understand
political life in ancient societies in something more that mechanistic typological
terms. In my view, it is critical that we begin to tackle issues of situated agency
and the role of practice in transforming structure if we wish to say anything of
insightful significance about the historical development of different forms of so
cial inequality. In this paper I have tried to present several theoretical constructs
based upon comparative analysis of ethnographic data that I believe hold some
promise in analyzing feasting ritual, and I have elsewhere tried to demonstrate
how they may be applied to archaeological cases in ways that yield fruitful new
insights (e.g., Dietler 1990, 1996, 1999a, 1999b ). However, this is by no means a de
finitive formulation, and I look forward to the emerging dialogue on these issues
that is promised by the convergence of perspectives in this volume.
AC K N OW L E DG M E NTS
My thinking on this issue has been evolving for over a decade, and it has greatly benefited
from the opportunity to receive challenging comments on different papers exploring this
theme from audiences in a variety of contexts, most particularly the "Food and the Status
Quest" symposium at the Ringberg Castle of the Max Planck Institute in Germany (1991),
the "Les Princes de la Protohistoire et !'Emergence de l'Etat'' symposium at the Centre
Jean Berard in Naples, Italy (1994), a seminar while a visiting professor at the Ecole des
Hautes Etudes en Sciences Sociales (1996), the SAA symposium that generated this book
(1998), and a symposium on "Consuming Power: Feasting as Commensal Politics" con
vened at Cornell University (1999) in which my work served as a keynote for lively critical
discussion. I am grateful to all the participants of these various sessions for their com
ments (including those with whom I disagree), and most particularly to Michel Bats, Mau
rice Godelier, Brian Hayden, Pierre Lemonnier, Jean-Paul Morel, Michel Py; Nerissa
Russel, Andre Tchernia, Terry Turner, and Polly Wiessner. Special thanks are due to In
grid Herbich for generous sharing of data and invaluable intellectual collaboration.
N OT E S
r . To avoid possible confusion, let me emphasize that I use the word commensal in its
original sense, rather than its peculiar biological adaptation. The word derives from
the Latin com mensalis, indicating the sharing of a table-hence, eating together.
Needless to say; many people around the world manage to eat together quite well
without using a table. Moreover, in a number of cases the sharing of food is
1 04
accomplished without the host and guests actually eating in the same space-in some
contexts it is actually considered impolite for the host to be present when his /her
guests consume their food (e.g. , see Richards l939:13y-136) . However, despite the minor
drawback of being grounded in a Eurocentric cultural trope, the term commensal does
provide a convenient way of indicating a range of forms of communal food consump
tion. Other possible alternative terms in common usage, such as adjectival versions of
companion (indicating the sharing of bread; from Latin) and symposium (the sharing of
drink; from Greek), have even more problematic semantic histories and associational
problems. And the game of inventing neologisms, such as co-alimentary, co-gustatory,
or the innumerable other possibilities, seems a needlessly pedantic exercise.
2. The comparative ethnographic focus of this chapter is limited to agrarian societies in
Africa, as this presents a more than sufficiently complex array of issues. Those
interested in the issue of feasting among "foragers" and "complex hunter-gatherers,"
including African examples, are directed to the works of Wiessner (1996) and Hayden
(1990, 1996).
3. All unreferenced descriptions of practices among the Luo people in this paper are
derived from research conducted by Ingrid Herbich and me in western Kenya from
1980 to 1983 (see, e.g., Dietler and Herbich 1993; Herbich 1987, 1991; Herbich and
Dietler 1991, 1993). Thanks are due to the National Science Foundation, the Wenner
Gren Foundation, the Boise Fund of Oxford University, the Office of the President of
Kenya, the National Museums of Kenya, and especially our Luo and Samia hosts and
our research assistants, Rhoda Onyango, Monica Oyier, and the late Elijah Ogutu.
4. I owe the "metaproduction" formulation to an insightful comment by Terry Turner.
5. Several French colleagues, in particular, have noted that the word "competition"
evokes a strongly agonistic struggle for dominance with markedly negative
connotations. Unfortunately; English lacks a convenient means to mark the subtle
distinction between competition and the more positively viewed concurrence. Hence,
my use of the English term "competition" should be understood to cover the entire
range of such possible relationships. My thanks to Pierre Lemonnier, Michel Py; and
Andre Tchernia, in particular, for challenging me to clarify this usage.
6. For the sake of simplicity, I use terms in this paper that are, in fact, specific to the
territories of several Luo subgroups in Siaya district (such as the joAlego). These
terms vary in o ther areas. Similarly; the siwanda is not a formally defined space in the
homesteads of all Luo groups.
7. To avoid cluttering the text with multiple citations of the same work, I will simply
point out here that the historical information in the following discussion is largely a
selective summary of parts of Michael Whisson's (1961) excellent paper "The Rise of
Asembo and the Curse of Kakia."
8. I use the words "he" and "man" here purposely to indicate the gender-specific nature
of these leadership roles. For one thing, Luo women are not members of the lineage
into which they marry. Hence they do not have the genealogical standing to acquire
authority in matters relating to the lineage of the area where they live after marriage.
1 05
Michael Dietler
REFERENCES
Ambler, C . H.
1991 Drunks, Brewers, and Chiefs: Alcohol Regulation in Colonial Kenya
1900-1939. In Drinking Behavior and Belief in Modern History, edited by S. Bar
rows and R. Room, pp. 16_5---183. Berkeley: University of California Press.
Amselle, J.-L.
1998
[1990] Mestizo Logics: Anthropology of Identity in Afiica and Elsewhere. Translated by
C. Royal. Stanford: Stanford University Press.
Anigbo, 0. A. C .
1996 Commensality as Cultural Performance: The Struggle for Leadership in an
Igbo Village. In The Politics of Cultural Peiformance, edited by D. Parkin, L. Ca
plan and H. Fisher, pp. 101-114. Oxford: Berghahn Books.
Appadurai, A.
1981 Gastropolitics in Hindu South Asia. American Ethnologist 8:494-5rr.
1986 Introduction: Commodities and the Politics of Value. In The Social Life of
Things: Commodities in Cultural Perspective, edited by A. Appadurai, pp. 3--63.
Apter, A.
1992 Black Critics and Kings: The Hermeneutics of Power in Yoruba Society. Chicago:
University of Chicago Press.
Bacon, M. K.
1976 Cross-Cultural Studies of Drinking: Integrated Drinking and Sex Differences
in the Uses of Alcoholic Beverages. In Cross-cultural Approaches to the Study of
Alcohol: An Interdisciplinary Perspective, edited by M. Everett, ]. Waddell and
D. Heath, pp. 23-33. The Hague: Mouton.
Barth, R
1967 On the Study of Social Change. American Anthropologist 69:661--670.
Bayart, J. R
1993 The State in Africa: Politics of the Belly. London: Longman.
Beattie, ].
1960 Bunyoro: An Afiican Kingdom. New York: Holt, Rinehart and Winston.
Bloch, M.
1987 The Ritual of the Royal Bath in Madagascar: The Dissolution of Death,
Birth, and Fertility into Authority. In Rituals of Royalty: Power and Ceremonial
in Traditional Societies, edited by D. Cannadine and S. Price, pp. 271-297. Cam
bridge: Cambridge University Press.
1989 Ritual, History, and Power: Selected Papers in Anthropology. London: Athlone.
Boserup, E.
1970 Women's Role in Economic Develupment. London: Allen and Unwin.
1 06
R I T U A L S O F CO N S U M PT I O N , C O M M E N S A L P O L I T I C S , A N D P O W E R I N A F R I C A N C O N T EXTS
Boston, ]. S.
1968 The Igala Kingdom. Ibadan: Oxford University Press.
Bourdieu, P.
1984 Distinction: A Social Critique of thejudgment of Taste. Cambridge: Harvard Uni
versity Press.
Bradbury; R. E.
1973 Benin Studies. London: Oxford University Press.
Brown, S. D.
1975 Ritual Aspects of the Maprusi Kingship. Cambridge: Afii.can Studies Center.
Cannadine, D.
1987 Introduction: Divine Rites of Kings. In Rituals of Royalty: Power and Ceremo
nial in Traditional Societies, edited by D. Cannadine and S. Price, pp. l-19.
Cannadine, D., and S. Price, eds.
1987 Rituals of Royalty: Power and Ceremonial in Traditional Societies. Cambridge:
Cambridge University Press.
Carlson, R. G.
1990 Banana Beer, Reciprocity; and Ancestor Propitiation among the Haya of
Bukoba, Tanzania. Ethnology 29:297-3n.
Child, I. L., H. Barry, and M. K. Bacon
1965 A Cross-Cultural Study of Drinking. 3. Sex Differences. Quarterly journal of
Studies in Alcohol (Supplement) 3:4!}-6r.
Clark, C. M.
1980 Land and Food, Women and Power, in Nineteenth Century Kikuyu. Africa
50: 357-370.
Cohen, A.
1974 Two Dimensional Man: An Essay on the Anthropology of Power and Symbolism in
Complex Societies. Berkeley: University of California Press.
1979 Political Symbolism. Annual Review of Anthropology 8: 87-n3.
Gwembe District, Zambia, 1950--1982. Stanford: Stanford University Press.
Comaroff, J. , and). Comaroff
1991 Of Revelation and Revolution: Christianity, Colonialism, and Consciousness in
South Africa. Vol. r. Chicago: University of Chicago Press.
1993 Introduction. In Modernity and Its Malcontents: Ritual and Power in Postcolonial
Africa, edited by J. Comaroff and]. Comaroff, pp. xi-:xxxvii . Chicago: Univer
sity of Chicago Press.
Counihan, C. M., and S. L. Kaplan, eds.
1998 Food and Gender: Identity and Power. Amsterdam: Harwood.
1 07
Michael Diet!er
Crumley, C .
1987 A Dialectical Critique of Hierarchy. In Power Relations and State Formation, ed
ited by T. C . Patterson and C. W Gailey, pp. 155-159. Washington, D.C . :
American Anthropological Association.
Dentzer, J.-M .
1982 Le motif du banquet couche dans le Proche-Orient et dans le monde grec du VIIe au
!Ve siecle avant].-C. Bibliotheque des Beales Frarn;:aises d'Athenes et de Rome,
246. Paris: Boccard.
Dietler, M.
1989 The Work-Party Feast as a Mechanism of Labor Mobilization and Exploita
tion: The Case of Samia Iron Production. Paper presented at 88th Annual
Meeting of the American Anthropological Association, Washington, D.C . ,
November.
1990 Driven by Drink: The Role of Drinking in the Political Economy and the Case
of Early Iron Age France. journal of Anthropological Archaeology 9:352-406.
Status in Prehistoric Europe. In Food and the Status Quest, edited by P. Wiess
1998 Consumption, Agency, and Cultural Entanglement: Theoretical Implications
of a Mediterranean Colonial Encounter. In Studies in Culture Contact: Interac
tion, Culture Change, and Archaeology, edited by]. Cusick, pp. 288-315. Carbon
dale: University of Southern Illinois Press.
l999a Rituals of Commensality and the Politics of State Formation in the
"Princely" Societies of Early Iron Age Europe. In Les princes de la protohistoire
et l' emergence de l' t!tat, edited by P. Ruby, pp. 135-152. Cahiers du Centre Jean
Berard, Institut Frani;:ais de Naples 17-Collection de l'Ecole Frani;:aise de
Rome 252, Naples.
l999b Reflections on Lattois Society during the 4th Century B.C. In Lattara 12:
Recherches sur le quatrieme siecle avant notre ere a Lattes, edited by M. Py, pp.
663-680. Lattes: Association pour la Recherche Archeologique en Languedoc
Oriental.
Dietler, M., and I. Herbich
1993 Living on Luo Time: Reckoning Sequence, Duration, History, and Biogra
phy in a Rural African Society. World Archaeology 25:248-260.
Dillon, R. G.
1990 Ranking and Resistance: A Precolonial Cameroonian Polity in Regional Perspective.
Stanford: Stanford University Press.
Donham, D.
1994 Work and Power in Maale, Ethiopia. New York: Columbia University Press.
[1979]
Douglas, M.
1984 Standard Social Uses of Food: Introduction. In Food in the Social Order, edited
by M. Douglas, pp. l-39. New York: Russell Sage Foundation.
1 08
Douglas, M . , and C . Isherwood

1979 The World of Goods: Towards an Anthropology of Consumption. New York: Norton.
Elias, N.
1978 The History of Manners. New York: Pantheon.
[1939]
Erasmus, C . J.
1956 Culture Structure and Culture Process: The Occurrence and Disappear
ance of Reciprocal Farm Labor. Southwestern journal of Anthropology
12:444-469.
Evans-Pritchard, E. E.
1949 Luo Tribes and Clans. Rhodes-Livingstonejournal T24-40.
1971 The Azande. London: Oxford University Press.
Firth, R.
1983 Magnitudes and Values in Kula Exchange. In The Kula: New Perspectives on
Massim Exchange, edited by J. W Leach and E. Leach, pp. 89-I02. Cambridge:
Fortes, M . , and E. E. Evans-Pritchard, eds.
1940 Afiican Political Systems. London: Oxford University Press.
Foucault, M.
1980 Power!Knowledge: Selected Interviews and Other Writings, edited by C. Gordon.
New York: Pantheon Books.
Friedl, E.
1975 Women and Men: An Anthropologist's View. New York: Holt, Rinehart and
Winston.
Friedman, J.
Garine, I. de
1996 Food and the Status Quest in Five African Cultures. In Food and the Status
Quest: An Interdisciplinary Perspective, edited by P. Wiessner and W Schiefen
hovel, pp. 193-218. Oxford: Berghahn Books.
Gefou-Madianou, D., ed.
1992 Alcohol, Gender, and Culture. London: Routledge.
Gero, ]. M.
wegian Archaeological Review 25:15-30.
Geschire, P.
1982 Village Communities and the State: Changing Relations among the Maka of South
eastern Cameroon since the Colonial Conquest. Translated by J. Ravell. London:
Kegan Paul.
Godelier, M.
1980 L'Etat: les processus de sa formation, la diversite de ses formes et de ses
bases. Revue International des Sciences Sociales 32:657-671.
1 09
Michael Dietler
Goldschmidt, W
1976 Culture and Behavior of the Sebei: A Study in Continuity and Adaptation. Berke
ley: University of California Press.
Goody; ].
1982 Cooking, Cuisine, and Class: A Study in Comparative Sociology. Cambridge: Cam
bridge University Press.
Gutmann, B.
1926 Das Recht der Dschagga. Munich: Beck.
Haggblade, S.
1992 The Shebeen Queen and the Evolution of Botswana's Sorghum Beer Indus
try. In Liquor and Labor in Southern Africa, edited by ]. Crush and C. Ambler,
pp. 395-412. Athens, Ohio: Ohio University Press.
Halperin, R., and]. Olmstead
1976 To Catch a Feastgiver: Redistribution among the Dorze of Ethiopia. Africa
46:146-165.
Hayden, B.
tion. journal of Anthropological Archaeology 9:31-69.
1996 Feasting in Prehistoric and Traditional Societies. In Food and the Status Quest:
An Interdisciplinary Perspective, edited by P. Wiessner and W Schiefenhi:ivel,
pp. 127-148 . Oxford: Berghahn Books.
Hayden, B., and R. Gargett
1990 Big Man, Big Heart? A Mesoamerican View of the Emergence of Complex
Society. Ancient Mesoamerica 1:3-20.
Herbich, I.
1987 Learning Patterns, Potter Interaction, and Ceramic Style among the Luo of
Kenya. The African Archaeological Review 5:193-204.
1991 The Flow of Drink in an African Society: An Ethnoarchaeological Perspec
tive. Paper presented at the 56th Annual Meeting of the Society for American
Archaeology; New Orleans, April.
Herbich, I., and M. Dietler
1991 Aspects of the Ceramic System of the Luo of Kenya. In Tiipferei- und
Keramikforschung, 2, edited by H. Ludtke and R. Vossen, pp. ro5-135. Bonn:
Habelt.
1993 Space, Time, and Symbolic Structure in the Luo Homestead: An Ethno
archaeological Study of "Settlement Biography" in Africa. In Actes du XIIe
Congres International des Sciences Prehistoriques et Protohistoriques, Bratislava,
Czechoslovakia, September 1-7, 1991, Vol. 1, edited by ]. Pavl!k, pp. 26-32. Nitra:
Archaeological Institute of the Slovak Academy of Sciences.
Hunter, M.
l96r Reaction to Conquest: Effects of Contact with Europeans on the Pando of Southern
Africa. London: Oxford University Press.
110
Karp, I.
1980 Beer Drinking and Social Experience in an African Society. In Explorations in
African Systems of Thought, edited by I. Karp and C. Bird, pp.83-n9. Bloom
ington: Indiana University Press.
Kelly, J. D. , and M. Kaplan
1990 History, Structure, and Ritual. Annual Review of Anthropology l9:ng-150.
Kertzer, D. I.
1988 Ritual, Politics, and Power. New Haven: Yale University Press.
Laburthe-Tolba, P.
1981 Les seigneurs de la foret. Paris: Publications de la Sorbonne.
Lemonnier, P.
1990 Guerres et festins: Paix, .!changes, et competition dans les Highlands de Nouvelle
Guinee. Paris: CID-Editions de la Maison des Sciences de l'Homme.
1996 Food, Competition, and the Status of Food in New Guinea. In Food and
the Status Quest: An Interdisciplinary Perspective, edited by P. Wiessner and
W Scheifenhovel, pp. 21g-234. Oxford: Berghahn Books.
Lentz, C .
1998 The Chief, the Mine Captain, and the Politician: Legitimating Power in
Northern Ghana. Africa 68:46-66.
Mair, L.
1934 An African People in the Twentieth Century. New York: Russell and Russell.
1977 African Kingdoms. London: Clarendon.
March, K S.
Power, edited by C. M. Counihan and S. L. Kaplan, pp. 45-80. Amsterdam:
Marshall, F.
1993 Food Sharing and the Faunal Record. In From Bones to Behavior: Ethnoarchae
ological and Experimental Contributions to the Interpretation of Faunal Remains,
edited by J. Hudson, pp. 228-246. Carbondale: Southern Illinois University.
Mauss, M.
1935 Les techniques du corps. journal de Psychologie 32:271-293.
1966 The Gift: Forms and Functions of Exchange in Archaic Societies. Translated by
[1925] I. Cunnison. London: Routledge and Kegan Paul.
Mburu, J. G.
1978 West Kenya Funerals: Expensive and Immoral. Umma 8:g-14, 18-19, 22-23.
McDonald, M . , ed.
1994 Gender, Drink, and Drugs. Oxford: Berg.
Meillassoux, C .
1975 Maidens, Meal, and Money: Capitalism and the Domestic Community. Cambridge:
111
Michael Dietler
Middleton, ]. , and D. Tait, eds.

1958 Tribes without Rulers. London: Routledge and Kegan Paul.
Modjeska, N.
1982 Production and Inequality: Perspectives from Central New Guinea. In In
equality in New Guinea Societies, edited by A. Strathern, pp. 50-rn8. Cam
Moore, M. P.
1975 Cooperative Labour in Peasant Agriculture. The journal of Peasant Studies
2:270-29!.
Moore, S. F.
1994 Anthropology and Africa: Changing Perspectives on a Changing Scene. Char
lottesville: University of Virginia Press.
Moore, S. F., and B. G. Myerhoff
1985 Introduction-Secular Ritual: Forms and Meanings. In Secular Ritual, edited
by S. F. Moore and B. G. Myerhoff, pp. 3-24. Assen/ Amsterdam: Van
Gorcum.
Morris, I.
1992 Death-Ritual and Social Structure in Classical Antiquity. Cambridge: Cambridge
University Press.
Murray, 0.
1990 Sympotic History. In Sympotica: A Symposium on the Symposion, edited by
0. Murray, pp. 3-13. Oxford: Clarendon.
Netting, R.
ogist 66:375-384.
Ngokwey, N.
1987 Varieties of Palm Wine among the Lele of the Kasai. In Constructive Drinking:
Perspectives on Drink from Anthropology, edited by M. Douglas, pp. n3-12i.
Obayemi, A. M. U.
1976 Alcohol Usage in an African Society. In Cross-cultural Approaches to the Study
of Alcohol: An Interdisciplinary Perspective, edited by E. Everett, ]. Wadell and
D. Heath, pp. 199-208 . The Hague: Mouton.
Pallier, G.
1972 Les dolotieres de Ouagadougou (Haute-Volta). Travaux et documents de geo
graphie tropicale 7:120-139.
Peristiany, ]. G.
1939 The Social Institutions of the Kipsigis. London: Routledge.
Perles, c .
1996 Les strategies alimentaires dans !es temps prehistoriques. In Histoire de
l'alimentation, edited by J.-L. Flandrin and M. Montanari, pp. 29-46. Paris:
Fayard.
112
R I T U A L S O F C O N S U M PT I O N , CO M M E N S A L P O L I T I C S , A N D P O W E R I N A F R I C A N C O N T EXTS
Polanyi, K.
1957 The Economy as Instituted Process. In Trade and Markets in the Early Empires,
edited by K. Polanyi, C . Arensberg, and H. Pearson, pp. 243-269. New York:
The Free Press.
Pryor, F. L.
1977 The Origins of the Economy: A Comparative Study of Distribution in Primitive and
Peasant Economies. New York: Academic Press.
Rehfisch, F.
1987 Competitive Beer Drinking among the Mambila. In Constructive Drinking:
Perspectives on Drink from Anthropology, edited by M. Douglas, pp. 135-45.
Richards, A. I.
1939 Land, Labour, and Diet in Northern Rhodesia. London: Oxford University
Press.
Sahlins, M.
1972 Stone Age Economics. London: Tavistock.
Sangree, W H.
1962 The Social Functions of Beer Drinking in Bantu Tiriki. In Society, Culture, and
Drinking Patterns, edited by D. ]. Pittman and C. R. Snyder, pp. 6-2r. New
York: Wiley.
Saul, M.
1981 Beer, Sorghum, and Women: Production for the Market in Rural Upper
Volta. Africa 51:746-764.
1983 Work Parties, Wages, and Accumulation in a Voltaic Village. American Ethnol
ogist IO :77-96.
Schapera, I.
1938 A Handbook of Tswana Law and Custom. London: Oxford University Press.
Schatzberg, M.
1993 Power: Legitimacy and "Democratisation" in Africa. Africa 63:445-46r.
Scott, ]. C .
1990 Domination and the Arts of Resistance. New Haven: Yale University Press.
Shipton, P.
1989 Bitter Money: Cultural Economy and Some African Meanings of Forbidden
Commodities. American Ethnological Society Monograph Series, l, Washing
ton, D.C .
1990 African Famines and Food Security: Anthropological Perspectives. Annual Re
view of Anthropology 19:353-394.
Southall, A. W
1952 Lineage Formation among the Luo. International African Institute Memoran
dum 26. London: Oxford University Press.
1956 Alur Society: A Study in Processes and Types of Domination. Cambridge: Heffer
and Sons.
113
Michael Dietler
Tambiah, S. ].
1985 Culture, Thought, and Social Action: An Anthropological Perspective. Cambridge:
Harvard University Press.
Turner, V.
1967 The Forest of Symbols: Aspects of Ndembu Ritual. Ithaca: Cornell University
Press.
1969 The Ritual Process: Structure and Anti-Structure. Ithaca: Cornell University
Press.
van Gennep, A.
1960 The Rites of Passage. Translated by M. B. Vizedom and G. L. Caffee. Chicago:
University of Chicago Press.
Vmcent, ].
1971 African Elite: The Big Men of a Small Town. New York: Columbia University
Press.
Whisson, M.
1961 The Rise of Asembo and the Curse of Kakia. In East African Institute of Social
Research Conference Proceedings, no pagination. Kampala: Mak:arere College.
Wiessner, P.
1996 Leveling the Hunter: Constraints on the Status Quest in Foraging Societies.
In Food and the Status Quest: An Interdisciplinary Perspective, edited by P. Wiess
1 14
4
OF FEASTING AND VA L U E
E N G A F E A S T S I N A H I S TO R I C AL P E R S P E C T I V E
( PA P U A N E W G U I N E A )
Polly Wiessner
Papua New Guinea is at once the land of feasting and the land of political in
trigue, 1 and the two are intertwined. Virtually every event of importance is ac
companied by feasting, and during feasts an array of strategies · are played out.
The strong political orientation of Papua New Guinea societies no doubt con
tributes to the fact that over 800 different languages are spoken in this small island
nation of some three and a half million inhabitants. A meaningful classification
of feasts by the social, economic, or political designs they entertain is difficult to
frame, particularly for the western highlands of Papua New Guinea where many
personal or group proj ects unfold under the umbrella of feasting, no matter what
the proclaimed purpose of the event.
Nonetheless, since feasting is a composite event in which certain conditions, so-
115
Polly Wtessner
cial behaviors, and provisioning activities coalesce to form the final occasion, a
breakdown of feasts by their components is useful for understanding when and why
certain strategies are deployed. Here I will briefly outline the essential components
of feasting, suggest some possible archaeological correlates of each, and mention
some of the social and political strategies they facilitate. Then, drawing on an eth
nohistorical study of feasting and exchange among the Enga of highland Papua
New Guinea, I will outline the complementary role of secular and sacred feasting in
the development of the great ceremonial exchange networks of Enga and go on to
focus on the role of feasting in the construction of the value and meaning of things.
The Enga case illustrates: (1) how material goods are differentially valued in
the context of feasting; (2) how the natural properties of things valued affects the
course of production and competition; and (3) how cultural constructions of
value are constantly tested against the realities of the surrounding world. When
contradictions occur, as they did in Enga, feasting can be called on to revalue
goods and redirect the course of change.
CO M PO N E NTS O F F EASTI N G
Of what is feasting composed? First, feasting requires the aggregation of people.
Aggregation in and of itself does not require feasting, though it is greatly facili
tated by it; if no food is available, crowds must disperse before long. Aggregation
should be detectable in the archaeological record, for example through the pres
ence of unusually large sites or a diversity of styles in artifacts found at a given site.
Second, feasting involves food sharing and food distribution. Food sharing ap
pears to have its roots in the parent-child relationship and thus can be a way of
expressing affection and extending familial behavior to distant or non-kin in
order to bond larger groups (Eibl-Eibesfeldt 1989) . By contrast, food distribution,
which often requires returns at a later date, creates temporary imbalance be
tween donors and recipients and permits the construction of inequality. Food
sharing and food distribution during feasting may be inferred in the archaeologi
cal record from faunal distribution or the collection of food remains from a broad
catchment area at one place. Sharing is more likely to leave remains at the site
than food distribution, where a good portion of the food may be carried away
from the distribution site.
Third, most feasts are held for a specific occasion: to appease ancestors, initiate
youth, marry, bury the dead, pay compensation, or assemble a labor force. The
goal of feasts may be inferred from the presence of certain archaeological fea
tures or artifacts.
Fourth, feasting usually involves some form of display; whether this is display
of food, objects, individuals, or groups. Artifacts or structures constructed for
display include special vessels, platforms, graves, or houses.
116
E N G A F E ASTS I N A H I STO R I C A L P E R S P E C T I V E
Fifth, and very importantly, feasting requires abundance. A party where the
food does not go round or runs out is not a good party. Abundance is perhaps the
single most important factor in determining which strategies can be played out
during a feast. Mere sufficiency permits a limited array of strategies; abundance
furnishes many more. In foraging societies and simple horticultural societies,
conditions of abundance may be difficult to achieve or to time, so that feasts take
place when food is available, constraining political strategies. Depending on what
is consumed in a feast or how food is presented, abundance may be more or less
archaeologically visible.
Finally, since feasts are about the consumption of abundance, they place de
mands on production. If these demands are accelerating, they may be reflected in
changing patterns of land use.
If all of these conditions for feasting are met, they coalesce to evoke excite
ment, warmth, and festivity, lending spirit to the occasion. What is experienced
in such an atmosphere usually makes deep impressions. The festive spirit may be
reinforced by emotionally laden aesthetic experiences during communal song or
dance.
The social components of feasting behavior, when combined, make possible a
number of political strategies. Table 4.1 breaks down the above elements of feast
ing and relates them to specific social or political strategies.
F E A S T I N G A N D VA L U E
All of the political strategies listed in Table 4.1 were played out during Enga
feasts. Here I will concentrate on the construction of value. It is well-established
in anthropology that the value of things is based on several criteria: capital
input, labor input, utility, abundance or scarcity, exchange rate, and "social"
value, including religious and political significance (Appadurai 1986; Parry and
Bloch 1989; Thomas 1991) . Although the first three criteria may be predictable
on the basis of economic models, exchange value and social value are culturally
constructed. 2 The cultural construction of value in material culture applies to
both gifts and commodities, for these categories often describe different stages
or uses in an object's life history. For instance, a pig can be purchased in barter
and then given as a gift. Feasting provides the ideal conditions for the valuation
of obj ects for a number of reasons. First, feasts gather people so that certain per
mutations of value and meaning can be broadly introduced. Second, during
feasting selected items are put in the focus of attention. For example, they may
figure prominently in cult procedures, be worn by central figures, exchanged
publicly, or placed in a culinary display. Studies in ethology in humans and other
primates have established that being the focus of attention is the primary means
of gaining status (Chance 1967; Hold 1976); the same principle appears to be ap-
117
TABLE 4.1
The Social Components of Feasting a n d Corresponding Pol itical Strategies
Social Components Aggregation Food Sharing Food Distribution Special Purpose Display Abundance
Bonding groups xxx xxx x xxx x x

Creating external ties
a. Between individuals x xx xx x x x
b. Between groups xxx xxx x xx xx xx
Establishing social position

Equality
a. Between individuals x xxx x x x
b. Between groups xxx xxx x xx xxx
Inequality
a. Between individuals x xxx xxx xx xxx xxx
b. Between groups xxx xxx xxx xxx xxx xxx
Production (work parties) xxx xxx xxx x

Manipulating social norms and values xxx xx xx xxx xxx xx
Constructing value of goods and valuables xxx x xx xxx xxx xx
:x:xx = very important xx = important x = less important

E N G A F E ASTS I N A H I S TO R I C A L P E R S P E C T I V E
plicable to material goods. Third, the special purpose of a feast is instrumental

in giving the objects utilized or displayed specific meanings. Fourth, the spirit
generated during the communal feasting, song, and dance turn fleeting impres
sions into lasting ones.
T H E E N G A : A C A S E S T U DY
The subjects of this case study, the Enga, are highland horticulturalists number
ing about 200,000 and well known in the anthropological literature through the
works of Meggitt (1965, 1972, 1974, 1977), Feil (1984), Wohlt (1978), Lacey (1975),
and Waddell (1972), among many others. Their staple crop, sweet potato, is culti
vated in an intensive system of mulch mounding to feed large human and pig
populations. The Enga population is divided into a segmentary lineage system of
phratries or tribes composed of some l,ooo to 6,ooo members, and their con
stituent exogamous clans, subclans, and lineages.3 The politics of land, social net
works, and exchange occupy much of men's time and effort, while women
devote themselves primarily to fainily, gardening, and pig husbandry: Frequent
and destructive warfare creates sharp divisions between clans (Meggitt 1977) that
are periodically repaired through ceremonial exchanges to reestablish peace. All
men are defined as potentially equal, though those who excel in public media
tion, organization of events, oratory, and the manipulation of wealth make
names for themselves as big-men. Women exert considerable influence over both
production and exchange in the private realm.
The Enga hold a rich body of oral traditions that include myth, song, poetry,
spells, incantations, and historical traditions (atome pii), which are distinguished
from myth or stories (tindi pii), in that they are said to be founded in eyewit
nessed events of the past. Historical traditions contain information on such mat
ters as tribal locations and subsistence practices some eight to ten generations
ago, wars and migrations, agriculture, the development of cults and ceremonial
exchange networks, leadership, trade, environmental disasters, and innovations
in song and dress. Genealogies of eight to twelve generations permit a rough
scheme of sequencing or dating (Table 4.2). The material on which I will draw
comes from a study of these historical traditions conducted with an Enga col
league, Akii Tumu, in rn8 Enga tribes (Wiessner and Tumu 1998) . Working on
such a broad scale made it possible to identify regional trends that occurred
within the span of Enga oral history and to check for consistency by using con
verging lines of evidence.
The period covered by Enga historical traditions begins just prior to the intro
duction of the sweet potato, continues through first contact with Europeans in
the 1930s, and onward into the present (Table 4.2). As the introduction of the
sweet potato is placed early within the span of historical traditions, it is beyond
1 19
Polly Wiessner
TABLE 4.2
Chronologica l Scheme of Events Discussed in Text
Prehistory
50,000 B.P. First immigrants arrive in Papua New Guinea by sea from Asia
12,000 B.P. Earliest archaeological evidence for Enga: Yuku cave in eastern
Enga, hunting and gathering site at ca. 1 ,300 m
10,000 B.P. Kutepa rockshelter in western Enga, hunting and gathering at
2,300 m
4000-2000 B.P. Pollen evidence indicates forest clearance for horticulture in
eastern Enga
400-250 B.P. Introduction of sweet potato to Enga and beginning of Enga
historical traditions
Historical Traditions
Generation before present

8 Population shift from high altitudes to lower valleys
Beginning of early Tee cycle
7 Kepele cult first practiced by horticulturalists of western Enga
6 Beginning of Great Ceremonial Wars
5 (ca. 1 855-1885?) Kepele cult imported into central Enga, called Aeatee
4 (ca. 1 8 85-1915) Tee cycle expanded to finance Great Ceremonial Wars
Aeatee cult expanded to coordinate Tee cycle and Great
Ceremonial Wars
Female Spirit Cult imported into eastern Enga
3 (ca. 1 9 1 5-1945) Tee cycle begins to subsume Great Ceremonial Wars
Aeatee/ Kepele cult used to organize the Tee cycle
First contact with Europeans in 1934
Last Great Ceremonial War fought 1938-1941
Tee cycle subsumes Great War exchange routes
Ain's cult 1941-1942
2 (ca . 1945-1975) Colonial period begins-late 1 940s in east, 1950-60s in west
Female Spirit cult spreads
1 (ca. 1 975-2005) 1975 Papua New Guinea's independence
the scope of reliable genealogical dating. Our best estimate is that it arrived be
tween 250 and 400 years ago, releasing many constraints on production owing to
its high yield per acre, ability grow well in poor soils and at high altitudes, resis
tance to blight, and superiority as pig fodder (Watson 1965a, 1965b, 1977). The
sparse population of Enga that had previously subsisted by taro gardening or
hunting and gathering was able to expand into higher niches, produce a substan
tial surplus, and store it "on the hoof" in pig populations. Another 200-350 years
passed before European gold prospectors and patrols entered the Papua New
Guinea highlands.
1 20
D E V E LO P M E N T S W I T H I N T H E S PA N O F
E N G A H I STO R I CA L TRA D I T I O N S
Descriptions from the earliest generations of Enga history portray a lifestyle very
different from that of more recent times. In the fertile valleys of eastern Enga,
where most of the land lies under approximately 1,900 m, sedentary agricultur
ists cultivated taro, yams, bananas, sugarcane, and other crops, supplementing
their diet with hunting. In central Enga at altitudes between 1,900 m and 2,200 m,
where taro and yams are not as productive, historical traditions describe shifting
horticulturalists who cultivated mixed gardens and hunted. In the less fertile and
more rugged valleys of western Enga, the population was divided by niche. Mo
bile groups who depended heavily on hunting and gathering inhabited regions
4
above approximately 2,100--2,200 m. To these "hunters" were attributed great
physical strength and possession of powerful ritual and magic. Shifting horticul
turalists, who subsisted on taro and other garden products, supplemented by
hunting, inhabited the steep, narrow valleys below. There is mention of regular
intermarriage and exchange between people of the high country and those of
the valleys, as well as a good deal of tension and misunderstanding. Throughout
Enga the population is described as sparse with long distances between settle
ments. Wars were fought and problems ultimately solved by dispersal in space;
compensation to allies was paid with land and gifts of food for feasting. At this
time, prized gifts and commodities included stone axes, salt, and items for self
decoration, the latter of which were displayed prominently at traditional feasts
and dances. The plots of historical traditions from this period often revolve
around hunting, trade, traditional dances, or warfare. Pigs were apparently kept
in small numbers, but enter only peripherally into the oral record.
Acceptance and utilization of the sweet potato varied by region, though his
torical traditions from all areas report shifts in population distribution, eventual
population growth, and the expansion of ceremonial exchange, religious ritual,
and feasting in response to increased social and political complexity (Wiessner
and Tumu 1998). The purposes, structure, and content of such responses in east
ern, central, and western Enga differed greatly. In eastern Enga the sweet potato
slipped into the garden regime with little note in historical traditions. Its initial
impact was only an indirect one-to stimulate the immigration of people from
higher regions to the south and east in pursuit of good garden land. These immi
grants, who came from clans closer to the sources of ax stone and other valu
ables, threatened the positions of big-men in the trade of eastern Enga. In
response, big-men of the Saka Valley took action and constructed a new system
of finance through which to raise the wealth necessary to maintain regional
prominence. As the legend goes, they sent messages and initiatory gifts to their
partners along established trade routes asking them to provide wealth in the
1 21
Polly Wiessner
form of pigs and valuables on credit, rather than by barter, and for them to en
courage their respective partners down the line to do the same. When wealth ar
riving along these chains of finance reached the Saka Valley, they invested it in
marriages and other relationships that would increase their spheres of influence.
Returns from these investments, together with wealth from home production,
were used to pay creditors in a public Tee festival. Thus through the concatenation
of trade partnerships based on kinship ties, chains of finance were constructed to
compose the skeleton of what was to become the Tee ceremonial exchange cycle.
The early Tee was conducted on a small scale, but nonetheless it introduced a
powerful innovation into Enga exchange--chains of finance that gave access to
the wealth of people who were usually beyond the bounds of kinship reckoning.
Some two to three generations after its conception, neighboring groups realized
the advantage of the Tee for obtaining wealth to finance bridewealth payments,
attract allies in warfare, pay war reparations, and secure partners along trade
routes. And so it spread throughout eastern Enga (Fig. 4.1).
0 10
km
...._____.
* Female spirit cult

Nebilyer VaUey
• Kepele sites
• Great Wars
� Opponents in Great Wars KauaeJ.Valley
• • • • Pre 1920 Tee routes
- - .Tee routes established after Great Wars * North Mendl
Figure 4.r. Networks of ceremonial exchange, warfare, and ritual that developed after the
introduction of the sweet potato.
1 22
As it grew, the Tee developed into a three-phase cycle (Feil 1984; Meggitt 1974) .
During the first phase, the saandi pingi, initiatory gifts of small pigs, goods, and
valuables were given on a private basis by individual families to Tee partners in
clans to their west, more often than not adj acent clans. Initiatory gifts flowed in
one direction and were intended to "pull" the tee pingi, the phase of main gifts.
When a sufficient number of saandi initiatory gifts had moved westward, big
men made efforts to launch the phase of the main gifts in which full-grown pigs,
cassowary, pearl shells, and other valuables were distributed by individual fami
lies during public clan festivals to partners who had given initiatory gifts. Some
days or weeks after the westernmost clan in the Tee had held its distribution, the
neighboring clan to the east followed suit and so the tee pingi worked its way
westward, clan by clan, with the main gifts flowing in the opposite direction to
the initiatory gifts. Tee pingi festivals were events of unsurpassed excitement,
pomp, and ceremony. Suspense and anticipation were high throughout, for to
give and receive large numbers of pigs, goods, and valuables was an indicator of
political victory in controlling the flow of wealth. Part of the wealth distributed
in the phase of the main gifts traveled the entire Tee network; part was pulled
out to finance local events.
When the phase of the main gifts reached the easternmost clan in the net
work, efforts were made to launch the final reciprocal phase of the Tee cycle, the
mena yae pingi or simply yae. The yae involved the distribution of butchered pork
that traveled in the opposite direction of the main gifts, retracing its steps.
Donors from the phase of the main gifts received approximately one side of
butchered pork for each pig they had given. It was during the yae phase that ex
tensive feasting took place, wealth in the form of pork was consumed, and de
mands for production renewed. However, pork received in the yae was not for
consumption only; but also used to pay a variety of debts and establish credit
with which to initiate a new cycle that would send the main gifts in the opposite
direction.
In central Enga, developments that took place after the introduction of the
sweet potato were quite different. Wars raged as tribes of higher altitudes sought
to take land in the fertile valleys after the introduction of the sweet potato (Fig.
4.1) and large segments of the population were displaced. Here a very different
system of exchange emerged: the Great Ceremonial Wars. These wars consisted
of a series carefully planned, large, seiniritualized tournaments lasting for weeks
or months fought recurrently two to three times a generation by entire tribes or
pairs of tribes. "Owners of the fight" were hosted by other tribes who supplied
them with a battlefield, food, water, and allied warriors for the duration of the
war. Their purpose was twofold. First, they displayed force and solidified al
liances to reestablish balance of power vis-a-vis enemy groups in the face of the
1 23
Polly Wiessner
population shifts mentioned above. Antagonism was limited to display on the

battlefield-no land could be gained or lost. Second, they were fought to brew
the massive exchanges and feasts that followed the Great Wars between "owners
of the fight," their hosts, and allies. These turned bonds of brotherhood estab
lished during the fighting into economic relations of exchange, putting displaced
groups back on the map of trade and exchange and creating networks between
four maj or valley systems (Wiessner and Tumu 1998) . Great War exchanges, such
as the Tee cycle, involved alternating distributions of live pigs and valuables and
those of butchered pork, with large feasts occurring during the latter.
In western Enga, groups who were formerly dependent on hunting and gath
ering moved into the Lagaip Valley and settled on land acquired from relatives re
lated through marriage. Such moves incited tension, which is evidenced in a
history replete with many small wars and migrations. The people of western
Enga turned to ritual life for integration. A large cult called the Kepele was
crafted to assemble the tribe, reaffirm unity, and draw attention away from local
disputes. Kepele cults of different tribes of western Enga were integrated into a
network bound together by the exchange of ritual experts, cult procedures, inno
vations, and attendance by relatives from other tribes (Fig. 4.1). During the five
days of Kepele celebrations, special feasts were held for those taking part in sa
cred events-the initiates, ancestors, ritual experts-as well as for celebrants who
remained outside the sacred area-women, children, and invited guests.
Furthermore, as the population grew and conflicts accelerated throughout
Enga, war reparations, which were traditionally paid to allies for men lost in bat
tle, were extended to enemies. While formerly opponents had dispersed after
warfare, one ousting the other, as the population grew and the land filled, con
flicting parties had to stay put and establish peace, a condition that was achieved
through wealth exchange. War reparations involved an initial payment of cooked
pork and subsequent feasting, followed by a series of exchanges of live pigs,
goods, and valuables. When successful, peace and balance of power were
restored, former exchange ties with enemies were reactivated, and new ones
created.
Initially; all of the above exchange systems depended heavily on wealth other
than pigs. The early Tee cycle involved the exchange of a small numbers of pigs,
but emphasis was on trade goods: salt, axes, cosmetic oil, plumes, woven net
wear, shells, and other items for self-decoration. Early Kepele feasts were heavily
provisioned with marsupials, cassowary; and forest products, as were the Great
War exchanges. War reparations and compensation were paid with land and food
for feasting. But as the population grew, the land filled, and competition acceler
ated, trade goods, game animals, and land no longer sufficed to fill the needs for
exchange. Pigs were the only form of wealth whose production could be readily
1 24
intensified to meet new demands, but their value was not high enough to en
courage intensification. A number of historical traditions directly state that many
people were simply not interested in the drudgery of raising more pigs. And so
big-men, who realized that their ambitions as well as the needs of their own clans
could only be met by increased pig production, sought to enhance the value of
the pigs. Some took promising young men into their households to teach them
more about pig husbandry or encouraged their daughters to demand larger num
ber of pigs as a part of their bridewealth. They also turned to more efficient
means-cults and feasting.
E N G A F EA S T I N G : S E C U LA R A N D SAC R E D
Feasting, which was so integral to the development of the Tee cycle, the Great
Ceremonial Wars, the Kepele cult, and other events, was held in two contexts:
the secular and the sacred. Sacred and secular feasts were distinctly separated,
though the two complemented one another. Most secular feasts were embedded
in exchange events composed of two- or three-phase wealth distributions sepa
rated by weeks, months, or years, similar to those discussed earlier for the Tee
cycle. Usually; two of these phases involved the exchange of live pigs, game,
goods, and valuables, and at least one involved the exchange of butchered pork,
vegetable foods, and wild game meat. Secular feasting adhered to the principles
of the economic exchange, that is to say, individual contributions of pork were
coordinated into a clanwide public distribution of wealth on the local ceremonial
ground. However, the actual donations were not pooled and redistributed by a
big-man, but given by individuals to their partners in the recipient group(s) with
obligations for future reciprocation. Individuals stood to gain status from giving
generously and from indebting others.
At the domestic level, secular feasts included: work party feasts to recruit as
sistance in household production, planning feasts for upcoming events, marriage
feasts, payments to maternal kin to mark the growth of children or compensate
for their injury; and funeral feasts in which maternal kin were compensated for
the loss of an individual who was born and raised by them. In all of these feasts,
except work feasts and planning feasts, contributions from the clan, subclan, or
lineage were given to affi.nal or maternal kin outside the clan to initiate or main
tain long-term exchange relationships. In addition to domestic feasts, larger sec
ular feasts were held to secure interclan relations as well as interfarnily ties-for
example, payments to compensate the clans of allies killed in warfare, payments
to enemies to compensate losses and terminate hostilities, or the distribution of
cooked pork in the final phase of the Tee cycle repaying wealth received in ear
lier phases and to demonstrate the prosperity and generosity of the donor clan.
Accordingly; large secular feasts usually involved display in a competitive con-
1 25
Polly Wiessner
text. Smaller secular exchanges and feasts were often timed to plan and finance
larger ones.
Interestingly, most secular feasts were not events that united the two parties in
communal consumption. Vegetable foods prepared to welcome guests were
eaten on the spot, sometimes with scraps of pork, but the substantial contribu
tions of pork or marsupials were handled as exchange items. In most secular dis
tributions, even weddings, recipients collected their shares and soon after set off
for home before darkness fell to distribute them within their extended families or
to those who had helped finance the exchanges. Public distribution permitted in
dividuals and groups to gain prestige, but the bonding of the two groups was
truncated by separate consumption in the absence of song, dance, and other fes
tivities. Though not eaten communally with donors, consumption of meat re
ceived in exchange was nonetheless seen as a form of indirect commensality, for
Enga did not consume meat stemming from groups who were considered to be
enemies. Moreover, consumption of pork among the Enga, whose daily diet is
one of vegetable foods, is an event of excitement and great pleasure that draws
praise, warmth, and gratitude for the food donors. And so, despite the absence of
communal consumption in Enga secular feasts, the addition of a round of ex
changes involving highly divisible foods that pleased many consumers added a
significant social element to secular exchange cycles.
Benefits obtained by ambitious men for arranging exchanges and accompany
ing secular feasts were substantial. They gained "name" for their organizational
efforts, and they gained wealth from successful exchange. Some of this was ap
plied to clan needs, but enough was left to pursue personal interests and social
position-to procure the wealth necessary to pay bridewealth for one or more
wives, to attract allies in warfare, to recruit disabled individuals or war refugees
to assist in production, and to secure trade partnerships.
Two features of secular feasting affected its potential for serving social and po
litical ends. The one was the lack of communal consumption, and the other, the
fact that the food distributed carried exchange obligations. The absence of com
munal consumption involving donor and recipient limited the potential of secu
lar feasts to generate feelings of unity. And maintenance of a certain distance
may have been intentional, for even though strong supportive ties were formed
outside the clan through exchange, these had to be suspended in times of war
fare. The fact that economic goals were embedded in secular feasting meant that
all moves of big-men to bring about changes were regarded as economically
driven and subject to scrutiny. Such an atmosphere was not conducive to manip
ulation of norms and values. Nonetheless, the incorporation of a phase of feast
ing into secular exchange allowed big-men to put increasing numbers of pigs in
1 26
the center of distributions and actively demonstrate their potential as a social and
economic currency.
Sacred feasts followed different principles than secular ones and offered differ
ent opportunities for the enterprising. At their core was commensality with hu
mans and the spirit world united through celebration, rites, incantations, song,
and food sharing. Sacred cult feasts were held at many levels from the consump
tion of a single pig at small healing ceremonies to cults for the ancestors that
drew together hundreds of people in festivities lasting up to five days or longer.
Sacred feasts were "nested"-smaller ceremonies were held first, and if unfavor
able conditions persisted, ever larger ones were held. Some tribes had a reper
toire of up to ten different cults and accompanying feasts that could be held for
the ancestors, "sky people," and other spirits. A tribe or clan's repertoire of cults
was periodically revised through the importation of new cults from other groups
who appeared to be thriving or through the invitation of ritual specialists from
other groups who introduced additional rites.
The largest sacred feasts, organized by big-men, gathered clan or tribal mem
bers for five days of communal consumption at the cult site and entertained in
vited guests from other tribes. Unlike secular feasts that involved competitive
provisioning and distribution, in sacred feasts every household contributed
equally and everybody received freely to reaffirm equality of group members-it
is said that even the dogs did not go hungry. Food produced for cult feasts was
grown specifically for that purpose and never used to incur or discharge obliga
tions related to secular exchange and feasting. The commensality of sacred feasts
allowed for strong bonding at the group level that did not occur in secular feasts.
As a result, entire tribes could be united.
Ritual experts who presided over cult rites were paid lavishly and stood to gain
economically, although they did not have the social networks to invest their earn
ings in such a way as to wield power. In contrast, the benefits for big-men who or
ganized these impressive events were other than financial ones. One benefit was
the prestige that accrued from their ability to coordinate the production, cooper
ation, and participation necessary to stage successful cult performances and
feasts. Prestige brought more than personal satisfaction-it was license to excel.
In this society where all men were potential equals and equality was staunchly
defended, name was only accorded to those who offered strong benefits to the
group as a whole. Prestige signaled a man's ability to do just that. Accordingly, a
man of renown won praise rather than protest for the number of wives he could
n;iarry, servants he could recruit, and wealth he could produce or obtain through
exchange. Moreover, big-men who could organize intergroup events were con
sidered too important to be kill ed senselessly in warfare. The protective umbrella
1 27
Polly Wiessner
of "good name" was not insignificant in a society where approximately 25 per

cent of men die in warfare (Meggitt 1977).
A second benefit open to big-men through participating in the organization of
cult performances was the opportunity to alter norms and values. Big-men par
ticipated in importing new cults from groups who appeared more successful,
summoning chosen ritual experts from other groups, and rearranging cult proce
dures to better communicate with the ancestors. In doing so, they were able to
align norms and values with their personal projects. Because all cult-related deci
sions were made in the name of group betterment and detached from economi
cally oriented exchange, the motives of big-men were not subj ect to the same
scrutiny as were their maneuvers in secular feasting and exchange. To give some
examples: through importing the praise poetry of bachelors' cults from other
groups and making additions in the process, big-men could update the ideal for
males; cults expressing equality of all men in the eyes of the ancestors could be
timed appropriately to appease grudges emanating from emergent economic in
equalities; and very importantly; goods and values were revalued through their
prominent presentation in association with the sacred.
Different though they were, the relation between sacred and secular feasts
was a complementary one. In a sense, sacred feasts laid the groundwork of rules
from which competitive exchange could proceed. They defined all males and
their family units as potential equal competitors for exchange, they displayed
what was valued before the eyes of the ancestors, and when exchange depended
heavily on group participation, they established a spirit of cooperation among
corporate groups. When these conditions were met, individuals could go out
and pursue their careers in secular exchange to benefit themselves and their
clans. In turn, success in secular exchange and feasting established the practical
value of items valued during sacred feasts and reaffirmed the importance of reg
ular cult feasts to secure the goodwill of the ancestors for economic enterprises.
Alternately; failures in exchange called for new cults, new rites, or in some cases,
new values.
In an archaeological context, secular and sacred feasting left different signa
tures. The absence of communal consumption in secular feasts had a profound
effect on the remains. When food was distributed and carried home, only traces
of structures for food display and distribution remained at the aggregation site
postholes from display platforms, small amounts of bone from meat consumed
during food preparation, and in some cases large earth ovens near the ceremonial
ground to steam pork, though much meat was prepared at individual houses.
The fact that meat was not consumed at the distribution site caused the majority
of faunal remains to end up around household steaming pits in other clans. For
sacred feasts,5 by contrast, prolonged aggregation required the construction of a
1 28
E N G A F E ASTS I N A H I S TO R I C A L P E R S P E C T I V E
feasting site with shelters for the participants; communal consumption of all
food left extensive faunal remains on the feasting site;6 the gathering of cele
brants and guests called for divided sacred and secular areas to separate insiders
from outsiders; and on larger Enga cult sites, special structures were constructed
for sacred rites, each of which had its own area for food preparation or con
sumption.
Significant also is the differential impact of competitive secular feasting and
noncompetitive sacred feasting on food production and associated utilization of
the landscape. Figure 4.2 compares the number of pigs given by big-men and
their wives in the main phase of the Tee cycle with the number contributed by
big-men to provision the Kepele cult by generation. It starts in the seventh gener
ation before present just as the Tee cycle was beginning and pigs were being in
troduced into the Kepele. It ends in the third to second generation (the beginning
of the colonial period) with a Tee held in the early 1950s and the last Kepele per
formances held before Christianity took hold in western Enga. Approximately a
third of the pigs given in Tee are produced by the household (see also Feil 1984)
and two thirds received as gifts; approximately half of the pigs distributed in Tee
were slaughtered in the final phase of feasting and the others invested into
bridewealth, compensation, and other payments. All of the pigs contributed to
the Kepele cult performances were produced at home and consumed during the
cult. As is clear in Figure 4.2, the overtly competitive distribution of wealth in sec
ular events and accompanying opportunities for individual investment and eco
nomic gain fostered rapid growth in comparison to the more gradual growth in
the noncompetitive Kepele. Although such growth might not be apparent from
the sparse remains left on secular feasting sites, its impact is apparent in patterns
and extent of land clearance and use.
Returning to the role of feasts in the valuation of material goods, below I will
give one example of an elaborate cult centered on feasting, the Kepele of west
ern Enga, through which pigs were made prominent in an area where pig pro
duction was arduous. Then I will go on to discuss the consequences of using
complementary strategies in secular and sacred feasting to value the pig and con
struct ever-larger exchange networks, the complexities that were generated, and
subsequent efforts to devalue the pig or shift value to other items.
T H E K E P E L E C U LT
The Kepele had its roots in a pre-sweet potato cult to assemble dispersed hunting
and gathering tribes of the high country in western Enga and initiate young men
into the secrets of the spirit world. 7 After the introduction of the new crop, when
these tribes acquired land from relatives and settled in the lower valleys, they
brought the Kepele with them. There it was restructured to fit needs of a horti-
1 29
250
20
T ee C yc e
c:
<ti
l
E 1 50
bi
:c
(I)
.....
a.
en
Cl
1 00
·a.
50
o l... """""",;
..... "' ;;
;;,_ ;;
;;; ;; :: K=e�p=e=le===
;;; ;:; �====:=::=:::::� C=u=
l t==;::::_�__,_
7 6 5 4 3 2
generation before present
Figure 4.2. Comparison of the growth of secular competitive exchange and feasting with
noncompetitive sacred feasting as measured in number of pigs contributed by big-men.
These figures are approximate. For the Tee cycle, number of pigs distributed is taken
from the historical traditions detailing the lives of major big-men from the Yakani and
Apulini tribes of central Enga. It is not possible to verify the accuracy of figures for ear
lier generations. However, since trees were planted at the end of lines of pigs distributed
to mark the accomplishments of each generation of big-men, the order of magnitude
should be correct. Prior to about the fourth generation, only wealthier clan members
participated in the Tee cycle; after the fourth generation, most families distributed at
least a few pigs as a matter of pride and sign of economic independence. Figures pre
sented here are the maximum for major big-men who were Tee cycle organizers; in the
second and third generations, big-men of less powerful clans might present 4 0-1 50 pigs,
minor big-men, 3 0-50 pigs, and ordinary men fewer than 10.
Figures fo r Kepele cult performances were calculated o n th e basis o f claims that early
cult performances were provisioned with marsupials and later ones increasingly with pigs.
From approximately the fourth generation on, it was stipulated that every male tribal
member should contribute one pig and one pig only. Big-men with more male children
would thus contribute more pigs as well as provide some pork for organizational activities,
but their contributions did not greatly outstrip those of other clan members.
E N G A F E ASTS I N A H I STO R I C A L P E R S P ECTIVE
cultural lifestyle, including pig production, and to integrate the new settlers and
their hosts (Wiessner and Tumu 1998).
The cult was called when poor environmental conditions, particularly the
poor growth of gardens, children, pandanus, and pigs, indicated that the ances
tors were discontented. Word was sent out to each clan to bring specific materi
als and foods from their own area for the construction of the tribal cult house
where rites for the ancestors would be held. On the appointed day; the clans con
verged on the cult site in full ceremonial dress, singing and dancing. Each clan or
subclan bore designated material for building the cult house, a perimeter post
representing their group, and special foods from their area for the feast. Men
from all clans of the tribe jointly constructed the cult house. Upon its comple
tion, a large feast was held for all contributors and their families.
While rites for building the cult house appear to have long been a feature of
the Kepele, formerly the feast had been provisioned with marsupials and other
forest products. After the introduction of the sweet potato, pigs or pork were
added to existing conventions and made integral to all new rites instituted. For
example, during the communal feast in the house construction stage, each tribal
segment was exp�cted to bring pigs of a specific color, stake them out, club them,
butcher them, and then divide the meat between all celebrants from the tribe. In
this way; pigs were used to underwrite the central political metaphor of the
cult-the division of the tribe into independent segments, on the one hand, to
gether with their mutual obligations to help one another, on the other. Cele
brants then feasted and went home, leaving the house to deteriorate for several
months to a couple of years until preparations for the major phase of the Kepele
had been made.
As time for the major phase of the Kepele drew near, word was sent out to cel
ebrants and guests to plant gardens and fatten pigs, gather materials, and to as
semble young men for initiation. Requirements stipulated that every male
tribesman provide one pig, whether he be adult or child. The sacred area for the
cult was fenced off, the Kepele ancestral cult house repaired and decorated, and
additional houses built for the initiation of boys, the steaming of pork, and the
housing of ritual experts, men, and boys who would participate. When the cele
brants had arrived, all of the pigs that had been brought for the ceremony were
staked out and clubbed before the throngs of spectators. For the larger cere
monies of this century; 300-500 pigs were provided. Three pools were con
structed at the site, one to collect the blood, another for the entrails, and a third
to wash the intestines. Pigs to be used in the most sacred rites were slaughtered
within the enclosed sacred area. In some cult performances, the skulls of the pigs
were taken to the major cult house where they were placed on shelves for all
131
Polly Wiessner
male celebrants to see (Gibbs 1978). The pork to be used in ritual procedures was
carefully separated from that given to women, children, uninitiated men, and rel
atives from other tribes, and pork was prepared very differently for each category
of celebrant: the initiates, initiated adult males, ritual experts, and old men who
would soon j oin the ancestors. Preparation of pork thus endorsed social divisions
within Enga society. The large pig kill marked the onset of five days of sacred
rites. Meanwhile women, children, and visitors remained outside the sacred area
singing, dancing, feasting, trading, arranging marriages, or engaging in other so
cial activities.
Men took designated pigs inside the sacred area and steamed them in pits for
the various rites. One pig was specially slaughtered for a ritual expert who retired
to a small hut and filled a gourd with the sacred water of life to be given to the
initiates. The fluid was composed of the condensed breath of the ritual expert
who blew on a cool stone ax blade and let the droplets roll into a gourd container
where it was supplemented by pork fat and sugarcane juice. All of these sub
stances were considered critical for the growth and long life of young men.
When night fell, the initiates were taken to the cult house where they witnessed
secrets of the spirit world and then were given the "water of life" from the gourd.
Boys who had been through these rites were considered to be weaned from their
mothers, on their way to becoming true men, and ready to witness the mating
and feeding of the sacred ancestral objects. Thereafter they could consume pork
that had been ritually prepared for the cult and no longer remain outside the sa
cred area eating ordinary pork like women and children.
The rites for the sacred objects, as far as we could determine, were added to
the Kepele in the post-sweet potato era, when western Enga settled in lower val
leys. The sacred objects consisted of stones in shapes reminiscent of male or fe
male genitalia and a basketwork figure (yupini) fashioned to look like a man who
represented the male ancestors. 8 The yupini was paraded to the Kepele site and
placed in the Kepele house, while the sacred stones were unearthed from their
resting place in the ground. In the cult house the yupini (or male ancestral stone)
was made to simulate copulation with the female stones, while spells to promote
fertility were recited. After copulation, the yupini was fed with pork fat, and the
sacred stones were greased in it. Meanwhile, the celebrants and initiates feasted
on pork that was prepared separately for each age group and category of partici
pant. On the last day of the ceremony, the stones were wrapped in pork fat and
buried. The ancestors were believed to have feasted with their human descen
dants during these rites and to be content to sleep thereafter, leaving human af
fairs to proceed unhampered and prosperous. At some cult sites, skulls of
deceased tribal members were deposited over the years in a tribal skull house
within the Kepele site where a pig was killed and small rites performed for each
1 32
to appease the ghost. During the Kepele ceremony, the skulls that had accumu
lated were placed in a pyre between layers of all edible plant foods, pig fat, and
firewood and ceremonially cremated. When cult rites were concluded, some of
the cult houses were destroyed; it was believed that the goodwill of the ancestors
had been evoked and fertility would prevail.
And so in the generations following the introduction of the sweet potato, pigs
were introduced into all feasts and rites of the Kepele cult that had formerly been
furnished by game and forest products. In this context, pigs and pork were given
a multiplicity of meanings. During the cult house construction pigs represented
tribal divisions; in subsequent feasts, these divisions were dismantled through the
sharing of pork. The consumption of pork fat, among other things, was put forth
as means of transforming boys to men. Pork was used to appease the ghosts of
the dead and symbolically shared with the ancestors to bring about good fortune.
The preparation of pork in different ways for different age groups and social po
sitions reaffirmed the structure of Enga society. The demand that each male pro
duce one pig for the Kepele reinforced equality of male tribesmen and gave
incentive for pig production. The great feasts for the public that took place out
side the sacred area gave a tangible reward for production to women, children,
and invited guests who were not involved in the sacred rites.
Location of Kepele sites conformed to no specific pattern predictable for ar
chaeologists. Sites were placed centrally in tribal territory so that they were ac
cessible to families whose houses were widely scattered over clan land. Some
Kepele sites offered magnificent views, while others were nestled on lower ter
races. All Kepele sites were planted with a wide array of trees and ceremonial
shrubs-some of the largest and older trees in the Enga landscape are on cere
monial grounds or cult sites. Remains left on Kepele sites would attest to the
gradual valuation of pigs and the multiple meanings conferred on them. Most
striking for archaeologists would be the number of structures and the presence
of pig bones in association with every one as well as the association of particular
pig bones, such as skulls, with specific structures (Fig. 4.3). The grand enclosure
of the site and highly differentiated areas within it might suggest a ritual of unifi
cation for a given social unit and the legitimization of internal divisions within.
P R O M OT I N G P I G S A N D C O M P E T I T I O N : T H E O U TC O M E
The upshot of enhancing porcine value though ritual, feasting, and exchange was
that by the fourth to fifth generation before present, eastern, central, and west
ern Enga offered diverse economic and ritual institutions that can all be counted
as networks of ceremonial exchange and that were all grounded in a common
currency-the pig. Big-men took full advantage of this similarity and diversity,
weaving competitive and cooperative elements of the exchange systems of east-
1 33
Polly Wiessner
PIG STAKES
x= pig bones
x x x x x x p= pig skulls
x
.. . . . . . .. .. . . . . . .. .. . . . . . .. . · · · · · · ·
. x
x
s= human skulls
$ hole for sacred objects
x STEAMING PITS D
·X xx x x
steaming pits
x x x
x
(lined with rocks)
PUBLIC FEASTING AND DANCING AREA
FENCE AROUND SACRED AREA
�------;M;,:OTE HOUSE ��

IN
n .xO
Cl(()MATA u x
x
rn
x
/8\X •••••
Q x
x
x x
x \§/ B
& rn
PENDOKO
X
·
EIDERS HOUSE
x
X
'e) x
X
X : RESIDENTIA L
X
CREMATION NEE NAPE AN A
-
�® :
' s s' NG
'U .'
AREA
SKULL H
Figure 4.3. Schematic drawing of a Kepele cult site. This diagram is based on descriptions
of informants while visiting Kepele sites, not the measuring and mapping of a single site.
Though Kepele sites vary, many have the features presented here. Large sites may cover
an area of 200 m by 200 m or more.
em, central, and western Enga to craft the final Tee cycle. Specifically, the thin
web of the early Tee cycle with its cooperative chains of finance was called on to
fuel the Great Ceremonial Wars, removing the limitation of finance by home
production. This began in the fourth generation before present around 1890
(Table 4.2). As the Great Wars expanded under the forces of dramatic intergroup
competition, big-men involved in the Great Ceremonial Wars constructed longer
Tee chains to tap into the wealth of the east and to reinvest the great mass of
wealth that flowed out of the Great War exchanges. The Tee was thereby trans
formed from a relatively discreet stream of finance to a network flooded with
wealth. The spheres of exchange carved out by alliance in the Great Ceremonial
Wars later provided the pathways for the Tee cycle to expand (Fig. 4.1) .
The cost, conflicts, and complexity of organization of the Great Ceremonial
Wars and Tee cycle then became formidable. Big-men in tribes straddling the Tee
cycle and Great Wars, aware of the different strategies that could be employed in
secular and sacred feasting and the benefits of their juxtaposition, imported the
Kepele cult from the west into central Enga. There it was called Aeatee, a term
1 34
used for the Kepele in song.9 The reasons for its import in the fifth generation are
not stated, but initial performances were said to be held on a small scale. In the
fourth generation, Aeatee rites to unite the tribe and express equality of mem
bers were elaborated into a six-stage performance spread out over five to ten years
with each stage involving extensive feasting. Aeatee feasts were interdigitated with
phases of the Tee cycle to facilitate its organi.Zation and timing (Table 4.3). The
first stage of the Aeatee assembled the tribe for a marsupial feast and established
unity within. For the second and third stages, when the ground was prepared and
the cult house constructed, big-men from clans to the east who organized the
Tee cycle were invited as spectators and guests for the accompanying feasts.
TABLE 4.3
The Six Stages of the Aeatee Cult of Central Enga
and Its Relation to theTee Cycle
Stage Feast Purpose Relation to Tee Cycle
Collection of building Communal hunt; Unites tribe None

materials for the cult marsupial feast
house
Trampling of the grass Marsupial feast Unites tribe; expresses Big-men come
on the building site tribal structure from the east to
plan the Tee
Construction of the Major pork feast; Unites tribe; expresses After feasts,
Aeatee house one pig per male tribal structure and initiatory gifts
tribesman complementarity of sent to launch
clans; reasserts equality Tee cycle
of tribal m�mbers
appeases ancestors
Cleaning of the house Communal hunt; Tribe united None
site, redecoration of the marsupial feast
walls
Rites for the sacred Large pork feast Ancestral stones fed and Big-men set off for
ancestral stones buried; tribal prosperity the east to
assured request major
phase of Tee
cycle
Burning of the Aeatee Largest pork feast Marks end to successful Main phase of Tee
house performance of the arrives at
Aeatee cult for the western end of
ancestors; rival tribes in Tee cycle; return
the Great Wars compete yae phase of
to burn the cult house butchered pork
initiated
1 35
Polly Wiessner
Upon witnessing the performance, they realized that tribes of western Enga
were unified, prosperous, and that it was time to "talk Tee." This was done on
the side in the quiet of men's houses. Following the great feast of the third phase,
big-men from the east were sent home with gifts to launch the saandi pingi (phase
of initiatory gifts) of the Tee cycle. In the fourth stage of the Aeatee, a large pork
feast was held for the occasion of the feeding and mating of ancestral stones. Af
terwards, big-men from the west set off with their Tee associates from the east to
try to launch the phase of the main gifts in the east. The fifth stage of the Aeatee
was held when the phase of the main gifts in the Tee had worked its way to the
westernmost point. At this time the largest pork feast for the ancestors was held
to mark the end of a successful cult performance that expressed equality, cooper
ation, wealth, and assured prosperity through communication with the ances
tors; in the sixth stage the cult house was burned. Once these goals were
achieved, individuals were allowed to go out to pursue their own interests. The
yae phase of the Tee cycle, that involving the distribution and consumption of
butchered pork, was initiated thereafter.
Throughout, Aeatee and Tee performances and feasts were separated in time
and space, no pork was passed from one event to the other, and knowledge of
their relationship was kept from the public-big-men concealed the role of the
Aeatee in facilitating their goals for the Tee cycle. In short, the feasts of the
Aeatee were used by big-men to mediate tensions, reaffirm old values, introduce
new ones, realign feelings of tribal members, and demonstrate to participants
and spectators that the tribe was strong, wealthy, and able to act as a corporate
unit. Once these goals were achieved and all men proclaimed as potential equals,
individuals were allowed to pursue their own interests to make a name in ex
change.
During the decades before first contact with Europeans, the Great Ceremonial
Wars collapsed under their own weight and their spheres of exchange were sup
planted by the expanding Tee cycle (Wiessner and Tumu 1998). By first contact
with Europeans in the 1930s the Tee cycle involved some 40,000 people and the
exchanges of tens of thousands of pigs. The Tee cycle recorded by ethnogra
phers and Inissionaries in the 195os-197os (Bus 1951; Elkin 1953; Meggitt 1972, 1974;
Feil 1984) was thus formed by forging three vast systems of feasting and exchange
in the decades before first contact with Europeans . .
The results of some five generations of increased agricultural potential, the
valuation of the pig, and the merging of exchange systems was that competition
in Enga accelerated at a runaway pace, with the sons of each generation outdo
ing their fathers in number of pigs distributed. Mounting pressure was put on
production, for example, while in the seventh generation pigs are said to have de
pended primarily on foraging for their sustenance, by the late 1960s Waddell
1 36
(1972) found that over 60 percent of the sweet-potato harvest was fed to pigs in
the Aruni tribe of eastern Enga. And now for the most interesting part: competi
tion and accelerated production cannot escalate indefinitely. Something must
give. On the eve of first contact with Europeans, a number of responses to curb
competition were initiated that involved, among other things, resituating the pig
in the Enga scheme of value and meaning.
R E S P O N S E S I N T H E W E S T : A I N ' S C U LT
The less fertile, rugged valleys of western Enga were the first to come up against
environmental constraints, for the shift from game animals and trade goods to
the pig as the primary currency of ritual and exchange imposed a heavy work
load on residents in western areas. By the early twentieth century, warfare had
become endemic, and nutritionally related diseases prevalent. Elders interviewed
voiced the dismay they felt at the time over demands on small household labor
forces to produce large numbers of pigs for ritual, social, and political occasions,
particularly for Kepele cult performances, war reparations, and to help finance
relatives within the sphere of the Tee cycle. Pressures for pig production in turn
incited conflict over pigs and good garden land, setting off a spiral of rampant
and destructive warfare that took the lives of many. So dire were the circum
stances that it was said that the lives of men seemed to have little more value
than the life of pigs. These perceptions corresponded to predictions from a pow
erful cult, the Dindi Gamu, that environment and society were on a downward
spiral and that the end of the world was imminent (Wiessner and Tumu 2000, in
press). Diseases introduced by Europeans and rumors of European intrusion
contributed to the misery and created a feeling that change was in the air.
In the early 1940s a new cult was launched by a family of big-men and ritual ex
perts from the northwestern corner of Enga-Ain's cult or the Mata Katenge. Its
goal was to stave off disaster through ritual intervention (Meggitt 1973; Gibbs
1977; Feil 1983). Among other things, the pig, perceived as being at the root of the
problems, was devalued by cult prescriptions. First, demands for pig production
were decreased by abolishing events that required or celebrated pigs-warfare
and thus war reparations and major cults such as the Kepele. Second, the greater
part of the pig population was removed from circulation by the slaughter of pig
herds and their sacrifice to the sun. Though feasting after sacrifice was a part of
the cult, disrespect for pork was shown through great wastage (Meggitt 1973:26),
a practice very atypical in Enga where pork is precious. As the cult swept through
western Enga it took on millenarian aspects-promises that new wealth would
come in the form of giant celestial pigs and pearl shells. Competition would
cease, for wealth would come freely and equally to all, requiring no productive
effort. When the promised wealth did not appear, the cult collapsed, but not be-
1 37
Polly Wiessner
fore the greater part of the pig population of western Enga was slaughtered.
Shortly afterwards, patrols and missionaries entered the area.
R E S P O N S E S I N T H E E A S T : T H E F E M A L E S P I R I T C U LT
The impact of accelerating competition and intensification of pig production
was not so heavy in eastern Enga, where the soil was richer and environment
more forgiving. However, an unintended consequence of valuing pigs, rather
than items with more limited availability, was that all able-bodied Enga could
produce pigs, distribute them, and thereby become competitors in exchange.
With a growing population and expanding networks of exchange, the number of
competitors became unmanageable-the plans of even the most powerful big
men were foiled by competitors on the rise. In this context, big-men of the fourth
generation introduced gold-lip pearl shells cut into crescent shapes as valuables
into the Tee cycle and other exchanges as part of an effort to consolidate power.
Pearl shells, unlike pigs, could not easily be obtained by ordinary men who did
not have a long-standing control of trade networks to the south and southeast
(see Clark 1991). Furthermore, they did not have to be fed or slaughtered, but
could be retained indefinitely, accumulated, and passed on to descendants
(Lemonnier 1990, 1996). Pearl shells were gradually accepted as valuable by tribes
of easternmost Enga because they were prized by their eastern and southern
neighbors. Daughters of wealthy men began to demand these valuables in their
requests for bridewealth. 10 However, the central Enga, key participants in the Tee
cycle, rejected the eastern valuation of pearl shells, preferring pigs, stone axes,
and cowry shells.
Big-men then attempted to influence the valuation of pearl shells by another
means-through importing the Female Spirit cult, which placed pearl shells at its
center and in association with prosperity and fertility. Initial motivations for im
porting the cult from non-Enga groups to the south (Strathern 1970, l979a), as
stated in historical traditions, were to enhance clan or tribal fortune as well as to
host an event similar to the Aeatee cult of central Enga, which would draw peo
ple from near and far to plan the Tee cycle. Historical traditions tell of voyages
taken by big-men from eastern Enga, laden with goods and valuables contributed
by themselves and fellow clanspeople, to purchase the sacred objects, spells, and
rites of the cult as well as the services of a ritual expert to institute the cult in
their clans.
The Female Spirit cult was directed at a spirit woman who came to men as a
bride but remained a virgin with a closed vagina, giving men protection against
the contaminating menstrual fluids of women but bringing fertility to them and
their families. Two houses were constructed, one male and one female, and seg
regated from public areas by three elaborately decorated fences. Male celebrants
1 38
were divided into two moieties, one representing males and one females to ex
press the cult's central theme: that male and female must be separated but indis
solubly linked (Strathern 1979a). This theme was timely for the expanding
ceremonial exchange networks that depended on women's work and cultivating
ties with affinal and maternal kin. After a suspenseful day of secretive rites in the
cult house for which thirty to sixty pigs were slaughtered, male participants
emerged from the cult house in a dramatic parade, holding pearl shells before
them, level with their eyes. They danced with stamping movements around the
ceremonial grounds to the awe of the crowds and then paraded back to the cult
house, reemerging with the net bags of pork for the communal feast. When the
feast was over, invited guests from central Enga talked Tee just as they had done
after Aeatee performances. No doubt the prominence of the shells in the cult and
their association with fertility and prosperity did not escape evaluation by these
shrewd men.
As the Female Spirit cult spread through clans of eastern Enga, many details of
performance and meaning were lost or altered as it was adapted to new situa
tions (Wiessner and Tumu 1998). The parade, which presented pearl shells as a
coveted valuable with a sacred dimension, was perpetuated. The acceptance of
pearl shells as items of great economic and symbolic value spread. However, be
fore their impact on the social and political order could be determined, Euro
peans intervened, airlifting literally millions of pearl shells into the highlands to
be used as a currency in dealings with indigenous populations. Inflation of pearl
shells occurred and their value subsequently declined. Whether the pearl shells
would have been accepted as the ultimate valuables, whether their circulation
would have narrowed the field of competition, relieved pressure for pig produc
tion, or led to institutionalized social inequalities through the inheritance of
wealth remain open questions (see Feil 1984 and Strathern 1979b).
By contrast to the case of pigs in Kepele feasts, no direct archaeological evi
dence would be likely to be found for the valuation of pearl shells via the Female
Spirit cult. However, indirect inferences could be made from the appearance of
cult sites with a new structure at the same time that pearl shell fragments, or the
plaques of hardened resin on which shells were mounted, showed up in the ar
chaeological record in greater numbers.
S U M M A RY
Within the span of Enga historical traditions, the oral record illustrates how new
value and meaning were conferred on pigs during sacred feasts, and reaffirmed
during secular ones. Pigs had some noteworthy attributes. That they could be
produced by all permitted unbridled competition on the one hand; on the other,
the fact that they must be slaughtered and consumed to profit from labor in-
1 39
Polly Wiessner
vested made them an inappropriate form of wealth for the consolidation and
transmission of power. So once pigs were endowed with sufficient value to be
come an all-purpose currency for social, political, and religious matters, competi
tion accelerated at a runaway pace. \Vhen such competition reached the point
where it introduced tensions and contradictions into Enga society and ran up
against constraints of the external world, cults and their accompanying feasts
were then sought by big-men to either devalue the pig or shift value to items
with limited availability and greater durability, for example, pearl shells.
To conclude, I would like to make the following points. First, the event of
feasting is a composite one and emphasis on its different components will have an
effect on strategies that can be played out during feasting, as well as the structure
of remains in the archaeological record. By juxtaposing feasts that permitted dif
ferent strategies, such as secular and sacred ones, complex institutions such as the
Enga Tee cycle can be formed. Second, a good portion of the value and meaning
of material goods is culturally constructed. Feasting, by combining a number of
social activities and behaviors, provides an optimal setting for brewing such con
notations. Third, strategies couched in feasting, including the valuation of
things, should be identifiable in certain archaeological contexts. And fourth, the
attributes of items for which value is promoted, such as requirements for pro
duction, availability, durability, divisibility, lifespan, and disposal, have a strong
impact on the course of social, economic, and political competition. Many social
strategies ride piggyback on the natural properties of goods and valuables as
demonstrated in the Enga case. Accordingly, altering the values and meanings of
things through feasting can direct or redirect the course of change.
NOTES
1 . There are some excellent publications on feasting in the New Guinea literature, to
mention a few: Rappaport 1968; Young 1971; Strathern 1971; Lemonnier 1990, 1996;
Kahn 1986; Knauft 1993.
2. Exchange value is an interesting issue for it involves mediating between values placed
on items by those in different segments of a society or different societies.
3. The term phratry, as used by Meggitt (1965) is perhaps the most accurate
anthropological term for tata andake in Enga-political units composed of aggregates
of clans united by an origin tradition and genealogy that link members to a common
ancestor. As discussed elsewhere (Wiessner and Tumu 1998), we have chosen to use
tribe, a less precise notion, in order to use a term comprehensible to the Enga
themselves.
4. History portrays groups in the high country as hunter-gatherers; apparently they
maintained a sense of identity as such. However, it is possible that they cultivated
small gardens as well.
5. Sites for smaller feasts, both sacred and secular, would not be visible archaeologically.
1 40
6. Enga use banana and breadfruit leaves fo r serving food, not wooden o r ceramic
vessels.
7. Procedures and rites for the Kepele varied from tribe to tribe and performance to
performance. The one given here is a general description of major activities
performed at the majority of cult houses.
8. The sacred stones used in Enga ritual were frequently mortars and pestles made by
former inhabitants of Enga, perhaps their distant forebears. Pestles represent male
ancestors and mortars female ones. The yupini figure appears to have been imported
from the Sepik region. Sacred stones had been used by surrounding groups in lower
fertile valleys prior to the introduction of the sweet potato, however they were new
to the Kepele cult of western Enga.
9. Many Enga names, such as clan or cult names, have a special term used to refer to
them in song. Aeatee was the song term for Kepele.
ro. Enga brides reside in their future husbands' residences for some weeks before
marriages are finalized. At this time they sing songs to their husbands' clansmen
asking for specific items to be paid as part of the bridewealth (Kyakas and Wiessner
1992) . Demands of brides can thus have an impact on the value of items. While stone
axes were requested by brides of central Enga within living memory, pearl shells
were requested in the east.
REFERENCES
Appadurai, A.
1986 The Social Life of Things: Commodities in Cultural Perspective. Cambridge: Cam
Bus, G.
1951 The Te Festival or Gift Exchange in Enga (Central Highlands of New
Guinea). Anthropos 46:813-824.
Chance, M. R. A.
1967 Attention Structure as the Basis of Primate Rank Orders. Man 2:503-518.
Clarke, ].
1991 Pearlshell Symbolism in Highland Papua New Guinea, with Particular Refer
ence to the Wiru People of Southern Highlands Province. Oceania
61:309-339·
Dietler, M.
Perspective, edited by P. Wiessner and W Schiefenhovel, pp. 87-126. Oxford:
Berghahn Books.
Eibl-Eibesfeldt, I.
1989 Human Ethology. New York: Aldine.
Elkin, A.
1953 Delayed Exchange in Wabag Sub-District, Central Highlands of New Guinea
with Notes on Social Organization. Oceania 3:161-2or.
1 41
Polly Wiessner
Feil, D.
1983 A World without Exchange. Anthropos 78:89-ro6.
1984 Ways of Exchange: The Enga Tee of Papua New Guinea. St. Lucia: University of
Queensland Press.
Gibbs, P.
1977 The Cult from Lyeimi and the Ipili. Oceania 48:1-25.
1978 The Kepele Ritual of the Western Highlands, Papua New Guinea. Anthropos
73:434-447.
Hayden, B.
Interdisciplinary Perspectives, edited by P. Wiessner and W Schiefenhovel,
Hold, B.
1976 Attention-Structure and Rank Specific Behavior in Preschool Children. In
The Social Structure of Attention, edited by M. Chance and R. Larsen. London:
Wiley.
Khan, M.
1986 Always Hungry, Never Greedy: Food and Expression of Gender in a Melanesian So
ciety. Cambridge: Cambridge University Press.
Knauft, B.
1993 South Coast New Guinea Cultures. Cambridge: Cambridge University Press.
Kyakas, A., and P. Wiessner
1992 From inside the Women's House: Enga Women's Lives and Traditions. Brisbane:
Robert Brown.
Lacey, R.
1975 Oral Traditions as History: An Exploration of Oral Sources among the Enga
of the New Guinea Highlands. Unpublished Ph.D. thesis, University of
Wisconsin.
Lemonnier, P.
1990 Guerres et Festins: Paix, echanges, et competition dans les Highlands de Nouvelle
Guinee. Paris: Maison des Sciences de !'Homme.
1996 Food, Competition and Status in New Guinea. In Food and the Status Quest:
An Interdisciplinary Perspective, edited by P. Wiessner and W Schiefenhovel.
Oxford: Berghahn Books.
Meggitt, M.
1965 The Lineage System of the Mae-Enga of New Guinea. New York: Barnes and
Noble.
1972 System and Sub-System: The "Te" Exchange Cycle among the Mae Enga.
Human Ecology l:n1-123.
1973 The Sun and the Shakers: A Millenarian Cult and Its Transformation in the
New Guinea Highlands. Oceania 44:1-37 and ro9-126.
1974 "Pigs Are Our Hearts!" The Te Exchange Cycle among the Mae Enga of New
Guinea. Oceania 44:165-203.
1 42
1977 Blood Is Their Argument. Palo Alto: Mayfield.

Parry, ]. , and M. Bloch
1989 Money and the Morality of Exchange. Cambridge: Cambridge University Press.
Rappaport, R.
1968 Pigs for the Ancestors: Ritual in the Ecology of a New Guinea People. New Haven:
Yale University Press.
Strathern, A.
1970 The Female and Male Spirit Cults in Mount Hagen. Man 5:572-585.
1971 The Rope of Maka. Big-men and Ceremonial Exchange in Mount Hagen, New
Guinea. Cambridge: Cambridge University Press.
l979a Men's House, Women's House: The Efficacy of Opposition, Reversal, and
Pairing in the Melpa Amb Kor Cult. Journal of the Polynesian Society 8:37-5I.
l979b Gender, Ideology, and Money in Mt. Hagen. Man 4:530-548 .
1994 Lines of Power. In Migration and Transformations: Regional Perspectives in New
Guinea, edited by A. Strathern and G. Stiirzenhofecker. pp. 231-256. Pitts
burgh: University of Pittsburgh Press.
Thomas, N.
1991 Entangled Objects: Exchange, Material Culture, and Colonialism in the Pacific.
Cambridge: Harvard University Press.
Waddell, E.
1972 The Mound Builders: Agricultural Practices, Environment, and Society in the Cen
tral Highlands of New Guinea. Seattle: University of Washington Press.
Watson, ].
l965a The Significance of Recent Ecological Change in the Central Highlands of
New Guinea. Journal of the Polynesian Society 74:438-450.
l965b From Hunting to Horticulture in the New Guinea Highlands. Ethnology
4:295-309.
1977 Pigs, Fodder, and the Jones Effect in Post-Ipomoean New Guinea. Ethnology
6:57-70.
Wiessner, P., and A. Tumu
1998 Historical Vines: Enga Networks of Exchange, Ritual, and Waifare in Papua New
Guinea. Washington, D.C . : Sinithsonian Institution Press.
1999 A Collage of Cults. Canberra Anthropology 22 (1): 34-65.
2000 Averting the Bush.fire Day: Ain's Cult Revisited. In Ecology and the Sacred:
(in press) Engaging the Anthropology of Roy A. Rappaport, edited by M. Lambek and E.
Messer. Ann Arbor: University of Michigan Press.
Wohlt, P.
1978 Ecology, Agriculture, and Social Organization: The Dynamics of Group Composi
tion in the Highlands of New Guinea. University Microfilms: Ann Arbor.
Young, M.
197! Fighting with Food: Leadership, Values, and Social Control in a Massim Society.
1 43
5
AKHA FEASTING
A N E T H N OA R C H A E O LO G I C A L P E R S P E C T I V E
Michael J. Clarke
This chapter presents the results of two field seasons (1996, 1997) of ethnoar
chaeological research on the feasting practices of the Akha of Northern Thai
land. The main argument that I will present is that the Akha place great
importance on feasting in their society because (1) feasts function as social
mechanisms that facilitate the creation and maintenance of a life-crisis support
network, and (2) feasts provide arenas for competition between extended family
groups for control over basic socioeconomic resources such as land, labor, and
political influence. I will discuss the nature of the Akha resource base and the
difficulties associated with it, some of the various forms of feast types that func
tion to cope with these difficulties, and lastly; archaeological indicators of feast
ing activity.
1 44
A K H A F E ASTI N G
The goals o f my research have been to document the ways in which feasting
interacts with social structure, to construct a typological system for feasting, and
to search for material correlates of feasting activity that might serve as archaeo
logical indicators of feasting activity in the past. For the purpose of this study, I
define a feast as any ritualized meal that is consumed by two or more people. By
'ritualized,' I mean that the meal is not eaten solely for sustenance, but rather, is
considered as only one facet of a greater social event.
E T H N O G R A P H I C B A C KG R O U N D
The Akha (also known as the Kaw or E-Kaw) are an ethnic minority living in the
northern mountainous regions of mainland Southeast Asia. Although their exact
population is not known, it is generally believed that they number somewhere
around 500,000 individuals (Tribal Research Institute 1995) . The vast majority of
Akha live in Yunnan Province, China though significant numbers have been mi
grating southward for the past century or longer, and many now live in Kentung
State, Myanmar, as well ,as Northern Thailand, Laos, and Vietnam. They are
members of the Lolo branch of the Tibeto-Burman linguistic group, and it is be
lieved that they (along with most other Hill Tribes) originated from the Tibetan
plateau and slowly migrated east and south toward their present homeland. The
exact date of the Akha's first entry into Thailand is not known, but it is probably
as recent as the late 1800s. Most Akha in Thailand live in Chiang Rai province, and
their total population is in excess of 48,500 people living in 258 villages (Tribal Re
search Institute 1995).
The majority of Akha in Thailand still prefer to live along the mountain ridges
at approximately 1,000 m elevation where they can practice a traditional method
of shifting cultivation. They grow dry rice, millet, corn, ginger, peppers, peanuts,
and a variety of other vegetables for consumption and sale. Recently, due to Thai
governmental efforts to halt migration and to eradicate opium production, many
Akha now tend sustainable fruit orchards, tea plantations, and cabbage farms.
Crop production is often inadequate for consumption needs, and people supple
ment their diet by gathering wild plants and hunting, and also by occasionally
working for wages as farm laborers. Hunting and gathering is not as common as
it was in the past. They raise a variety of domestic animals for use in their many
sacrifices and ceremonies, as well as for sale. These include many varieties of
fowl (chicken, duck, turkey), in addition to cattle, water buffalo, pigs, goats, and
dogs. The Akha are pantheists who place great emphasis upon ancestor worship,
spirit placation, and rice fertility rituals.
Each Akha village is an independent political unit composed, ideally, of mem
bers of at least three patricians (Lewis 1992:208). The most basic social unit is the
extended fainily, with the oldest male in each family acting as the head of that
1 45
Michael]. Clarke
family. He functions as family priest in regular offerings made to the ancestors.

The patrilineages (groups of related extended families within a village) mediate
all relationships involving kinship ties, marriage, residential patterns, and rights
of succession.
Although the Alffia are nominally egalitarian, there are institutionalized ad
ministrative positions in each village. A description of these positions is germane
to this discussion because the people who hold these offices are key players in the
feasting complex. It is their favor which is often courted in feasts held by less in
fluential persons, and they themselves are very active feast givers because their
positions are very often based on proof of spiritual potency which is demon
strated through economic success. Oftentimes, their major role in office is to be
the focal point of lineage-wide or village-wide feasting.
Foremost among these administrators is the Dzuma, or Village Founder
Leader (Kammerer's [1986] term) . Each Alffia village must have a Village
FoundercLeader, and his role is primarily a religious one. He is responsible for all
village-wide religious events, of which there are a great many. However, in Alffia
culture, the realm of the sacred is not clearly distinguished from that of the pro
fane, and consequently, it is often the Village Founder-Leader's duty to arbitrate
in disputes, to fine people for social infractions, and to validate decisions made by
the council of elders regarding secular issues. He is considered the 'Father of the
Village,' which has both familial and authoritarian connotations. Ideally, the Vil
lage Founder-Leader is a ritually pure man who was the first settler in a new lo
cation. Once his position is established, it is considered hereditary. However, it is
not unknown for men to be removed from this office by the assembled village
elders. If the Village Founder-Leader is a strong-willed man with many relatives
in his village, he can be extremely powerful and influential. He is always invited
to feasts as it is in everyone's best interest to court his favor.
There are two more quasi-political positions that are recognized by the Alffia ,
both of which are religious in nature. The first is a ritual specialist called a Bu
Moe. These men are Ritual Reciters (Kammerer's [1986] term), and they are re
sponsible for performing rites for individuals in times of sickness, death, and
other such matters relating to the spirit world. This position carries a great
amount of respect, and in any village where a Reciter resides, he will be an influ
ential man. Furthermore, almost all religious services that he provides involve
the sacrifice of animals and a subsequent feast. He is a key player in the Alffi a
feasting complex, and it is often through using his sacred services that an Alffia
family can host a feast that will have significant social consequences in the secular
realm of village politics.
The third official vill age position is that of the Baji, the village Blacksmith. In
traditional times, the Blacksmith would have been an important person to have
1 46
A K H A F E A ST I N G
in a village, and thus, he was treated with great respect and ritualized deference.
Today, however, the Akha import their tools from Thai market towns, and the
role of the Blacksmith has been reduced to its ceremonial minimum. Neverthe
less, he is still an influential man, and in terms of Akha cosmology, he is next only
to the Village Founder-Leader in importance.
All villages also have a less important political headman who tends to deal with
more secular village-wide issues such as relations with the Thai government. The
headman is usually appointed by a council of elders in whose hands the real
power lies. The elders are all of the men in the village over the age of about fifty.
They, as grandfathers, are the paternal heads of multiple households, and they
each look out for the interests of their respective families in terms of the village
dedsion-making process. For example, when the village must decide whose land
to appropriate in order to build a new school, the elders will hold council and
make the final decision. In the villages in which I have conducted research, I have
noted that certain elders will form coalitions and essentially dominate village pol
itics via their joint control of the majority of the population. These coalitions are
formed and maintained through vigorous and ongoing feasting between the par
ties involved. The elders are also the figureheads of any individual's main support
group, the lineage. As such, their households are often the location of lineage sol
idarity feasts, and it is through the elders and their indulgence that junior mem
bers may move up the ranks of influence and security within the lineage. Elders
play a pivotal role in the feasting complex, and, by tradition, all elders are wel
come to attend any feast that is given within the village (although they usually
only attend those feasts given by friends and relatives). This tradition serves to il
lustrate the dynamic relationship between sociopolitical power and feasting in
Akha society.
T H E O R E T I C A L B A C KG R O U N D A N D D I S C U S S I O N
The Akha are a transegalitarian society, as discussed by Hayden (Chapter 2). This
means that, for the most part, their society maintains an egalitarian ethos, but
that there are sizable differences in wealth and power between individuals and
family groups. These material-resource and social-power inequities result in the
emergence of subtle class-like distinctions. However, the nature of the resource
base of many traditional transegalitarian societies (which will be discussed
presently) generally does not allow social-class distinctions to develop fully, and
there are many instances, of which Leach's (1956) study of the Kachin is the sem
inal example, where there is actually a regression toward greater social equality
and individual self-autonomy. Table 5.1 gives some examples of the material in
equities that exist between three of the richest men and three of the poorest men
in my main study village of Mae Salep.
1 47
Michael]. Clarke
TABLE S . 1
Com parison of Rich and Poor Fa m i l ies
House Land Sales

z z
Family Heads No. in Family Producers (m ) (m ) Buffalos ($US)
Rich Men
Larche (a) 5 121 . l 60,800 12 340
Larche (b) 27 15 236.25 84,800 0 3, 600
Larche (c) 7 4 1 75 43,200 19 4,504
Poor Men
Larche (x) 5 4 34.8 20,800 0 0
Labu (y) 5 3 70 12,800 0 0
Larche (z) 3 2 77 1 1 ,200 0 0
Note: "Producers" is the number of workers in the home. Buffalos are included because they are the pinnacle of
wealth accumulation. "Sales" only regards animal sales for the years 1995 and 1996. Latche (b) has no buffalos only
because he sold a large herd the year previously; and he owns many cattle.
In Akha society; socioeconomic inequity; competition, and aggrandizing be

havior emerges primarily between the lineages and clans as a whole, rather than
between individuals, and this, more than anything, serves to define transegalitar
ianism in the Akha case. Individual households in Akha society; even if they are
rich, do not become socially dominant in the way that individual households be
come dominant in more complex transegalitarian or stratified societies where
power is more easily centralized. Particular lineages, on the other hand, can and
do become socially dominant. For the purpose of this study, I follow Hayden and
Maneeprasert (1996) and define clans as patrilineally related families living in sep
arate communities, and lineages as patrilineally related families (smaller seg
ments of clans) living in the same community.
Clan/lineage relations among the Akha are structured in a systematic way
through long tradition and cultural norms. They are both the explicit village so
ciopolitical structure, and the implicit supraregional sociopolitical structure
(Kammerer 1986). Feasting is integral to the dynamics of clan/lineage relations,
and as such, is an "institution" (in the same sense as parliament, or the demo
cratic system) that can be used and manipulated by individuals and groups to ef
fect changes in the sociopolitical sphere.
The nature of the Akha economy affects both social structure and feasting
practices. Hill Tribe swidden horticulture allows for a relatively high population
density; compared to hunting and gathering, and consequently, presents unique
organizational challenges. Hill Tribe villages, which may have up to several hun
dred people in residence, are, of necessity; segmented into various allied factions.
In Akha society; the prime dividing line seems to be lineage I clan membership.
1 48
A K H A F EASTI N G
However, because the ownership of swidden fields (in traditional times) was only
temporary, and families eventually had to change residence once the land was no
longer fertile, the long-term (multigenerational) monopolization of the resource
base by one allied group was not possible. Furthermore, because of the availabil
ity of new land, socially weaker and subdominant groups always had the option
of moving out of a village and farming in a new locale. Consequently, the forma
tion of a wealthy land-owning gentry class was not possible. No one could de
pend on long-term economic security, and even mere economic survival was a
precarious and transitory thing for every family. Hence, there was, and still is, a
need for a structured and dependable life-crisis support network.
All individuals in Akha society are aware of the precariousness of their eco
nomic situation, and most people realize that they may have to depend on the
help of friends and family when they are faced with hard times (i. e . , crop failures,
sickness, disputes, or other misfortunes) . Furthermore, at certain points in the
agricultural year it is necessary to organize cooperative work groups (for field
clearing, for example), which are essentially based on friends and family recipro
cating labor. It is important to maintain these relationships.
There are many ways in which a drastic change of fortune can occur. The litera
ture makes ample reference to bandits and armies robbing entire villages (i.e.,
Lewis 1969, 1970; Kammerer 1986); rich individuals are sometimes held for ransom;
crops can receive inadequate or excessive rain; and disease or insects can destroy en
tire fields. In the past, famine was not unknown among the Hill Tribes. For in
stance, some of the richest families in Sam Soong village lost almost all of their pigs
to disease in the past two years; one family lost nearly So pigs (see Fig. 5.1).
The Akha resource base is in some ways similar to that of a generalized hunt
ing and gathering people's resource base: it is fluctuating, to some degree un
predictable, and not strictly nucleated. In other ways, however, it is similar to
the resource base of more complex societies: it is labor intensive; increased ex
traction labor does not make resources vulnerable to overexploitation but in
creases total output; the economy allows for the production of occasional
surpluses; not all land is of the same quality; and land that is cultivated is effec
tively private property.
On the basis of the field data collected, it is clear that to cope with these
unique resource qualities, the Akha employ a mixture of hunter-gatherer and
complex-society social strategies, especially feasting techniques. The typical
hunter-gather problems (the precariousness and fluctuations of the food supply)
are dealt with by alliance formation and mutual assistance created in part by
feasting within the clan and lineage, and to some extent, within the vill age. There
is little or no aggrandizing or competitive behavior at these feasts. Some exam
ples of intralineage solidarity feasts are the thirteen annual family ancestor offer
ings, the sickness-curing feasts, and newborn naming feasts.
1 49
Michael]. Clarke
250
l/J 200
Cl
·a.
- 1 50
0
... 111111 Recently Dead
Q)
.c 1 00
E
:::s 0 Currently Owned
z 50
0
Q) .c .c ctS
.c Q) Q; Q;
0 :::s � ·:;, .c
ctS c E
i:O
0
E
:::J
E
Q) :::J 5
i:O Q) :::J >-
.c Cl..
() :2
Figure 5.r. Compound bar graph illustrating the number of pigs that each clan in Sam
Soong village owns, and how many of those pigs recently died of disease.
The typical complex-society problems that the Akha encounter (the competi
tion for labor, spouses, land, and political control in general) are dealt with to a
great extent by feasting between the clans and lineages (e.g., feasts that foster al
liances between groups such as weddings, or which are grandiose enough to val
idate a claim to power such as funerals). These larger feasts serve to advertise the
wealth and productivity of the host lineages. These kinds of feasts exhibit some
mildly aggrandizing attributes (such as food delicacies and an abundance of alco
hol and narcotics) and have the emic purpose of being impressive. The more os
tentatious Akha funeral feasts fall into this category.
Advertising household and/ or lineage success is adaptive and advantageous
for at least two reasons. First, because it attracts and maintains a large and reli
able work force in the form of healthy, hard-working wives from influential fam
ilies marrying into the family and relatives moving into the village. Since village
political matters (especially litigations) are often decided by the votes of the elder
men who represent the various lineages in the community, it stands to any per
son's advantage to have as many relatives as possible in their village. Further
more, hard-working wives increase a farm's production level. Second, success
advertisement is advantageous because it encourages deference. Wealthy men
are seen as both blessed and intelligent. Their opinions hold more weight than an
average person's.
It is more appropriate to conceptualize the above division of feasting activity
1 50
-hunter-gather-type solidarity feasting and complex-society competitive feast

ing-as two degrees, rather than as two categories of feasting activity. They are
not mutually exclusive divisions. A wedding feast, for example, will tend to be
promotional, and therefore competitive in nature, but will also always foster a
certain degree of intraclan solidarity and alliance enhancement. A poor family's
wedding will emphasize intralineage solidarity (most guests being from within
the lineage) whereas a rich family's wedding will emphasize the promotional
aspects (many guests being from other lineages). This is why it is so difficult to
create feasting classification categories based on function alone; although each
feast category may share a structural homogeneity (i.e., all weddings have a cer
tain prescribed ordering of ritual and food consumption), most feasts have a vari
ety of functions, which are all manifested to varying degrees. The degree to
which any function is made manifest, whether it be social bonding, wealth adver
tisement, or outright debt fostering, is contingent upon many factors. The pri
mary factors seem to be the host's wealth, relative sociopolitical position, and
ambitions.
AKHA F EAST TYPES

Akha life is rife with feasting. Some feasts are small, involving only a few people
and the ritualized consumption of a very small amount of food, whereas other
feasts are enormous, involving many hundreds of people and the consumption of
several water buffalos, pigs, and chickens over a period of weeks. There is no sig
nificant secular or religious event that is not accompanied by a feast of some
kind, except for possibly a few political meetings. For instance, the Akha hold
feasts for purifications, divorces, menopause, field ceremonies, and all initiations
of office.
Feast quality (i.e., expensive, overtly public, or modest and subdued) is in most
cases a function of the distance of the guests (genetically, geographically, and
physically) from the host. Greater, more intense, and more symbolic measures
are required to communicate social messages to people who are farther apart. In
most cases, it is only the larger types of feasts, such as weddings and funerals,
that have guests participating in them who are not in some way already intimate
with the host.
A large variety of occasions merit a feast in Akha society. What all feasts have
in common, however, is that there is always some sort of meat consumed. Be
cause the daily Akha diet consists mostly of rice and vegetables, meat of any kind
is considered a treat. In this sense, all feasts contain a desirable delicacy and enti
tle the host to a certain degree of gratitude and respect.
Feasts are generally held inside a home. The young men do most of the cook
ing, except for the rice, and the middle-aged men act as servers. Women and men
151
Michael]. Clarke
sit in opposite halves of the house, and both male and female elders are given
seats of honor upon a raised sleeping platform. There is very little ostentation
displayed at Akha feasts by the host. It is never considered appropriate to brag or
to boast. However, some of the larger Akha feast "types" can be characterized as
promotional in nature because they contain a great variety of food delicacies, al
cohol, and other desirable goods. It is often at these larger feasts where the
women will don their full silver ornamental headdresses and accompanying ban
gles, which are a family's main repository of surplus wealth (Fig. 5.2). They liter
ally display their family's wealth upon the top of their heads (although they see it
simply as formal wear).
Table 5.2 is a summary of most of the important and more common Akha
feasts. It is in no way exhaustive (for a more detailed account see Clarke 1998).
Many of the feasts listed, such as purification and penalty feasts, have a wide vari
ety of possible forms and functions. Some purifications are simply small cere
monies meant to cleanse a holy place, others are larger events that relate to
specific problems involving an individual, whereas still others operate more in
the social realm and are concerned with social events such as divorce and adul
tery. Furthermore, some of the larger feast types, especially the funeral, are actu
ally a series of small and large feasts spread over a period of time.
Figure 5.2. Elder lineage women and peer guests feasting at the women's table of honor
during an Akha marriage feast. Note the elaborate and valuable headdresses and bead
work worn for this occasion. (Photo by B. Hayden)
1 52
AKHA F EASTI N G
TABLE S.2
Akha FeastTypes
Cost
Feast Type Attendance (US$) Function
Ancestor offering 3-10 $5 Family solidarity

Newborn naming 3-7 $5 Family solidarity
Sickness curing 10--25 $20--$ 5 0 Solidarity I allies
Butchers 5-15 $ 1 0--$ 1 5 Solidarity I allies
Workmen's 5-30 $ 1 0--$ 5 0 Acquire labor
Penalty 5-1 0 $2-$ 150 Social control
Purification 5-50 $5-$300 Solidarity I control
Gate rebuilding 5-1 0 $ 1 0--$ 3 0 Village solidarity
Lords of Barth 5-10 $ 1 0--$ 3 0 Village solidarity
Harvest Bach household $5-$20 Village solidarity
New Year's Bach household $5-$75 Village solidarity
Dzuma's annual (?) 1 0-50
Wedding 20-200 $25-$500 Promotion
New house 20-300 $25-$600 Promotion
Menopause (?) 20-1 00 $ 1 0 0-$250 Promotion
Funeral 15-1,000 $25-$ 1 ,000+ Extreme promotion
The feasts shown in Table 5.2 have been arranged in loose order of size, start
ing from the smallest and most intimate at the top, and :finishing with the largest
and most public at the bottom. The first two feasts, ancestor offerings and new
born naming, are small household events that are held fairly frequently, the an
cestor offerings being made thirteen times a year. A chicken, along with some
rice, whiskey; and tea is consumed at these events.
Butchers', curing, and workmen's feasts are medium-sized events that operate
mostly at the lineage level. Curing feasts can be given to people, usually children,
whether they are ill or not. Curing feasts are an opportunity for lineage elders to
meet, reassert their spiritual potency (as part of the ritual), and offer the younger
clan members an opportunity to ingratiate themselves (Fig. 5.3). Butchers' feasts
are given to the elders when an animal (almost always a pig) is slaughtered for
sale. Theoretically, all village elders are welcome, but in my field observations, I
have noted that only those elders related to, or closely allied to, the hosting
butcher attend. Workmen's feasts are purely secular. They are a form of recogni
tion for men who have helped with a building project (usually a house). These co
operative work feasts generally include lineage members, and they are
characterized by the serving of dog meat (a delicacy not appropriate for ritual
consumption) and copious amounts of whiskey.
1 53
Michael]. Clarke
Figure 5.3. Lineage elders (the three men at the left), a Ritual Reciter (center rear), and
the younger lineage host (right, holding child) at a curing feast. One of the lineage elders
is tying a string around the child's wrist for spiritual protection as the Reciter looks on.
(Photo by B. Hayden)
Purifications and penalty feasts have a wide range of forms. Most are small in
size, and they are held infrequently. These feasts are a means of enforcing con
formity to social norms and of asserting the power of the ruling elder males. It is
the council of elders who decide the size, and hence the cost, of penalty and pu
rification feasts.
There are five annual village-wide feasting events: the ceremonial gate rebuild
ing, the offering to the Lords (i.e., Spirits) of the Earth, the harvest festival (also
known as the swinging festival), the New Year's festival, and the annual feast
given by the Village Founder-Leader (Yo la la). The first two annual feasts, the
gate rebuilding and the offering to the Lords of the Earth, are small feasts that the
elders celebrate at the location of the ceremonial gate and the village water
source, respectively. The expense for these feasts comes out of a village fund. The
harvest and New Year festivals are typified by each family feasting in their own
home, with possibly the elders invited to a special meal at the Village Founder
Leader' s house. For the harvest festival, the villagers will purchase shares of a
1 54
water buffalo, which they will take away and consume at home. For the New
Year festival they generally eat pork and drink copious amounts of whiskey. Some
households will have many guests visiting them during the New Year festival and
may be inclined to spend a fairly large sum of money. The annual Village
Founder-Leader's feast is considered a payback to the villagers for the support
they have given him, and it is also a time to thank the special protective spirit of
the Village Founder-Leader.
The last four feasts outlined in Table 5.2, wedding, new house, menopause,
and funeral feasts, tend to be the largest and most costly for individuals. It is at
these feasts where obvious delicacies, such as candy; betel nut (a narcotic), and
beer start to appear on a regular basis. It is not uncommon (in recent years) for
printed invitation cards to be delivered to guests. All of these characteristics point
to the fact that these feasts are intended to be grandiose and, hence, promotional.
Wedding feasts can vary greatly in size, depending upon the wealth of the
host, but in general they are quite expensive (relative to income) and last for three
days. One or two large pigs and several chickens are usually consumed. New
House feasts are very similar to wedding feasts. They tend to have more guests
but are shorter in duration, lasting for only one full day and night. Menopause
feasts are events that initiate women into a new ceremonial role, that of the
White-Skirted Woman, and they are not given to every woman, only to a select
few who qualify and can afford the necessary ceremonial sacrifices.
Menopause feasts can be very costly: several large pigs, numerous chickens, and
in some areas, goats are required in order to perform the ceremony. If the cere
mony is widely publicized, then the family may also have to kill a buffalo in order
to feed the many guests who will arrive. There is no compulsion to perform this
ceremony once menopause is reached, and, in fact, Kammerer (1986) has noted
that it is usually the husband who insists that the ceremony be completed. It is very
honorable to have a White-Skirted Woman as a member of one's household; it is
also thought to bring great blessing and fortune, but more importantly; from an
ethnoarchaeological perspective, it is an opportunity for a family to display its suc
cess, wealth, and desirability of affiliation to the people of the general region.
Akha funerals are by far the most emically important of all ceremonial events.
The Akha people attach great religious importance to them, but funerals also ful
fill very practical social functions. They are a venue at which successful and pow
erful households can reassert their claim to distinction and deference, and they
are an avenue along which lineages and clans can compete for status and recogni
tion. A wealthy family's funeral for an elder will consist of the sacrifice of 1-7
water buffalo, 5-20 pigs, and numerous other animals. The family will pay the
presiding ritual specialist and his assistants in silver. Funerals are incredibly ex
pensive relative to the income of the people. The cost of a funeral can easily be
1 55
Michael]. Clarke
the equivalent of several years' income. However, donations for the funeral very
often are collected from a wide range of clan and lineage members.
A F EASTI N G EXAM PLE

I would now like to relate the events surrounding a specific Akha wedding feast,
which will illustrate some of the points that I 1have made thus far. The wedding
took place in an Akha village called Mae Salep, situated near the Myanmar bor
der. There are nine different clans represented in Mae Salep, and the total village
population is 267. One clan, the Latches, forms the majority of the population,
and it is this clan that essentially governs village life. The Latche families have
formed a close alliance with the second most populous clan, the Labus, and to
gether the elders of these two clans dominate village politics. This alliance was
created by, and is maintained by, reciprocal feasting between the various families
within these two clans.
This particular wedding feast was held for the son of the most prominent Labu
man in the village, Mae Salep's Ritual Reciter. This is a very traditional household,
and this wedding was in many ways a very typical example of a large Akha feast.
The feast took place in the home of the father of the groom and lasted for
three days. On the day preceding the actual celebrations, the groom and the bride
performed a brief ritual involving the passing of an egg back and forth between
them. Although this ritual is considered the actual initiation of the marriage, it is
given relatively little importance. No one is expected to pay attention to this cer
emony; instead, the members of the household were busy making the prepara
tions for the feast. The feast itself is considered the public announcement and
recognition of the union. This illustrates the role that ritualized public consump
tion of food plays in the formation of the Akha community structure.
In order to feed the guests, the father had to provide a large, fat, corn-fed, cas
trated boar, as well as several chickens. This in itself is a considerable economic
investment, but there were also large expenditures made on other delicacies such
as whiskey, candy, cigarettes, and on staples such as rice, cabbage, and tea. The
various meals at the three-day event were truly excessive in terms of the amount
and variety of food when compared to daily fare. Meat itself is a delicacy in Akha
society. The wedding also provided people with an opportunity to gamble and to
smoke opium.
Over the three days, more than 75 people attended, although many did not
participate for the full duration. It is significant to note that elders from other vil
lages came to participate. These men were representatives of related lineages liv
ing in nearby villages, and it was important to maintain contact with them
because the Akha generally relocate several times during their lifetime, fre
quently to relatives' villages (Kammerer 1986). The host also made a point of
inviting the generally disliked Thai school administrators. These men look down
1 56
on the Akha because o f their poverty; but the Akha can do little about this be
cause they often depend on the head administrator's cooperation in regards to
matters of citizenship and land rights. If the wedding had been considered a
purely festive event, then the host would probably not have invited these Thai ad
ministrators, but because these men are very powerful, the host used the wed
ding as an opportunity to court their favor.
In order to cook such cqpious amounts of food, the family had to have (or bor
row) a large assortment of cooking and serving vessels. These included one very
large wok 50 cm in diameter, and four pots, one 55 cm in diameter, one 45 cm, and
two 30 cm in diameter (Fig. 5.4). They also had about 50 serving bowls, 25 drink
ing glasses, and 27 blue ceramic whiskey cups. What should be noted is that these
figures are far in excess of the average Akha family's daily cooking/ serving vessel
assemblage. Normally, a family of this size will need only one or two woks of
about 30 cm in diameter, and two or three pots ranging from 15 cm to 30 cm in di
ameter. On average, they will own ten to fifteen serving vessels and around five
drinking glasses. This difference is significant because the relative size and num
ber of cooking and serving vessels is a strong archaeological indicator of feasting
Figure 54 Large cooking pots and woks used for preparing large
quantities of food for large feasts. (Photo by M. Clarke)
1 57
Michael]. Clarke
activity. My research has determined that families who are actively engaged in
ongoing feasting will own a disproportionate amount of these vessels, and that
the remains of these vessels will be deposited in the middens surrounding their
house.
The wedding feast served as a solidarity-reinforcing event on many levels.
Firstly, in terms of the greater community support group, all of the elder males
representing the various lineages in the village were in attendance, with the ex
ception of those who had converted to Christianity and thus had opted out of the
traditional village cooperative unit. Their absence was very evident, and made a
very clear statement of disassociation. At the lineage level, it is worth noting that
Latche clan members assumed prominent roles as organizers, workers, and
servers in this Labu family wedding, and this was a very symbolic act of friend
ship and cooperation, which is a reflection of the sociopolitical alliance that the
two clans have formed in Mae Salep.
On an individual level, this particular feast served to integrate the young
groom and his new household into the greater lineage support group, and it also
served as a platform on which his father, the host, could advertise his family's suc
cess and revalidate his position as the village Ritual Reciter. For instance, at one
point in the wedding, the young groom walked among the assembled guests and
shared a small drink of whiskey with the men present. These men then made a
small cash donation to the household. This is considered "begging for blessing,"
and the money is to help defray the cost of the feast and to help the young couple .
set up a household. However, it is also a symbolic act of inclusion within the
greater lineage and, as such, the people are expressing their cooperative ethos
while the young groom is acknowledging his indebtedness to his lineage.
A R C H A E O LO G I C A L I N D I C AT O R S O F F E A S T I N G A C T I V I T Y
I have alluded to a number of archaeological correlates of feasting activity in the
preceding sections, but I would now like to discuss them in a more systematic
fashion. Although the following observations were made in a culturally specific
context, and no ethnoarchaeological generalization can be considered a cross
cultural archaeological truism, the relationships that I present have been struc
tured in a manner that is largely based on the material requisites for feeding large
groups of people simultaneously using traditional technology. These relation
ships are not intended to be a checklist of evidence for archaeologists to consult
when they find similar data at their sites, but rather, it is hoped they will provide
insights that will help archaeologists explain hitherto unexplained phenomena,
as well as being complementary to other data that is used in archaeological analy
ses of feasting. The systematic occurrence of large-scale feasting is evident in the
multi.modal size distribution of cooking vessels in large assemblages such as
those of entire villages (Fig. 5.5).
1 58
30
25
20
>-
u
c
!l 1 5
er
I!!
u..
10
0
8" 1 0" 1 2" 1 4" 1 6" 1 8" 20" 22" 24" 26 (+) "
Size of Woks
70
65
60
55
50
>-
45
g 40
!l 3 5
er
!!! 3 0
u..
25
20
15
10
5
0
6" 8" 1 0" 1 2" 1 4" 1 6" 1 8" 2 0" 22· 24 " 26" 2 8(+)"
Size of Pots
Figure 5.5. Histograms showing the multimodal size distribution of woks and pots in an
Akha village. Sizes of both woks and pots in the 6- to 1 4-inch diameter range represent
normal food-preparation vessels. Modal peaks in the 20- to 26-inch range represent food
preparation vessels for feasts of increasing sizes.
Michael]. Clarke
COO KI N G VESSELS
As already noted for wedding feasts, the size and number of cookmg vessels can
be used as a measure of feasting activity. Vessels greater than a certain size are re
ally only practical to use when one is cookmg for large numbers of people and
large volumes of food must be produced. Furthermore, daily meals generally
consist of a maximum of two to three dishes (excluding rice) though it is not un
common for feasts to consist of five to ten different dishes. In order to be able to
prepare such a wide variety of dishes at the same time, several cookmg vessels
are necessary for each dish. The average Akha household consists of approxi
mately ten people, and a family will usually provide for its daily cookmg needs
with one or two small woks, three or four small pots and one large pot approxi
mately 50 cm in diameter to cook pig food in. In all cases, households that ex
ceeded this norm in the cookmg vessel assemblage did so because they owned, or
borrowed, vessels for feasting purposes.
S E RVI N G V ES S E LS
The relative number of serving vessels per household (or other corporate group)
may be indicative of feasting. It would seem obvious that everyone would have a
functional minimum of these items, but families with excessive numbers must
use them for other purposes. In traditional Akha villages, no families acted as
merchants for such vessels. In some societies, special elaborate serving vessels
have a specific feasting function; however, this is not the case for the Akha. The
only thing that is indicative of feasting for the Akha, in terms of serving vessels, is
their number.
An average Akha family may own 10-15 serving bowls and plates, 1-10 drinkmg
glasses, and 4-6 rice baskets. For some of the larger feasts, hundreds of these ves
sels are required. Not only are more vessels required because of the greater num
ber of guests, but also because of the larger variety of food and drink available.
No one ever owns the hundreds of vessels needed for the very large feasts. In
stead, they borrow what is needed from their relatives. However, those house
holds that engage in regular small- to medium-scale feasting always own an
adequate number of serving vessels for these events, and this is always greater
than the village average.
TROPHY BONES
Very often after a feast, the host will save the jaw or horns of a large sacrificed an
imal for display. It must be remembered that meat is not a daily component of
the diet, and the offering of an expensive animal for sacrifice is reason enough to
gain prestige. The display of water buffalo horns is particularly prestigious, as the
sacrifice of one or more water buffalo at a funeral is a very auspicious event.
1 60
AKHA F EASTI N G
Much more common however, is the display of pig mandibles; some homes can
have as many as twenty displayed on the wall (see Fig. 2.8).
COOKI N G F I RES
The relative size and number of hearths is also a good indicator of feasting activ
ity. This may prove difficult to recognize archaeologically. Families that actively
participate in traditional feasting often have large kitchens with extra hearth
rings and braziers. It is frequently necessary to expand the area of the hearth for
a feast. In the case of the Akha, very traditional homes, with the division between
male and female halves, have discrete ash dumps for each hearth in the home.
This is because women and men maintain their own hearth, and both men and
woman have a separate entrance to the house. Each sex dumps the ash outside of
their respective entrances, indicating the two major cooking areas.
T E M P O R A R Y K I TC H E N S
The construction of temporary kitchens may also be archaeologically recogniza
ble. These would only be needed for very large feasts. Their construction is
sometimes necessary if the existing kitchen is not large enough to accommodate
the feasting preparation. In the Akha case, and I suspect in most other cases, tem
porary or extra kitchens are added on to original permanent kitchens. This was
because most of the implements needed for cooking could be shared between
the two, and also because the preparations are easier to coordinate that way.
G A R B AG E F I R E S
Garbage fires with many feasting remains in them are very indicative of feasting
in the Akha case. Normally, small bits of refuse are tossed away indiscriminately,
or swept through the space in the floorboards. Large amounts of trash are tossed
away, downslope, from the house into a toft area. There are only two occasions
that seem to merit the effort of a garbage fire. One is the generation of a large
amount of waste produced from some kind of plant-processing activity, husking
corn for example. These types of garbage fires, because of the singularity of their
contents, produce a very homogeneous type of ash. The second occasion where
a garbage fire is necessary is following a large feast, and the ash from these fires
always contains a very large variety of contents, from bone and food remains, to
whiskey bottles and broken ceramics. Garbage fires were always lit outside, in
front of the house.
FA U N A L R E M A I N S
The Akha predilection for dicing up their meat completely, bones and all, created
some problems in terms of the analysis of faunal remains. Pigs and chickens are
1 61
Michael]. Clarke
almost completely consumed. The people chew and swallow the well-cooked
bones, and most of those that they cannot manage to ingest are generally eaten
by the numerous dogs and pigs that are always at hand. This consumption pat
tern, of course, must change when people are eating large animals such as cows
or water buffalos. Although I have never had the opportunity to witness a water
buffalo being butchered, I am sure that they would not have attempted to com
pletely dice up and eat such large, thick bones.
The remains of a small cow that was eaten at a new house feast provided some
minor insight into large-mammal taphonomic processes. It was only at these large
feasts that an entire animal was ever eaten all at once and in one location. Conse
quently, these are the only times when large portions of the skeleton of an animal
are ever deposited all in one dumpsite. In all other Akha cases, and I suspect in
many tribal societies everywhere, portions of butchered animals are distributed
among friends and family members throughout the village. Consequently; the
bones are deposited in different middens throughout the village.
It is possible that the Akha might choose to butcher and preserve (i.e., smoke
or salt) a large animal, and then deposit its bones all in one dumpsite, but I have
never seen this done (although Lewis [1969] claims that in Burma they salt and
dry large game animals). The practical necessity of immediate consumption
aside, it also serves a greater social function to divide up the meat. In one in
stance, a family caught a gopher and rather than just eating the small animal
themselves, they divided it up into quarters and spread them around to their rel
atives in the village.
Finally; faunal remains in human graves have sometimes been cited as evi
dence of feasting. In the Akha case, this seems to hold true. According to Lewis
(1969), at different points in the funeral, feasting food is offered to the deceased,
and some sacrificed animals are hung over the coffin. They are subsequently
buried with the body.
S U M M A RY
I believe that feasting is a social phenomenon of great significance that often
leaves recogruzable traces in the archaeological record. There is a growing body
of evidence that suggests that promotional or competitive feasting may have
been of central importance in the initial transition of early humans from simple
hunter-gatherers to complex horticulturalists (see Bender 1978; Hayden 1990,
1992a, 1992b; Keswani 1994; Liden 1995). The ethnographic record is able to lend
insight into the various ways in which feasting can play a dynamic role in the cre
ation and maintenance of social structure.
Studying the Akha, as one specific example of complex horticultural-society
feasting practices, has provided us with some rudimentary insights into the role
1 62
that feasting can play in traditional nonmoney economies. I t is evident that in

Akha society, feasting functions primarily as a social mechanism that creates and
maintains a life-crisis support group, bringing people together in solidarity
enhancing events. Similarly, feasting can be used as an articulation point between
allied corporate groups (in this case clans and lineages) . The feast provides po
tential allies with a venue to express their desire to ally and to prove the sincerity
of that desire through gifts of highly desired food, drink, and narcotics. Feasting
can also be an avenue for people to jockey for power positions within their
greater support group.
Related to feasting as a way of maintaining a support group is the Akha's use
of feasting as a form of wealth and success advertisement. Promoting the eco
nomic success of the household allows families to attract and hold high-quality
labor and spouses, to create and maintain trading contacts, and to engender so
ciopolitical deference within their community. In fact, it is considered neces
sary for officeholders (those who make special claims based on spiritual
potency) to constantly revalidate their social position through the hosting of
regular feasts. In Akha culture, economic success is seen to be based on spiri
tual purity and power (similar to the idea of 'mana' in the South Pacific) . Bless
ings are given by the ancestors and are a sign of sacred connectiveness.
Therefore, people who make claims to official political power positions must
be able to demonstrate their ritual potency through economic success. Similar
concepts pervade many other Southeast Asian cultures and feasting complexes,
as argued by Friedman (1975).
Much work still needs to be done in order to advance our understanding of
feasting among the Hill Tribes. There are several o ther groups that inhabit the
mountainous regions of mainland Southeast Asia that also have highly developed
feasting complexes. As yet, little is known about the feasting of these other
groups. The practice of feasting is closely tied to the economics of production,
and although a great deal of data have been gathered on the specifics of Akha
feasts, much more data is still required on the economy which makes this feasting
possible.
REFERENCES
Alting von Geusau, Leo
r983 Dialectics of Al<hazan: The Interiorization of a Perennial Minority Group.
In Highlanders of Thailand, edited by Jon Mckinnon and Wanat Bhruksasre,
pp. 243-277. Kuala Lumpur: Oxford University Press.
Appadurai, A.
r986 The Social Life of Things: Commodities in Cultural Perspective. Cambridge: Cam
1 63
Michael]. Clarke
Arnold, ).
1993 Labor and the Rise of Complex Hunter-Gatherers. Journal of Anthropological
Archaeology 12:7.5---II9.
Blitz, John
American Antiquity 58 (2): 8CJ--96.
Bender, B.
1978 Gatherer-Hunter to Farmer: A Social Perspective. World Archaeology
10:204-222
Blau, P.
1964 Exchange and Power in Social Life. New York: John Wiley.
Blumberg, Rae Lesser
1978 Stratification: Socioeconomic and Sexual Inequality. University of California, San
Diego: Wm. C. Brown.
Cashdan, E.
1980 Egalitarianism among Hunters and Gatherers. American Anthropologist
82:n6--1 20.
Clarke, Michael).
1998 Feasting among the Akha: An Ethnoarchaeological Case Study. M.A. thesis,
Department of Archaeology; Simon Fraser University, Burnaby, B.C.
Cooper, Robert
1984 Resource Scarcity and the Hmong Response. Singapore: Singapore University
Press.
Crabtree, Pam
1990 Zooarchaeology and Complex Societies: Some Uses of Faunal Analysis for
the Study of Trade, Social Status, and Ethnicity. In Advances in Archaeological
Method and Theory, vol. 2, edited by M. Schiffer, pp. 155-160. Tucson: Univer
sity of Arizona Press.
Dietler, Michael
Case of Early Iron Age France. Journal of Anthropological Archaeology 9:
352-406.
Douglas, M.
1984 Standard Uses of Food: Introduction. In Food and the Social Order, edited by
M. Douglas, pp. 1-39. New York: Columbia University Press.
Earle, T.
1997 How Chieft Come to Power: The Political Economy of Prehistory. Stanford: Stan
ford University Press.
Falvey, Lindsay
1977 Ruminants in the Highlands of Northern Thailand. Thai-Australian High
land Agronomy Project, Tribal Research Institute, Chiang Mai University,
Chiang Mai.
Fiddes, Nick
1991 Meat: A Natural Symbol. New York: Routledge.
1 64
Friedman, Jonathan
ogy, edited by M. Bloch, pp. 161-202. London: Malaby Press.
Hanson, Rev. 0.
1913 The Kachins: Their Customs and Traditions. New York: American Baptist Mis
sion Press.
Hayden, Brian
1992a Ecology and Complex Hunter I Gatherers. In A Complex Culture of the British
Columbia Plateau, edited by B. Hayden. Vancouver: University of British Co
lumbia Press.
1992b Models of Domestication. In Transitions to Agriculture in Prehistory, edited by
Anne Birgitte Gebauer and T. D. Price, pp. 273-299. Santa Fe: Prehistory
Press.
1995a Pathways to Power: Principles for Creating Socioeconmnic Inequalities. In
Foundations of Social Inequality, edited by T. Douglas Price and Gary Feinman,
pp. 15-85. New York: Plenum.
1995b A New Overview of Domestication. In Last Hunters-First Farmers, edited by
T. D. Price and A. Gebauer, pp. 273-299. Santa Fe: School of American Re
search Press.
An Interdisciplinary Perspective, edited by P. Wiessner and W Schiefenhi:ivel,
1998 The Dynainics of Wealth and Poverty in the Transegalitarian Societies of
Southeast Asia. Unpublished paper, Dept. of Archaeology, Simon Fraser Uni
versity; Burnaby, Canada.
Hayden, Brian, and Ralana Maneeprasert
1996 Feasting among the Akha in Thailand. Unpublished report to the National Re
search Council of Thailand.
Janse, Olav
1944 The Peoples of French Indochina. Sinithsonian Institution War Background
Studies Number Nineteen, Washington, D.C.
Kammerer, Cornelia
1986 Gateway to the Akha World: Kinship, Ritual, and Community among High
landers of Thailand. Ph.D. dissertation, Anthropology Department, Univer
sity of Chicago, Chicago.
Keswani, Priscilla Schuster
1994 The Social Context of Animal Husbandry in Early Agricultural Societies:
Ethnographic Insights and an Archaeological Example from Cyprus. journal
of Anthropological Archaeology 13:255-277.
Kim, Seung-Og
1994 Burials, Pigs, and Political Prestige in Neolithic China. Current Anthropology 35
(2): 119-131.
1 65
Michael]. Clarke
Kirsch, A. T.
r973 Feasting and Social Oscillation: A Working Paper on Religion and Society in Upland
Southeast Asia. Southeast Asian Data Paper No. 92. Ithaca: Cornell University,
Dept. of Asian Studies.
Leach, Edmund
r956 Political Systems of Highland Burma. Boston: Beacon Press.
Lefevre, E.
r995 Travels in Laos: The Fate of the 'Sip Song Pana' and 'Muong Sing' (r894-r896).
Bangkok: White Lotus.
Lewis, Paul
r969 Ethnographic Notes on the Akhas of Burma: Vols. r-2. Human Relations Area
Files, New Haven, Connecticut.
r970 Ethnographic Notes on the Akhas of Burma: Vols. 3-4. Human Relations Area
Files, New Haven, Connecticut.
r99r Thai Hill Tribes Phrase Book: Lonely Planet Language Survival Kit. Victoria, Aus
tralia: Lonely Planet Publications.
r992 Basic Themes in Akha Culture. In The Highland Heritage: Collected Essays on
Upland North Thailand, edited by Anthony Walker. Singapore: Double-Six
Press.
Lewis, Paul, and Elaine Lewis
r984 Peoples of the Golden Triangle. London: Thames and Hudson.
Liden, Kerstin
r995 Megaliths, Agriculture, and Social Complexity: A Diet Study of Two Swedish
Megalith Populations. Journal of Anthropological Archaeology I4:404-4I7·
Maneeprasert, Ralana
r988 Ban Mae Salaep Akha: Chiang Mai. Report on file at the Tribal Research In
stitute, Chiang Mai.
Marshall, F.
r993 Food Sharing and the Fauna! Record. In From Bones to Behaviour: Ethnoarchae
ological and Experimental Contributions to the Interpretation of Fauna! Remains,
edited by J. Hudson, pp. 228-246. Carbondale: Southern Illinois University.
McGuwere, R. , and R. Paynter
r99r The Archaeology of Inequality. Oxford: Blackwell.
Schrock, John
r970 Ethnographic Study Series: Minority Groups in Thailand. Washington, D.C . : Cul
tural Information Analysis Center, Center for Research in Social Systems.
Schubert, Bernd
r986 Proposals for Farming Systems-Oriented Crop Research of Wawi Highland Agricul
tural Research Station in Northern Thailand. Center for Advanced Training in
Agricultural Development: Technical University of Berlin, Berlin.
Smith, M. E.
r987 Household Possessions and Wealth in Agrarian States: Implications for Ar
chaeology. journal of Anthropological Archaeology 6 (4): 297-335.
1 66
Tai Landa, Janet

1994 Trust, Ethnicity, and Identity: Beyond the New Institutional Economics of Ethnic
Trading Networks, Contract Law, and Gift-Exchange. Ann Arbor: University of
Michigan Press.
Tribal Research Institute
1995 The Hill Tribes of Thailand: Fourth Edition. Chiang Mai, Thailand: Technical
Services Club, Tribal Research Institute, Chiang Mai University.
Visitpanich, Theera, and Lindsay Falvey
1980 A Survey of the Highland Pig Industry. Thai journal of Agricultural Science
13:259-267.
1 67
6
P O LYNESIAN FEAST I NG I N
E THNOH ISTOR I C , E THNOGRA PH I C ,
AND AR CHAEOLOG I CA L CONTEX TS
A C O M PA R I S O N O F T H R E E S O C I E T I E S
Patrick V. Kirch
What could be more visually evocative of a "classic" Polynesian scene than a

feast? In the public mind, the Hawaiian lu'au has indeed become a potent symbol
of Polynesian-ness. Yet the critical role of the feast in Polynesian societies is
ethnographically and ethnohistorically verifiable (Bell 1931), not merely a late
capitalist invention of the multinational tourist industry. Just as the last Hawaiian
king, Kalakaua, used the lu'au as a political stage to entertain such international
figures as Robert Louis Stevenson, so his chiefly (and priestly) predecessors ritu
ally incorporated the first European explorer to the archipelago, Captain James
Cook, in a symbolic act of feeding upon the temple platform of Hikiau (see
Sahlins 1995). For generations before Cook burst through the boundaries of
Kahiki, Hawaiian chiefs had elaborated the feast as a social nexus wherein the
1 68
P O LY N E S I A N F E ASTI N G : A CO M PA R I S O N O F T H R E E S O C I ET I E S
surplus fruits o f production-extracted from the populace by cooperation o r by

coercion (both martial and ideological)-were put to work furthering chiefly po
litical aspirations. To invoke the Marxist conception of a social formation, the
feast occupies strategic terrain at the interface between infrastructure and super
structure. Through the complex act of feast giving, the structures of heterarchy
and hierarchy were continually renegotiated.
Douglas Oliver (1989:291-292), in his definitive comparative ethnography of
Oceanic societies, advances several criteria differentiating feasts from ordinary
domestic eating. First, feasts are quantitatively distinguished by larger numbers of
participants, incorporating consumers from more than a single household unit.
Second, feasts involve "larger amounts of food per intended eater." Third, there
are important qualitative differences, such as the inclusion of delicacies or ritually
marked foods in feasts. In Polynesian societies, these special foods include, for ex
ample, pork, dog, or fowl, prized species of fish (such as pelagic game fish), sea
turtle, "fancy puddings" (usually incorporating an emollient such as coconut oil),
or "well-aged fermented breadfruit." In some Polynesian societies, human flesh
was also a component of feasts. 1 Fourth, there are spatial differences in the ways
and places in which feast foods were consumed, further differentiating them
from domestic eating. In many Polynesian societies there were spatially defined
feasting places, although these vary in the extent to which they were architecton
ically marked by permanent structures. Finally; the disposal of the remnants of
feasts sometimes differed from the ways in which ordinary household food re
mains were disposed of. The detritus of religious feasts might be given special
treatment, for such ritually charged garbage could be dangerous to those who
came into contact with it.
One of the great strengths of the Polynesian ethnographic record lies in the
opportunities for controlled comparison among a range of societies all charac
terized as chiefdoms, yet displaying a remarkable range of variability in degree of
hierarchy or stratification, in size of populations and scale of political units, in
the levels of production intensification, and similar variables. The classic compar
ative studies of Marshall Sahlins (1958) and Irving Goldman (1970) exploited these
opportunities to derive from the Polynesian record some general models of cul
tural evolution in pre-state, ranked societies.
The ethnographic and ethnohistoric literature on Polynesian feasting is vast, 2
grist for a monograph of its own. Here I limit myself to the role of feasts in three
Polynesian societies, each representative of one of the categories in Goldman's
sociopolitical classification (Table 6.1). These societies are: (1) Tikopia, a small
scale "Traditional" Polynesian chiefdom; (2) the Marquesas, an exemplar of mid
range "Open" societies given to fluidity in their social structures; and, (3) Hawai'i,
a chiefdom so complex and stratified that it is sometimes characterized as an "ar-
1 69
Patrick V. Kirch
TABLE 6.1
Key Contrasts between T i kopia, Ma rquesas, a n d Hawa iia n Case Stud ies
Attribute Tikopia Marquesas Hawai'i
Goldman's class Traditional Open Stratified

Island size (km') 4.6 1 ,057 16,692
Total population 1 ,250 50,000 300,000+
Population of
maximal political unit 1 ,250 1 , 500-3 ,000 30,000-50,000
Degree of Minimal (2 levels, Intermediate High (class
stratification little status marking) endogamy between
chiefs and commoners)
chaic state." I explore the extent to which variation in traditional feasting within
these three Polynesian societies displays regularities-or signal distinctions
with respect to differences in scale, stratification, or other indices of sociopolitical
complexity. I do this by examining three aspects of feasting: (1) the fanctions of
feasts, whether emically or etically construed; (2) the foods prepared for and con
sumed in feasts, especially as these are differentiated from everyday cuisine; and
(3) the architectonic space within which feasting was practiced. These latter two
features of feasts may be of particular salience in developing an "archaeology of
feasting" within Polynesia.
T I KO P I A : F E A S T I N G A N D K I N S H I P
Although Tikopia qualifies as a "chiefdom" society, it is governed more by cultur
ally ingrained concepts of kinship than by the political pronouncements of its
four ariki, or hereditary chiefs (Firth 1936). 'We, the Tikopia," Matou Nga Tikopia,
is an indigenous slogan encapsulating this notion of a closely knit community in
which everyone is bound to everyone else through consanguineal or a:ffinal ties.
The role of food in Tikopia was summed up Sir Raymond Firth in these words:
"Food serves as a most important material manifestation of social relationship,
and through it kinship ties, political loyalty, indemnity for wrong, and the canons
of hospitality are expressed" (1939:38).
The cover term for "feast" in Tikopia is anga, meaning either the feast itself, or
the "specific assemblage of food for [a] stage in ceremony" (Firth 1985 :11); how
ever, there are many specific indigenous lexical categories of feast, not all of
3
which are called anga (Table 6.2). Functionally, we may subdivide Tikopia feasts
into three main groups: (1) domestic or secular feasts; (2) chiefly feasts; and (3) rit
ual feasts. The latter are associated with particular religious ceremonies that
make up the annual cycle known as the "Work of the Gods" (Firth 1967).
1 70
P O LY N E S I A N F E A STI N G : A C O M PA R I S O N O F T H R E E S O C I ET I E S
TABLE 6.2
Pri ncipal Ki nds of Feasts i n Tikopia
Functional Category Indigenous (Lexical) Category and Subtypes
Secular I domestic feasts 1. Initiation of young men: te umu lasi ("the great oven")
2. Marriage: te umu tanakianga ("the oven of joining"); te anga
Political/ chiefly feasts General term: anga

1 . Moringa, feast of newly elected chief
2. Pungaumu (anga.fi), rnidcareer feast of chief
3. Aroarorima (anga soro), feast as chief ages
4. Fakatangata (anga soro), feast late in chief's career
5. Fakamatua, final feast of chief
Ritual I religious feasts Associated with all key components of the "Work of the Gods"
cycle, e.g.:
1. Recarpeting ancestral temples
2. Dance to Quell the Wind (Taomatangi)
3 . Proclamation at Rarokoka
Domestic feasts typically celebrate key stages-rites de passage-in the life cycle
of members of households. Important feasts of this kind are those of new male
initiates at the time of circumcision, and marriage feasts. In both cases more than
a single household unit is involved, and the exchange of food between house
holds is a key social function of these events.
Major anga feasts are also given by chiefs at particular points in their careers.
Of these chiefly feasts, Firth writes that:
From the point of view of the chief the anga marks a stage in the progression of his
reign. It gives an opportunity to display his food resources and to assert his rank; it
secures for him ceremonial expressions of thanks from his chiefly guests and of loy
alty from his clanspeople; and in the later stages demonstrates his own fidelity to his
gods and thereby ensures their continued interest in him (1939:222) .
There are, in theory, a progression of such anga throughout the life of a chief,
beginning with the te moringa feast held soon after his election, through midca
reer feasts of a type known as anga soro because of the immense quantities of
grated taro (soro) involved, to the fakamatua, or "making elderly" feast, held
when the chief determines that the end of his life is approaching.
Feasts marked virtually all of the many rites and ceremonies held throughout
the course of the ritual year (Firth 1967), such as the major island-wide rituals called
the "Dance to Quell the Wind" (Taomatangi.), and the "Dance of the Flaming Fire"
(Urangaji), or the week-long ritual extraction of the sacred turmeric dye (Nuanga).4
1 71
Patrick V. Kirch
In the foods accumulated, prepared, and consumed in these various kinds of

anga among the Tikopia, we can distinguish differences from ordinary domestic
consumption, although the main distinction is quantitative rather than qualita
tive. Because the Tikopia do not keep dogs or pigs for food, there is no particular
emphasis on special flesh foods for their feasts. 5 Rather, feasts are characterized
by the preparation of special kinds of "fancy puddings," concoctions of starch
staple mixed with an emollient, usually coconut cream or coconut oil. 6 One such
feasting food is the roi, a pudding made from "taro, breadfruit or ripe banana
mixed with sago flour and coconut cream, then cooked overnight" in the earth
oven (Firth 1985: 403). This roi is a key component in many Tikopia rites, such as
the Proclamation of Rarokoka (Firth 1967=280). 7 In the annual Work of Somo
somo, for example, all of the men who have married women from a particular
clan during the previous year made a gift of roi to the clan chief. In quantity; such
gifts could be prodigious, as much as 40 baskets from a single man, requiring
"preparations . . . begun months in advance" (Firth 1967:393). Archaeologically,
however, none of these special foods are likely to be preserved, with little proba
bility of a qualitatively distinctive signature for feasting in the Tikopia archaeo
logical record.
The architectonic context of Tikopia feasting varies depending upon the type
of feast. Secular feasts such as those for initiation and marriages take place within
the normal domestic quarters. Indeed, all food preparation for both domestic
feasts and chiefly feasts utilized the same cookhouse facilities as for ordinary
household food preparation. Ritual feasting, however, was spatially associated
with the lineage and clan temples, located in the sacred district of Uta, along the
inner shore of the lake. In most respects, these temples (which also had associ
ated sacred cookhouses) were like ordinary domestic dwellings, and indeed, his
torically they were the former dwellings of ancestors, which had become sacred,
'boly houses" through generations of use and through the burial of ancestors
within them (see Kirch 1996, 2000). Whether such ritual spaces could be sepa
rately distinguished from domestic spaces in the archaeological record-without
the aid of the ethnographic evidence-is a serious question, again raising the
possibility that a prehistoric record of feasting in Tikopia may well be refractory
to archaeological analysis.
T H E MARQU ESAS: F EASTI N G A N D

CO M P ET I T I V E I N VO L U T I O N
In marked contrast to Tikopia, the Marquesas Islands and their population never
comprised a single, integrated community linked by bonds of kinship, or even by
politically induced integration. Although the Marquesan islanders shared a com
mon language and many aspects of culture (all of which marked them as distinct
1 72
P O LY N E S I A N F E ASTI N G : A C O M PA R I S O N O F T H R E E S O C I ET I E S
from other, related Polynesian groups), politically they were divided into a larger
number of independent tribal units, each typically focused on a single valley, or at
most uniting several adjacent valleys (Handy 1923; Thomas 1990 ). These political
units were frequently at war with each other, with acts of warfare ranging from
limited raiding to obtain sacrificial victims, to more extensive raids aimed at the
destruction of an enemy's productive resources, to outright wars of territorial
conquest (Thomas 1990: 87 passim) .
Proto-historic Marquesan society falls into a group (along with such other
Polynesian societies as those of Mangaia and Rapa Nui) characterized by Irving
Goldman (1970 ) as "open." In contrast to "traditional" societies such as Tikopia,
the open societies were marked by more fluidity and negotiation among compet
ing status positions of chief (haka'iki), warriors (toa), and priests (tau'a). In Gold
man's words, these systems were "more strongly military and political than
religious," and stability was "maintained more directly by the exercise of secular
powers" (1970:20 ) . Elsewhere (Kirch l99I:I44) , I have characterized proto-historic
Marquesan society as marked by "an involuted cycle of prestige rivalry and com
petition." As ethnohistorian Nicholas Thomas (1990 ) observes, the two main are
nas in which such competition took place were those of waifare and .feasting. In
such a highly competitive society, it is not surprising that Greg Dening describes
the Marquesan ko 'ina-the feast-as a kind of "market" :
Enata had no wealth that could be accumulated. What they accumulated were the obli
gations they were owed by the distribution of their wealth. Koina, the feast, was their
marketplace. All the fuods, all the ornamentation, all the energy used in dance and
song were expended in a short time. What remained was a lien on tomorrow. (1980:63)
The Marquesan cover term for "feast" is ko 'ina, with many specific kinds of feast
designated by an adjectivally modified form of ko 'ina, as in ko'ina tupapa'u, "feast
for the dead" (Table 6.3) . In addition, the specific terms mau, for "memorial
feast," and heana, for "cannibalistic feast," were primary lexical designations for
two significant kinds of feast /festivals (Thomas 1990; Dening 1980) . As in
Tikopia, there was in traditional (pre-mission) Marquesan life a considerable di
versity of feasts, as outlined in Table 6.3. Because of the severe collapse of Mar
quesan society after European contact, the nature of many of these feasts is
incompletely known, since they were abandoned well before systematic ethno
graphic observation commenced in the early twentieth century, and can be re
constructed only through early visitors' accounts and oral testimony of older
Marquesans. Here I follow the account of Handy (1923:203-223 ) who carried out
"salvage ethnography" in the archipelago in 1920.
Among the main domestic or secular feasts were ko 'ina for betrothal and mar
riage, and to celebrate the tattooing of a young man. Marquesan society was con-
1 73
Patrick V. Kirch
TAB LE 6.3
Pri ncipal Kinds of Feasts i n the Ma rquesas
Functional Category Indigenous (Lexical) Category and Subtypes
Secular I domestic feasts 1. Betrothal feasts (ko'ina tuia)

2. Marriage feasts (ko'ina hunona)
3 . Tattooing feasts (ko'ina tuhi tiki)
Chiefly I political feasts 1. Competitive entertainment feast (ko'ina hakahiti)
2. Memorial feasts for deceased chiefs (mau)
3 . Memorial feasts for deified priests (ko'ika oke, ko'ika vaihopu, and
ko'ika u'upua)
4. Harvest festivals (ko'ina tapavau)
War I cannibalistic feasts Ko'ina heana
siderably more hierarchically differentiated than that of Tik.opia, and only

"wealthy" families would ordinarily command the resources to carry out such
feasts. These included the households of chiefs (haka'iki) and other elites, such as
priests (tau'a), warriors (toa), and various kinds of experts (tuhuna).
Of greater significance were the memorial festivals "held in honor of the de
parted spirit of a man" (rarely were these celebrated for women), called mau, or
in the case of festivals to deify a deceased priest, ko'ika (Handy 1923:212 passim).
Mau and ko'ika feasts were impressive displays, both of prestige foodstuffs such as
numerous baked hogs, and of special starch and fermented breadfruit prepara
tions, as well as of bodily ornamentation for elite participants and dancers.
Handy writes that "the greatest of all feasts were the memorial festivals cele
brated long after the actual death of chiefs and chiefesses, inspirational priests, or
ceremonial priests for the purpose of deifying their spirits" (1923:216). Another
category of feast associated with chiefs and held under their aegis was the harvest
festival (ko 'ina tapavau), held after the completion of the breadfruit harvest and
the successful filling of the silos of fermented breadfruit. 8
Finally; there were feasts held at the successful conclusion of a war against an
enemy tribe, "celebrating the capture of human victims" (Handy 1923 :218), called
ko'ina heana, or simply heana. According to Handy's all-too-brief account, these
heana were rather solemn affairs, lacking the elaborate dancing and singing that
typically accompanied other feasts. Rather, "the ko'ina heana consisted merely of
a feast at the dance area [tohua], where human flesh, pig, and popoi [fermented
breadfruit paste] were consumed" (1923 :219) . The bodies of sacrificial victims in-
1 74
tended a s offerings t o the gods were apparently not consumed, but those o f other
victims were eaten by the chiefs, warriors, and priests.
Marquesan feasting was thus distinguished qualitatively as well as quantita
tively from ordinary eating, not only in the preparation of special starchy pud
dings (as in Tikopia), but in the emphasis on prestige flesh foods, particularly pigs
but also in the heana feasts, human flesh. As elsewhere in Polynesia, the pig had a
particular association with rank and status, and may have been reserved prima
rily, if not exclusively, for consumption by males.
Turning to the architectonic or spatial context of feasting, the contrast with
Tikopia is striking. Beginning-on archaeological evidence-during the Expan
sion Period (ca. A.O. noo-1400), the Marquesans began to construct unique feast
places, called tohua. Essentially, these consisted of large, rectangular dance
plazas, usually a leveled terrace, surrounded by platforms (paepae). These plat
forms included foundations for temples (me'ae), chiefs' houses, and viewing plat
forms for spectators who witnessed the elaborate dances that accompanied
feasting. There were also associated cookhouse structures where the baking of
pigs and other food preparations were carried out. 9 Tohua were often of consid
erable size, with individual platforms of "megalithic" dimensions; the terrace of
Tohua Vahangeku' a in Taipi Valley (Nukuhiva) measures 174 m long by 26 m
wide, and is surrounded by a complexity of ancillary structures and platforms
(Fig. 6.1; see also Suggs 1961:162-163) . Tohua were often architecturally orna
mented with relief or freestanding sculpture of anthropomorphic character, in
some cases commemorating individuals who had been deified in mau or ko'ika
feasts (Linton 1925:85).
An exemplar of Goldman's "open" class of Polynesian societies, the Marque
san pattern of feasting displays important differences from that of "traditional"
Tikopia. Despite similarities in categories of feasts (as in the domestic marriage
feasts), there are critical differences, with Marquesan feasts playing an important
role in the overall emphasis on competition and status /prestige rivalry (the com
memorative, deification, and cannibalistic feasts). Whereas in Tikopia there was
little if any qualitative distinction between ordinary and feast foods, the Marque
san feasts emphasized the provision of prestige flesh foods, especially pig, and in
the case of heana feasts to celebrate success in war, of human flesh itself. 10 In the
architectonic sphere, the great emphasis accorded feasting as an aspect of so
ciopolitical competition is matched by the Marquesan development of a special
feast place, the tohua, construction of which required substantial labor invest
ments. Archaeologically, both the emphasis on qualitatively different (and tapho
nomically preservable) animal foods, and the construction of permanent and
architecturally elaborate feast spaces, greatly enhance the likelihood that a pre-
1 75
Figure 6.r. The main ceremonial plaza of Ta' a' oa tohua on the island of Hivaoa, Marque
sas Islands. Informants state that the platform in the foreground was for the cooking and
display of human sacrificial victims. (Photo by P. V. Kirch)
historic record of the long-term development of feasting behavior might be pos

sible for the Marquesas.
The key differences in Tikopian and Marquesan practices of feasting reflect
fundamental differences in the two societies. In the small-scale community of
Tikopia where all members of society are linked by bonds of kinship, the feast
1 76
serves a s a mechanism fo r assuring social solidarity, binding the community to

gether in commonly perceived purpose. In the larger-scale, multitribal society of
the Marquesas, the feast had become an instrument of prestige rivali:y and com
petition, a means of 'fighting with food.' As Thomas writes, "the medium of
food, its movements, and the differing qualities and quantities of particular pre
sentations must have, as a totality, provided an index to virtually everybody's sit
uation of power or dependence" (1990:97) .
H AWA l ' I : F E A S T I N G A N D K I N G S H I P
In turning to the Hawaiian Islands, we move from the realm of Goldman's
"open" level of chiefdom society into the "stratified." Indeed, on a variety of cri
teria, the proto-historic polities of Hawai'i 11 are among the most complex chief
doms ethnohistorically or ethnographically documented, arguably described as
"archaic states." In these Hawaiian polities, the chiefly class had elaborated itself
into as many as nine grades of chiefs, practiced virtual class endogamy (as well as
sibling marriage to maintain the purity of the blood line), had alienated direct
control of land and other means of production from the commoner class (insti
tuting a system of usufruct rights subject to regular tribute payments), supported
a body of craft specialists as well as warriors and household retainers, and were
ideologically legitimated by a religious system based on the cults of Ku (god of
war) and Lono (god of dryland agriculture), the former requiring human sacri
fice. With genealogies that proclaimed them to be direct descendants of the
gods-indeed, god-like "raging blazes" who walked on earth-it is a matter of
semantics whether we choose to designate the highest ranked, paramount lead
ers of Hawai'i as "chiefs" or "kings." 12 I shall here adopt the latter terminology;
and refer to the political system of proto-historic Hawai'i as one of kingship. The
question of immediate moment, then, is how the Polynesian patterns of feasting
that we have examined for Tikopia and the Marquesas were further transformed
as Hawaiian society moved from a structure based on kinship to one based on
kingship.
One of the most salient features emerging from a perusal of the Hawaiian
ethnographic and ethnohistoric literature, with regard to feasting, is the relative
dearth of references to feasting among commoners (the maka'ainana), and the con
comitant richness of description for feasts among the chiefly class (the ali'i). The
consumption of vast quantities of food-even on a daily basis-had become virtu
ally a defining attribute of the ali'i, marked reciprocally by their corpulence. 13 As
the Russian explorer von Kotzebue observed of the Waialua chiefess Pi'ia in 1824, "I
can . . . bear testimony to another qualification, of equal importance in her
estimation-she has certainly the greatest appetite that ever came under my obser-
1 77
Patrick V. Kirch
vation" (183o:vol. 2, 223; emphasis added). In their endomorphism, the Hawaiian

chiefs had carried to an extreme the widespread Polynesian cultural logic that
chiefs should naturally be bigger than their subjects.14 But it appears they had also
co-opted the feast as a distinctly chiefly form of consumption, yet one more means of
distinguishing and differentiating themselves from the common folk.
In Tikopia and the Marquesas, domestic feasts such as those accompanying
initiation, marriage, or tattooing, comprise a major category in the functional
classifications of feasts. For Hawai'i, however, such domestic feasts are but little
remarked, even by such indigenous ethnographers as Kamakau (1964:25-26). Far
more important for the maka'ainana, were small feasts (usually focused on the
sacrifice of a pig) to celebrate the cutting of a new canoe, or the successful har
vest of a set of irrigated taro terraces (Kamakau l976:rr9, 36-37), both critical as
pects of production and both largely male-focused arenas.
The chiefs, in contrast, could virtually be found feasting every day of their
lives. From many early-nineteenth-century documents pertaining to the chiefly
estate at Waialua, O'ahu, Kirch and Sahlins (1992) chronicle the ceaseless de
mands put upon the common people of that district for taro, poi, pigs, fish, and
other foodstuffs to support the bloated chiefly establishment in Honolulu. Quali
tatively distinctive foods, which in other Polynesian societies might be reserved
for special feast occasions, had become the daily fare of ali'i households. While
only the male chiefs ate pork (on a daily basis), the female chiefesses developed a
great fondness for puppies, and hundreds of small poi-dogs (fattened upon the
starchy taro paste) were raised in pens for the latter's regular enjoyment.
When the chiefs and their retainers were not resident in one of their principal
local� s (such as Waipi'o or in Kona on Hawai'i, or after European contact in La
_
haina or Honolulu), they traveled about their domains literally feasting off the
produce demanded up of their subjects. Thus in September of 1833, the king
Kauikeaouli (Kamehameha III) traveled with his substantial household to Wai
alua, descending as the resident missionary Emerson put it, "like a company of
locusts" upon the common folk. Emerson's wife penned in her journal: "I shall be
glad when it's over. Pigs, dogs, fish and fowl have been slaughtered in large num
bers, and ever so many calabashes of poi have been prepared" (both Emerson
quotes in Kirch and Sahlins 1992:145-146) .
In the proto-historic Hawaiian polities, and continuing up until the abandon
ment of traditional religious practice during the later reign of Kamehameha I
(and ceasing completely after his death in 1819), the other great arena for feasting
was at the great cult temples dedicated to Ku and Lono (see Valeri 1985). Here the
feasts were formal components of highly elaborated temple rituals, presided over
by a cadre of priests (kahuna) and the chiefs, and in which (some) male common
ers were allowed to participate as observers and at times as consumers of certain
1 78
P O LY N E S I A N F E A S T I N G : A C O M PA R I S O N O F T H R E E S O C I ET I E S
feast foods, but from which women were wholly excluded. I cannot here go into
the complex matter of gender differentiation in Hawaiian food consumption, ex
cept to note that this was highly systematized, and included prohibitions (kapu)
on the eating of such ritually marked items as pork, certain kinds of bananas, cer
tain species of fish, and so forth.
The quantitative scale of such feasts could be grand indeed, as the nineteenth
century Hawaiian scholar David Malo described for a feast to conclude the dedi
cation of a new luakini heiau, a temple dedicated to the war-god Ku:
That night a large number of hogs, as many as 800 (elua lau), were baked; and the
priests were separated into two divisions, one on this side and one on that side of
the mana [house], each division taking part in the service alternately.
The pork was also divided into two portions, 400 of the hogs being assigned to
the priests seated at one end of the building and 400 to the priests seated at the
other end (kala). The priests and their men ate the flesh of the swine and continued
their prayers, without sleep, until morning.
The next morning, which was Kupau, the kuili service was kept up, and contin
ued without intermission all day. That day 400 pigs were served out to the worship
pers, 200 (e!ima kanaha) to those at one end of the temple and 200 to those at the
other end.
The service was still kept up during the ensuing night, 240 pigs being baked and
served out-120 to priests at this end of the temple and 120 to those of the other
end of the temple. The service continued all night.
During the next day, Olekukahi, the kuili service still went on, and 400 pigs were
baked and divided out equally between the priests at the two ends of the temple.
(Malo 1951:172)
All this (1,440 pigs, to be exact), moreover, was merely preliminary to what Malo
calls the "great feast" following the girding of the Ku image with a sacred loin
cloth, at which the number of slaughtered hogs is unfortunately not specified. 15
Clearly, for two of the three attributes of Polynesian feasting upon which I
have focused-function and the nature of feast foods-the Hawaiian pattern had
been considerably transformed in the context of a hierarchically differentiated
society. No longer a significant aspect of commoner life, feasting was a virtually
mundane, ubiquitous aspect of the chiefly habitus. Prestige foods, especially pigs,
were essential to chiefly feasts, whether for daily consumption or for ostentatious
display during temple rituals. Hawaiian feasting was likewise differentiated archi
tectonically, since it was spatially concentrated in two kinds of venue: the chiefly
residence, and the principal temples.
As with the Marquesas, the possibilities for recovering an archaeological
record of feasting in Hawai'i seem good. Although Hawai'i lacked formal feast
ing centers such as the tohua, the main Hawaiian temples (heiau) were principal
1 79
Patrick V. Kirch
loci of feasting, and refuse pits associated with these structures should contain
the remains of temple feasts. Likewise, chiefly residences ought to be differen
tially marked from commoner habitations by the presence of prestige food items,
especially the bones of pigs, dogs, and certain kinds of fish (such as pelagic game
fish). There is indeed support for such differential faunal assemblages in the
Hawaiian archaeological record (Weisler and Kirch 1985) .
CO N C LU S I O N S
I conclude this all-too-brief comparison of feasting in three Polynesian societies
with two general observations. The first concerns variation in feasting behavior
in relation to key structural characteristics of these societies. At risk of sounding
like an old-time Malinowskian structural-functionalist, the case can be argued
that differences in feasting between Tikopia, the Marquesas, and Hawai'i do cor
relate closely with such attributes as sociodemographic scale, degree of stratifi
cation and hierarchy; and extent of aggressive competition (war). In a small-scale
society such as Tikopia, where kinship underwrites the Durkheimian sense of
community; feasts are largely organized at the immediate suprahousehold level,
and are partaken of by commoners and chiefs alike. Indeed, in Tikopia chiefs are
expected to give feasts for the common people at key stages in their careers. In the
Marquesas, some differences emerge, correlating with the larger scale of Mar
quesan polities, but more particularly with the increased hierarchy of Marquesan
society and with the involutionary emphasis on overt competition. In the Mar
quesas, the feast became less of a medium for regular social communion than an
instrument of political competition. The Hawaiian case takes us the farthest
from the Durkheimian mode, in which the feast was co-opted as a virtual prerog
ative of the ruling class, less an instrument of power than a pervasive, daily re
minder of the immense gulf of social distinctions that separated ali'i from
maka'ainana.
It is only an illusion that Tikopia, the Marquesas, and Hawai'i appear to repre
sent ideal stages along some kind of evolutionary continuum for chiefdom soci
eties. None of these is actually ancestral to the other, and all are ethnohistoric
endpoints preceded by millennia of history. To trace the real history of feasting
in Polynesia requires that we pull back the ethnohistoric tapestry; and enter the
messy realm of the "archaeological record." In this chapter, I have alluded briefly
to the potentials for an archaeology of feasting, and I have been mixed with re
gard to my assessments of these possibilities. For Tikopia, I frankly think that the
likelihood of developing a robust archaeological record of feasting behavior over
time is not very good. Tikopian feasting has little to distinguish it qualitatively
from ordinary eating, nor is there any elaboration of the architectonic contexts
within which feasting occurs. Indeed, although I spent two long field seasons ex-
1 80
cavating in Tikopian sites (Kirch and Yen 1982), I cannot say that I dug up the re
mains of anything specifically identifiable as a feast.
For the Marquesas and Hawai' i, and other Polynesian societies of their
"types,'' the possibilities for an archaeology of feasting seem more feasible. This
reflects both the emphasis on qualitatively different kinds of feast foods ( espe
cially pig), and the architectural elaboration of feasting places (whether tohua in
the Marquesas, or chiefly residences and temple sites in Hawai'i). Indeed, for
Hawai'i there are tentative archaeological identifications of chiefly residential
sites based on distinctive faunal assemblages (Weisler and Kirch 1985), as well as
great potential for the reconstruction of feasting behavior associated with tem
ples. In short, an archaeology of feasting in Polynesia is likely to be-to adopt a
metaphor from contemporary '1ocal" Hawaiian eating-a "mixed plate." But
then, such has always been the nature of our historical science.
N OT E S
I . The revisionist stance of Arens (1979), which argues that cannibalism is entirely a
Western construction of the indigenous Other, does not stand up to the ethnographic
or archaeological records for various Pacific Islands. Sound documentation-either
ethnographic or archaeological, or both-for the consumption of human flesh exists
for Fiji, the Marquesas, Mangaia, Easter Island, and probably New Zealand.
2. Oliver observes that for Oceania generally; "feasting occupies a larger part of many
Islands' ethnographies than any other kind of activity . . . and perhaps deservedly
so-not because of the drama usually associated with it but because of the political
relations it represented and revealed" (1989:291).
3. Elsewhere, Firth writes that the term anga applies both to feasts given by chiefs, and
also "to the accumulations of food for initiation and marriage rites" (1939:222).
4. In 1978 I participated in the Nuanga of the Ariki Tafua, the second-ranked chief of
Tikopia. This is one part of the Work of the Gods that had not been abandoned as a
result of the conversion to Christianity. Although the extraction and purification of
the turmeric dye was the focus of this ritual activity; and not feasting per se, the
Nuanga met all of the key criteria for feasting, such as the participation of large
numbers of people, the accumulation of significant quantities of food, the
preparation of special foods (such as taro puddings), and the consumption of food in
a special precinct, that of the chief's house and adjacent compound.
5. One exception to this is the marine turtle (Chelonia mydas), a food associated with
chiefly status. When turtles are taken, they are usually brought to the chief, and may
be kept for days or even weeks before being slaughtered and consumed on a feast
occasion. Their flesh, however, is not exclusively reserved for the chief, and will be
shared out with other participants.
6. For another ethnographic case of the importance of such puddings ifaikai, literally
"made food") in another traditional Polynesian society; see Kirch (1994:9s---1 00,
18!}-213) on Futuna.
1 81
Patrick V. Kirch
7. Firth writes that "the sacredness of the roi depends upon its being used to provide
offerings in the most important religious ceremonies. It is said then to be made by
the chiefs' elders for their respective deities. 'Each makes it for his god"' (1967:280).
8. Marquesan subsistence depended more than virtually any other Polynesian society
(except perhaps that of the Society Islands) on arboriculture or orchard gardening,
with specific emphasis on the breadfruit (Artocarpus altilis). Since the breadfruit
harvest occurs in a short period, the vast yield of starch is preserved through a
method of semi-anaerobic pit fermentation and ensilage (see Kirch 1984:132-135).
9. Suggs (1961:72-73, fig. 25b) reports excavating such an "oven complex" at the
Hik:ok:u' a tohua site in Hatiheu Valley, Nuk:uhiva.
ro. Of course, quantity was also important in Marquesan feasting, as Thomas points out:
"Prestige depended not simply upon returning the equivalent of what had previously
been received, but upon giving more, and particularly more than the receivers could
eat" (199o:ro1).
n. There were four main, competing chiefdom polities in the archipelago at the time of
European contact, centered on the principal islands of Kaua'i, O'ahu, Maui, and
Hawai'i.
12. Historical ethnographers Sahlins (1995) and Valeri (1985) have. used the term king, I
believe justifiably.
13. On the corpulence of Hawaiian chiefs in the early nineteenth century, see Kirch and
Sahlins (1992:vol. l, 78-79).
14. As one Tongan informant put it to the ethnographer E. W. Gifford, "Can't you see he
is a chief? See how big he is" (1929:124).
15. Human sacrifice was also an essential aspect of such luakini ritual, but unlike the
Marquesan situation, human flesh was not generally consumed, other than the
symbolic eating of the victim's eye (see Valeri 1985).
REFERENCES
Arens, W.
1979 The Man-Eating Myth. New York: Oxford University Press.
Bell, F. L.
1931 The Place of Food in the Social Life of Central Polynesia. Oceania 2:117-35.
Derring, G.
1980 Islands and Beaches: Discourse on a Silent Land, Marquesas 1 774-1 880. Honolulu:
University of Hawaii Press.
Firth, R.
1936 We, the Tikopia. New York: American Book Company.
1939 Primitive Polynesian Economy. George Routledge and Sons.
1967 The Work of the Gods in Tikopia. New York: Humanities Press.
1985 Tikopia-English Dictionary. Auckland: Auckland University Press.
Gifford, E. W.
1929 Tongan Society. Bernice P. Bishop Museum Special Publication 6r. Honolulu.
Goldman, I.
1970 Ancient Polynesian Society. Chicago: University of Chicago Press.
1 82
Handy, E. S . C .
1923 The Native Culture in the Marquesas. Bernice P. Bishop Museum Bulletin 9.
Honolulu.
Kamakau, S. M.
1964 Ka Po 'e Kahiko: The People of Old. Bernice P. Bishop Museum Special Publica
tion No. 51. Honolulu.
1976 The Works of the People of Old. Bernice P. Bishop Museum Special Publication
6I. Honolulu.
Kirch, P. V.
1984 The Evolution of the Polynesian Chiefdoms. Cambridge: Cambridge University
Press.
1991 Chiefship and Competitive Involution: The Marquesas Islands of Eastern
Polynesia. In Chiefdoms: Power, Economy, and Ideology, edited by T. Earle,
pp. II9-45· Cambridge: Cambridge University Press.
1994 The Wet and the Dry: Irrigation and Agricultural Intensification in Polynesia.
Chicago: University of Chicago Press.
1996 Tikopia Social Space Revisited. In Oceanic Culture History: Essays in Honour
of Roger Green, edited by J. Davidson, G. Irwin, F. Leach, A. Pawley, and
D. Brown, pp. 257-274. Dunedin: New Zealand Journal of Archaeology Spe
cial Publication.
2000 Temples as "Holy Houses" : The Transformation of Ritual Architecture in
Traditional Polynesian Societies. In Beyond Kinship: Social and Material Repro
duction in House Societies, edited by R. Joyce and S. D. Gillespie, pp. 103-n4.
Philadelphia: University of Pennsylvania Press.
Kirch, P. V., and M. Sahlins
1992 Anahulu: The Anthropology of History in the Kingdom of Hawai'i. Chicago: Uni
versity of Chicago Press.
Kirch, P. V., and D. E. Yen
1982 Tikopia: Prehistory and Ecology of a Polynesian Outlier. Bernice P. Bishop Mu
seum Bulletin 238. Honolulu.
Kotzebue, 0. von
1830 A New Voyage Round the World in the Years 1 823, 24 , 25, and 26. 2 vols. London:
Colburn and Bentley.
Linton, R.
1925 Archaeology of the Marquesas Islands. Bernice P. Bishop Museum Bulletin 23.
Honolulu.
Malo, D.
1951 Hawaiian Antiquities. Bernice P. Bishop Museum Special Publication 2. Sec
ond Edition. Honolulu.
Oliver, D. L.
1989 Oceania: The Native Cultures of Australia and the Pacific Islands. Honolulu: Uni
versity of Hawai'i Press.
Sahlins, M.
1958 Social Stratification in Polynesia. Seattle: American Ethnological Society,
1 83
Patrick V. Kirch
1995 How "Natives" Think: About Captain Cook, for Example. Chicago: Universily of
Chicago Press.
Suggs, R. C .
1961 The Archaeology of Nuku Hiva, Marquesas Islands, French Polynesia. Anthropo
logical Papers of the American Museum of Natural History, No. 49. New
York.
Thomas, N.
1990 Marquesan Societies: Inequality and Political Transformation in Eastern Polynesia.
Oxford: Clarendon Press.
Valeri, V
1985 Kingship and Sacrifice: Ritual and Society in Ancient Hawaii. Chicago: University
of Chicago Press.
Weisler, M., and P. V Kirch
1985 The Structure of Settlement Space in a Polynesian Chiefdom: Kawela,
Moloka'i, Hawaiian Islands. New Zealandjournal of Archaeology 7:129-158.
1 84
7
FEASTING FOR PROSP ERIT Y
A S T U D Y OF S O U T H E R N N O RT H W E S T COAST F E A S T I N G
James R. Perodie
Feasts, potlatches, and related activities may have been significant factors in the
transition from egalitarian to ranked societies (see Chapters 2, 3 and 9; also Clark
and Blake 1994:17, 25, 28, 29; Hayden 1990:32, 37, 1993:225, 1995:24, 25, 74) . Thus,
any inquiry into the emergence of socioeconomic inequality should include a
study of feasting, potlatching, and related activities. Indeed, questions naturally
arise as to why individuals of indigenous cultures expended significant time, en
ergy; and resources on feasting and potlatching. The Northwest Coast region is a
good area for studying these phenomena because feasts and potlatches were cen
tral elements of the indigenous cultures, and because abundant ethnographic in
formation provides valuable emic data for analyzing Northwest Coast feasts and
potlatches. The Kwakiutl, Nootka, Coast Salish, and Twana were chosen for this
1 85
James R. Perodie
analysis because they are sufficiently similar culturally to treat as one group for
the purpose of analyzing feasting behavior. Due to the incomplete nature of ob
servations in individual ethnographies it is necessary to examine feasting as a re
gional phenomenon. No single ethnography or ethnographic collection has
relevant information on all cultural aspects pertinent to this investigation. Other
researchers have assumed or argued that some or all of these groups can be con
sidered culturally similar for general analytical purposes (Barnett 1968:10, 21;
Boas l89T317; Codere 1990:370; Drucker 1951:456; Elmendorf 1971:356; Ferguson
l984b:267; Rosman and Rubel 1971:6).
Feasts and potlatches can be differentiated by the nature of the distribution
that occurred on each occasion. According to Drucker (1951:370, 372), feasts were
occasions when an individual (the sponsor) would distribute food to four or more
invited guests, whereas at potlatches nonfood property was also distributed to in
vited guests (Fig. n). By stipulating that "four or more" invited guests must be
present, Drucker is probably more concerned with distinguishing potlatches and
feasts from simple nuclear family meals or informal visits than he is with defining
rigid attendance requirements.
Figure 7.I. A potlatch at Alert Bay prior to 1 9 1 4 . Of particular note are the sacks of flour
and pots of food that are to be given away as well as the full ceremonial regalia of the
major participants standing at the far end of the display area. (Photo by]. Welsh RBCM
2 3 0 7-b, courtesy Vancouver Public Library)
1 86
S O U T H E R N N O RT H W E S T COAST F E A S T I N G
Because there is a wide variety of feasts and because the definition o f potlatch
varies widely from ethnographer to ethnographer, indeed, because almost any
type of feast has been called a "potlatch" by various authors, it is important to
distinguish carefully between different types of feasts as discussed below. Feast
will be used as a comprehensive term for feasts and potlatches. In some cases, the
term potlatch will be retained for convenience if it is used by original ethnogra
phers. The distinctions between potlatches and other feasts, in any event, are not
critical to the new analytical approaches that follow. Of much greater impor
tance is whether the distributed materials were returned to the sponsor, and how
and when this occurred.
Many anthropologists have studied Northwest Coast feasts and come up with
widely different opinions on the role or function of feasts in Northwest Coast so
cieties. For example, Boas states: "The underlying principle [of the potlatch] is
that of the interest-bearing investment of property" (1966:77). According to Boas,
"the principal motivation in the behavior of the [Kwakiutl] Indians is the desire to
obtain social prestige" (1966:51). Acquiring and maintaining high prestige re
quired correct marriages, and wealth accumulated via industry and potlatch in
vestments (Boas 1966:51).
Codere (1966:127-129) concludes the Kwakiutl potlatch was a form of nonviolent
"warfare" utilized to increase social prestige. Codere states that the Kwakiutl were
characterized by "their limitless pursuit of . . . social prestige" (1966:118) and that at
one time warfare and potlatching were probably "interchangeable means of gain
ing prestige" (Codere 1966:122). However, at least partly as a result of European in
fluences, warfare subsided and potlatching became the only significant method of
establishing or maintaining high social prestige (Codere 1966:118, 127-129).
Drucker (1951:377, 386; Drucker and Heizer 1967:134) and Barnett (1938b:351,
1955:253, 256, 1968 :123) argue that the potlatch announced and reasserted the
hereditary claims (i.e., seats, dances, properties) of the sponsor and his heirs, and
therefore their claims for social status. Barnett (1938b:353, 357, 1955:250, 1968:123)
adds that Northwest Coast cultures emphasized liberality and generosity; or their
simulation, and that the potlatch was "characterized by . . . an implied equation
of social worth with institutionalized liberality" (Barnett 1938b:357). Drucker
notes that "the Nootkan data are in very close harmony with Barnett's . . . ap
praisal" (1951:386).
Suttles (1960:302---g o4) argues that the potlatch was a redistribution mechanism
that enabled a network of interacting communities, each of which had varying
econoinic opportunities and capabilities, to maintain high levels of food produc
tion and equal food consumption among all members of the network. Suttles
states that "the drive to attain high status is clearly not the explanation of the pot
latch" (1960:304). Piddocke (1965:244, 245, 258) also argues that the potlatch was a
redistribution mechanism. However, he (1965:258) states that the potlatch had
1 87
James R. Perodie
multiple functions, and that in addition to redistributing food and wealth, pot
latch distributions were the potlatch sponsor's opportunity to secure and validate
his social prestige by demonstrating his generosity (Piddocke 1965:245, 257, 258).
And Rosman and Rubel (1971:205, 206) argue that potlatches were individual
and group rites of passage that occurred at critical social junctures when position
holders would turn over prerogatives to successors. They assume that prestige
and status were prime motivators.
The explanatory framework used in the present analysis differs because (1) it
assumes that time-consuming and costly behavior that persists over time ought
to be associated with some practical benefit, as Hayden argues in Chapter 2; (2) it
assumes that sponsors of the most elaborate and expensive feasts were ambitious
individuals and that their supporters utilized feasts for advancing their own ambi
tions, and (3) it identifies the specific practical benefits ambitious individuals were
seeking with various types of feasting activities. This perspective is similar to
Boas's in that Boas acknowledges one practical benefit of some feasts as increas
ing wealth. Codere may have viewed increased wealth and control as practical
beneficial outcomes of warfare, and hence potlatching. However, she (1966:n8)
maintained that the potlatch was an agent for increasing social prestige and did
not detail any specific practical benefits of this prestige in the manner that I dis
cuss in the following pages.
I suggest that ambitious individuals who want to maintain and increase their
economic and political control, control over labor, and wealth, need to achieve
several specific goals. In order to control labor, labor must somehow be attracted
and held. Increasing economic and political control requires a multitude of al
liances with other groups in order to increase access to resources, security, and
other support. Ambitious individuals need a consistently available mechanism for
increasing their wealth and political power. All of these goals can be accomplished
with feasting to create alliances, to indebt others, to escalate the production of
surpluses, and to disproportionately concentrate control of debts, alliances, and
surpluses in the hands of the organizers of feasts-in the hands of aggrandizers.
Most of the remaining paper is dedicated to detailing why these goals are impor
tant to ambitious individuals, and how they are achieved with feasting. Brian Hay
den has developed the theoretical feasting model that is used in this chapter in
several unpublished manuscripts (e.g. , Hayden and Maneeprasert n.d.).
A S S U M PT I O N S
There are five assumptions underlying the model used in analyzing Northwest
Coast feasting. The assumptions and their justifications are:
I. Ambitious individuals, or "aggrandizers" (after Clark and Blake 1994), on

the southern Northwest Coast participate in feasting activities for their
1 88
own benefit. Numerous ethnographers note that a feast distribution "was

no free and wanton gift" (Codere 1966:69), rather it was given to further
the donor's own interests (Barnett 1938b:350, 1968 :77-78, 126; Boas 1966:51;
Kamenski 1917:48; Mauss 1967:3; Sproat 1987:80). One of the main goals of
this chapter is to identify the specific advantages that aggrandizers derive
from various types of feast giving. In recent years, George Buchanan was
awarded a Nobel Prize for developing economic models in which individu
als participating in a system of exchange are assumed to act for their own
self-interest (Lewin 1986:941) .
2. When individuals systematically invest significant amounts o f time, en
ergy, and resources in an activity, they expect to gain specific, practical
benefits. This assumption extends the first assumption by stipulating that
terms such as status and prestige are too vague to account for an individ
ual's ambitious program of feast giving (see Chapter 2). Specific, practical
benefits being sought must be identified (Hayden 1995:21; Young 1971:211).
3. The majority of support group members (usually lineages, corporate
groups, 1 or communities) must at least tacitly approve of the aggrandizer' s
activities (Hayden 1995:21). Intragroup conflict would erupt otherwise (and
undoubtedly did on occasion), inhibiting the efforts of aggrandizers.
There is abundant ethnographic evidence that sponsoring a feast required
a high degree of group cooperation among members of support groups
(e.g. , Barnett 1968:123; Boas 1921:1341; Codere 1966:78; Druck.er 1951:268). It
seems unlikely that any individual could successfully sponsor a large feast
if faced with a concerted opposition.
4. The elaboration of feasting on the Northwest Coast was a function of the
region's environmental richness. This richness originally enabled the
amassing of food and property for distribution (Barnett 1938 :118, 1968:76;
Druck.er 1951:37, 59).
5. An aggrandizer's main objective is to increase control by indebting others.
In discussing the Kwakiutl, Walens (1981:13) states that a "nobleman" at
tained success only by indebting others. Hayden (1995:24, 69) argues that
ambitious individuals increased their power by indebting others and creat
ing return obligations. Gosden (1989), Mauss (1967:73), Wohlt (1978:100),
and Strathern (1971:215-216, 219) all make similar points: power and pros
perity resulted from indebting others.
F EAST I N G T Y P E S
Hayden (Chapter 2-see also 1995:27; also Hayden and Maneeprasert n.d. :5-9)
proposes that all practical benefits or goals of feasting can be categorized into (at
least) eight general types that can be conceptualized as ideal feast types. Ambi
tious southern Northwest Coast individuals maintained and increased their eco-
1 89
James R. Perodie
nomic and political control, control over labor, and wealth, by sponsoring feasts
of these eight general types, although purposes often overlapped or were com
bined in a single event.
1. Solidarity. Feasts promoting cooperation within a group (e.g. , household,

lineage, corporate group, village) are considered solidarity feasts. Solidar
ity within a group will enhance economic productivity and security, pro
mote mutual support between group members in conflicts with other
groups, and ensure support for group leaders. In solidarity feasts, hierar
chical differences should be downplayed, food should be the main product
distributed, contributions from all participants frequently occur, no debts
or return obligations should result, and participation and attendance
should be widespread.
2. Reciprocal. Reciprocal feasts initiate and maintain alliances between groups
for the purposes of security, marriage, and economic benefits (e.g., in
creased access to resources or exchange partners). Two or more groups or
their representatives must be involved, and each group generally tries to
impress the other(s) without being overtly competitive so as to maintain
an amiable atmosphere.
3. Solicitation. Solicitation feasts are feasts given to solicit favors or support
from more powerful individual(s) . Solicitation feasts are probably unidirec
tional exchanges where the solicited powerful individual receives food
and/ or gifts without incurring the expense of a specific return feast.
4. Promotional. Promotional feasts advertise group success and prosperity in
order to attract desirable potential labor, allies, exchange partners, and
other supporters. Group success and prosperity is promoted by making
public distributions to important and specifically invited individuals with
whom a relationship is desired (e.g., families with potential marriage part
ners, or potential security allies). Distributions are also made to other mis
cellaneous guests and supporters in order to promote the host group to as
wide a circle as possible.
5. Competition. Competitive feasts create material profit for the feast sponsor
because invested feast distributions are returned at a future feast with in
terest. Competitive feasts are large, lavish events where the sponsor en
deavors to maximize his contractual obligations and profit by distributing
enormous amounts of food and property.
6. Political Support. Political support feasts are sponsored in order to obtain or
increase political support.
7. Acquisition of Political Positions. Feasts that function as a formal criterion
for political advancement are categorized as feasts for acquiring political
positions.
1 90
S O U T H E R N N O RT H W EST COAST F E A S T I N G
8 . Work-Party Feasts. Work-party feasts obtain labor for specific labor

intensive projects such as house building. Individuals provide their labor in
exchange for participation in feasting activities.
9. Child-Growth Feasts. A ninth type of feast might be added to Hayden's pro
posed ideal feast types. Child-growth feasts are feasts that invest surpluses
in children in order to increase their worth so that wealth exchanges at
marriages will be maximized, and alliances between families for other pur
poses will be strengthened.
Separating feasts in this manner is artificial because the feast types obviously
overlap, and a single feast will probably involve a combination of these proposed
types. However, identifying these specific practical benefits of feasts facilitates
this analysis by clarifying the components.
FRAM EWO RK
In order to simplify the analysis of feasting in the study area, the Kwakiutl,
Nootka, Coast Salish, and Twana feasts will be separated into three more basic
categories: (1) feasts where the sponsor does not expect the recipients of food or
gifts to invite him to a future feast, which I will call no-return feasts; (2) feasts
where the sponsor expects the main recipients of his food or gifts to invite him to
a future feast and return equivalent amounts to him which I will call equal
,
return feasts; (3) feasts where the sponsor expects to receive an increased return
at a future feast, which I will call greater-return feasts. It is important to empha
size that Drucker (1951:290-291, 381), Drucker and Heizer (1967:79-80), Elmen
dorf (1960:332, 343), Boas (1966:51), Barnett (1938a:132, 195p90, 255, 257, 1968:30, 84,
100 ), Sproat (1987:80 ), and Birket-Srnith (1967:12, 13, 15, 35) all state that on at least
certain occasions potlatch and feast distributions obligated the recipients to re
turn the feast and gifts.
The return expected by a feast sponsor helps indicate which of the nine gen
eral purposes for holding feasts is involved, which in turn identifies the practical
benefit sought by the feast sponsor. How the different feast types fit into this
framework, and how aggrandizers used them to increase their wealth and con
trol, will be discussed in the coming sections.
In order to understand feasting goals and dynamics, it is necessary to examine
the question of who the main feast sponsors are. Details will be provided
throughout the chapter; however, the principal individuals who sponsor feasts,
whether they are no-return, equal-return, or greater-return feasts, appear to be
highly motivated hereditary elites ("chiefs") with the power and capability to
amass the required resources for lavish or larger feasts. "No one but a chief of
high rank could potlatch for himself" (Drucker 1951:182; also see Barnett 1955:134;
Drucker 1939:64, 1951:72, 193, 141, 143, 243, 247, 286, 370, 376, 377, 439; Elmendorf
1 91
James R. Perodie
1960:322, 327, 338, 343, 353, 361--63, 388, 404-405, 410, 443, 1971:361). Barnett (1955:141)
and Elmendorf (1960:333, 427, 433) state that elite individuals were verbally trained
from an early age to be ambitious. Barnett states that the potlatch "expresses
purely personal intentions and ambitions" (1968: 126). Boas (1897:343; also in Bar
nett 1968 :5) makes the same points. Although sponsorship is sometimes attrib
uted to a single individual such as a household head, the principal sponsor can be
regarded as representing his own interests, the interests of the corporate group's
upper-class members, and to a lesser extent the interests of other free corporate
group members (see Barnett 1938b:350; Drucker 1939:64).
Daughters could inherit titles and property (e.g., Barnett 1938a:131, 1955:251;
Boas 1925:67--69, 91, 105; Codere 1990:367; Rosman and Rubel 1971:135), but this oc
curred less frequently; if at all, before contact (Barnett 1995:251; Codere 1990:367).
For this reason I will refer to feast sponsors and aggrandizers in the masculine
third person.
N O - R ET U R N F E A STS
No-return feasts have been defined a s those feasts in which the sponsor does not
necessarily expect a return feast. Nevertheless, as will be seen, even on these oc
casions the sponsor can be viewed as pursuing his own interests.
Solidarity Feasts
In order to attain their ambitions, aggrandizers needed control over a supply of
labor. Resource sites were definitely privately or corporately owned (Barnett
1955:59, 241, 250, 251, 1968:78; Boas 1921:1345, 1966:35-36; Drucker 1951:42, 43, 47,
247, 248, 256--257, 454; Suttles 1960:300; Walens 1981:13, 71). Hereditary chiefs who
owned such productive resource sites needed workers to collect and process the
resources at their hunting, fishing, or gathering sites (Drucker 1951:279, 454; Sut
tles 1960:300 ). The output from these resource sites fed the chiefs' families and re
tainers (Drucker 1951:244) and enabled aggrandizers to accumulate surplus food
for feasting and exchanging.
Control over labor secured several other benefits for aggrandizers: first, it in
creased security for both the group and its leaders. Second, it secured control over
production of nonfood goods such as canoes (Boas 191p338; Drucker 1939:64),
which could be used for exchanging or distributing. Third, controlling labor pro
vided the opportunity for acquiring privately owned 'benchmen" (Barnett
1938:130; Drucker 1951:251; Walens 1981:38) who enforced the aggrandizers' orders.
The two crucial points are: (1) that aggrandizers needed control over labor in
order to prosper (Barnett 1955:247, 1968:113; Boas 1921:1333, 1334; Drucker 1951:244,
280, 323, 454; Elmendorf 1971:372); and (2) that aggrandizers required a means of
controlling the individuals who fished, hunted, gathered, or otherwise supported
1 92
S O U TH E R N N O RT H W E S T COAST F E ASTI N G
�, them. Aggrandizers indebt the members of their group by providing food for
subsistence, and other perks such as giving them economic rights, naming their
children at feasts, giving them their own prestigious names such as war chief
names, and lending them ceremonial privileges such as hair ornaments for their
daughters' puberty feasts so that they are able to sponsor their own small feasts
(Drucker 1951:47, 140-141, 245, 257, 270, 273, 280, 381, 383-384; Elmendorf 1960:324;
Walens 1981:13). Some retainers might even receive a wife as a reward for their
service, which the retainer might not be able to afford otherwise (Drucker
1951:273; Mozino 1970:32-33). In his discussion of indigenous economic systems of
the Northwest Coast and elsewhere, Herskovits (1952:462-463, 470, 482) notes
that the supporters of aggrandizers may be rewarded with prestigious titles or
given assistance in accumulating their marriage gift. Indebting, rewarding, and
controlling group members in these or other manners results in the group mem
bers becoming the ret�ers of the group elites (for additional descriptions of
group members as tenants or retainers see Barnett 1968:45; Boas 1921:1333, 1334;
Drucker 1951:43, 271-272; Elmendorf 1960:325).
Solidarity feasts were one way aggrandizers kept their group members con
tent, productive, and supportive. The way feasts were employed to attract labor
will be discussed under promotional feasts. According to Drucker, young chiefs
"were told that they must 'take care of' their people . . . providing them with
food, giving them feasts, winning the good will and affection of the commoners,
for 'if your people don't like you, you're nothing"' (1951:131). Young chiefs were
advised "to 'be good to their people, treat them kindly, and give them many
feasts to make them happy"' (Drucker 1951:454; also see Drucker 1951:439-440).
And according to Barnett, a chief "gave frequent feasts and entertainments to the
members of his family group to maintain their good will (1955:246). That is,
"
there were no guests from other groups, or at least no essential guests. Similar
considerations may have been responsible for the community feasts held in more
complex_ chiefdoms and early states such as described by Kirch (Chapter 6),
Junker (Chapter rn), and Schmandt-Besserat (Chapter 14).
Due to the private or corporate ownership of productive resource sites, people
who did not own property could only survive by aligning themselves with a
group in order to gain access to subsistence resources (Barnett l938a:130, 1955:244;
Boas 1921:1345-1347; Drucker 1939:59; Elmendorf 1960:268; Walens 1981:71). Exam
ples are cited stating that individuals caught using a resource site without the
proper right might be killed (Barnett 1955:244; Boas 1921:1345-1347; Boas
1966:35-36). "Service and deference were traded for economic security" (Barnett
l938a:130; also see Rosman and Rubel 1971:79).
Moreover, for free individuals, group membership was not static; members
were able to choose the group with which to affiliate according to the relative
1 93
James R. Perodie
benefits they would receive (Barnett 1938b:350-351, 1955:193; Drucker 1951:71, 279;
Elmendorf 1971:358-359; Rosman and Rubel 1971:173; Walens 1981:39). The treat
ment retainers received determined whether they remained with the current
house group, or aligned themselves with another (Barnett 1938a:129, 1955:246,
1968:45; Boas 1921:1333; Drucker 1951:453, 454). And as Drucker (1951:131, 280,
439-440, 454) and Barnett (1955:246) indicate, feast participation was a deciding
factor in whether an individual continued to align with a particular group. Feast
participation included the benefits of celebrating group membership, accessing
abundant or special feasting foods, and enjoying a party atmosphere. Drucker's
informants told him that individuals did not mind working for a chief because
they knew they would receive a feast in return (Drucker 1951:251).
The majority of the food distributed at solidarity feasts appears to have come
from the output of group members (Boas 1921:1334; Drucker 1951:251, 371). Boas
cites several examples of tribute amounts. Salmon fishers and berry pickers
might give 20 percent of their catches (Boas 1921:1335), bear hunters gave one for
every three caught (Boas 1921:1338), and mountain goat hunters gave five out of
ten goats obtained (Boas 1921:1334). However, ethnographic accounts do not
seem to detail specific distributions at solidarity types of feasts. On the other
hand, there are more detailed accounts for larger, more conspicuous feasts. Many
of the commonly known feasts such as naming feasts or puberty feasts were
probably not solely solidarity feasts. Many feasts would have had solidarity fea
tures because the sponsor's group would have participated. It seems likely that
feasts that were held for strictly solidarity purposes occurred regularly, were
small and informal with rank differences being downplayed (Barnett 1955:246).
For example, Drucker and Heizer (1967:37) cite an occasion when a chief re
turned from a potlatch and provided an informal feast to "his people" who did
not accompany him relating to them the festivities that occurred. Barnett
,
(1955:266) notes an instance when a feast sponsor thanked his household support
ers by making a small distribution to them the day after all invited guests had de
parted. These are probably typical examples of small household solidarity feasts.
Work-Party Feasts
Work-party feasts are also given by wealthy elites without an expectation of a re
turn feast. The feast sponsor provides the feast in exchange for labor on a specific
project such as building a house: "Building a house . . . [required] control of a
considerable amount of manpower, which in turn depended on economic re
sources to support the people while they worked, with enough surplus to give
feasts and other diversions to entertain them" (Drucker 1951:72).
Barnett (1938b:352), Boas (1925:339 ), Elmendorf (1960:152, 312), and Rosman and
Rubel (1971:163) observed that feasts were exchanged for labor on projects such as
1 94
S O U T H E R N N O RT H W E S T COAST F E ASTI N G
house building, post carving, and post raising. Details specifying where the in
vited workers were from, and the amounts of food and property that were dis
tributed, are minimal. In the instance Boas (1925:313) cites the house was being
built by one of the Fort Rupert septs (i.e., "subtribes") and the laborers were
members of the other three Fort Rupert septs. Barnett (1938b:352) states that la
borers were not from the house builder's own group. Elmendorf (1960:312) indi
cates that labor, and the distributed food and property, was provided by both the
house builder's group and other households. Work-party feasts were probably in
tracommunity events, and most of the distributions were probably produced by
the house builder's retainers. Alternatively, the distributed goods might have
been accumulated with credit exchange transactions. Loans were utilized for fi
nancing many types of feasts, including work-party feasts. Loans were a critical
mechanism for amassing resources and indebting others, and will be examined in
the equal-return section.
It seems likely that at least some lower-class members of the house builder's
group participated in the feasting activities, either as a result of providing labor
for the construction project or as a result of helping orchestrate the feast (see
Barnett 1938b:352). Therefore, some solidarity aspects were probably involved
with large work-party feasts. It is even possible that work-party feasts reinforced
alliances. For example, the relatives of the house builder's spouse might have
been among those who provided labor and participated in the feasting activities,
thus solidifying relations between the two families or groups.
Solicitation Feasts
It seems likely that on occasion individuals made solicitation feasts and distribu
tions in order to elicit favors or support from wealthy or powerful individuals.
These feasts were probably on a very small or medium scale with no return feast
expected. However, the only evidence I encountered came from Elmendorf who
notes that a feast distribution might have been made to an individual "whose
'heart one wants to grease' preparatory to a request" (1948:626 note). Solicitation
feasts were evidently not major occasions, and were probably subsumed with
other feast types (e.g., reciprocal feasts) or were too small to be of much impor
tance for most ethnographers.
E Q U A L - R ET U R N F E ASTS
Some preliminary observations on equal-return feasts should b e made before dis
cussing specific types of feasts in this category. Ambitious hereditary elites were
also the principal organizers and sponsors of equal-return and greater-return
feasts. However, because these feast types have different goals than no-return
feasts do, it is more likely that the principal recipients of equal-return and
1 95
James R. Perodie
greater-return feasts were themselves hereditary elites rather than retainers-in

contrast to many no-return feasts where the main recipients of food and gifts
were retainers. Thus, the principal recipients of equal-return or greater-return
feast distributions were hereditary elites from outside the sponsor's group (Bar
nett 1955:255, 266; Boas 1925:353-357; Drucker 1951:377; Elmendorf 1960 :337, 338,
443-444) . Depending on the occasion, guests might have included powerful cor
porate heads who were prospective defensive allies (Ferguson 1984:315) , families
or corporate leaders of groups with prospective marriage partners for the spon
sor's children (Elmendorf 1960:443) , or potential wealth exchange partners who
were judged by the sponsors to be good credit risks (Barnett 1968 :81-82) . With
feast invitations it was implicit that an invited guest could also bring other mem
bers of his household or village to the feast (Elmendorf 1960:339) . Hence, inter
group feasts often included numerous miscellaneous guests in addition to the
specifically invited guests. The choice of which household members accompa
nied chiefs to these feasts was undoubtedly another coveted reward that chiefs
used to control labor and supporters within their groups.
Furthermore, the distribution of food and gifts to the principal guests at inter
village feasts was given in ranked order and amounts (Barnett 1955:255, 1968:27;
Codere 1966:65; Drucker 1951:379, 1948:232; Elmendorf 1960:324, 341-342) . Intravil
lage feasts, such as solidarity feasts, apparently did not have ranked distributions
(e.g., Barnett 1955:255; Boas 1925:119-125) . However, it seems probable that ranked
distributions occasionally occurred (depending on the situation) if more than one
family or household group was involved.
In addition, as Hayden (1995:52) acknowledges, the return furnished at subse
quent return feasts might not always meet the original sponsor's expectations.
Drucker and Heizer (1967:79-80) write that a difficulty (at least in aboriginal times)
in assessing returns was caused by the exchange mediums: no two canoes, slaves,
native blankets, pelts, or olachon containers were identical (also see Barnett
1968: 84) . Consequently; an exact comparison between what was given, and what
was returned, was impossible. Therefore, an individual expected to provide an
equal return must have been obligated to return an "approximately equal" amount.
Promotional Feasts
Promotional feasts on the southern Northwest Coast can be considered either as
no-return or equal-return feasts. No-return promotional distributions included
the incidental food and gifts given to miscellaneous guests and participants who
were not principal guests, such as retainers accompanying the principal guests.
These distributions were part of the overall expense of sponsoring a feast, and
ensured the feast was a "good show," dissipated envy; guaranteed that everyone
felt involved, rewarded supporters, and so on (Hayden 1995:46, 68 ) . For example,
1 96
S O U T H E R N N O RT H W E S T COAST F E A ST I N G
Elmendorf (1960:324, 342) noted that commoners received token gifts of low
value after the ranked distributions to the main guests (also see Boas 1925:231-233;
Drucker 1948 :232, 1951:299, 438).
However, many equal-return feasts (including maturation and "life-crisis"
feasts) were equally; if not more, important as promotional opportunities be
cause they also involved distributions to specific prospective allies, marriage part
ners, or other important contacts. For example, potential suitors from a girl's
own village, and from other villages, were invited to her puberty feast and in
formed of her availability for marriage (Elmendorf 1960:443-444; also see Barnett
1955:151). The privileges to be inherited by her future children, and to be given at
the time of her marriage, might also be displayed at a girl's puberty feast
(Drucker l95r:r41) .
Aggrandizers also publicized their heirs, and simultaneously advertised group
success to prospective partners and supporters by sponsoring lavish naming
feasts for their children (Drucker 1951:124-125, 247; Elmendorf 1960:385-386) .
Naming feasts were expensive (Barnett 1955: 132), and the sometimes extravagant
distributions (e.g., 300 silver dollars [Elmendorf 1960:387]; $100 at one individual's
naming feast and $160 at another [Barnett 1955:139]) were intended to advertise
group success and the benefits of affiliation (see Barnett 1955:142, 143). In many; if
not most cases, the sponsors of these maturation, or other promotional feasts
would themselves be invited to similar feasts hosted by the guests who were
within the group of potential future in-laws or allies. Much of the distribution
costs of promotional feasts would be recouped at such subsequent feasts (Barnett
1955: 134, 255, 257) . Barnett does not specify a time span for such return invitations,
but return invitations were probably received within a few years. As a result of
the above considerations, I have decided to examine Northwest Coast promo
tional feasts in the equal-return section.
Promotional feasts were opportunities for aggrandizers to advertise group
prosperity; and the benefits of group affiliation, in order to attract productive or
skille d labor, allies, marriage and exchange partners, and other supporters with
critical abilities (Hayden 1995 :52) . This goal is recognized by Schulting (199s:r5,
183) who observed that some feasts were used to attract supporters. Barnett
(1938b:353-354), too, reported that feasts gained the sponsors publicity and recog
nition ("prestige" or "status") outside of their group and that this recognition was
required to achieve their aims. Barnett does not specify what these aims were, or
how publicity helped the aggrandizers achieve them, but the publicity probably
helped aggrandizers to attract labor, marriage, and exchange partners, and other
supporters, in the manner outlined below.
Skilled and powerful supporters were especially important for success in con
flict situations. The lack of powerful allies or supporters resulted in a weak pub-
1 97
James R. Perodie
lie image, and "perceived weakness invited attack" (Ferguson l984b :308). There
fore, promoting the group as being prosperous and well connected increased se
curity. In fact, Ferguson states that warfare might even be motivated by the desire
to obtain resources that could be redistributed to "attract individual followers
and ally groups" (1984a:49) .
As Drucker puts it, an aggrandizer's goal was "to attract lower rank people to
his house . . . [which] he did by good treatment, generosity (giving many feasts
and potlatches)" (1951:280) . Promotional feasts demonstrating the benefits of
group membership attracted the labor that aggrandizers required to harvest re
sources at sites they owned or controlled, to produce exchange goods, and to
support their feasting activities. Some aggrandizers also sought to attract follow
ers for military campaigns (Ferguson l984b :272), and promotional feasts would
have helped here as well. Aggrandizers were sometimes even willing to accept
slothful retainers because rejecting these individuals might provoke their more
industrious relatives to also leave (Drucker 1951:280).
Families of desirable workers, hunters, or craftsmen "would be courted [by a
chief] to the extent of giving them economic and ceremonial rights, to entice
them to associate" with him (Drucker 1951:280 ). And prospective tenants took ad
vantage of their position to promote their own self-interest. Residence "choices
are made on the basis of a type of strategy involving rank of seat obtainable,
rank of group involved, and its size and power" (Rosman and Rubel 1971:173; see
also Hayden 1995:66). Aggrandizers' promotional displays of wealth and prosper
ity influenced tenants' and lower elites' perceptions of which groups were most
successful, most powerful, and most likely to support tenants' aspirations.
Child-Growth Feasts
Wealthy hereditary elites sponsored the most lavish maturation and life-crisis
feasts (Barnett 1955: 134; Drucker 1951:139, 141, 143; Elmendorf 1960:38 8, 410, 443).
Maturation and life-crisis feasts were probably primarily promotional opportuni
ties for aggrandizers to advertise themselves and their groups. The leading guests
who were invited from other villages and the amounts of property distributed
varied with the family's wealth and ambitions (Barnett 1955:255, 257, 261; Drucker
1951:141, 148; Elmendorf 1960:385, 386, 443, 460 ). Poor families either kept these oc
casions private and did not have feasts larger than the nuclear family, or held very
small gatherings where they invited the chiefs of their household and delivered a
small gift (Barnett 1955:132, 143; Drucker 1951:124-125, I4I, 149; Elmendorf
1960:409, 410).
However, the puberty and naming feasts mentioned above also exemplify an
other aspect of maturation feasts: "child-growth" investments (Hayden 1995:44,
45, 54, 55, 59; Mauss 1967:6, 26; Oberg 1973:33, 35, 81, 121; Rosman and Rubel
1 98
l97r:r74, 187, 195, 198) . Maturation feasts for children such as naming (Barnett
1955 :134, 138, 139; Boas 1925:119-130; Drucker 1948 :208, 1951:124-125; Elmendorf
1960:385-386), ear piercing (Drucker 1951:124-125; Elmendorf 1960:410, 418), and
puberty (Barnett 1955:151; Boas 1966:370; Drucker 1948 :212, 213, 1951:139-145; El
mendorf 1960:410, 418) were essentially investments made by the parents with the
expectation of increased compensation or levels of wealth exchange at future
marriage and perhaps funeral exchanges (Hayden 1995:44, 45) . With every expen
sive maturation feast given for a child, the level of wealth exchange at their mar
riage undoubtedly increased, providing a means of recouping the initial feast
expenses. Barnett (1955:180) reports that wealthy families enhanced the value and
desirability of girls by secluding them to the extent that they could hardly walk
properly. Drucker (1951:143) makes similar observations.
In terms of investing wealth in children through maturation feasts, increasing
the cost of marriages serves the needs of aggrandizers by stimulating surplus
production and creating increasingly large debt relationships throughout the
community (Hayden 1995:44, 45, 54, 55)-a topic to be taken up again shortly. To
prosper at utilizing wealth to control others (Mauss 1967=73) aggrandizers need
competitors. Rivals also engage in production of surpluses, increasing the total
wealth available for manipulation and escalating the scale of contractual obliga
tions. Rosman and Rubel say it succinctly: "one needs one's rivals" (1971:159, 172) .
That is one reason "lazy men" are despised (Barnett l938a:131, 1955 :248; Drucker
1951:328; Hayden 1995:46, 48). Lazy people thwart the aggrandizer's ambitions by
not producing surpluses, by not exchanging wealth, and by "mooching" off elite
surpluses when in need. Child-growth investments probably at least occasionally,
and eventually, yielded a surplus return; however, due to the lack of immediate
returns or compulsory returns from the full range of elites invited (only some of
whom might actually enter into more formal reciprocal exchanges), as well as the
more uncertain nature of the returns, child-growth investments are examined
here under equal-return feasts.
Previously it was mentioned that production by retainers was one manner in
which food and property were amassed for feast distribution. A second important
method involved credit exchanges or loans. Individuals from all four of the North
west Coast cultures examined here amassed property by borrowing it (Barnett
l938a:135, 1955:135, 258-259, 1968: 81; Boas 1897=341, 196678-79; Codere 1966:69-7!;
Drucker 1948:233; Elmendorf 1960:330, 338, 358). The borrowed property was then
distributed to supporters of aggrandizers with a contractual agreement that an
equal or greater value to the goods loaned out would be given back to the ag
grandizer within a given amount of time or when needed. The initial loan was
then repaid when the ambitious borrower received his return goods from sup
porters at a future feast (e.g., Boas 196679; Elmendorf 1960:358). Some groups re-
1 99
James R. Perodie
quired that interest be paid on loans. The northern Coast Salish groups asked for
100 percent interest on a loan (Barnett 1938 :135, 1955:258-259), the southern Coast
Salish (Barnett 1955:258-259) and the Twana (Elmendorf 1960:330) did not require
interest, and the Kwakiutl had variable interest rates that depended on the term:
20 percent for less than six months, 40 percent for six months, and 100 percent for
12 or more months (Boas 1897:341, 1966:78-79; Codere 1966:69-71) . Furthermore,
an individual's "credit rating" (i.e., reputation, "status," past performance) affected
his borrowing capabilities and the interest he paid where applicable (Barnett
1968: 81; Boas 1897:341, 1966:78; Elmendorf 1960:330; Rosman and Rubel 1971:30 ). In
dividuals burdened with poor credit had to pawn their name for a year, and paid
exorbitant interest rates, even over 200 percent, to get reestablished (Boas 1897:341,
1966:78; Codere 1966:70 ). As Hayden (1995:52) states, it seems likely that loans were
not always returned according to these specified rates.
Even without interest on loans, there were obvious advantages in this system
for borrowers and lenders. Borrowers were able to amass property and indebt
others with their secondary distributions. Lenders increased their claims over
others, forcing borrowers to actively engage in ongoing production and wealth
exchanges. Lenders who were faced with defaulting borrowers might have been
able to "foreclose" on resource sites or other property that belonged to the bor
rower either in lieu of payment or until late payments and "late charges" were
paid. In some cases defaulters may even have been enslaved, or served with some
other potential penalty such as being forced to direct members from their house
hold to labor for their creditors. Furthermore, lending property and distributing
wealth at feasts enabled aggrandizers to maintain full production of food and
goods without worrying about storage, security, or spoilage (Hayden 1995 :45).
Importantly; an individual's ability to lend or borrow, his credit rating, the inter
est rates he paid, and the network of borrowing and lending contacts he had,
were all undoubtedly affected by the promotional feasts he sponsored, his past
record of transactions, and the degree of support from successful, hereditary cor
porate groups.
The contractual claims created over an individual by indebting them via loans
or feast distributions were exacting. An individual who failed to repay his debts al
most certainly lost allies, supporters, the chance for a good marriage, and his
credit rating or "prestige." More seriously; warfare or the threat of warfare was
often the result of defaulting on obligations (Hayden 1995:35, 40, 60; Mauss 1967:3,
11). If ordinary tenants had to pawn their "name" (i. e., reputation) for a year to as
sure repayment (Boas 1897:341, 1966:78; Codere 1966:70 ), elites must have had sim
ilar liens on each other to give force to these contractual debts. In extreme cases,
violence could be used. Ferguson states that "war is an exchange gone bad" (Fer
guson 1984a:17-18, 41), and that unfulfille d "social" obligations motivated war-
200
S O U T H E R N N O RT H W E ST COAST F E A S T I N G
fare. Mcllwraith (1948 :I:320, 11:376; Ferguson l984b:307) reports an instance when a
Bella Bella group destroyed a Rivers Inlet Kwakiutl village for failing to repay
their potlatching debts. Boas (1897:366, 1966:54) gives an example of an unveiled
threat to a father-in-law who was trying to evade repaying the marriage gift: the
son-in-law drowned a wooden effigy of his wife by putting a stone around the
image's neck and dropping it in the sea. And Drucker states that an individual's
name might be "seized if a debt was not paid" (1939:63). This is an extreme
penalty, considering that an individual's name represented all his economic and
ceremonial rights (Barnett l95s;r34; Drucker 1939:62) . Presumably an individual
was not enslaved when it was his name, not his person, that was seized although
this might have varied according to the amount involved in defaulting. Presum
ably; he minimally lost the valuable ceremonial songs, dances, and other privi
leges connected with his name, as well as ownership of productive resource sites
the name bestowed. Thus, potentially; a person's power and influence may have
been destroyed by not fulfilling his contractual feasting obligations.
Reciprocal Feasts
Reciprocal feasts were ongoing reciprocal wealth exchanges between groups that
initiated and maintained alliances between groups for the purposes of security,
marriage, and economic benefits (e.g. , access to resources or exchange partners) .
The ongoing nature of reciprocal feasts fulfilled several important needs for the
self-interest of aggrandizers (Hayden 1995:42-44, 47) : first, it reinforced the con
centration of wealth and power in the hands of those who already had it.
Wealthy and powerful groups tended to reciprocate and marry with each other,
further distancing themselves from lesser groups. Second, surplus production
was maintained, increasing the wealth aggrandizers could manipulate. Third,
continuous exchanges provided a perpetual opportunity for the creation of con
tractual obligations between elite groups, and also enabled aggrandizers to in
debt less wealthy individuals who borrowed wealth in order to marry; or to
establish other types of alliances that would advance elite sponsors and their in
terests.
Intergroup marriage feasts, and the ongoing feast exchanges they initiated are
good examples of reciprocal feasts (Barnett 1955:182, 183, 184, 190, 194; Drucker
1951:275, 286, 291-292; Elmendorf 1960:349, 358, 362) . The Twana, Nootka, and
southern Coast Salish had equal-return marriage feasts; the Kwakiutl, Pentlatch,
and Comox (northern Coast Salish) expected a surplus return and therefore these
feasts will be examined under greater-return feasts (Barnett 1938 :132, 1955:188, 190,
207; Boas 1966:51, 53; Codere 1966:63, 68-69; Drucker 1951:377, 381) .
In addition to wealth-exchange benefits, marriage feasts provided other bene
fits. First, they initiated and maintained security alliances between groups,
201
James R. Perodie
protecting one or both groups from attack and slave raids (Barnett 1955:182, 183,
270; Drucker 1951:302; Elmendorf 1960:349, 561; Ferguson 1984b :285). Second,
marriages with significant wealth exchanges established resource alliances that
enabled access to another group's resource sites (Drucker 1939:59, 1951:267, 268;
Elmendorf 1960:267-268; Ferguson 1984b:288; Suttles 1960:299). Drucker
(1951:248) cites an example of a rich salmon stream obtained by a chief through
marriage. Third, marriage feasts with wealth exchanges increased credit-ex
change opportunities because the spouse's relatives became additional contacts
for lending, borrowing, or other support (Barnett 1968:50, 51; Drucker 1951:430;
Elmendorf 1960:361-362). Barnett notes that if a groom's father did not have
enough property for the marriage gift to the bride's family "he called upon . . . his
wife's brothers . . . for their contributions to swell the total of the marriage gift"
(1955:185). Fourth, marriage feasts with wealth exchanges forged alliances that ex
panded an individual's political connections and helped increase political control
within and between groups (Barnett 1938:133, 1955:182; Jewitt 1987:19) . And fifth,
marriage alliances secured the transfer of titles and privileges from the bride's
family to the groom and his family. Among the Twana, Nootka, and most Coast
Salish groups, these titles and privileges were intended for the couple's future
children (Barnett 195p89, 1968:31; Drucker 1951:291; Elmendorf 1960:381, 384) . If
no children resulted, the titles and privileges reverted to the bride's family
(Drucker 1951:291) . However, despite the claim that the titles and privileges were
solely for the couple's future children, it seems likely that the groom or his fam
ily controlled the titles and privileges at least until the couple's children reached
independent maturity (Drucker 1951:266, 269; Barnett 1955:251).
Kwakiutl, Pentlatch, and Comox marriages contrasted with those of the
Twana, Nootka, and other Coast Salish groups, because the groom himself
overtly received and controlled the titles and privileges given by the bride's fam
ily (Barnett 1955:189, 290, 294, 1968 :31; Boas 1966:62-63, 71; Codere 1990:367; Ros
man and Rubel 1971:173) . As a result, the creation of a family was often a
secondary consideration compared to acquiring wealth, allies, resource rights,
privileges, and titles (Boas 1966:51, 56). Multiple (sequential monogamous and
polygamous) marriages were sought in order to accumulate titles and privileges
(Barnett 195p94; Codere 1990:368; Drucker 1948 :277) . Sham marriages between
two men, a man and a household dog, or one man and part of another's body,
were practiced solely for acquiring titles and privileges and presumably as a pre
text for exchanging wealth as well (Barnett 1938a:133, 1955:203; Boas 1897:359,
1966:55; Codere 1990:368; Drucker 1948:215) . The amounts of property exchanged
at large, elite marriages between corporate groups or villages clearly indicate
that it was the well-connected wealthy elites who were reaping the benefits listed
above (Drucker 1948:279, 1951:286; Elmendorf 1960:353, 361-363, 1971 :361) . Elmen-
202
S O U T H E R N N O RT H W E S T C O A S T F E A S T I N G
dorf (1960:550) states that group exogamy; not necessarily village exogamy; was
the principal concern behind elite marriages. Thus, in addition to establishing a
variety of alliances with other villages, elite marriages potentially brought in
creased access to allies, labor, and other supporters within a village. At one south
ern Coast Salish marriage exchange the groom's party delivered a marriage-feast
gift of $200 cash, 20 Indian blankets, 70 trade blankets, and 400 loaves of bread.
Three years later the bride's family returned l,ooo loaves of bread and $mo cash
(Barnett 1955:200) . The return was seemingly less than the original gift, thus this
exchange also illustrates that in reality an "equal return" might have been less
than the original gift, or equal to it, or a little more. Usually only food was ex
changed in the subsequent reciprocating feasts (Barnett 1955:192, 199; Drucker
1951:296-297).
The subsequent reciprocal feasts could go on indefinitely for as long as the
bride and groom wished to remain married, or for as long as their respective
groups wish to remain allied. It seems likely that an exchange occurred every few
years in order to affirm ties between the two groups involved (Barnett 1955:190,
194; Elmendorf 1960:358). After each complete gift and return cycle either party
was free to "divorce" the other without further payment, except of course for the
privileges that reverted to the bride's family if the couple was childless (Barnett
1955 :195; Drucker 1951:291, 302) . To perpetuate the marriage and the alliance, a
new cycle of exchanges had to be initiated by the groom or his representatives
(Barnett 1955: 195; Boas 1966:54-55) .
A groom's marriage gift might be returned immediately (Barnett 1955 :190, 200;
Drucker 1951:291), in one year (Barnett 1955:188, l9l, 199) or in several years, for
example after the birth of children (Barnett 1955:189, 200; Drucker 1951:291, 297;
Elmendorf 1960:355). An immediate return could only be made by the wealthiest
individuals, and would publicize their wealth and power. Alternatively; by delay
ing the return, the bride's father (or family head) was able to amass resources for
returning the marriage gift by distributing and investing the bridewealth receipts
among his own network of influential partners (Barnett 1955:198, 200, 1968:59;
Drucker 1951:291) .
The groom or his family did not always make the initial distribution in a recip
rocal wealth exchange. If a suitor was poor 'but skilled professionally and gave
every assurance that he would be an asset to the [bride's] family" (Barnett
1955:193), the bride's father might accept the suitor without a marriage gift. Alter
natively; a prospective father-in-law might discreetly loan a poor but desirable
suitor a modest sum for a marriage gift. In both of these cases the groom paid his
debt to the father-in-law by laboring for him, often staying in the father-in-law's
house as a member of his household (Barnett 1955:193, 203). Drucker and Heizer
(1967:72) even report an instance when a groom desired a larger marriage gift re-
203
James R. Perodie
turn than his father-in-law could or would provide. Ultimately, the groom him
self provided his father-in-law with the resources to make an enormous marriage
gift return. These situations do not seem to fit the typical pattern of ongoing re
ciprocal wealth exchanges between two groups, however, it is clear that mutually
beneficial alliances were created. These illustrations are good examples of norms
being bent to accommodate practical benefit considerations.
Distributing the receipts from the marriage gift accomplished at least three ob
jectives: first, the bride's father paid off any existing debts. Doing so perpetuated
his reputation as a good credit risk, ensuring he could borrow again. Second, the
bride's father created additional claims over individuals by indebting them with
his distributions. And third, production of surplus goods for the marriage gift re
turn could proceed without the burden of storing the marriage gift items. The
opportunity for both the bride's and groom's families to create contractual obli
gations was probably a primary reason for the lavish nature of elite marriages.
The feast given by a groom when he distributed the marriage gift payback was
"often the greatest of his career" (Drucker 1948:279) .
Considering the numerous benefits that resulted from reciprocal feasts, it is
easy to agree with Rosman and Rubel (197r:r73-174) that marriage involved a
strategy in which the resources, military strength, credit rating, positions, and in
fluence of a prospective spouse's family were scrutinized in order to maximize
the practical benefits that would accrue to the participating families (Drucker
1951:287-288; Elmendorf 1960:335). The fact that marriages were arranged for
elite children, frequently of a very young age, substantiates the inference that this
strategy was being followed by families (Boas 189?:362; Drucker l95r:r43, 287; El
mendorf 1960:353). Moreover, polygamous marriages enabled the wealthiest,
most ambitious individuals to multiply the advantages of this strategy (Barnett
1955: 193; Drucker 1951:301; Elmendorf 1960:367-368; Hayden 1995:43).
The marriages of commoners contrasted markedly with elite marriages, al
though of course, there was a continuum of intermediate marriages (Drucker
1951:286, 288-289, 292, 293; Elmendorf 1960:370). Lower-class individuals typically
did not marry outside of their own community (Elmendorf 1960:353, 404, 550).
Some commoners "just woke up in bed together" (Elmendorf 1960:370) indicat
ing little or no formal marriage ceremony, feasting, or ongoing wealth exchanges.
In other cases the parents made an informal agreement, invited a few guests to a
meal and distributed token gifts (Barnett 1955:192; Drucker 1948:279, 1951:292, 299;
Elmendorf 1960:370, 1971:361) .
Marriage feasts were not the only type of reciprocal feasts. This discussion has
focused on them because of the abundant and detailed ethnographic records of
marriage feasting and wealth exchanges. However, other feasts such as the
Twana Intervillage Eating Festivals (Elmendorf 1960:139--141) achieved similar re-
204
sults. I t seems likely that less formal and less well-documented feasting between
elites from different houses or corporate groups occurred and also achieved sim
ilar results.
Feasts for Political Positions and Support

Political support feasts and feasts to acquire political positions will be grouped
together because they were similar (although not identical), and because increas
ing political support was so tightly connected with many feasting types (e.g. , sol
idarity; promotion, reciprocal) that it is difficult to distinguish feasts solely for this
purpose.
The main feast that functioned as a formal condition for political advancement
is the funeral feast. Funeral feasts occurred when an elite heir announced he was
taking over the positions and property of the deceased (Barnett 1955:220; Birket
Smith 1967:20; Drucker 1951:148; Drucker and Heizer 1967=132), as well as the de
ceased' s credit and debt obligations (Drucker and Heizer 1967=132; Codere
1966:69; Hayden 1995:65) . Thus, the funeral feast was the heir's statement assur
ing other elites that he would keep the debt and exchange cycle operating (Hay
den 1995:65) . Importantly, the heir did not necessarily continue the deceased's
specific contractual relationships by distributing property to the deceased's for
mer exchange partners (Drucker and Heizer 1967=131). However, the heir could le
gitimately affirm his intention to maintain the debt and exchange cycle by
destroying property and by distributing property to commoners who were pres
ent at the funeral feast (Drucker and Heizer 1967:131) . (Drucker and Heizer do
not specify whether or not the commoners who received property distributions
were related to the deceased, but the heir's goals could probably be attained
whether the commoners were related or not.) Even in cases when a new con
tractual obligation was not initiated with specific elite individuals, the heir would
still be more or less reimbursed for his distribution when he became a leading
guest at future funeral feasts for other community members (Barnett 1955 :259,
261; Hayden 1995:66). Therefore, funeral feasts functioned as equal-return feasts.
Rosman and Rubel (1971:170) state that funeral feasts were important occasions
of succession among some northern tribes (e.g. , Tsimshian, Haida, Tlingit), but
in the south (e.g., Kwakiutl) the inheritance had already been received at feasts
during the lifetime of the original holder (Rosman and Rubel 1971:170, 171, 175) .
But the potential for competition between several rival heirs at funeral feasts con
tradicts Rosman and Rubel. Drucker and Heizer (1967:131, 132) state that competi
tion for a deceased's property and positions could occur, and Boas (1925:89)
implies that competition erupted if another individual doubted a prospective
heir's claims. In addition, although an heir's father or uncle might give the heir
nominal ownership of important seats and privileges, the father or uncle retained
205
James R. Perodie
control over the seats and privileges until their death (Barnett 1955:251; Drucker
1951:266, 269) . It was only after the funeral feast that the "apprenticeship" ended
and the heir began exercising control (Drucker 1951:145) . In order to permit heirs
to firmly establish their network of credit and debt, elite funeral feasts were gen
erally held one to two years after the actual death of a chief or high-ranking elite,
although interim funeral feasts might also be held to reassure rivals and creditors
of the heir's intentions and capabilities (Barnett 1955:219, 220; Drucker
1951:148-149).
At the funeral feast the heir needed to prove he was worthy of succeeding the
deceased. Sponsoring an impressive feast indicated that he had the necessary in
dustry, ambition, group support, wealth, and resources to do so. Boas
(192575-89) cited a situation in which an important Kwakiutl chief had died and
his nephew was to replace him . The men from thirteen Kwakiutl tribes were in
vited to the funeral feast, fed, and then 2,000 blankets were distributed among
the 658 seat holders of the thirteen tribes. Sponsoring this feast would have been
an enormous undertaking for anyone, and the heir must have had considerable
assistance from his supporters. The stakes were high. The heir's uncle's seat, a
house, two valuable coppers, four feasting dishes, and various other songs and
privileges could all have been forfeited if the heir had failed to display the re
quired control over the economic and social resources of the corporate group.
Depending on the deceased's influence, guests might be from other tribes
(Boas 1925 :75-89) or strictly from the deceased's village (Elmendorf 1960:459) . A
claimant to a more powerful or contested position needed to sponsor a more ex
travagant feast than a claimant to a lesser position. Drucker (1951:148, 149) reports
that for lesser chiefs and wealthy commoners the heir invited only the local chiefs
to a feast and gave them gifts. For lesser individuals, funeral feasts probably oc
curred shortly after death, and they probably exhibited more traits of solidarity
feasts for nuclear or extended families. It is doubtful that slaves had any funeral
feasts at all.
G R E AT E R - R E T U R N F E A S T S : T H E C O M P E T I T I V E F E A S T
Competitive feasts were the only feast type in which greater returns were re
quired on initial feast distributions, thus creating material profits for sponsors and
supporters alike. "Competitive feasting involving interest payments is clearly one
of the most common strategies used for extending personal power and wealth . . .
[on] the Northwest Coast" (Hayden 1995:58-59) . Investments and wealth ex
changes were necessary to increase wealth beyond the limited surpluses provided
by retainer labor (Hayden 1995:59). Boas (189r341, 1966:77) states that the under
lying principle of the potlatch distribution "is that of the interest-bearing invest-
206
S O U T H E R N N O RT H W E S T COA ST F E ASTI N G
ment of property." In addition to increasing wealth, competitive feasts enabled

aggrandizers to increase the scale of their contractual obligations with larger and
larger exchanges (Hayden 1995:52) .
Competitive feasts were sponsored by wealthy individuals (Elmendorf
1960 :338, 343) and involved enormous property distributions with hundreds or
thousands of people in attendance. Eells (Castile 1985:3rn, 323) documented 425
people and 1,200 people attending two specific feasts of this kind. The pretext
might be any of the common feast occasions such as marriages (Boas 1897:359;
Barnett 1955 :257), secret society initiations (Elmendorf 1960:337, 343), maturation
feasts (Barnett 1955:263), or spring intertribal meetings (Barnett 1955:255, 257).
Codere (1966:68-69, 70), Boas (1897:343, 359), and Barnett (1938a:132, 1955:188, 202)
all state that competitive feasting returns were expected to be twice the initial dis
tribution, although short-term loans within a household or extended family
might carry interest rates as low as 25 percent or none at all (Boas 1966:78).
As an example of a highly attended competitive wealth exchange with vast
amounts of property being transferred, Boas (1925 :236-357) recounts a succession
of wealth exchanges initiated by a marriage involving two very high-ranking
Kwakiutl families from two different Fort Rupert septs (see Table 7.1 [also sum
marized in Rosman and Rubel 197I: I61-163]). The ratio of property given to prop
erty returned seems obscure because they are not in identical currencies.
However, Boas (1966:54) assures readers that "the value of the goods paid (at a
marriage gift return) is far in excess of what the bride's father has received."
Sponsoring a competitive feast required years of preparation (Barnett 1968:4;
Boas 1898:682, 1925:236-357; Elmendorf 1960:338) with time periods of five to ten
years or longer generally being cited (e.g., Castile 1985:3rn; Hayden 1995:60). The
aggrandizer spent these years amassing property by financing smaller-scale loans
and feasts (Barnett 1955:258-259; Codere 196670 ), as well as by demanding sur
plus production from retainers. Ultimately, or ideally, all outstanding credits were
collected to finance one enormous feast distribution that would be invested for a
few years until being returned with a profit (Barnett 1955 :258, 259; Boas 1925:237,
347; Codere 1966:70 ).
Surplus returns enabled aggrandizers to pay off their supporters with a profit
in addition to increasing their wealth in terms of material property; titles, privi
leges, and the scale of their contractual obligations. Providing profits to support
ers motivated supporters to continue producing surpluses that aggrandizers
could manipulate, and attracted the surpluses of other parties interested in prof
itable exchanges (Hayden 1995:52) .
In addition, a profitable feast return of more property than could be amassed
by small exchanges or tenant labor provided the aggrandizer with a rare oppor-
207
James R. Perodie
TABLE 7.1
Ma rriage Wea lth Exchanges between Two High-Ran king Kwa kiutl Fa m i l ies
Occasion Property Distributed Invited Guests Comments
Groom's marriage gift 500 blankets 12 tribes Groom amasses blankets

by calling in outstanding
loans.
Father-in-law's 300 blankets; a Groom sells copper two

immediate return ceremonial title; the years later for 600 tins
"great copper" of grease. Grease used
named Sewa to finance a feast for
several tribes.
Groom's second 1 ,000 blankets Occurs four years after

distribution to the initial exchanges.
father-in-law
Father-in-law's return 300 blankets; 50 shawls; 10 tribes Occurs four years after in
20 pairs gold earrings; Groom's second
20 gold bracelets; distribution. Bride's
25 silver earrings; father amasses
25 silver bracelets; property by calling his
50 pair abalone shells; outstanding loans.
50 silk kerchiefs; Groom immediately
10 phonographs; distributes the goods
35 sewing machines; to the important
250 wooden boxes; chiefs of the 10
40 dressers; 50 masks; invited tribes.

250 dancing aprons;
25 canoes; 15 boats;
a box of crests
Source: Boas 1925:236-357. Presumably this exchange cycle continued until economic limits of increase were
reached and it became impossible to exchange escalating amounts.
tunity-the massive quantities of returned property could be distributed to

wealthy and powerful guests from other villages with whom the aggrandizer
wanted to indebt or establish a relationship for purposes of resource access, secu
rity, or political support. In Boas's (1925:236-357) account of the marriage between
two high-ranking Fort Rupert families, the groom used the proceeds from the
father-in-law's first return to sponsor a feast for several tribes (Boas 1925 :289 ), and
he distributed the father-in-law's second return to the ranking chiefs of the ten
guest tribes attending the event (Boas 1925 :353-357; also see Barnett 1955:255, 266,
1968 :59; Elmendorf 1960 :337, 338).
208
S O U T H E R N N O RT H W E S T C O A ST F E ASTI N G
Finally; the Kwakiutl took competitive feasting to extreme forms with rivalry
feasts "in which sums to be given back and forth were pyramided until one rival
was broken" (Drucker 1951:381) and significant property destruction occurred. In
some of these feasts, blankets were destroyed, grease and canoes were burned,
and slaves were kille d (Boas 1897=353-354, 1966:93, 98; Codere 1966:77-78,
1990:369) . No other groups destroyed significant amounts of property in compet
itive feasts (as opposed to funeral feasts), and only the northern Coast Salish
Comox approached the intensity of the Kwakiutl rivalry feasts (Barnett l938b:357,
1955 :265, 265-266, 1968 :81; Drucker 1948:232, 1951:377, 381, 383). Rivalry feasts were
probably the ultimate competitive arena for attracting and maintaining labor, po
litical support, exchange partners, and other supporters (Mauss 1967=72) . As Boas
(1897=353-354) reports, the broken rival lost his influence and undoubtedly his sup
porters and many economic rights. He was, in essence, bankrupt. Although often
characterized as the "typical potlatch," these feasts with increased surplus returns
were clearly extreme forms on the Northwest Coast, and reciprocal and no
return feasts were much more common.
C H A N G E S W I T H C O N TA C T
At this point there i s n o evidence t o indicate that the feasting activities described
by ethnographers are fundamentally different in principle from precontact feast
ing activities, except perhaps for the more extreme forms of competitive and de
structive feasting (Barnett 1968 :104, 105; Drucker 1951:457) . This is supported by
Codere (1966:v; 61) who argues that postcontact changes reflected existing ten
dencies in Kwakiutl culture. Thus, this analysis should provide a useful model for
the indigenous development of social inequality; rather than just documenting
European-induced adaptations in Northwest Coast culture.
Many differences between precontact and postcontact feasting have been cited.
Before European arrival (1) the goods distributed were native-manufactured
(Barnett 1955 :256-257; Codere 1966 :94, 1990:369 ); (2) property was distributed less
often and in smaller amounts due to difficulties amassing large quantities of na
tive goods (Barnett 1968:105; Codere 1966:124, 1990:369; Drucker 1951:376-377);
(3) fewer guests from other villages were invited, due to more hostile relations
(Barnett 1955:256-257; Codere 1990:369); and (4) only the wealthiest hereditary
elites could afford to sponsor a feast (Barnett 1968 :105; Codere 1990:368; Drucker
1939) . Depopulation, wage labor, and merchandise brought by Europeans in
creased the scale of feasting and enabled more people, including more common
ers and women, to be involved in sponsoring feasts (Barnett 1968:105, Codere
1966:125, 1990:363, 369, 371; Drucker l95I:I3, 181). However, the underlying motiva
tions and fundamental principles of feasting activities do not seem to have
changed from precontact to postcontact.
209
fames R. Perodie
CONCLUSION
Feasts were sponsored on the southern Northwest Coast to serve the sponsors'
interests and ambitions. They constituted important strategies and techniques
for promoting self-interest via the acquisition of wealth and power. Ambitious in
dividuals striving to increase their own power and prosperity needed to accom
plish several goals.
First, they needed to attract and bind labor to themselves or their group. High
quality labor supported the aggrandizers and their activities such as house build
ing and harvesting resources at privately or corporately owned resource sites.
Good supporters also increased group security and provided surplus food and
other products that aggrandizers could manipulate.
Second, ambitious individuals needed to form alliances with other powerful
groups in order to increase political support, increase security, and gain access to
greater exchange networks and marriage partners.
And third, aggrandizers needed a means to invest initial surpluses so that their
wealth and prosperity could cycle and multiply
Various forms of feasting accomplished all of these goals. Throughout this
paper references and examples of distributions were provided to demonstrate
that (r) the principal feast sponsors were wealthy hereditary elites, and (2) the
principal recipients of equal- and greater-return feast distributions were powerful
individuals capable of mutually assisting the sponsors in pursuing their goals for
potential exchange partners, political supporters, and security allies. One of the
key foundations underlying feasting distributions was contractual debt. Aggran
dizers' feasting distributions indebted retainers, allies, exchange partners, and
other recipients with harsh consequences for unfulfilled obligations.
Many proposals for the "function" of the potlatch exist, but as we have seen,
there is no single potlatch phenomenon. Rather, there is a wide range of feasts
with different purposes. Therefore, it is illusory to search for a general meaning
beyond that of promoting the practical self-interest of organizers and supporters
under conditions of extractable resource abundance. Social Structural explana
tions such as Rosman and Rubel's approach neglect the practical benefits that
provided the impetus for the Northwest Coast feasts. Suttles's argument that the
potlatch was a redistribution mechanism is inconsistent with Northwest Coast
ethnographic data documenting that individual motivations, competition, group
solidarity, and the control of surpluses were important aspects of potlatching.
Explanations that focus solely on the investment aspect of feasting provide an in
complete analysis of feasting because they do not address the numerous other
practical benefits or goals of feasting. Similarly, equating feasting with warfare is
misleading. Equating feasting with warfare may imply that at least some feasting
210
S O U T H E R N N O RT H W E S T COAST F E ASTI N G
participants were motivated by their own self-interest, but such explanations fail
to identify and differentiate the solidarity, promotional, investment, and other as
pects involved in many feasts and potlatches. Explanations based on "status" or
"prestige" are deficient because they emphasize vague psychological yearnings
for approval without acknowledging practical consequences. Thus, explanations
citing warfare, prestige, hereditary claims, or investments, offer incomplete or
unspecific answers, which by themselves create misleading impressions concern
ing the psychological motivations, strategies, and practical goals associated with
feasting. I suggest that the present surplus-based model of self-interested aggran
dizers integrates a much greater variety of theoretically and ethnographically
important factors and observations in a more satisfactory fashion than previous
models. It is only regrettable that even more detailed ethnographic observations
are not available on the critical variables that have been discussed. Perhaps the ul
timate utility of this approach will only be demonstrable by other cross-cultural
studies that support the same kinds of interpretations that I have offered here.
N OT E
r. Corporate groups can be defined as cooperative groups that are enduring and hold
joint rights and responsibilities in resources or property ownership and also maintain
sociopolitical relationships with other groups or individuals (see Hayden and Cannon
1982) .
REFERENCES
Barnett, H . G.
l938a The Coast Salish of Canada. American Anthropologist 4o:n8-14r.
l938b The Nature of the Potlatch. American Anthropologist 40:349-358 .
1955 The Coast Salish of British Columbia. Eugene: University of Oregon Press.
1968 The Nature and Function of the Potlatch. Eugene: University of Oregon Press.
Birket-Srnith, K.
1967 Studies in Circumpaci.fic Culture Relations, Vol. I. Kobenhavn: Munksgaard.
Boas, F.
1897 The Social Organization and Secret Societies of the Kwakiutl Indians. U.S.
National Museum Annual Report 1895, pp. 3n-738 .
1898 Final Report on the Northwestern Tribes of Canada. British Associationfor the
Advancement of Science Reportfor 1898, pp. 628-688.
1921 Ethnology of the Kwakiutl. Thirty-fifth Annual Report of the Bureau of Ameri
can Ethnology, 1913-1914: I and IL Washington.
1925 Contributions to the Ethnology of the Kwakiutl. New York: Columbia University
Press.
1966 Kwakiutl Ethnography. Chicago: University of Chicago Press.
21 1
fames R. Perodie
Castile, G. P., ed.

1985 The Indians of Puget Sound: The Notebooks of Myron Eells. Seattle: University of
Washington Press.
Clark, ]., and M. Blake
1994 The Power of Prestige: Competitive Generosity and the Emergence of Rank
Societies in Lowland Mesoamerica. In Factional Competition and Political De
velopment in the New World, edited by E. Brumfiel and ]. Fox, pp. 17-30. Cam
Codere, H.
1966 Fighting with Property: A Study of Kwakiutl Potlatching and Waifare 1 792-1 930.
Seattle: University of Washington Press.
1990 Kwakiutl: Traditional Culture. In Handbook of North American Indians, Vol. 7,
Northwest Coast, edited by W Suttles, pp. 359-377· Washington, D.C . : Smith
sonian Institution.
Drucker, P.
1939 Rank, Wealth, and Kinship in Northwest Coast Society. American Anthropolo
gist 41:55-65.
1948 Culture Element Distributions XXV!: Northwest Coast. Anthropological Records
Vol. 9, No. 3. Berkeley: University of California Press.
1951 The Northern and Central Nootkan Tribes. Smithsonian Institution Bureau of
American Ethnology Bulletin 144· United States Government Printing Office,
Washington, D.C.
Drucker, P., and R. F. Heizer
1967 To Make My Name Good. Berkeley: University of California Press.
Elmendorf, W W
1948 The Cultural Setting of the Twana Secret Society. American Anthropologist
50:625-633.
1960 The Structure of Twana Culture. Pullman: Washington State University Press.
1971 Coast Salish Status Ranking and Intergroup Ties. Southwestern journal of An
thropology 27:353-380.
Ferguson, R. B.
l984a Introduction. In Waifare, Culture, and Environment, edited by R. B. Ferguson.
New York: Academic Press.
1984b A Reexamination of the Causes of Northwest Coast Warfare. In Waifare, Cul
ture, and Environment, edited by R. B. Ferguson, pp.267-328. New York: Acad
emic Press.
Gosden, C .
1989 Debt, Production, and Prehistory. journal of Anthropological Archaeology
8:355-387.
Hayden, B.
1993 Archaeology: The Science of Once and Future Things. New York: Freeman.
212
S O U TH E R N N O RT H W E ST COAST F E A S T I N G
1995 Pathways to Power: Principles for Creating Socioeconomic Inequalities. In

Foundations of Social Inequality, edited by T. Douglas Price and Gary M. Fein
man, pp. 15-86. New York: Plenum.
Hayden, B., and A. Cannon
1982 The Corporate Group as an Archaeological Unit. journal of Anthropological
Archaeology 1:132-158.
Hayden, B., and R. Maneeprasert
n.d. Feasting among the Akha in Thailand. Ms. in possession of B. Hayden,
Simon Fraser University; Burnaby; British Columbia.
Herskovits, M. J.
1952 Economic Anthropology. New York: Alfred A. Knopf.
Jewitt, ]. R.
1987 TheAdventures and Sufferings of]ohnR.]ewitt. Vancouver: Douglas and Mcintyre.
Kamenski, A.
1917 Tlingit Indians of Alaska. Fairbanks: University of Alaska Press.
Lewin, R.
1986 Self-Interest in Politics Earns a Nobel Prize. Science 234:941--942·
Mauss, M.
1967 The Gift : Forms and Functions of Exchange inArchaic Societies. New York: Norton.
Mcilwraith, T. F.
1948 The Bella Coola Indians. 2 vols. Toronto: University of Toronto Press.
Mozino, ]. M.
1970 Noticias de Nutka: An Account of Nootka Sound in 1 792. Translated by I. H. Wil
son. American Ethnological Society Monograph No. 13. Seattle: University of
Washington Press.
Oberg, K.
1973 The Social Organization of the Tlingit Indians. Seattle: University of Washing
ton Press.
Piddocke, Stuart
1965 The Potlatch System of the Southern Kwakiutl: A New Perspective. South
western]ournal of Anthropology 21:244-264.
Rosman, A., and P. G. Rubel
197! Feasting with Mine Enemy. New York: Columbia University Press.
Schulting, R. J.
1995 Mortuary Variability and Status Dijferentiation on the Columbia-Fraser Plateau.
Burnaby; B.C.: Archaeology Press of Simon Fraser University.
Sproat, G.
1987 The Nootka. Victoria, B.C.: Sono Nis Press.
Strathern, A.
197! The Rope of Moka. Cambridge: Cambridge University Press.
Suttles, W
1960 Affinal Ties, Subsistence, and Prestige among the Coast Salish. American An
thropologist 62:296-305.
21 3
James R. Perodie
Walens, S.
1981 Feasting with Cannibals. Princeton: Princeton University Press.
Wohlt, P.
1978 Ecology, Agriculture and Social Organization: The Dynainics of Group
Composition in the Highlands of Papua New Guinea. Unpublished Ph.D.
dissertation, University of Minnesota, Minneapolis.
Young, M.
197! Fighting with Food. Cambridge: Cambridge University Press.
214
8
THE B I G DR I NK
F E AST A N D F O R U M I N TH E U P P E R A M AZO N
Warren R. DeBoer
In [the] pre- 1 849 period, the Kwa kiutl "potlatch" does not seem to be so developed or stri king
a n i n stitution that there is a need for a d isti nctive term for it. There is noth i n g rem a rkable
a bout the giving of gifts or even of ceremo n i a l ized giving i n human society; a n d ma rriages,
com i n g s-of-age, a n d meeti ngs with other vi l lagers or outsi ders a re so freq uently the occasion
for g ifts that Kwa kiutl practices . . . seem less particu l a rly Kwa kiutl than h u m a n .
Codere 1 96 1 :445
The feast-or its Spanish cognate la fiesta, which perhaps has a richer and more
encompassing connotation-has long been a focus of anthropological inquiry.
Yet feasts have been viewed in contrasting ways: as mechanisms for leveling or for
accumulating wealth and power; as devices for aggregating dispersed popula
tions or for reaffirming social .distance in concentrated populations; as ploys for
promoting solidarity or for surfacing festering divisions; as forums for accentuat
ing or for easing rites of passage; as contests in which an often ineffable "pres
tige" is accrued; and as the classic arenas in which otherwise latent worldviews
are exposed and dramatized. Given its protean character, the feast would appear
to be a delightfully ambiguous category, one whose members are linked by faint
resemblances but otherwise lack common properties. In its fuzziness, the feast
21 5
Warren R. DeBoer
could be likened to the category "game" (Lakoff 1987:16) . Codere's observation,

therefore, is apposite.
However ambiguous a general category, the feast may provide a useful entree
into the workings of specific cultural systems. Its very unboundedness and rami
fying nature resist interpretive closure and may thereby open a wide window into
what is revealed and concealed . in human behavior, the structures of meaning
that scaffold such behavior, and, as will be argued, archaeological residues of
both. Here I attempt to marshal the available evidence concerning the major tra
ditional fiesta of the Conibo and Shipibo, denizens of the central and upper
Ucayali basin in the Amazon region of eastern Peru. This fiesta, the ani shri!ati 1
("big drinking" in Conibo-Shipibo), was practiced as recently as the Inid-197os
( d'Ans 1994; Heath 1991; Roberta Campos and Joan Abelove, personal communi
cation, April 18, 1975) but, as far as is known, has been discontinued since that
time. The evidence is of three kinds: (1) interviews with Conibo-Shipibo inform
ants who remember the fiesta (the "memory culture" approach); (2) a fairly large
but unevenly reliable corpus of references from Inissionaries, explorers, and a
2
few early ethnographers; and (3) material correlates of the ani shri!ati that can be
tracked in the archaeological record.
In pursuing this project, I will focus on those central themes proposed by Hay
den and Maneeprasert, namely: "(1) the purpose of the feast; (2) the group in
volved; and (3) the amount of surplus that can be assembled for feasting
purposes" (1996:18). This tripartition, however, cannot fully enfold the complexi
ties of the ani shri!ati, and various sorties into other aspects of Conibo-Shipibo
culture will contribute context.
P U R P O S E O F T H E F I E S TA
The manifest purpose of the ani shri!ati was to mark and celebrate the period
when a girl or cohort of girls reached, or approached, marriageable age. This
straightforward statement, however, needs clarification. The fiesta, although
technically a "puberty rite," did not necessarily correspond to the onset of pu
berty as signaled by first menstruation. In part, this is due to the custom of
arranged infant betrothals, today rare but formerly more common, in which a
girl as young as six years can be proinised to an older, usually adolescent male
(Eakin, Lauriault, and Boonstra 1980: 80-83) . This practice possibly accounts for
the great variability noted for the ti.Ining of the ani shri!ati, ranging from six to ten
(Izaguirre 1928:234-236) through the more realistically pubescent ages of eleven
to fourteen years reported by several nineteenth-century observers (Pallares and
Calvo in Larrabure i Correa 1905-1909, 9:60; Sabate in Izaguirre 1922-1929,
rn:194-195) . The fiesta, therefore, celebrated marriageability more than puberty.
The word celebrate is perhaps somewhat infelicitous given that the focal event
216
F E AST A N D F O R U M I N T H E U P P E R A M A Z O N
o f the ani shri!ati was a clitoridectomy performed b y specialized female surgeons.

In the Spanish literature, these specialists are variously called ancianas ("old
women"), madrinas ("godmothers"), or sacerdotas ("priestesses") . 3 Rationale for
the operation is complex and leads to a consideration of Conibo-Shipibo mythol
ogy, sexual politics, precarious gender roles and, by extension, the entire fabric of
culture (Odicio Roman 1969; Roe 1982; Gebhart-Sayer 1987).
For archaeologists who often seem resistant to, or at least wary of, ethno
graphic sensibilities, let me point out that much of the following is explicated by
the Conibo-Shipibo themselves. That is, it is not totally a "deep-structural" con
coction on the part of foreign observers. A fundamental theme in Conibo
Shipibo thought is the opposition between "male culture" and "female nature."
As procreative beings of nature, females are regarded as oversexed. This excessive
concupiscence allies females with the disruptive forces of nature. In contrast,
males are the guardians of culture, moderation, and order. As contextually diag
nosed by Roe ( 1982, l988:n8-120) and more specifically targeted by Bertrand
Ricoveri (1996) , this theme has both conceptual and spatial referents that can be
visualized in terms of concentric circles (Fig. 8 .1) . An inner core constitutes the
male domain of culture and consists of houses flanking a plaza kept immacu
lately clean (by women!). The outermost circle constitutes the exotic, a zone of
libidinous females and male hunting. An intermediate circle has liminal proper-
GARDEN FOREST
Figure 8.1. Schema contrasting male-associated core with

female-associated periphery.
21 7
Warren R. DeBoer
ties: here plants are cultivated; "garden hunting" (Linares 1976) is pursued oppor
tunistically; and sexual intercourse, both conjugal and illicit, takes place.
Whatever else culture is about, it has to do with corning to terms with nature
and history and, in some sense, taming both. As with other Amazonian groups
(Murphy and Murphy 1974:230; cf. Bamberger 1974), Conibo-Shipibo mythic his
tory recounts a time when women were in control, possessed phallic flutes, or,
more literally; had penises. For the Conibo-Shipibo-correctly sensing the
anatomical correlates of this history-the clitoris is a vestige of an era when male
and female roles were reversed. Periodic excision is necessary to reverse the
vengeful resurfacing of female control and the concomitant reversion to nature
and disorder that it portends. This is not a worldview without angst. Accordingly;
vagina dentata symbolism and castration anxiety are rife among the Conibo-Ship
ibo as well as other Panoan-speaking groups (e.g. , Kensinger 1995 :237-246;
Siskind 1973). That these are ancient South American, if not pan-human, beliefs is
attested clearly in Chavin iconography of the Andean Formative (Burger
1992:figs. 178-179, 207) .
P R E PA R AT I O N
All sources agree that hosting an ani shreati was a long and arduous process.
Preparations could begin as much as two to three years before the event and en
tailed the clearing of gardens and the planting of manioc, camote (sweet potato),
and sugarcane (Samanez y Ocampo 1980: 80-81, Karsten 1955:156-157, Roe
1982:97-g8). From these cultigens, prodigious quantities of manioc beer and cane
4
liquor were brewed. As the fiesta approached, many additional demands had to
be met.
Men hunted in order to provide meat. As meat storage in the humid tropics is
short-term, captive peccaries, monkeys, juvenile tapirs, and curassows were
raised and fattened as pets in anticipation of the corning feast (Odicio Roman
1969:49-60, Illius 1985:586). Even manatees were caught, penned, and fed a special
diet of water plants (Heath 1991:7) . New pottery vessels, especially large beer
storage jars and beer-serving mugs, had to be manufactured (Samanez y Ocampo
1980:44; Tessmann 1928 :206-207; Roe 1982:97-g8, 100-101; Heath 1991:5) . Women
wove new clothing and beaded new ornaments, as protocol demanded that hosts
and guests alike be dressed appropriately (Sabate 1922-1929, rn:26g-271).
In addition, a large guest house was often constructed (Diaz Castaneda
1923:407-408; Illius 1985:584). Obviously; the core sponsors of an ani shreati (said
to number from one to five men, including fathers of the girls undergoing the
ceremony) had to be able to marshal and orchestrate considerable labor-labor
that, in turn, had to be fed and served beer, the basic lubricant of Conibo-Shipibo
social life.
218
I N V I TAT I O N
Conibo-Shipibo settlements range from single households (today, usually two ad
jacent structures, one a domicile, the other a kitchen) to communities number
ing in the hundreds. Historical and archaeological records indicate that this
variability in settlement size is an old one (DeBoer 1981) . Clearly, large settle
ments were in a better position to host large ani shreati, although smaller com
munities hosted diminutive versions of the same. In the 1880s, Samanez y
Ocampo (1980:80--81) estimated that a fiesta attracted two hundred to three hun
dred guests. My Conibo-Shipibo informants said that this was a typically sized
gathering. At the large Shipibo community of San Francisco de Yarinacocha,
guests arrived from as far as the Pisqui to the north and the Pachitea to the south,
a catchment of a few hundred kilometers by river (Roe 1982:98).
Odicio Roman (1969:49-60) furnishes otherwise lacking information on the
mechanics of invitation. Special heralds, called chaniti in Shipibo, canoed up
stream and downstream to announce the upcoming event. These emissaries
were received with ritualized hostility. The shoreline invitees would taunt, 'We
want to go see our other women," to which the chaniti would reply, "You brag
garts will get your necks cut" (as will be seen, not an idle threat) .
The geography of invitation implicates another deep-seated scheme in Conibo
Shipibo culture that, although cast as concentric circles in Figure 8 .2, should be
E N E MY
ZO N E
N EUTRAL
ZONE
Figure 8.2. Schema contrasting friends, affines, and

enemies.
219
Warren R. DeBoer
viewed in terms of river distance. The ring of extralocal invitation encompasses

potential affines and allies. This proximate, "neutral" zone can be contrasted with
a more distant periphery inhabited by nahua, the generic Panoan term for "for
eigners" or "enemies." That locality and the gradients of distance, both spatial and
social, that surround it should be so partitioned is not unfamiliar. Siskind
(1973:49-50) sketches a parallel model for the Panoan Sharanahua as does Erikson
(1994:26) for the Mayoruna; much farther afield, comparable examples are legion
(e.g. , Cole 1945:131 for the Ifugao; Boddy 1982:694 for a Sudanese village).
F I RST E N CO U N T E R S
Feasts can b e viewed as drama (Goldman 1964, Gregor 1977). The arrival of
canoe-borne guests to an ani shri!ati certainly resembled an act in a well-staged
play. The reception of guests was steeped in etiquette. Hosts, bedecked in finery;
went to the shoreline with mugs filled with welcoming manioc beer or cane
liquor. After this initial encounter, visitors, amidst a chorus of drums, flutes, and
panpipes, were escorted to the newly constructed guest house where more liba
tions were served. Confrontation was now choreographed. Host and guest males
formed facing lines, each male bearing a hardwood club (called macana in local
Spanish, huino in Conibo-Shipibo); light head taps were exchanged between the
contestants (Sabate 1922-1929, rn:269-271; Roe 1982:98---9 9 ) . Host and guest
women, in a parallel case of ritualized conflict, wrestled (Heath 1991:8). Such an
tagonistic "opening ceremonies" are a widespread feature in South American
ethnography (e.g., Clastres 1998:225-226) . In the Conibo-Shipibo case, it is inter
esting that such publicly acted h�stilities involve females against females, males
against males. Intersexual berating and bantering are evident but seem to play a
less focal role than among the Sharanahua (Siskind 1973:rn5) or the Cashinahua
(Kensinger 1995:57-58) who have elevated such ribaldry to high art.
Guest males then engaged in an archery contest in which the targets were a
white-lipped peccary; a curassow, or a spider monkey tethered to a large cross of
balsa wood (Odicio Roman 1969:49-60; Roe 1982:99-1m; cf. Eakin 1990:12-13).
Some of the older accounts state that the best archers had first claim to marry the
girls to be circumcised (Diaz Castaneda 1923:407-408; de Uriarte 1982:240) . In
contrast, Heath (1991:6--7 ), who actually witnessed one of the last ani shri!ati,
places the killing of these animals raised as pets later in the fiesta. In addition, she
states that it is the host males who speared the captured game, the final killing
being done by the very women who were to oversee the clitoridectomy. This was
done by clubbing the animal with flutes, otherwise male instruments (compare
Gebhart-Sayer 1987:254-259).
In terms of the schema in Figure 8 .2, the animals chosen for ritual sacrifice
were not a random set. Monkeys are regarded, perhaps with warrant, as shame-
220
FEAST A N D F O R U M I N TH E U P P E R AMAZO N
less fornicators. Peccaries are classic Amazonian symbols of unrestrained sexual

ity (Roe 1982:223) . Manatees and dolphins are viewed as the aquatic seducers of
bathing women (Roe 1982:51-52; Kendall, Trujillo, and Beltran 1995:27) . 5 Further
more, in the context of the ani shri!ati, these wild animals had been raised as pets,
that is, they had been domesticated and thereby had entered the "neutral" zone
of affinity. It is these animal affines that were first shot by males and then clubbed
to death by women wielding phallic flutes. As Heath tersely puts it: "The death of
the wild animals symbolises the subduing of woman's animal-like instincts"
(1991:8 ) .
S C H E D U L I N G A N D D U R AT I O N
Sources state that the ani shri!ati was ideally held under a full moon, although
how such convenient night lighting could be guaranteed is unclear (Girard
1958:244; see Roe 1982:104-105 for the symbolic signilicance of the moon). With
respect to duration, there are two modes that either represent a chronological
shift (considered to be the more likely) or a distinction between the Conibo and
their linguistic and cultural brethren, the Shipibo. For the Conibo, the ani shri!ati
is said to have lasted eight days with the clitoridectomy performed on the last
night (Pallares and Calvo, cited by Rippy and Nelson 1936:39; Diaz Castafieda in
Izaguirre 1928 :234-236) . For the Shipibo, the fiesta is said to have lasted three days
with the clitoridectomy taking place on the second night (Tessmann
1928:206-207; Karsten 1955:156-157; Roe 1982:40; Gebhart-Sayer 1987:243-259;
Heath 1991) . Samanez y Ocampo (1980:80-81) was certainly off the mark when he
claimed that the fiesta lasted two to three months.
DANC I N G A N D FIGHTING
The ani shri!ati was a multimedia festival with drum, panpipe, and flute music
specific to the ceremony, singing (Tschopik 1954) , feasting on the game slaugh
tered for the event, drinking to excess, and dancing until dead drunk (Pallares and
Calvo in Larrabure i Correa 1905-1909, 9:60) . Two stages in these festivities can
be recognized. The first involved rather affable unisexual dancing, the men in one
circle, the women in another (Diaz Castaneda 1923:407-408; de Uriarte 1982:241) .
At some point, the young girls to be initiated, wearing "dazzling" attire (Heath
1991:5 ) , joined as dance-masters, orchestrating the performance of the whole.
This brief appearance opened a second stage in which behavior, undoubtedly
fueled by inebriation, became more frenzied and ultimately more sinister. After
the public announcement of the clitoridectomy, men were said to go berserk, as
saulting each other with macanas, and attempting to down an opponent in order
to subincise the nape of his neck with a knife that was traditionally bladed with a
serrated toucan beak, more recently with filed metal. 6 This knife, customarily
221
Warren R. DeBoer
suspended around the neck for ready access, is called huishtiti (Samanez y
Ocampo 1980: 80-81; Diaz Castaneda 1923:407-408; Karsten 1955:157-158; Roe
1982:100-101; cf. Lauriault 1952) . Accusations of adultery were said to be the pri
mary instigation of these confrontations. Watching women at first encouraged
these scuffles but then attempted to intervene before more serious, but rarely
lethal, outcomes ensued (de Uriarte 1982:242; Karsten 1955:157-158; Odicio
Roman 1969:49-60). As an aside, I should note that in the early 1970s, when con
ducting archaeological work on the Ucayali, several of my older Shipibo work
men bore multiple nuchal scars, proud tokens of their virility, remembered or
otherwise. While men attempted to hold center stage in drunken histrionics at
tendant to the ani shri!ati, things were happening among women as well.
C L I TO R I D E C TO M Y
The actual operation, the nominal purpose of the fiesta, deserves a separate
paper, which is another way of saying that details are sparse when it comes to se
cretive aspects of Conibo-Shipibo culture, although both female and male anthro
pologists have considered the "problem." As indicated earlier, the operation was
performed by women skilled as surgeons. Called shatoti7 (lllius 1985:586), it took
place in a special structure situated well away from the main plaza fille d with danc
ing and otherwise hysterical men (the "culture" zone). This removed structure
was variously called quischiquepiti (Pallares and Calvo in Larrabure i Correa
1905-1909, 9:60) or pU.Shuva (Karsten 1955=155-156; Girard 1958:240) . Note that this
structure (essentially a medical hut) was situated in the "neutral" zone. In this
sense, it resembles the tanpo (a term of possible quechua origin), or potter's shed,
where raw clay and other materials are converted into cultural form. The young
girls were anesthetized with ample draughts of cane liquor, a bit stronger than
manioc brew: Surgical implements were heated and presumably sanitized in a spe
cial clay pot, of which more later. The excision of the clitoris was done with the
sharpened nails of the sacerdota (Izaguirre 1928 :234-236), with a sharp cane knife
(Eakin, Lauriault, and Boonstra 1980:79-80), or, as one of my male informants
averred, with the metal lid of a discarded tuna can. The wound was washed in a
heated solution of piri-piri (various plants of the Cyperaceae family-see Tournon
1984, Eakin, Laurialt, and Boonstra 1980:63-64), and then a fired clay object, the
shi!rvi!nanti (males call it bushi, or penis), was inserted between the vaginal labia to
prevent cicatricial complications (d'Ans 1994 relays some typically out-of-focus
photographs; also see Karsten 1955:157-158; Roe 1982 gives a different slant).
THE MORN ING AFTER

"Concluido el masato y los viveres, todo el mundo se dispersa" (Samanez y
8
Ocampo 1980:80-81). When the beer and food are gone, everybody leaves.
222
Samanez y Ocampo got i t right b y placing beer first. Food seems a lesser concern.
Like the Cubeo (Goldman 1966: 83-84) , drink, not food, fueled the ani shri!ati. If
we look at the Panoan-speaking landscape hugging the Ucayali basin, we find
that manioc beer tends not to be produced among interfluvial groups such as the
Cashinahua, Yaminahua (Townsley 1994:280 ) , or Cashibo (Frank 1994:211-212) .
Among the latter, for instance, meat sliced from the captively raised tapir is a cen
tral reward of the major feast. In contrast, among the Conibo-Shipibo, food-to
play with a famous dictum-is good to think, while beer, and the altered states it
can induce, is good to drink For hosts, not to be able to offer more than guests
.
can drink is cause for mortification and entails a consequent lowering of social
and political status registered by host and guest alike (cf. Goldman 1966: 83-84 for
the aptly named Cubeo "drinking party") .
. T H E A R C H A E O LO G I C A L S I G N AT U R E
With typical aversion to understatement, the late Donald Lathrap argued:
On surveying the average size of each of the functional classes of pottery used by
the various surviving Tropical Forest groups, one is led to the conclusion that a ves
sel with a maximum diameter of more than 40 cm is very likely to be a fermenta
tion vessel. The presence of such large containers in an archaeological deposit can
be taken as evidence that beer was present and that the whole fiesta pattern was in
full swing. (1970:55)
Lathrap was clearly basing this judgment on work among the Shipibo, a project
continued by several of his students. As documented elsewhere (DeBoer and
Lathrap 1979; DeBoer in press), the Conibo-Shipibo ceramic assemblage reduces
to four functional classes and three size modes. As illustrated in Figure 8 .3, the
functional classes are cooking pots (or, more technically, pots placed over a fire
and therefore unpainted), liquid-storage jars, food-serving vessels, and beer
mugs. The size modes are: (1 ) large (ani, which can be glossed as "big" or "grow
ing") vessels that are typically, but not exclusively, used in fiesta contexts; (2) "or
dinary" -sized vessels (sometimes marked by the ranked subordinate terms anicha
and anitama) used in quotidian activities of cooking, eating, and drinking; and (3)
small vessels (vacu, or "kid-sized") serving as portable containers for canoe travel
ers. The resultant twelve-celled matrix, diagrammed in Table 8.1, not only sum
marizes recent Conibo-Shipibo pottery but also can be applied to the
archaeological ceramics of the Cumancaya phase, dating to about A.O. 900-rnoo
(Raymond, DeBoer, and Roe 1975) . As the Cumancaya site proper is identified as
an ancestral settlement by the Combo-Shipibo (Loriot and Hollenbach 1970 ) , the
case for historical continuity would appear to be strong. The case is made
stronger still by the co-occurrence of items specific to the ani shri!ati in both mod-
223
- kenti a n i
ken t i a n i tama
I
"
kenti vacu
f/W4I W/4 I
o cm 20
kenpo a n i ken po
flLJ
kenpo vacu
kencha vacu
P a s c u a kencha kencha
Figure 8.3. Conibo-Shipibo vessel forms. Unpainted cooking pots are at top, painted serv
ing vessels below.
F E A S T A N D F O R U M I N T H E U P P E R AM AZO N
TABLE 8 . 1
Classificatio n of the Con ibo-Shipibo Cera m i c Assemblage,
with Average Ca pacities in Liters
Serving Cooking
Food Bowl Beer Mug Jar Olla
Small for transport kencha vacu kenpo vacu chomo vacu kenti vacu
0.7 1.2 2.4 5.0
Medium for quotidian kencha kenpo chomo kenti

1.7 2.8 anitama anitama
16.5 11.1
kenti anicha
29.0
Large for fiesta pasqua kencha kenpo ani chomo ani kenti ani
3.8 7.5 103.3 80.2
Italics indicate Conibo-Shipibo terms.
em and archaeological contexts. These include rattle-based beer mugs (Ray

mond, DeBoer, and Roe l975::fig. 45k; Heath l991:n) as well as several supplemen
tary items that are activated specifically in a fiesta context.
Notable are the toncoate that were suspended on strings in order to withdraw
beer from the bottom of a chomo ani. These vessels, alternatives to tipping beer
kegs as they approach empty, took a symbolically charged form. They are mod
eled as dolphins (Fig. 8.4-A)-notorious riverine rapists-as testicles (Fig. 8 .4B),
or, as shown in Tessmann (1928:Tafel 58), as testicle- and penis-crunching turtles
(Kensinger 1995:237-246) . Although toncoate have not been identified in prehis
toric archaeological sites, serving vessels bearing dolphin adornos are a conspicu
ous feature at Cumancaya, notably in ceremonial deposits composed of smashed
pottery vessels (Roe 1973).
Paraphernalia either specific to-or activated during-the ani shreati include
the cane knife for excising the clitoris (Fig. 8.4C), the men's knife for subincising
the neck of accused adulterers (Fig. 8 -4D), and the clay plug, or shervenanti, in
serted into the girl's vagina after the surgical removal of the clitoris (Fig. 8 .4E) .
The latter has been found in recent Conibo-Shipibo middens as well as prehis
toric ones.
Other accoutrements attendant to the ani shreati include a small painted olla,
the shervenanti kenti (Fig. 8 .5A) . This form, known only ethnographically, was
used to sterilize implements used in the clitoridectomy as well as the water for
cleaning the wound. As a painted cooking pot, this form violates conventions in
225
Warren R. DeBoer
o
r/A !VJ
cm
rm
5
�E
�
Figure 8-4. Artifacts associated with the ani shreati. A. Toncoate, or suspended beer dipper,
in the form of a river dolphin. American Museum of Natural History, Catalogue Num
ber 4 0.0/ 4770. B.]obosko toncoate, or beer dipper modeled as testicles. This specimen, un
painted and unfired, was recorded in San Francisco de Yarinacocha in 1 97 1 . C. Paca kiintsii
(Roe 1 9 82: 3 2 7), or cane knife used in clitoridectomy. Specimen formerly on display at the
American Museum of Natural History, New York. D. Traditional huishtiti, or men's knife
with toucan beak blade (after Tessmann 1 928:fig. 1 4). E. Clay plug, a probable sherve
nanti, found at the Cumancaya site (after Roe 1 973 :fig. 63 f-g).
which a pot placed over a fire should not be so decorated. It can be treated as a
standard case of sacred reversal, as widely attested in world ethnography (the
"contraries" of the Cheyenne-Grinnell 1923, 2:7!)--86-and the "opposite
talk.ers" of the Zuni-Tedlock 1979:503-come to mind) . Another vessel form, a
small, shallow bowl called nane ati in Shipibo, held the black pigment used in
painting the faces of girls undergoing the clitoridectomy. Roe (1973=173-174) has
identified what is likely to be an archaeological example of this form at the site of
Cumancaya (Fig. 8 .5B) .
226
Figure 8.5. Pottery vessels associated with the ani shreati. A. Shervi!nanti kenti, or vessel
used for heating water for the clitoridectomy. Recorded at the Conibo hamlet of Iparia in
1 97 1 . B. Shallow bowl with resist decoration on interior, a probable nane ati or paint
bowl. From the Cumancaya site (after Roe 1 973 :fig. 63 a-b).
Given the vagaries of preservation in a humid, tropical environment, the

above list of artifacts shared between contemporary fiesta-associated assem
blages and imputed fiesta contexts of the past is impressive. The shervenanti,
rattle-based beer mug, and nane ati would appear to close the case for continuity,
at least as expressed in specific artifacts. The case can be strengthened further,
however, by looking at patterns. As Lathrap noted in the passage opening this
section, large vessels in the tropical forest are often feast vessels. Well, how large?
227
Warren R. DeBoer
Samanez y Ocampo (1980: 80-81) estimated that 30 to 40 "enormous" beer jars

had to be manufactured for the ani shreati. Farabee (1922:85-86) noted that the
mother of the girl undergoing the operation made a large jar holding 20 or more
gallons (So liters or so), while Roe's (1982:97-101) informants indicated that 15 to
20 such jars were needed for a typical ani shreati. Although emphasizing the sym
bolic rather than the volumetric, Gebhart-Sayer (1987: 86-91) casts similar ver
dicts. As given in Table 8.1, the mean volume of large beer jars (chomo ani)
manufactured by the Conibo-Shipibo in 1971 was about 100 liters. Thus volumet
ric estimates of beer produced for an ani shreati might range from 1,500 to 4,000
liters, figures that correspond to Alvarez's (1970:50) comment that 3,000 liters of
manioc beer could be consumed readily at a Piro festival within a few days (cf. Il
lius 1985:584) . In size and volume, jars and serving vessels from the thousand
year-old Cumancaya and Sonochenea sites conform to their modern descendants
(DeBoer in press) .
No fossilized ani shreati, needless to say, has been "dug up"; however, ceremo
nial deposits, in addition to common middens, have been sampled in excavations
conducted along the Ucayali. Ceremonial deposits include linear arrays of
smashed ceramic vessels, conjoinable into whole pots, and urn features contain
ing either primary or secondary burials. Table 8.2 presents the occurrence of beer
vessels (kenpo, chomo ani, and their prehistoric "cognates") and "other vessels" in
a range of ethnographic and archaeological contexts. For contemporary Conibo
Shipibo as well as for prehistoric midden and burial contexts, beer vessels consti
tute about one-sixth of all ceramic containers. In "smashed vessel features,"
however, this ratio is essentially one. Although these smashed vessel features
TAB LE S.2
Distribution of Beer and Non beer Vessels i n Ethnographic
and Archaeological Contexts
Beer Vessels Other Ratio of Beer I Other Vessels
Conibo-Shipibo vessels in use 42 220 - 1 :5
Sonochenea midden (vessels) 29 134 - 1 :4.5
Cumancaya midden (sherds) 703 2,127 - 1 :3
Shahuaya midden (vessels) 33 1 72 - 1 :5
Urn burials, Sonochenea and 7 25 - 1 :3 . 5
Cumancaya
Smashed vessel features, 14 16 -1:1
Sonochenea and Cumancaya
Data from DeBoer (1971), Roe (1973), Raymond, DeBoer, and Roe (1975), and DeBoer and Raymond (1987).
228
F E AST A N D F O R U M I N TH E U P P E R A M AZO N
. have been interpreted as mortuary deposits covering extended primary burials,

there is little conclusive evidence to support this interpretation (Raymond, De
Boer, and Roe 1975:59) . Their ceramic "signature" clearly indicates a feasting de
posit in which beer-related vessels predominate.
The smashing of beer mugs and other vessels as ceremonial endings to fiestas
such as the ani shri!ati runs as a motif throughout Conibo-Shipibo ceremonialism
(DeBoer and Lathrap 1979:135, note 8). The fact that such ritual closings, domi
nated by beer vessels, should be so conspicuously represented at archaeological
sites such as Cumancaya and Sonochenea is further evidence for the antiquity of
the Conibo-Shipibo feasting pattern. And there is more.
Decoration, specifically painted or incised designs (qui!ni!a), plays a basic role in
Conibo-Shipibo worldview, and it would be surprising if such designs did not fig
ure in such salient and culturally central events as the ani shri!ati-and they do, al
beit in contradictory ways. On the one hand, decoration was made flamboyant
and otherwise visually impressive in such public events (DeBoer and Moore
1982) . On the other hand, design applied to the faces of girls undergoing the
clitoridectomy was reduced to dots and crosses, the same designs adorning the
interior of the nani! ati, or paint bowl, for the initiates (Roe 1982:322; Gebhart
Sayer 1985 :594). Design simplification also characterized the shi!rvi!nanti kenti (Fig.
8 .5A) and the crudely incised shi!rvi!nanti proper from Cumancaya (Fig. 8 .4E) . In
terms of Conibo-Shipibo aesthetics, these designs are depauperate; they allude to
an uncivilized and designless world, a peripheral netherland occupied by various
brands of subhumans exemplified by the Cashibo and Amahuaca; the intermedi
ate or interfacial zone once again serves as a middle ground, a kind of buffer in
which designs are amplified to secure boundaries and are simultaneously simpli
fied to acknowledge the permeability of all boundaries. 9
Over the last thousand years, the Ucayali basin-a zone stretching a thousand
kilometers north to south along the eastern base of the Andes-yields convinc
ing ethnographic, ethnohistorical, and archaeological evidence for the focal role
of feasts. This combined data set is somewhat privileged, in that archaeologists
have routinely worked with informants whose ancestors produced the remains
in question. Without such guidance, archaeologists are likely to succumb to uni
versal and obj ectivist accounts that merely subsume or explain away specific
facts.
D I S C U S S I O N A N D CO N C LU S I O N
When seeking the latent functions of the ani shri!ati, one confronts a veritable si
lence in ethnohistoric and ethnographic accounts (but see below). If one were to
ask a Conibo-Shipibo man what the "evolutionary consequences" of this fiesta
might be, a likely response would be "what's more important than women?"-an
229
Warren R. DeBoer
answer that would please the sociobiologically inclined but would leave unad
dressed the specific form that the ani shri!ati takes.
In pursuing a more satisfactory account, it is instructive to cite a casual com
ment offered by de Uriarte: "From this competitive fiesta emerged future leaders,
future spouses, and future rivals. If the hosts wanted to attract skillful sons-in
law; the male guests were equally intent on obtaining wives" (Ortiz 1982:240, my
translation) . To appreciate this insight, additional ethnographic context is re
quired. The Conibo-Shipibo generally follow matrilocal residence. Therefore, the .
ani shri!ati not only acted as a demographic signal of the availability of nubile fe
males but also served as a vehicle for attracting economically essential sons-in
law-men who fished and hunted, cleared gardens, constructed houses, and
would come to sponsor the next generation of ani shri!ati. Yet another factor in
tervened.
Until the 1930s and even 1940s, the Conibo-Shipibo were avid raiders, pillaging
backwoods groups such as the Amahuaca, Isconahua, Campa, and Cashibo for
slaves, booty; and wives. At least for the Conibo, such raiding was already com
monplace in the late seventeenth century (DeBoer 1986) . Wives captured in these
raids constituted an obvious exception to the otherwise general practice of mat
rilocality: That is, men not only married-in, but also made up a military corps that
could and did abduct wives from the outside. Even today; genealogies record the
"recruitment" of such foreign women and can trace their pedigree to living de
scendants.
From the Conibo-Shipibo perspective-one shared by their mestizo neigh
bors-these foreign women, although often treated as second-class citizens and
10
drudge labor, were nonetheless exotic, untamed, and decidedly sexy. Whatever
sexual fantasies were involved, both inside and "recruited" forms of female labor
produced the prodigious quantities of beer needed for fiestas and gave birth to
children-preferably females-who attracted another cohort of sons-in-law. Dur
ing this replicative process, ideally timed to a full moon, females lost their collec
tive clitoris for the effort. Something has got to be missing.
These ethnographic addenda force reconsideration of earlier schemata in
which male culture, centered on the house plaza, was contrasted with an outer
female zone of nature and wilderness. In quotidian terms, the reverse situation
holds! Women ruled the home front and were part of enduring residential units
of related females, while men were in-marrying strangers typically lacking sup
portive kin, were obliged to observe mother-in-law avoidance, and otherwise oc
cupied precarious social positions. As Brown (1963) observed some time ago,
matrilocality can be a rather formidable rite of passage for men. Warfare and the
female booty it provided were, among other things, male palliatives in this
11
predicament as it is rendered in Figure 8.6.
230
SAVAG ES
' AN D SLAVES
I
,
EXTRA
ORDINARY ORDI NARY
SEX SEX
Figure 8.6. Schema for marriage and sex. Dotted lines in
dicate two major flows of marriage partners: (1) the
customary practice of matrilocality whereby a husband
moves in with his wife and her family; (2) the capture of
women from enemy groups.
In Table 8 .3, the schemata used to order a welter of disparate data are con
densed in terms of polarities and intervening gradients. By and large, these tri
adic sets are less than astonishing (which accounts for their ready translation) . At
least one pair of seeming contradictions, however, stands out. How can fe
male(ness) simultaneously be unbridled nature and the epitome of domesticity?
Or, how can male(ness) be the custodian of culture and, at the same time, the
TABLE S.3
Synopsis of Schematic Pol a rities:The Con ibo-Shipibo Point of View
Inner (Civilization) Outer (Savagery)

house garden forest
culture culture-nature nature
friend affine enemy
above sex sex dangerous sex
design design amplified or simplified designless
male male /female female
female male
231
Warren R. DeBoer
perpetrator of rapine? Such ambiguity, however-anathema to logicians and to

objectivist notions of science-is the way in which humans work and the manner
by which human phenomena are played out. This would seem to be particularly
the case when sex--or its cultural representation, gender-is confronted. In his
oft-cited yet untired verdict, Giddens put it this way: "Don't look for the func
tions social practices fulfill, look for the contradictions they embody" (1979:131) .
One only has to reread Naven (Bateson 1936), or this essay for that matter, to be
reminded that human bodies and behavior are infected by transfiguring and oth
erwise queer properties.
Another matter that needs to be taken into account is the seeming longevity of
the ani shri!ati, a full millennium if our reading of the archaeological record is
correct. Over this millennium, Conibo-Shipibo society and environment have un
dergone major changes including the wrenching dislocations effected by Euro
pean colonialism in its various sequent forms (San Roman 1975 provides a
review). In fact, whether one can speak of the Conibo or Shipibo a thousand
years ago is quizzical. As ethnic appellations that mark real-world entities, these
labels are clearly inventions forged, developed, and negotiated during the period
in question. Yet the ani shri!ati survived these storms until the 1970s. The interest
ing issue, therefore, is not the "evolutionary potential" of this particular institu
tion so much as its perdurance in the face of worlds being remade. The
remarkable stasis of the ani shri!ati suggests an institution that is anchored se
curely to the deepest values of Conibo-Shipibo culture in its becoming and main
tenance. Perhaps stasis and replication are as much evolutionary processes as
change itself.
The ani shri!ati remains somewhat fuzzy with respect to the driving themes of
this volume. This fiesta surely played a role in social and biological reproduction,
prompted production of drink and other material goods that otherwise would
not have been produced (although such "surplus" was dissipated in the gorman
dism of the fiesta), attracted labor essential to the planned well-being of hosting
communities, but less clearly acted as a crucible leading to increased cultural
complexity, however measured. Other than promised bodies, exchange of gifts or
other material goods was not emphasized. Valued and increasingly essential
items such as machetes, axes, chainsaws, or outboard motors simply were not
written into the script of the fiesta, although the ani shri!ati may have formed (or
thwarted) social relations by which such transactions could be pursued in other
contexts.
In short, the ani shri!ati appears to have had primarily an equilibrating rather
than transformative role in Conibo-Shipibo culture. The demise of this focal in
stitution in the 1960s and 1970s was a symptom of emergent opportunities for a
new brand of aggrandizers who stepped outside the traditional fold and its lin-
232
F E AST A N D F O R U M I N T H E U P P E R A M AZO N
gering and increasingly embarrassing vestiges of "the primitive." This new breed,
as schoolteachers or political activists, increasingly turns to interfacial roles with
the national economy, the international market with its rapacious desire for
primitive art and extollations of the native point of view, and the entailed need to
deal with an onslaught of meddling visitors, including archaeologists and eth
nographers. If not all-purpose fulcra for weighing the course of culture histories,
feasts like the ani shreati nonetheless provide gauges for sensing the meaning con
gealed in the special moments that aggregate human populations, expose the ag
onies that accompany human existence, and underscore attempts to reset a
world always in danger of going off course. As explored here, feasts-and the
bodies and brains behind them-leave detectable traces along the great water
way of the Ucayali, as they do elsewhere.
NOTES
r . Also known as pishta (Odicio Roman 1969) and wake honeti (Karsten 1955=15y-156). As
evident in citations, the present study of the ani shri!ati leans heavily on the masterful
and imaginative synthesis given by Roe (1982). Dr. Roe has also made available the
important paper by Carolyn Heath (1991), who witnessed and recorded one of the
last ani shri!ati held at the settlement of Sinuya in 1976. Unfortunately, this significant
document remains unpublished.
2. These accounts are often difficult to assess, as it is clear that authors were regularly
using earlier sources, often without citation. I have tried to be critical in selecting
what appear to be original documents, but, on another occasion, it would be useful
to construct a detailed phylogeny of original and derived accounts.
3. In 1975, Roberta Campos and Joan Abelove, both of whom were then working
among the Shipibo of the Pisqui, told me that the last clitoridectomy had been
performed three years before when the last knowledgeable female surgeon
("maestra") was still alive.
4. What portion of agricultural production was devoted to these periodic obligations is
unknown. For the Cubeo, Goldman estimates that as much of one-third of a manioc
crop could be destined for beer necessary for hosting ceremonies (1966:86) .
5. Heath observed women wading into the enclosure in which a manatee was kept
(1991:7). Before killing the captive animal by plugging up its nostrils, to facilitate
drowning, and then clubbing its head, again with flutes, "they played and danced
with the manatee for about half an hour, imitating the sexual act." Outside of the
context of the ani shri!ati, I have seen men similarly club to death shore-hugging
dolphins because of the danger (and temptation) that these animals may have posed
to bathing women.
6. The cutting of the ti!sho, or neck, can be viewed as a male counterpart to
clitoridectomy. In this context, the neck is phallic, and the threatened but curtailed
act of ''head-taking'' (ti!shti!a) is a fairly convincing representation of castration (see
Bettelheim 1954 for additional arguments of this kind).
233
Warren R. DeBoer
7. In d'Ans's 1970 reissue of Navarro's 1903 dictionary, the entry for clitoridectomy is
maspijana chicai or, when parsed, maspi ("female genitalia" ) -jtina ("tongue" )-chicai
("cut-out") . The pervasiveness of vagina dentata imagery is here reinforced by
likening the clitoris to the tongue.
8. The legacy of any given ani shri'ati is well encapsulated by Lathrap's general
diagnosis of Tropical Forest feasting:
The crucial mechanism by which one Tropical Forest village could achieve or maintain a
position that would impress its neighbors was to give a fiesta which lasted longer, ex
pended more beer, and unleashed more drunken brawls than any other fiesta in memory.
The drunken brawls during these fiestas afforded a culturally sanctioned opportunity for
discharging all the tensions and interpersonal aggression which had built up in the course
of day-to-day living. In its form and functions, the Tropical Forest fiesta pattern was not
unlike that of the modern cocktail party, but it lasted far longer and, on the average, more
crockery and heads got broken. (1970:54)
Cocktail party may be a bit antiquated as a convincing allusion, but the diagnosis is
still telling.
9. In this light, the claim made by some Conibo-Shipibo that the clitoridectomy is a
custom adopted from the Cashibo-a people whom they otherwise regard to be
abject savages-assumes some comprehensibility, if not historicity (Gebhart-Sayer
1985:592).
rn. In 1968 at the combined Campa-Conibo mission of Shahuaya, Pancho, a mestizo
patron who had several Campa "wives," remarked that the latter were "muy
ardiente" in contrast to the Conibo who were "muy frio." Whatever the case, his
testimony is of notional interest.
11. It is not my intention here to explain why the Conibo-Shipibo practice matrilocality.
One might cite the statistically framed hypothesis forwarded by Ember and Ember
(1971) that such a residence rule disperses fraternally bound males and thereby
discourages internal aggression, while favoring external aggression. One could just as
well argue that male absenteeism-promoted by mercenary activities during the
rubber boom era, employment in lumbering or other extractive industries, or, more
recently, military service-could encourage "matrifocal" tendencies. From a
Darwinian perspective, however, one can assess consequences without specifying
origins. Given matrilocality, Murphy (1957) still provides one of the best accounts of
the resultant "predicament" in an Amazonian society: Mutatis mutandis, his argument
could be applied to the Conibo-Shipibo.
REFERENCES
Alvarez, Ricardo
1970 Los Piros: Hijos de Dios. Lima: Santiago Valverde.
Bamberger, Joan
1974 The Myth of Matriarchy. In Woman, Culture, and Society, edited by M. z. Ros
aldo and Louise Lamphere, pp. 263-280. Stanford: Stanford University Press.
Bateson, Gregory
1936 Naven. Stanford: Stanford University Press.
234
FEAST A N D FORUM I N THE U P P E R AMAZON
Bertrand-Ricoveri, P.
1996 Un aspecto de la dialectica masculino-ferninino en la mitologia Shipibo. An
thropologica 14:81-104. Pontificia Universidad Cat6lica del Peru.
Bettelheim, Bruno
1954 Symbolic Wounds. Glencoe: The Free Press.
Boddy, ).
1982 Womb as Oasis: The Symbolic Context of Pharaonic Circumcision in Rural
Northern Sudan. American Ethnologist 9 (4): 682-698.
Brown, Judith K.
1963 A Cross-Cultural Study of Female Initiation Rites. American Anthropologist
7J: 571-594.
Clastres, Pierre
1998 Chronicle of the Guayaki Indians. New York: Zone Books.
Codere, Helen
1961 Kwakiutl. In Perspectives in American Indian Change, edited by Edward H.
Spicer, pp. 431-516. Chicago: University of Chicago Press.
Cole, Fay-Cooper
1945 The Peoples of Malaysia. New York: D. Van Nostrand.
d'Ans, Andre-Marcel
1970 Materiales para el estudio del grupo lingaistico Pano. Lima: Universidad Nacional
Mayor de San Marcos, Serie Lexicos r.
1994 L'Initiation et l'excision des filles chez les Indiens Shipibos d' Amazonie. In
formations Ethnographiques Provenant de Documents d'Amateurs. L'Eth
nographie 90 (2): 9-29.
DeBoer, Warren R.
1971 Archaeological Explorations on the Upper Ucayali River, Peru. Ph.D. disser
tation, Department of Anthropology, University of California, Berkeley.
1981 Buffer Zones in the Cultural Ecology of Aboriginal Amazonia: An Ethnohis
torical Approach. American Antiquity 46 (2): 364-377.
1986 Pillage and Production in the Amazon: A View through the Conibo of the
Ucayali Basin, Eastern Peru. World Archaeology 18 (2): 231-246.
in press Ceramic Assemblage Variability in the Formative of Ecuador and Peru. In
Archaeology of Formative Ecuador, edited by J. Scott Raymond and Richard
Burger, Dumbarton Oaks, Washington, D.C.
DeBoer, Warren R., and Donald W Lathrap
1979 The Making and Breaking of Shipibo-Conibo Ceramics. In Ethnoarchaeology:
Implications of Ethnography for Archaeology, edited by Carol Kramer, pp.
102-138. New York: Columbia University Press.
DeBoer, Warren R., and James A. Moore
1982 The Measurement and Meaning of Stylistic Diversity. Nawpa Pacha
20:147-162. Berkeley: Institute of Andean Studies.
DeBoer, Warren R., and). Scott Raymond
1987 Roots Revisited: The Origins of the Shipibo Art Style. journal of Latin Ameri
can Lore 13 (1): n5-132.
235
Warren R. DeBoer
Descola, Philippe
1996 The Spears of Twilight: Life and Death in the Amazon jungle. New York: The
New Press.
de Uriarte, Buenaventura L.
1982 La Montana del Peru. Annotated by Fr. Dionisio Ortiz. Lima: Gcifica 30.
[1942]
Diaz Castaneda, Cesar
1923 Kunibo. Inca: Revista trimestral de estudios antropologicos l (2): 398-409. Lima:
Museo de Arqueologia de la Universidad Mayor de San Marcos. (Much the
same material written in 1912 appeared as "Sohre los Indios Cunibos" in Iza
guirre 1922-1929, 1:301-320.)
Eakin, Lucille
1990 Nuevo Destina: The Life Story of a Shipibo Bilingual Educator. Dallas: Interna
tional Museum of Cultures, Summer Institute of Linguistics.
Eakin, Lucille, Erwin Lauriault, and Harry Boonstra
1980 Bosquejo etnografico de los Shipibo-Conibo del Ucayali. Lima: Ignacio Prado
Paster.
Ember, Melvin, and Carol R. Ember
1971 The Conditions Favoring Matrilocal versus Patrilocal Residence. American
Anthropologist 73:571-594.
Erickson, Philippe
1994 Los Mayoruna. In Guia ethnografica de la alta Amazonia, edited by Fernando
Santos and Frederica Barclay; vol. 2, pp. l-128 . Quito: FLACSO-Sede.
Farabee, William C .
1922 Indian Tribes of Eastern Peru. Papers of the Peabody Museum of American
Archaeology and Ethnology X. Harvard University.
Frank, Edwin H.
1994 Los Uni. In Guia etnografica de la alta Amazonia, edited by Fernando Santos
and Frederica Barclay; vol. 2, pp. 129-237· Quito: FLACSO-Sede.
Gebhart-Sayer, Angelika
1985 Notizen zur Madchenbeschneidung bei den Shipibo-Conibo, Ost Peru. In Die
Braut, Geliebt-verkauft-getauscht-geraubt: Zur Rolle der Frau im Kulturvergleich.
Edited by G. Volger and K. von Weick. Special issue. Ethnologica n (2): 592-597.
1987 Die Spitze des Bewusstseins: Untersuchungen zu Weltbild und Kunst der Shipibo
Conibo. Hohenschiiftlarn: Klaus Renner.
Giddens, Anthony
1979 Central Problems in Social Theory: Action, Structure, and Contradiction. Berkeley:
University of California Press.
Girard, Rafael
1958 Indios selvaticos de la Amazonia peruana. Mexico City: Libra Mex Editores.
Goldman, Irving
1964 The Structure of Ritual in the Northwest Amazon. In Process and Pattern in
Culture: Essays in Honor ofJulian Steward, edited by Robert A. Manners, pp.
ln-122. Chicago: Aldine.
236
F E A ST A N D F O R U M I N T H E U P P E R A M AZO N
1966 The Cubeo: Indians of the Northwest Amazon. Urbana: University of Illinois
Press.
Gregor, Thomas
1977 Mehinaku: The Drama of Daily Life in a Brazilian Indian Village. Chicago: Uni
versity of Chicago Press.
Grinnell, George Bird
1923 The Cheyenne Indians. 2 volumes. New Haven: Yale University Press.
Hayden, Brian, and Ralana Maneeprasert
1996 Feasting among the Akha: The 1996 Report. Manuscript presented to the
symposium "The Archaeological Importance of Feasting," organized by
Michael Dietler and Brian Hayden, Annual Meeting of the Society for Amer
ican Archaeology, Seattle.
Heath, Carolyn
1991 An ani sheati in the Shipibo Village of San Pablo, Rio Sinuya, Lower Ucayali
River, Peru. Unpublished manuscript translated from the Spanish by Peter Roe.
Illius, Bruno
1985 Die grosse Trinken: Heirat und Stellung der Frau bei den Shipibo-Conibo,
Ostperu. In Die Braut, Geliebt-verkauft-getauscht-geraubt: Zur Rolle der Frau im
Kulturvergleich, edited by G. Volger and K. von Weick. Special issue. Ethnolog
ica 11 (2): 584-591.
Izaguirre, Bernardino
1922- Historia de las misiones franciscanas y narracion de los progresos de la geogra.fia:
1929 Relatos originales y producciones en lenguas indigenas de varios misioneros,
161<}-1921. 14 vols. Lima: Talleres Graficos de la Penitenciaria.
1928 Descripci6n hist6rico-etnografico de algunas tribus orientales del Peru. Bo
letin de la Sociedad Geografica de Lima 45:196-236.
Karsten, Rafael
1955 Los Indos Shipibo del Rio Ucayali. Revista del Museo Nacional 24:154-173. Lima.
Kendall, Sarita, Fernando Trujillo, and Sandra Beltran
1995 Dolphins of the Amazon and Orinoco. Bogota: Fundaci6n Omacha.
Kensinger, Kenneth M.
1995 How Real People Ought to Live: The Cashinahua of Eastern Peru. Prospect
Heights, Illinois: Waveland Press.
Lakoff, George
1987 Women, Fire, and Dangerous Things. Chicago: University of Chicago Press.
Larrabure i Correa, Carlos
1905- Collecion de leyes, decretos, resoluciones i otros documentos oficiales referentes al
1909 Departamento de Loreto. 18 vols. Lima: Imprenta de "La Opinion Nacional."
Lathrap, Donald W
1970 The Upper Amazon. New York: Praeger.
Lauriault, E. H.
1952 El Hushati Chama. Peru Indigena 2:56-60.
Loriot, James, and Barbara Hollenbach
1970 Shipibo Paragraph Structure. Foundations of Language 6:43-66.
237
Warren R. DeBoer
Murphy, Robert F.
1957 Intergroup Hostility and Social Cohesion. American Anthropologist 59 (6):
IOl8-ro35.
Murphy, Yolanda, and Robert F. Murphy
1974 Women of the Forest. New York: Columbia University Press.
Odicio Roman, Francisco
1969 Mitos y Leyendas: Mitologia Chama. Lima, privately published.
Pallares, Fernando, and Vicente Calvo
1975 Historia de las misiones del Convento de Santa Rosa de Ocopa. Jose Amich
[1870] y Continuadores. Lima: Editorial Milla Batres. (Also appears in Larrabure i
Correa 1905-1909, 9:1-205 .)
Raymond, J. Scott, Warren R. DeBoer, and Peter G. Roe
1975 Cumancaya: A Peruvian Ceramic Tradition. Occasional Papers No. 2, Depart
ment of Archaeology, University of Calgary.
Rippy, J. Fred, and Jean Thomas Nelson
1936 Crusaders of thejungle. Chapel Hill: University of North Carolina Press.
Roe, Peter G.
1973 Cumancaya: Archaeological Excavations and Ethnographic Analogy in the
Peruvian Montaiia. Ph.D. dissertation, Department of Anthropology, Uni
versity of Illinois-Urbana.
1982 The Cosmic Zygote: Cosmology in the Amazon Basin. New Brunswick: Rutgers
University Press.
Sabate, Luis
1922- Viaje de los padres misioneros del Convento del Cuzco a las tribus
1929 salvajes de las Campas, Piros, Cunibos, y Shipibos por el Padre Fray Luis
Sabate en el afio 1874. In Historia de las misiones franciscanas y narracion de las
progresos de la geografia: Relatos originales y producciones en lenguas indigenas de
varios misioneros, 161<}--1921, by Bernardino Izaguirre, ro:19-304. Lima: Talleres
Graficos de la Penitenciaria.
Samanez y Ocampo, Jose B.
l98o Exploracion de las rios peruanos Apurimac, Eni, Tambo, Ucayali y Urubamba hecha
parJose B. Samanez y Ocampo en 1883 y 1884. Lima: Sesator.
San Roman, Jesus
1975 Peijiles historicos de la Amazonia peruana. Lima: Ital-Peru.
Siskind, Janet
1973 To Hunt in the Morning. London: Oxford University Press.
Tedlock., Dennis
1979 Zuni Religion and World View. In Handbook of North American Indians, vol. 9,
pp. 499-508. Washington: Smithsonian Institution.
Tessmann, Giinter
1928 Menschen ohne Gott. Stuttgart: Strecker und Schroder.
Tournon, Jacques
1984 lnvestigaciones sobre las plantas medicinales de los Shipibo-Conibo del Ucay
ali. Amazonia Peruana 5:91-u8.
238
Townsley, Graham
1994 Los Yaminahua. In Guia etnogrtifica de la alta Amazonia, edited by Fernando
Santos and Frederica Barclay, vol. 2, pp. 241-358. Quito: FLACSO-Sede.
Tschopik, Harry
1954 Notes for Indian Music of the Upper Amazon. Ethnic Folkways Library Album
No. FE4458 . New York: Folkways Records and Service Corp.
239
9
FEASTS AND LA B OR M O BI LIZATION
D I S S E C T I N G A F U N D A M E N TA L E C O N O M I C P R A C T I C E
Michael Diet/er and Ingrid Herbich
The use of feasts to mobilize collective labor has been a widespread and funda
mental economic practice of societies around the world. In fact, variants of the
practice are so strikingly omnipresent in the ethnographic and historical litera
ture that a good case can be made for acknowledging it both as virtually a uni
versal feature among agrarian societies (e.g. , see Erasmus 1956; Moore 1975;
Uchendu 1970) and as the nearly exclusive means of mobilizing large voluntary
work projects before the spread of the monetary economy and the capitalist
commoditization of labor and creation of a wage labor market.
This fact is of enormous potential significance to archaeologists in their at
tempts to understand ancient societies, particularly in terms of grappling with is
sues such as the role of labor control and exploitation in the development of
240
F E A STS A N D LA B O R M O B I L I ZAT I O N
social inequality. However, the realization of that potential has been hampered
by the lack of a fully theorized understanding of the specific range of practices
that enable voluntary labor to be mobilized on a scale above the household level,
how the possibility for labor exploitation inheres in some of these practices, and,
crucially; the ways that feasting operates as a mechanism of conversion within
this realm. Despite frequent programmatic statements by archaeologists about
the importance of understanding the means of controlling labor (e.g. , Webster
1990) and various attempts to do such things as quantify labor inputs in public
projects (e.g. , Renfrew 1973; Trigger 1990) , little serious consideration actually has
been given to developing a theoretical explanation of collective labor-mobilization
practices other than slavery. However, we would argue that a nuanced under
standing of the complex and intimate relationship between feasts and labor is
crucial. That is because, although feasts have a nearly universal role in this do
main among agrarian societies, this does not mean that they operate everywhere
in exactly the same fashion or that they have the same potential for exploitation
in all cases. Hence, the devotion of a chapter in this book to an explicit theoreti
cal dissection of this fundamental issue.
The following discussion is based upon cross-cultural analysis of ethnographic
cases derived from both the anthropological literature (especially building upon
the seminal works of Erasmus 1956; Moore 1975; and Uchendu 1970) and our own
ethnographic and historical research in western Kenya. We begin with an at
tempt to define a working vocabulary and a set of analytical concepts for expos
ing the relationship between feasts and labor and we propose a model of
"collective work events" that serves as a basis for understanding both the "con
version" functions of feasts and their potential for exploitation. We then move
from this more abstract discussion to some ethnographic examples that illustrate
the points we develop. Finally; we explore the implications of this analysis for ar
chaeology; and especially for the understanding of labor exploitation and the de
velopment of social inequality in ancient societies.
COLLECTIVE WORK EVENTS (CWEs)

Work feast is the term we use to describe a particular form of the "empowering
feast" mode of commensal politics (see Dietler, Chapter 3) in which commensal
hospitality is used to orchestrate voluntary collective labor. That is, the work
feast is an event in which a group of people are called together to work on a spe
cific project for a day (or more) and, in return, are treated to food and/ or drink,
after which the host owns the proceeds of the day's labor. The work feast, as the
term is defined here, actually constitutes one pole in a range of labor-mobilization
practices that we call collective work events (CWEs). At the other pole of the CWE
1
range is what we call the work exchange (see Fig. 9.1).
241
Michael Dietler and Ingrid Herbich
!iqj'
"
m,
Work E rI Work Feast

:P'!
+
Labor Reciprocity Obligations
+ + >
+
Temporal Finality of Exchange Transaction
< + +
+
Lavishness of Hospitality
<
+ +
+
Size of Work Group
< + +
+
Social Distance of Labor Recruitment
< + +
+
Potential for Exploitation
< + +
Figure 9.1. Schematic representation of Collective Work

Events (CWEs) showing the correlated inverse-trend rela
tionships among work-group size, the degree of recipro
cal labor obligations, the scale of hospitality required, and
the social distance of workers capable of being mobilized.
Several analytical points need to be established in order to understand the op

eration of CWEs. Most immediately, it must be emphasized that, from a com
parative analytical perspective, work feasts and work exchanges are not binary
oppositional categories, but rather terms used to describe polar tendencies along
a continuum (Moore 1975). This continuum is defined by several factors (work
group size, reciprocal labor obligations, scale of hospitality, social distance of
workers, etc.) that vary in a fairly predictable relationship to each other (see Fig.
9.1). Perhaps the most important defining characteristic is the inverse relationship
between the degree of reciprocal labor obligation and the scale of hospitality
required.
At the extreme work-exchange end of the scale, little if any food needs to be
provided (often simply a little ordinary refreshment), but the moral obligations
to reciprocate by participating in the work-exchange events of those who have
participated in one's own event are very strong and explicit. This reciprocation
may be either in person, or by sending a member of one's household as a substi
tute. At the extreme workfeast end of the scale, reciprocal labor obligations may
be very weak (and vaguely implicit) to completely nonexistent, but the lavishness
of the hospitality expected is quite significant.
In other words, crudely stated, the difference is basically one between an ex
change of labor for labor versus and exchange of labor for hospitality. That is,
work exchanges operate through a kind of delayed reciprocity, where the host of
the event assumes a labor debt to all the participants that must be repaid at a later
242
F E ASTS A N D L A B O R M O B I L I Z AT I O N
date. Work feasts, o n the other hand, operate more a s a temporally finite ex
change transaction: lavish hospitality is "exchanged" directly for labor, and no
further obligations exist between host and guest. In some cases the event is at
least partially acknowledged by the participants as a kind of exchange, and judg
ments about the quality and quantity of the beer and food provided affect the
amount of work done and the size of the group that participates (e.g., see Barth
l967a; Colson 1949; Donham 1994; Goldschmidt 1976; Netting 1964; Saul 1983;
Hunter 1961) . However, as Karp has noted for the Iteso of East Africa, the partic
ipants do not necessarily consciously envisage this as an exchange transaction; in
stead, the feast may be seen simply as "the vehicle through which cooperation is
achieved" (1980:88). In other words, labor relations are constituted through, and
euphemized as, relations of commensality.
In work exchanges, the size of the work group able to be mobilized is limited
to fairly small collectives of usually less than fifteen people; and these groups are
very often organized through kinship or friendship networks (Erasmus 1956) . As
Moore (1975) has noted, the rather precisely and explicitly reckoned reciprocal
labor obligations create certain scheduling constraints that both limit the size of
the groups and generally result in one of two organizational arrangements for re
ciprocation. In what he calls the "individual exchange" pattern, individuals will
participate in several different work-exchange networks, and will thus have mul
tiple individual obligations with several other persons who will have their own
sets of obligations that only partially intersect. In the "group exchange" pattern,
a consistent group of individuals will work together in a consecutive circuit on
the projects of each of its members. As Moore (1975) further noted, where insti
tutionalized differences in social rank exist, recruitment to either of these kinds
of work-exchange events is normally confined to individuals of comparable sta
tus with comparable landholdings to be worked.
With work feasts, on the other hand, labor can be mobilized on a much larger
scale (up to several hundred people) and projects can be undertaken that would
not be possible with work exchanges. Work feasts are also more effective at re
cruiting workers from a wider social radius, without reference to kinship, neigh
borhood affiliation, or social status. It is the scale of hospitality-the copiousness
and quality of the drink and food provided (and the reputation of the host for pro
viding these things in abundance)-that draws people to participate rather than
close social relationships. Furthermore, work feasts are ad hoc events that are
mounted for specific projects and they do not form part of a permanent cyclical
organizational structure of labor relations. This means that, as noted, work feasts
(of the ideal polar form) do not entail lingering obligations on the part of the host
to participate in the work feasts of his or her guests, and this feature becomes even
more marked as the size of the work feast expands (Erasmus 1956; Moore 1975) .
243
One can actually distinguish two variants of the work feast, one voluntary and
one obligatory. In the voluntary work feast, people are drawn to the event simply
by the prior reputation of the host for providing lavish feasts. The obligatory
form, which is often referred to as corvee labor, exists only where there is institu
tionalized central authority in the form of religious leaders, chiefs, kings, or
other types of state apparatus. In these cases, people are drawn to participate be
cause a ruler or public institution has the moral authority to require their pres
ence as a form of labor tribute. However, as will be discussed in more detail later,
rulers who fail to orchestrate corvee projects through the same work-feast idiom,
by providing a generous quantity of food and drink for the workers, will soon
meet grumbling and resistance. Rulers cannot rely on coercive force to motivate
participation: any stable long-term system of labor tribute must rely on the con
tinual production of consent-which means operating through and playing upon
the same practices that have symbolic resonance within the population as a
whole: Hence, there is very good reason to view corvee labor simply as a variant
of the work feast in which the composition of the labor force is predetermined
by an ideology of obligation and authority. In fact, approaching this relationship
from the opposite direction, Bourdieu (1990 :118) sees the voluntary work feast as
simply "a covert exaction of corvees"-a perspective that will be better appreci
ated after the discussion of exploitation below.
A second important point that needs to be made about these CWE modes is
that they are not mutually exclusive and it is not possible to characterize particu
lar societies by one or the other of them. Rather, in most societies both modes
will be employed in different contexts for different purposes. For example, work
exchanges may be the normal pattern for small groups performing routine agri
cultural tasks (e.g., weeding, harvesting, field clearing, transport of crops from
field to home), while work feasts will be employed for projects requiring a larger
work group (e.g. , house building and repair, fence or rampart construction, road
building and repair, waterworks construction and maintenance, mining, agricul
tural work on very large plots of land, mounting trade expeditions) or one con
vened for an urgent task. Moreover, within given local contexts, these different
forms of CWE are often categorically marked; that is, many societies distinguish
different kinds of CWE by name and conceptualize them as distinct categories
(see below). In other cases, the differences are more clinal than categorical (e.g.,
see Chibnik and de Jong 1990; Saul 1983) .
Another important related point that needs to be established is that the specific
forms of CWE found in real ethnographic contexts do not necessarily closely ap
proach the polar extreme versions used here to define the abstract analytical con
tinuum. Nor need there be only two categories. In other words, a local version of
what would be classified as a work feast on our abstract comparative scale might
244
F E A STS A N D LA B O R M O B I L I ZATI O N
still involve sonie implicit reciprocal labor obligations (often only for participat
ing km); and, in the opposite direction, there might well be several categories
within the work-exchange end of the continuum with decreasing hospitality ob
ligations and decreasing size. For example, the Maale of Ethiopia have three cat
egories of CWE: the d.abo, the mol?o, and the helma. The d.abo is a moderately
large (up to thirty people) ad hoc work feast of variable participation at which the
provision of much beer is mandatory and the quantity of beer determines the
length of the workday, but at which the host also acquires some implicit recipro
cal work obligations. The molfo is a large (between about six and fifteen people)
rotating work exchange with a formal organization and durable composition at
which some beer is usually provided (but is not obligatory) . The helma is a very
small work exchange (three to four people) following a fixed cycle of rotation at
which beer is not necessary; but some beer will be served if one wants the work
to extend beyond the standard half day (Donham 1994; cf. Barth 1967a; Mayer
1951; Nadel 1942:248-251; Tosh 1978 :41 for a range of different local classifications) .
Hence, from a n analytical perspective, i t i s important t o recognize that such
culturally specific forms of CWE may be located at various points along the ab
stract continuum according to local expectations about the relative degree of
labor reciprocity obligations and scale of requisite hospitality. Moreover, in a
given society, there may be no forms that closely approximate the polar ex
tremes. As will be discussed later, this fact has important implications for under
standing the potential of these practices for labor exploitation. However, for the
moment, let us simply establish the semantic point that, in applying a term such
as workfeast to empirical ethnographic or archaeological cases, we are merely sig
nifying those arrangements that tend toward the work-feast end of the abstract
analytical continuum of CWEs.
It should additionally be noted that CWEs are used to congregate groups of
workers performing identical tasks of a relatively unskilled nature. They are use
ful in contexts where the simple multiplication of the number of hands brought
to bear on a task is effective in reducing the time of completion or in enabling
certain feats (such as the movement of heavy objects or the construction of large
structures) that could not be accomplished by members of a household alone.
Tasks for which the specialized skills of an individual are more important than
the multiplication of the number of workers are accomplished through other
means. The example discussed later of iron working among the Samia of Kenya
nicely illustrates this distinction by contrasting the mining of ore through work
feasts with the compensation of smiths through payment with a part of their
production (although, of course, such specialists may also be treated to some
commensal hospitality) . Another related feature of CWEs commented upon in
many ethnographic contexts (e.g., Barth 1967a; Donham 1994; Kennedy 1978) is
245
that the quality of work performed at work feasts is generally less good than at
work exchanges, and the quality often gets worse as the size of the feast expands.
This is frequently due in part to the effects of the alcohol that often accompanies
the work at work feasts. However, it also stems from the different sense of recip
rocal obligation that attaches to small work exchanges, as well as the increasing
difficulties of supervision with increasing size.
As a final preliminary observation, it should be noted that, contrary to older
conceptions of an idealized "domestic mode of production" composed of self
contained household production and consumption units, CWEs are fundamental
to the operation of the agrarian economy because they mobilize the essential in
terhousehold communal labor flows that, in fact, sustain domestic units (see
Donham 1994) . Moreover, work feasts, in particular, are extremely important in
the political economy because of the context they provide for the acquisition and
conversion of symbolic and economic capital (to employ Bourdieu's
[199o:n7-n9] useful terminology). In the first place, as with all other types of
feast, they provide an opportunity to make public statements about prestige and
acquire symbolic capital (Chapter 3). It should not be forgotten that, for the par
ticipants, a lavish work feast is, above all else, a festive social occasion (rather than
simply a day of wage labor). Such an event not only mobilizes labor: it augments
the reputation of the host for generous hospitality in the same way that, for ex
ample, hosting a large marriage ceremony or sponsoring a communal ritual does
(e.g. , see Chibnik and de Jong 1990; Colson and Scudder 1988 :77; Kennedy 1978).
However, its peculiar characteristic among feasts in general is that it simultane
ously provides a means of harnessing the labor of others in order to acquire eco
nomic capital that subsequently can be converted to additional symbolic capital
by several other means. As will be explained below, work feasts, in effect, act as a
mechanism of indirect conversion between spheres of exchange in multi-centric
economies and thereby provide a potential catalyst for increasing inequality in so
cial relations.
WORK F EASTS I N ETH N O G RAP H I C CONTEXTS

At this stage, it i s necessary to ground this schematic discussion o f collective
work events, especially work feasts, in some brief ethnographic examples in
order both to give a sense of the everyday lived experience behind these abstract
concepts and to expand upon the theoretical points made.
The importance and ubiquity of work feasts can hardly be doubted by anyone
familiar with the ethnographic and historical literature.2 In fact, work feasts have
been such a common feature of agrarian societies that they have often been taken
for granted as part of the expected cultural background-a self-evident feature
mentioned in a various passages but not singled out for detailed analysis (with
246
F E ASTS A N D LA B O R M O B I L I Z AT I O N
some fortunate exceptions). However, even some of the briefer analyses are quite
revealing. Goldschmidt, for example, wrote of the mayket work feast of the Sebei
of East Africa that "no institution is so central to modern Sebei life" (1976:156); and
among the Kofyar of Nigeria, the concept of God is even phrased in terms of a
wealthy farmer who distributes beer in return for labor (Netting 1964).
As noted earlier, work feasts were, and are, used to perform a wide variety of
tasks for which the sheer multiplication of hands either allows a project to be
done in a short space of time or enables a project that could not be undertaken
otherwise. Most often these include agricultural tasks and construction or main
tenance / repair projects, but such things as the organization of game drives, min
ing, and trade expeditions are also recorded. For example, the genesis of the
extensive trading network of the Kamba of Kenya was due in large measure to
the innovative adaptation of the traditional mwethya work-feast system to organ
ize transport (Cummings 1976:92-93).
What is more, the motivational effectiveness of CWEs in inducing people to
participate in working together is evident in the fact that, despite some obviously
important effects on such practices stemming from the spread of wage labor and
the transformation of labor into a marketable commodity under colonial
regimes, work feasts and work exchanges have continued to persist alongside
wage labor in many areas (Colson and Scudder 1988; Erasmus 1956; Moore 1975;
Saul 1983) . Indeed, during the colonial period, long after many other elements in
an indigenous economy had been accepted as commodities with a monetary
value, there often persisted a lingering negative feeling about exchanging labor
for money (as also was the case with land). Barth (1967a), for example, noted that
among the Fur of the Sudan, government workers were for some time unsuc
cessful in recruiting local workers by offering money wages-even when those
wages were calculated to be twelve times the value of the millet beer demanded
by an individual in the context of a work feast. Similarly, on Samoa and in the
Cook Islands of the Pacific, Lemert (1979) noted that natives could not be re
cruited for agricultural work on plantations by offering money; only the promise
of mea miti ("something to sip" : i.e., several bottles of drink for a feast after the
work is done) would provide a sufficient incentive.
Hunter's description of work feasts (amalima, singular: ilima) among the
Pondo of South Africa nicely illustrates and highlights several of the typical fea
tures noted in the earlier discussion. These amalima can mobilize up to two hun
dred people for a particular task. Word is spread by telling the neighbors that an
ilima will be held in a particular field on a given day, and the news quickly spreads
through the district by word of mouth. There is no obligation for anyone to par
ticipate, but people are drawn by the prospect of the feast. Moreover, there is a
hierarchy of preference between beer feasts and meat feasts:
247
an ilima with beer always draws more than an ilima with meat . . . It is known and
discussed in the community beforehand how many barrels [of beer] have been pre
pared, and so the number of people is in some ways commensurate with the
amount of beer provided, but other considerations such as the scarcity of beer at
the time, the number of other festivities on, the occupation of people with their
own lands, and the reputation of the owner of the ilima for generosity; or stinginess,
affect the number attending. (Hunter 1961: 89)
As Hunter describes one such event:
At one ilima to take mealies off the cob, five barrels of beer were provided and an
average of forty people were present at one time, the number of men and women
being about equal. People came and went, assisting in the work for as long as they
stayed. Beer was passed round at intervals, so the amount they got depended upon
the length of time they stayed. (1961: 89)
As she further noted, people work hard at such events, but they come to them be
cause they enjoy participating: ''An ilima is a party. The crowd of people, the mix
ing of the sexes, and the refreshments, give even to hard work the atmosphere of
play. There is conversation, and songs, and flirting" (1961:90) . This means also that
"Usually an ilima gets through a considerable amount of work, but the quality of
the work, particularly in a weeding party, is apt to be poor" (1961:90 ) .
Bohannan and Bohannan (1968) provide another illuminating description of a
work feast among the Tiv of Nigeria:
The biggest [yam-] mounding party we saw was called by a man of about forty-five,
Ytlaun of MbaDuku / MbaYar. He called his best friend from the neighboring lineage
. . . who gathered about twenty of his own agnatic kinsmen about him and came to
MbaDuku late one evening. They brought drums and hoes, a hurricane lamp, and
their best clothes. They danced at several compounds along the way, to songs com
posed for the occasion and to well-known work songs. When they got to Yilaun's
compound, they danced until about l:oo A.M. They began work the next morning at
daybreak. This group was joined by Ytlaun's agnatic kinsmen: all the agnatic descen
dants of his father's father, and a few youths with more attenuated agnatic links.
Drawn by dancing and the promise of rich food, several young men of Ytlaun's
mother's lineage came, and two of his wives' brothers also arrived.
This group made yam mounds at a feverish pitch for about four hours. They
were then given food, prepared and brought to the fields by all of the women of Yi
laun's compound. They ate in groups of about half a dozen; each group was given a
huge calabash of yam porridge with three different sauces. There was so much food
they could not eat it all. Work was suspended during the heat of the day. They
began again about four in the afternoon and worked until dark, and were fed again.
That night they danced until well after midnight. The next day the procedure was
repeated, and all the farms of Yilaun's compound were finished. That night they
248
F E ASTS A N D LA B O R M O B I L I ZAT I O N
began the biggest dance of all: Ytlaun furnished pot after pot of millet beer and
killed two small goats. They again danced far into the night, and the next day
danced for almost two hours at the local market. (Bohannan and Bohannan
1968:73-74)
By hosting this large work feast, Ytlaun was actually able to use this labor to ex
pand his farmland into disputed territory in a region where land was becoming
increasingly scarce. "In addition, Yilaun reaped great prestige from his lavish
treatment of the relatives of his best friend: during the dancing at the market
place, his name was on the lips of all" (Bohannan and Bohannan 1968:74).
As pointed out earlier, com�e labor also is generally organized through the
idiom of the work feast (e.g. , see Dillon 1990:129; Goody 1982:67; Richards 1939;
Washburne 1961:140). In making this generalizing observation, there is, of
course, some potential danger in using cases from colonial contexts because in
many instances "chiefs" were created by decree in societies where they did not
previously exist (e.g. , see Chapter 3). Hence, such neo-chiefs may simply have
continued the labor-mobilization practices used by wealthy influential men who
did not previously have the authority to command labor tribute. However, this
danger is mitigated by the fact that even cases with long-established kingdoms
seem to follow this pattern. As noted earlier, Bourdieu (1990:118) considers the
voluntary work feast as simply a camouflaged version of corvt!e. While this state
ment obscures important distinctions within the range of events that constitute
work feasts (outlined in the initial discussion of CWEs), it does quite correctly
point to the potential for some work feasts to develop into a mechanism of labor
exploitation.
SAM I A W O R K F EASTS A N D T H E I S S U E
O F L A B O R E X P L O I TAT I O N
Let us now turn to a brief consideration of an ethnohistorical study on precolo
nial iron production conducted among the Sarnia people in western Kenya as a
means of focusing particularly upon this issue of labor exploitation and exploring
the closely related conversion function of the work feast. 3 The Sarnia are an
agrarian society with a traditionally acephalous form of political organization.
They speak a language of the Bantu family and are neighbors of the Nilotic
speaking Luo living north of the Winam Gulf in the northeast corner of Lake
Victoria. Until the influx of European industrially produced iron in the 1920s, all
of the iron used over a several-thousand-square-kilometer area encompassing
both Sarnia and large parts of the neighboring Luo territory was derived from a
single ore source in the Sarnia hills. This source, composed of hematite deposits
in pockets of about 7 meters thickness, constitutes the richest iron ore source in
Kenya (Brown 1995:43). The importance of the precolonial exploitation of these
249
deposits is reflected in both the considerable quantities of iron slag found at the
base of the hills and in the Luo name for the principal object produced from this
ore, a large iron hoe blade called Kwer Nyagot, or "Hoe, Daughter of the Hills"
(Fig. 9.2). The production of these hoes was based upon a system fueled by large
work feasts, which was an elaborated version of the same mechanism used to
mobilize labor for a variety of other projects. A wealthy man (that is, one with a
large number of wives capable of raising a copious supply of millet, brewing it
into beer, and preparing a generous supply of food) would call together all the
willing men of the area on a given day to mine ore from the Samia hills. There
was no obligation to participate, but men were drawn to do so by the prior repu
tation of the host for generous hospitality. After spending the day gathering ore,
these men were treated to a great feast, after which they would return home and
the host was left with a large supply of iron ore. No further compensation was
required, and the host was considered to own the proceeds of the day's labor. A
smelter and a smith were then called to convert the ore first into blooms and then
into hoes, respectively: Each of these craftsmen was compensated for his labor by
being given some of the hoes produced from the ore.
These hoes, some of which still survive as heirlooms in Luo and Samia homes,
were extremely valuable. Although used for utilitarian agricultural purposes,
they formed part of a prestige sphere of exchange in a multi-centric economy:
their acquisition required the giving of livestock and they were even used along
with cattle as bridewealth in marriage transactions. So important were they to
the neighboring Luo that Samia travelers were exempt from attack for fear of en-
Figure 9.2. Iron hoe blade (length = 56 cm) from the precolonial
era in western Kenya. These hoe blades were made in the Samia
Hills and exchanged widely throughout the region. They could
be used as bridewealth and to exchange for livestock. (Photo by
M. Dietler and I. Herbich)
250
dangering exchange links , and giving a daughter to b e married t o a Samia man

was considered highly desirable because of the exchange relationships it engen
dered. Furthermore, the still iron-rich slag (the smelting process was not very ef
ficient) was often collected by those Samia who could not manage to host mining
work feasts on their own, and small chunks of iron were hoarded until there was
enough to make a hoe.
While this work-feast method of engaging in iron production was, in princi
ple, open to all Samia men, in practical terms its effective manipulation was lim
ited to those wealthy enough to provide sufficiently large feasts to mobilize large
work groups. As noted elsewhere (Chapter 3), the agricultural and culinary labor
required for such an event is formidable, and in Samia and Luo societies this re
quired many wives. However, acquiring wives required the accumulation of
wealth (i.e., cattle and iron hoes) for bridewealth. Moreover, there is an obvious
link between subsistence production, marriage, and iron production that would
insure that an initial position of advantage in access to this process would tend to
have a spiraling effect in augmenting wealth and prestige (see Fig. 9.3). That is,
wealth in cattle was necessary for the bridewealth to obtain the multiple wives
whose labor could produce a large feast. But once achieved, the hoes gained
through the institution of the work feast could be used to obtain more wives
(through conversion to stock used as bridewealth or used directly in marriage
transactions); and the increased productive capacity represented by these women
could be used to more effectively and frequently amass the supplies for a large
feast and engage again in iron production. Those men without the initial "capi
tal" (in terms of cattle, crops, and wives) to produce a large feast were effectively
excluded from the cycle and were reduced to being regular guests I workers at the
work feasts of the wealthy.
This example serves to illustrate how feasts may act as a means of indirect
conversion between spheres of exchange (or regimes of value) in a multi-centric
economy (Fig. 9.4). Such economies, in which different classes of goods circu
lated in separate exchange regimes (of variable number and kind depending
upon the culture) and in which there were strong moral sanctions against con
verting between the spheres, were a very common feature of pre-monetary eco
nomic systems that did not have a uniform and universal scale of value (Barth
1967a; Bohannan 1955; Piot 1991; Salisbury 1962); indeed, Kopytoff (1986) sees
them as a universal feature, even in the Western capitalist context. In pre-mone
tized western Kenya, for example, one could exchange grain for pots or a range
of other craft items and agricultural produce. However, grain was so low on the
scale of value that no one would be willing to accept even a huge quantity of it in
direct exchange for prestige-sphere goods (except, perhaps, during a famine); yet
its conversion into beer and food in the context of a feast was a prime means of
251
A ricultural Production
WIVES
IRON HOES
Figure 9.3. Schematic representation of the link between subsistence production, marriage,
and iron production in Samia society showing how an initial position of advantage in ac
cess to this process would tend to have a spiraling effect in augmenting wealth and prestige.
both acquiring prestige and mobilizing the labor by which prestige-sphere ex
change objects could ultimately be obtained. A ritual form of commensal hospi
tality (the work feast) was able to perform the apparently impossible function of
converting lowly grain into valuables and prestige, of linking separate spheres of
exchange via an indirect route involving the mobilization of labor (both culinary
and productive) .
The Samia example also demonstrates the way in which work feasts serve as a
conduit for reciprocal conversions of what Bourdieu (1990) calls economic and
4
symbolic capital. People are drawn to participate in such events by the reputa
tion of the host for generous hospitality. This reputation is an aspect of symbolic
capital acquired through the expenditure of material capital in previous feasts.
But through the institution of the work feast, this symbolic capital is used to har
ness the labor of others for the acquisition of further material capital, while at
the same time augmenting the symbolic capital of the host by generating further
prestige and embellishing his reputation for generosity.
Finally, this example also illustrates how the manipulation of this practice can
sometimes lead to increasing social and economic inequality even in the context
of ideologically egalitarian precapitalist societies. When one segment of a com
munity becomes adept at managing this entrepreneurial device and begins to act
consistently as hosts of large work feasts while others find themselves continually
serving as guests / workers, then one has the beginnings of a pattern of labor ex-
252
Cattle
Prestige Sphere
t
Labor
I
FEASTS
Figure 9.4. Schematic representation of the role of feasts as a privileged mechanism of indi
rect conversion between spheres of exchange in a multi-centric economy. This is a highly
simplified rendition of one possible set of spheres of exchange; comparative ethnographic
surveys show many possible permutations of numbers of spheres and valuations of objects.
ploitation by which some individuals or groups are able to dominate the system
and extract wealth and prestige from the labor of others.
Barth (1967a, 1967b) noted a similar potential for "growth spirals" among the
mountain Fur villages of the Sudan. There the people of the village had gradu
ally separated into two strata: a moderately prosperous group of work-feast
hosts and a less fortunate fraction who, through their frequent participation as
guest/ workers at the feasts of others, effectively formed a labor reserve for the
more wealthy. Some of the wealthy individuals had even begun to further aug
ment their relative advantage by expanding the use of work feasts to enter the
cash-crop market.
Similarly, among the Tarahumara of Mexico, Kennedy (1978 : 86) noted that
wealthier men were able to give more frequent and more lavish tesquinadas
(work feasts fueled with maize beer), and thereby to gain the labor necessary to
maintain larger areas under cultivation. This allowed them not only to hold more
beer feasts, but also to amass surplus maize for distribution to others in times of
famine. Both of these things generated considerable symbolic capital. As he
noted, the beer feast operates as a "prestige showcase" that serves "the important
function of publicizing rank and power," and this feature was even more marked
during the lean season before the next harvest (Kennedy 1978 : 118).
253
A similar situation was noted among the Maale of Ethiopia by Donham

(1994:146--152). He calculated that in one context about two thirds of the net sur
plus of work-feast labor that richer villagers enjoyed came from people in the
poorest categories of the village hierarchy. These were households that, on aver
age, sponsored no work feasts at all during the entire period of his fieldwork.
Moreover, what he identified as "middle villagers" (i.e., those in the middle eco
nomic stratum) on average enjoyed a small net surplus of work-feast labor, but
"still showed a negative deficit in cooperation with richer villagers" (Donham
1994:148).
Hence, a potential for labor exploitation clearly inheres in some forms of the
work feast. But a marked pattern of exploitation is by no means a universal result
of the operation of work feasts; nor do all exploitative forms necessarily result in
escalating inequalities.5 Consequently, it is important to ask what conditions
favor the development of such an asymmetrical pattern and what kinds of con
straints might inhibit its spiraling expansion?
One immediate potential constraint stems from certain specific organizational
arrangements of the relationship between workers and food providers. In some
contexts, certain public projects may be undertaken by groups of people who
share not only the work but also the provisioning of the feast. For example,
among the Wamira of Papua New Guinea, Kahri (1998) notes that all men must
work together to construct and periodically clear out common irrigation canals.
This involves work feasts of perhaps thirty men for which food is supplied jointly
by the wives of all the men. Hence, a potential for exploitation depends upon the
existence of work feasts hosted by individual households, or at least by a group
smaller than the collection of assembled workers / guests. Given this arrange
ment, Barth (1967a) located another potential brake on the system in the fact that
the quality and productivity of the workforce tends to deteriorate as the group
expands. As noted earlier, this feature has very frequently been remarked upon in
other cases (e.g. , Donham 1994; Kennedy 1978). While this may constrain the size
of individual work feasts, it does not explain limitations on the expansion of the
system. However, more important constraints are imposed by the cultural defini
tion of the transaction that constitutes a work feast.
What becomes immediately clear is that those events which, by local cultural
convention, conform most closely to the extreme work feast pole of the contin
uum of CWE possibilities are the most susceptible to exploitation. This is be
cause the event is · treated by the participants as a finite exchange transaction
where lavishness of hospitality replaces any delayed obligations for reciprocal
labor. Hence, the host is not limited by the time and energy constraints of taking
on multiple labor-debt obligations. Moreover, this form of work feast comes clos
est to the concept of a labor market, while at the same time the fact that the ex-
254
change is orchestrated through a ritual of commensal hospitality subtly serves to

euphemize the nature of the transaction-the exchange is represented as an act
of generosity and made palatable even in societies with a strongly egalitarian
ethos. Finally, the fact that recruitment of the labor pool may expand well be
yond very local kin and friendship networks means that, even in those cases
where some weak and implicit labor obligations might accrue, the transaction
occurs between the host and a large number of individuals who are not already
linked by close social relationships and who stand outside the moral community
within which such obligations might be recognized (see Sahlins 1972).
As the culturally construed proper form of the work feast is located farther
away from the polar type (hence as reciprocal labor obligations increase and are
more explicitly and precisely calculated), such asymmetrical exploitation be
comes more difficult. The accruing labor obligations restrict the future activities
of the host and the size of the work group that can be assembled through a work
feast. As noted earlier, the precise location on the CWE continuum of a given
local version of the work feast is determined by local cultural conceptions. This
can, of course, shift over time as practice responds to changing problems and his
torical contexts. But what is important to note is that a potential for exploitation
is inherent in the upper range of the work-feast end of the continuum. This po
tential can be powerfully magnified through certain linkages-when work feasts
are used to produce goods or materials that can be employed in external ex
change circuits to increase economic and symbolic capital (for example in the
case of producing iron hoes for cattle exchanges, or the use of work feasts to pro
duce cash crops).
It is also necessary to briefly discuss another form of labor exploitation that is
a more variable aspect of work feasts-that is, an intrahousehold one based upon
gender. In very many cases, it is women who constitute the productive base for
the agricultural and culinary labor that underwrites the operation of work feasts,
yet have little or no claim to the benefits of the event (except indirectly, as mem
bers of the household). For example, in the Samia case discussed earlier, it was
the multiple wives of a wealthy man who provided the crops and the culinary
labor necessary to produce a feast. Yet it was largely the husband who benefited
from the mining labor organized through the feast-both in terms of the pres
tige generated and the subsequent use of the iron hoes (for the acquisition of cat
tle and more wives). Hence, there was a double system of exploitation in
operation. A similar pattern of the gendered division of labor is found in many
agrarian societies, especially in Africa (e.g., see Clark 1980; Geschire 1982). This is,
in fact, one of the attractions of polygyny and one reason that wealthy men and
chiefs in patrilineal polygynous societies tend to have many more wives than
other men.
255
It would be misleading, however, to suggest that this is a universal pattern of

exploitation. In the first place, there are many cases in which men are responsible
for agricultural production and, perhaps more rarely, culinary labor for their own
feasts. There are also cases in which the division of agricultural labor is more bal
anced, and gendered patterns of crop ownership and responsibilities can, in fact,
be rather complicated (e.g., see Barth 1967a:151). Moreover, in some instances
women are capable of mounting work feasts on their own and have proprietary
rights over the proceeds of the event. It is very common that women organize
work exchanges, but somewhat less common that they are the official hosts of
work feasts, especially the larger types. However, it does happen in some con
texts. Moreover, CWEs of various types can be turned to novel uses that actually
give women the possibility for increased economic independence and power:
among the Luo of Kenya, for example, during the 1980s women began to expand
the traditional work exchanges for agricultural tasks into new kinds of coopera
tive work-exchange associations (among some potters, for example) that enabled
them to pool resources, build collective capital, and ameliorate individual risk
(see Herbich and Dietler 1989 ). Hence, work feasts and other CWEs cannot auto
matically be viewed simply as a mechanism of gendered exploitation of labor.
However, one can observe that gendered exploitation, where it exists, does tend
to involve exploitation of female labor, whereas an inverse pattern (that is, in
which men provide the bulk of the agricultural and culinary labor that supports
women's work feasts) is rather rarer. It is extremely important to recognize the
potential permutations of these various intersecting patterns of exploitation in
understanding the labor-mobilization practices of a given society and their po
tential for creating and transforming structures of inequality.
CO N C LU S I O N
As archaeologists become increasingly interested in pursuing questions concern
ing the transformation of social relations, the organization and institutional set
ting of labor is a topic that demands to be explored in much greater depth. In
particular, it is extremely important to develop a theoretical understanding of the
practices that enable the mobilization of voluntary collective labor.
Archaeologists have expended a good deal of energy attempting to identify;
classify, and understand "craft specialists" and systems of "specialized produc
tion" (e.g., Childe 1936; Clark 1995; Clark and Parry 1990; Costin 1991; Herbich
1987; Peacock 1982). However, they have been far less engaged with exploring the
forms of collective labor (voluntary and involuntary) that underlie in a more fun
damental way the operation of the agrarian political economy. Within the fields
of history and anthropology there exists a considerable literature analyzing
forms of involuntary labor classified under the rubric of "slavery" (e.g., Finley
256
F E ASTS A N D LA B O R M O B I L I Z AT I O N
1985; Garlan 1988; Kopytoff 1982; Meillassoux 1991; Miers and Kopytoff 1977); al
though, to date, this has been of far less concern to archaeologists. The major ex
ception to this pattern of neglect has been archaeologists working on the highly
idiosyncratic form of slavery found on the plantations of North America (see Sin
gleton 1995), although scholars working in a few other regions have also made
important contributions (e.g. , Arnold 1988; Daubigney 1983; Peschel 1971) .
This chapter has attempted to focus an analytical lens on the even less well ex
plored, yet extremely important, process of voluntary collective labor mobiliza
tion. A theoretical understanding of such labor-mobilization practices is crucial
to archaeologists for several reasons. In the first place, a fuller awareness of the
range and operation of such practices exposes the inadequacies of assumptions
that have guided archaeological interpretation in the past, such as simplistic cor
relations between the existence of large-scale earthworks and the necessity of
centralized political organization. The idea that such projects must be the result
of tributary corvee labor is simply not warranted, as it is clear that work feasts can
mobilize voluntary work groups on a similar scale for similar kinds of projects.
Indeed, it should be evident at this point that corvee labor can only be understood
when it is properly situated in the context of the full range of voluntary "collec
tive work events" because it operates as a kind of variant of the work feast. Even
large state-directed proj ects, at least those that depend upon the labor of free sub
jects rather than slaves, will usually take the organizational form of work feasts.
A theoretical dissection of work feasts also reveals how they serve the impor
tant role of operating as mechanisms of conversion between economic and sym
bolic capital. Like other feasts, they convert agricultural produce into immediate
prestige for the host. However, they also have the advantage of simultaneously
harnessing labor that can be used to generate further materials for future feast
events or to produce goods that can be used to acquire other forms of symbolic
capital or enlarge the household productive base (e.g., through marriage transac
tions). As explained earlier, this is an extremely important function of indirect
conversion in premonetary multi-centric economies where there are moral sanc
tions against direct conversions between spheres of exchange. It is important
precisely because it allows goods of little value to be used to create and acquire
valuables and prestige. It provides a venue for building a career through strategic
"investment."
Such an exploration of CWEs also exposes the potential that exists for some
forms of work feast to enable the increasing exploitation of labor even in the con
text of societies with an egalitarian political ethos. Regardless of the formal ide
ology of a society, large work feasts that are viewed as a finite exchange
transaction with no reciprocal labor obligations can result, in the course of prac
tice, in asymmetrical labor flows, such that some individuals or households de-
257
Michael DUtler and Ingrid Her!Tich
rive wealth and prestige from the labor of others. This fact has profound signifi
cance for the long-term development of social relations and economic structures.
Obviously, it also has crucial implications for archaeologists attempting to under
stand the development and exacerbation of conditions of social inequality. Fi
nally, it dramatically underscores yet one more reason why it is essential for
archaeologists to recognize the importance of feasts in social life.
ACKNOW L E D G M E NTS
Portions of this discussion have been presented previously at various symposia and semi
nars (e.g. , Dietler 1989, 1996; Herbich 1991). We benefited particularly from discussion at
the E cole des Hautes Etudes en Sciences Sociales in Marseille and the Universite de Paris I
(Sorbonne-Pantheon). Special thanks are due to Anick Coudart, Jean-Paul Demoule, and
Andre Tchernia.
For making possible our ethnographic research in western Kenya, thanks are due to the
National Science Foundation, the Wenner-Gren Foundation, the Boise Fund of Oxford
University, the Office of the President of Kenya, the National Museums of Kenya, and es
pecially our Luo and Sarnia hosts and our research assistants, Rhoda Onyango, Monica
Oyier, and the late Elijah Ogutu.
NOTES
r. The distinction between "work feasts" and "work exchanges" corresponds
approximately to categories with slightly different names developed in several earlier
comparative analyses of collective labor practices. For example, Erasmus (1956)
denoted these practices with the terms "festive labor" and "exchange labor,"
respectively; whereas Moore (1975) called them "non-reciprocal co-operative labour"
and "reciprocal co-operative labour." While owing an obvious major debt to these
seminal works, we employ an alternative terminology in order to signify a particular
analytical focus on the role of the feast event in orchestrating labor projects and to
mark certain subtle differences from these earlier works.
2. Work feasts were/ are found from Africa to Latin America, from the Caribbean to Asia,
from European peasants to North American colonists and their barn-raisings. In
addition to the works cited elsewhere in this text, see, for example, Cancian 1972;
Eguchi 1975; Guillet 1980; Herskovits 1937:70-76; Kahn 1998; Kenyatta 1938:59--6 0 ;
Lomnitz 1976:183; March 1998; Moerman 1968; Provinse 1937; Salisbury 1962. This is, of
course, merely a minuscule sample cited in order to suggest the geographical extent
of the practice.
3. All unreferenced descriptions of practices among the Luo and Sarnia peoples in this
paper are derived from research conducted by us in western Kenya from 1980 to 1983
(see, e.g., Dietler and Herbich 1993; Herbich 1987, 1991; Herbich and Dietler 1991, 1993).
4. Bourdieu's (1990) The Logic of Practice is a theoretical reformulation of concepts, such
as symbolic capital, presented earlier in his (1978) Outline of a Theory of Practice.
258
Although the latter text is better known in the Anglophone anthropological

community, it has been largely superseded by the later work. The adoption of
Bourdieu's metaphorical use of "capital" to translate various forms of socially
constructed power poses certain dangers-not least the fact that it may be easily
misconstrued as advocating an extension of the simplistic and inappropriate
neoclassical economic perspective to the analysis of social and cultural phenomena or
as naturalizing the cultural logic of capitalism. Hence, although we have certain
misgivings, the term symbolic capital is used here largely because it is the most useful
trope we have found for illuminating the operation of the work feast. However, we
issue the caution that it is used in the specific sense stemming from Bourdieu's
theoretical program for the analysis of symbolic domination, and not in a neoclassical
economics sense or rational actor theory sense.
5. For example, Donham ( 1994:46-152) provides an example of two very different
patterns in the same society in Ethiopia. The Bola faction had a classic pattern of
labor exploitation by the wealthy as described earlier. However, among the Dofo
group, there was a very different flow of work-feast labor due to the influence of a
political struggle by elders attempting to preserve their factional position. In this case,
with a few exceptions, the most politically influential households had deficits of labor
flows in work feasts and there was a net transfer of labor from elders' households to
middle-aged men's households (the elders sent younger members of their very large
households to the work feasts of junior men rather than coming themselves). In this
case, "elders appeared to be transferring labor to middle-aged men in return for
support and deference" (Donham 1994:151) .
REFERENCES
Arnold, B.
1988 Slavery in Late Prehistoric Europe: Recovering the Evidence for Social Struc
ture in Iron Age Society. In Tribe and Polity in Late Prehistoric Europe, edited by
D. B. Gibson and M. N. Geselowitz, pp. 179--192. New York: Plenum Press.
Barth, F.
l967a Economic Spheres in Darfur. In Themes in Economic Anthropology, edited by
R. Firth, pp. 49--174· London: Tavistock.
l967b On the Study of Social Change. American Anthrapologist 69:661--670.
Bohannan, P.
1955 Some Principles of Exchange and Investment among the Tiv. American An
thrapologist 57:60-70.
Bourdieu, P.
1986 The Forms of Capital. In Handbook of Theory and Researchfor the Sociology of Ed
ucation, edited by). G. Richardson, pp. 241-258 . New York: Greenwood Press.
259
Brown, ].
1995 Traditional Metalworking in Kenya. Cambridge Monographs in African Ar
chaeology 38, Cambridge.
Cancian, F.
1965 Economics and Prestige in a Maya Community. Stanford: Stanford University
Press.
Chibnik, M., and W de Jong
1990 Agricultural Labor Organization in Riberefio Communities of the Peruvian
Amazon. Ethnology 29:75-95.
Childe, V G.
1936 Man Makes Himself. New York: Mentor.
Clark, C. M.
1980 Land and Food, Women and Power, in Nineteenth-Century Kikuyu. Africa
50:357-370.
Clark, ]. E.
1995 Craft Specialization as an Archaeological Category. Research in Economic An
Clark, ]. E., and W J. Parry
1990 Craft Specialization and Cultural Complexity. Research in Economic Anthropol
ogy 2:289-346.
Colson, E.
1949 Life among the Cattle-Owning Plateau Tonga. Livingstone, Zambia: The
Rhodes-Livingstone Museum Occasional Papers 6.
Gwembe District, Zambia, 1950-1982. Stanford: Stanford University Press.
Costin, C. L.
1991 Craft Specialization: Issues in Defining, Documenting, and Explaining the
Organization of Production. Archaeological Method and Theory p-56.
Cummings, R. J.
1976 The Early Development of Akamba Local Trade History; c.1780-1820. Kenya
Historical Review 4: 85-no.
Daubigney; A.
1983 Relations marchandes mediterraneennes et proces des rapports de depen
dance (magu- et ambactes) en Gaule protohistorique. In Modes de contacts et
processus de transformation dans les societes anciennes, pp. 659-683. Rome: Ecole
fran<;aise de Rome.
Dietler, M.
1989 The Work-Party Feast as a Mechanism of Labor Mobilization and Exploita
tion: the Case of Sarnia Iron Production. Paper presented at 88th Annual
Meeting of the American Anthropological Association, Washington, D.C . ,
November.
260
F E A STS A N D LA B O R M O B I L I Z AT I O N
9:352-406.
Perspective, edited by P. Wiessner and W. Schiefenhovel, pp. 87-125. Oxford:
Berghahn Books.
Dietler, M., and I. Herbich
1993 Living on Luo Time: Reckoning Sequence, Duration, History, and Biogra
phy in a Rural African Society. World Archaeology 25:248-260.
Dillon, R. G.
1990 Ranking and Resistance: A Precolonial Cameroonian Polity in Regional Perspective.
Stanford: Stanford University Press.
Donham, D.
1994 Work and Power in Maale, Ethiopia. New York: Columbia University Press.
[1979]
Eguchi, P. K.
1975 Beer Drinking and Festivals among the Hide. Kyoto University A.fiican Studies
9: 69-90.
Erasmus, C. ).
1956 Culture Structure and Culture Process: The Occurrence and Disappearance
of Reciprocal Farm Labor. Southwesternjournal of Anthropology 12:444-469.
Finley; M. I.
1985 The Ancient Economy. London: Hogarth Press.
Garlan, Y.
1988 Slavery in Ancient Greece. Ithaca: Cornell University Press.
Geschire, P.
1982 Village Communities and the State: Changing Relations among the Maka of South
eastern Cameroon since the Colonial Conquest. Translated by J. Ravell. London:
Kegan Paul.
Goldschmidt, W.
1976 Culture and Behavior of the Sebei: A Study in Continuity and Adaptation. Berke
ley: University of California Press.
Goody; ).
1982 Cooking, Cuisine, and Class: A Study in Comparative Sociology. Cambridge: Cam
Guillet, D.
1980 Reciprocal Labor and Peripheral Capitalism in the Central Andes. Ethnology
19:151-167.
Herbich, I.
1987 Learning Patterns, Potter Interaction, and Ceramic Style among the Luo of
Kenya. The A.fiican Archaeological Review s;r93-204.
261
1991 The Flow of Drink in an African Society: An Ethnoarchaeological Perspec

tive. Paper presented at the 56th Annual Meeting of the Society for American
Archaeology; New Orleans, April.
Herbich, I., and M. Dietler
1989 River-Lake Nilotic: Luo. In Kenyan Pots and Potters, edited by]. Barbour and S.
Wandibba, pp. 27-40. Nairobi: Oxford University Press.
1991 Aspects of the Ceramic System of the Luo of Kenya. In Tiipferei- und Keramik
forschung, 2, edited by H. Ludtke and R. Vossen, pp. 105-135. Bonn: Habelt.
1993 Space, Time, and Symbolic Structure in the Luo Homestead: An Ethno
archaeological Study of "Settlement Biography" in Africa. In Actes du XIIe
Congres International des Sciences Prehistoriques et Protohistoriques, Bratislava,
Czechoslovakia, September 1�, 1991, Vol. 1, edited by ]. PavUk, pp. 26---3 2. Nitta:
Archaeological Institute of the Slovak Academy of Sciences.
Herskovits, M. ].
1937 Dahomey: An Ancient West African Kingdom. New York: J. ]. Augustin.
Hunter, M.
1961 Reaction to Conquest: Effects of Contact with Europeans on the Pondo of Southern
Africa. London: Oxford University Press.
Kahn, M.
1998 "Men Are Taro" (They Cannot Be Rice): Political Aspects of Food Choices in
Wamira, Papua New Guinea. In Food and Gender: Identity and Power, edited by
C. M. Counihan and S. L. Kaplan, pp. 2g--44. Amsterdam: Harwood Academ
ic Publishers.
Karp, I.
1980 Beer Drinking and Social Experience in an African Society. In Explorations in
African Systems of Thought, edited by I. Karp and C . Bird, pp. 83-n9. Bloom
ington: Indiana University Press.
Kennedy, ]. G.
1978 The Tarahumara of the Sierra Madre: Beer, Ecology, and Social Organization. Ar
lington Heights, Ill . : AHM.
Kenyatta, ].
1965 Facing Mt. Kenya: The Tribal Life of the Gikuyu. New York: Vintage.
Kopytoff, I.
1982 Slavery. Annual Review of Anthropology n:207-30.
1986 The Cultural Biography of Things: Commoditization as Process. In The So
cial Life of Things: Commodities in Cultural Perspective, edited by A. Appadurai,
pp. 64--9!. Cambridge: Cambridge University Press.
Lemert, E.
1964 Forms and Pathology of Drinking in Three Polynesian Societies. American
Anthropologist 66:361-374.
Lemonnier, P.
1990 Guerres et festins: Paix, echanges et compt!tition dans !es Highlands de Nouvelle
Guinee. Paris: CID-Editions de la Maison des Sciences de !'Homme.
262
Lomnitz, L.
1976 Alcohol and Culture: The Historical Evolution of Drinking Patterns among
the Mapuche. In Cross-Cultural Approaches to the Study of Alcohol: An Interdisci�
plinary Perspective, edited by M. Everett, J. Waddell and D. Heath, pp. 177-198.
The Hague: Mouton.
March, K. S.
Power, edited by C. M. Counihan and S. L. Kaplan, pp. 45-80 . Amsterdam:
Mayer, P.
1951 Agricultural Co-operation by Neighborhood Groups among the Gusii. In
Two Studies in Applied Anthropology in Kenya, edited by P. Mayer. London:
HMSO.
Meillassoux, C .
1991 The Anthropology of Slavery: The Womb of Iron and Gold. Translated by A. Das
nois. Chicago: University of Chicago Press.
Miers, S., and I. Kopytoff, eds.
1977 Slavery in Africa: Historical and Anthropological Perspectives. Madison: Univer
sity of Wisconsin Press.
Moerman, M.
1968 Agricultural Change and Peasant Choice in a Thai Village. Los Angeles: Univer
sity of California Press.
Moore, M. P.
1975 Cooperative Labour in Peasant Agriculture. The journal of Peasant Studies
2:270--29!.
Nadel, S. F.
1942 A Black Byzantium: The Kingdom of Nupe in Nigeria. London: Oxford Univer
sity Press.
Netting, R.
ogist 66:375-384.
Peacock, D. P.
1982 Pottery in the Roman World: An Ethnoarchaeological Approach. London: Long
man.
Peschel, K.
1971 Zur Frage der Sklaverei bei den Kelten wlihrend der vorromischen Eisenzeit.
Ethnographisch-Architologische Zeitschrift 12:527--s39.
Piot, C .
1991 Of Persons and Things: Some Reflections on African Spheres of Exchange.
Man 26:405-424.
Provinse, J.
1937 Cooperative Ricefield Cultivation among the Siang Dyaks of Central Bor
neo. American Anthropologist 39:77-102.
263
Renfrew, C.
1973 Monuments, Mobilization, and Social Organization in Neolithic Wessex. In
The Explanation of Culture Change: Models in Prehistory, edited by C. Renfrew,
pp. 539-558 . London: Duckworth.
Richards, A. I.
1939 Land, Labour, and Diet in Northern Rhodesia. London: Oxford University Press.
Sahlins, M.
1972 Stone Age Economics. London: Tavistock.
Salisbury, R. F.
1962 From Stone to Steel: Economic Consequences of a Technological Change in New
Guinea. Melbourne: Melbourne University Press.
Saul, M.
1983 Work Parties, Wages, and Accumulation in a Voltaic Village. American Ethnol
ogist rn:77-96.
Singleton, T.
1995 The Archaeology of Slavery in North America. Annual Review of Anthropology
24:II9-l40.
Tosh, ].
1978 Clan Leaders and Colonial Chiefs in Lango: The Political History of an East Afiican
Stateless Society. Oxford: Clarendon Press.
Trigger, B.
1990 Monumental Architecture: A Thermodynamic Explanation of Symbolic Be
haviour. World Archaeology 22:n9-132.
Uchendu, V.
1970 Traditional Work-Groups in Economic Development. In Annual Conference
Proceedings, East Afiican Universities, Dar es Salaam, vol. 5. Kampala: Makerere
Institute of Social Research.
Washburne, C .
1961 Primitive Drinking: A Study of the Uses and Functions of Alcohol in Preliterate So
cieties. New York: College and University Press.
Webster, G.
1990 Labor Control and Emergent Stratification in Prehistoric Europe. Current An
264
Pa rt 2
A R C H A E O LO G I C A L
P E RS P E CT I V E S
10
THE EVO L U TION OF RIT UA L FEASTING SYSTE MS
IN PREHISPANIC PHILIP PINE CHIE FDOMS
Laura Lee Junker
At the time of initial Spanish colonization, the Philippine archipelago had a polit
ical landscape composed of numerous coastal and riverine chiefdoms of varying
scale and complexity that interacted through maritime trade. Archaeological evi
dence for status-related differences in burials and regional settlement hierarchies
indicate that these complex societies existed for at least a millennium prior to Eu
ropean contact (Hutterer 1977; Jocano 1975; Junker 1994, 1999). Historical sources
suggest that these societies were ruled by hereditary chiefs. However, remarkably
low population densities relative to productive agricultural land, and the conse
quently high value placed on control of labor, favored political units composed of
shifting, alliance-structured coalitions rather than the more permanent, territori
ally defined constituencies characteristic of chiefdoms in many other regions of
267
Laura Leejunker
the world (Junker 1998; Kiefer 1972; also see Reid 199J:J-6; Winzeler 1976; Wolters
1982:1-15). Ethnohistoric analysis shows that these political coalitions were main
tained through the strategic redistribution of resources obtained through tribute
mobilization, maritime trading, raiding, and sponsored craft production to both
allies and subordinates in the contexts of ritualized exchange and competitive
feasting. Because of the tenuous nature of political ties, competition for wealth
and followers was perpetual and fierce, with chiefs continually striving to obtain
new sources of politically manipulable wealth, elite knowledge, and power.
Some time at the end of the first millennium A.D. , and intensifying just prior to
early sixteenth-century Spanish contact, Philippine chiefs became involved in
long-distance maritime trade for prestige goods with the Chinese empire and
other Southeast Asian polities (Hutterer 1977; Junker 1994). Chinese porcelain,
silks, metal goods, and other exotic luxury items from outside the archipelago be
came key symbols of social status and political power for the Philippine chiefly
elite. Archaeological and historic evidence suggest that this foreign luxury-good
trade reached its height in terms of volume and interpolity trade competition in
the fifteenth and sixteenth centuries, with increasing use of foreign imports as
"wealth" objects in elite households and as grave accompaniments to high-status
burials (Fox 1964, 1967; Junker 1993a, 1993b; Nishimura 1988). Historic sources in
dicate that foreign porcelains were in high demand as status-enhancing serving
assemblages in ritual feasting and they were circulated as valuables in elite ex
changes associated with ritual feasts. Not surprisingly, the emergence of larger
and more complex chiefdoms in a number of regions of the Philippines in the
two centuries prior to Spanish contact is associated with an increased involve
ment in the foreign porcelain trade and an expanded scale of ritual feasting (Hut
terer 1973, 1977; Junker 1999) .
Here, I focus on the role of ritual feasting in the evolving political economy of
Philippine chiefdoms between the tenth and nineteenth centuries. Because mar
itime trade for porcelains, silks, metal weaponry, and other mainland Asian pres
tige goods became an important component of chiefly political economies in
many Philippine island polities by the late first millennium A.D. , their literate Chi
nese trade partners have provided us with nearly a millennium of historic writ
ings on Philippine culture. These Chinese sources include a few references to
ritual feasting (Junker 1998). However, the richest documentary evidence for rit
ual feasting and its role in chiefly political economies comes from contact-period
(sixteenth-century) Spanish texts, many of which provide considerable detail on
"pagan rituals" as part of Christian evangelizing efforts. Another remarkable
source of information on ritual exchange and celebratory feasts are ethnographic
accounts of Philippine chiefdoms in the island interiors (such as the Bagabo,
Magindanao, and Bukidnon of Mindanao), complex societies that were still
268
R I T U A L F E ASTI N G S Y ST E M S I N P R E H I S PA N I C P H I L I P P I N E C H I E F D O M S
largely intact and independent o f colonial control in the early twentieth century.
In his 1913 Wild Tribes of the Davao District, ethnographer Fay Cooper Cole pro
vided a particularly vivid description of human sacrifice and ritualized exchanges
as part of an annual "harvest" feast among the Bagabo of southeastern Min
danao. Finally; archaeologists have recently begun to look at material evidence
relevant to the long-term evolution of ritual feasting in the Philippines, studying
household cerainic assemblages and subsistence remains at a number of coastal
chiefly centers (Junker, Mudar, and Schwaller 1994; Mudar 1997; see Fig. rn.1).
H ISTORICALLY WELL
KNO\M\I POLITIES
I 1. P'u-tuan (Butuan)
I
I 2. Manila
3. Ma-i (Mindoro)
�
4. Cebu
5. Sul u
'
- , 6. Magindanao
ETHN OGRAPHICALLY
KNOWN CHIE FDOMS
5. Tausug (Sulu)
6. Magindanao
7. Bagabo
8. Bukidnon
ARCHAEOLOGICAL S ITES
2. Santa Ana/Manila
4. Cebu
9. Calatagan
- -
1 0 . Tanjay
I
o 200 km
Figure ro.r. The location of tenth- to nineteenth-century

Philippine chiefdoms known through Chinese accounts,
Spanish records, ethnographic study, and archaeological
research.
269
Laura Leejunker
"Feasts of merit" in prehispanic Philippine chiefdoms, like ritual feasts in many

complex societies, served to reproduce social relations. Both community cohe
sion and social rank differentiation were reaffirmed and continually renegotiated
through exchange of valuables, chiefs' oral narratives, animal sacrifice, food
prestations, and ancestor-invoking ritual. What is particularly significant about
the Philippine case is that ritual feasting was highly competitive and socially
transformative, with each feast realigning social and political power relations
within alliance and patronage networks. Chiefs, nobles, and those with elite pre
tensions created the social debt fundamental to expanding power asymmetries by
distributing increasingly more elaborate status foods and valuables to an ever
widening circle of participants. In many chiefdom-level societies, ritual feasting
tends to focus less on overt status competition and negotiation of political rela
tionships than on ideological reinforcement of existing power hierarchies be
tween hereditary chiefs and their constituencies. In the Philippines, and
elsewhere in island Southeast Asia, the development of strongly centralized poli
ties with powerfully integrated political hierarchies and relatively permanent,
territorially defined power bases may have been constrained by certain ecologi
cal, geographic, and demographic factors. In these weakly integrated polities,
political power was transient and relied, not on control over fixed territorial bases
or well-defined unilineal descent groups, but on cultivating ties of personal loy
alty and expanding one's power base through frequent and elaborate gift-giving
and ceremonialism. The first section of this chapter examines the complex role
of ritual feasting in Philippine chiefly politics through an analysis of historic and
ethnographic sources.
In the second half of the chapter, I consider evolutionary aspects of the feast
ing system and how it relates, in a dynamic way; to other components of an ex
panding chiefly political economy Archaeological evidence for ritual feasting at
prehispanic Philippine settlements documents changes in the intensity, scale, and
social dynamics of these ceremonial events as more complex chiefdoms emerged
in the millennia prior to European contact. The distribution of faunal and floral
remains that might be associated with ritual feasting, the presence of specialized
food-preparation and serving assemblages at prehispanic settlements, and the ar
chaeological study of specific spatial contexts associated with feasting suggest
some ways in which these feasting systems may have transformed over time. In
creased production of high-value foods such as pig, water buffalo, and rice, and
their differential distribution across elite and non-elite residential zones at some
fifteenth- and sixteenth-century chiefly centers suggests widening social partici
pation in "feasts of merit" and an inflationary scale of material inputs compared
to earlier periods. Archaeological evidence from a number of chiefly centers also
suggests that foreign porcelain serving assemblages become an increasingly im-
270
R I T U A L F E A S T I N G S Y ST E M S I N P R E H I S PA N I C P H I L I P P I N E C H I E F D O M S
portant component of elite feasting paraphernalia i n the few centuries prior to

European contact. At the same time, expanded production of local status wares
may be tied to the emulation of elite feasting rites by lower-ranked individuals
who did not have access to elaborate foreign porcelain assemblages.
Although the archaeological data documents these material trends, we must
turn once again to the ethnohistoric evidence from the Philippines and elsewhere
in Southeast Asia to develop more specific ideas about how an expanding system
of competitive feasting might relate to broader transformations in chiefly politi
cal economies. Integrating archaeological and ethnohistoric data, I conclude the
paper with a discussion of how ritual feasting is tied to changing chiefly produc
tion strategies, escalating competition over access to foreign porcelains, and in
creasing interpolity warfare and slave-raiding as larger-scale chiefdoms emerged
in the fifteenth and sixteenth centuries. What is clear is that material exchanges
associated with ritual feasting are central to political alliance building and to the
expansion of chiefly political economies. Therefore, ethnohistoric and archaeo
logical analysis of ritual feasting is key to any study of long-term political
processes in the prehispanic Philippines.
RITUAL F EASTI N G I N P H I L I P P I N E COMPLEX SOCI ETI ES:

T H E ETH N O G RA P H I C EVI D E N C E
What the sixteenth-century Spaniards described as "feasts of ostentation and
vanity" or "feasts of merit" (Colin 1975:175 ) were a central feature of the political
economy of virtually all Philippine complex societies. Ritual feasts are described
by Spanish observers and early ethnographic accounts as the cornerstone of so
cial, political, economic, and religious life in these societies. These feasts were
generally associated with elite life-crisis events (birth, marriage, illness, death)
and events critical to the political economy (e.g., chiefly succession, trading expe
ditions, warfare and maritime raiding expeditions, political alliances or "peace
pacts," "harvest festivals" and other ritual events associated with the agricultural
cycle; Boxer manuscript 1975:190, 201, 213-214; Chirino 1904:262-271; Loarca
1903:149-151; Perez l95r:ro2-ro3, no; Pigafetta 1975:65-66; Plasencia l903b:190-195;
Santa Ines 199078-79) .
Ethnohistoric evidence suggests that feasting in the Philippines simultaneously
transacted values of sociopolitical integration and asymmetries, wealth accumu
lation and generalized redistribution, socially restricted ritual potency, and super
naturally reinforced community well-being. The ability of a chief to draw on the
resources of his constituency to finance a feast demonstrated a status-enhancing
power to mobilize productivity. His generosity in distributing food to both nobles
and commoners attending the ritual feast emphasized his role as a superior kins
man and strengthened the often tenuous bonds that held together political coali-
271
Laura LeeJunker
tions. Lavish gifting of valuables such as porcelain bowls or metal gongs to elite
participants served to both maintain politically significant reciprocal exchange re
lationships (the material glue of alliance building) and to symbol overtly a partic
ular chief's rank in a social hierarchy vis-a-vis other elites (as measured through
his control of wealth).
Although contact-period Philippine chiefdoms diverged widely in terms of
scale and complexity, ceremonial feasting was ubiquitous and was characterized
by certain pan-archipelago features. These features include (1) sponsorship
(though not always exclusively) by elite individuals (most frequently; chiefs); (2)
the performance of sacrificial rites using animals (usually pigs, chickens, and/ or
water buffalo), other foodstuffs, people, and/ or manufactured goods contributed
by individuals in a tributary or subservient role to the sponsor; (3) elite exchanges
of valuables (e.g., porcelain, silks, gold jewelry) as part of ongoing reciprocal ex
change partnerships; (4) reallocation of meat and other feasting foods for con
sumption according to kinship ties and social rank relations with the sponsoring
elite; and (5) the conferring of social prestige on the feast's sponsor in accordance
with the feast's lavishness and the social debt created through the sponsor's
prestations.
F E A S T I N G A N D T H E R I T U A L I Z E D N E G OT I AT I O N
O F S O C I A L - P O L I T I C A L R E L AT I O N S
In an ethnohistoric analysis of the use of food in ritual contexts amongst the na
tive American Oglala, Powers and Powers distinguish "feasts" from mundane
consumption as "a food event which is somehow commemorative or celebratory
of perhaps a historical or religious occasion . . . [the feasts] having some intrinsic
social value which transcends the nutritive function of eating" (1984:83) . Feasts
are thus generally associated with ceremonial or ritual events held either periodi
cally (or on an ad hoc basis), they generally involve commensal units larger than
the usual domestic units, and they are often highly structured in terms of what is
served and how it is served. Often, the most significant aspect of a feast is its rit
ual context, within which are generally embedded specific social goals, although
Hayden (1994) points out that feasts do not invariably involve religious compo
nents or social competition (e.g. , work feasts). Ritual feasts are occasions when
social relations are negotiated and reaffirmed (Rosman and Rubel 1978). Because
ritual feasts in the Philippines were most often sponsored by chiefs and tied
closely to aspects of chiefly political economy; they were also important in rein
forcing the ideology of chieftainship. To be effectively promulgated, a ruling
class ideology must be materialized in widely witnessed ceremonial events and
publicly displayed symbolic objects (DeMarais, Castillo, and Earle 1996; Earle
l99Tl43-192). Since Philippine chiefdoms did not construct a monumental land-
272
scape a s a permanent representation o f political and social power relations

(Junker 1999), repetitive ceremonial events were particularly important in institu
tionalizing power asymmetries.
Sixteenth-century Spanish sources and early ethnographic accounts of Philip
pine chiefdoms persisting into the modern era suggest that ceremonial feasts
were held frequently throughout the year in a variety of social contexts that in
cluded both calendrical and life-crisis ritual. Life-crisis events requiring the spon
sorship of a ritual feast by the kin group included the birth of a child,
achievement of puberty, marriage, illness, or death of a household member (e.g. ,
Anon. 199oa:164; Anon. 199ob:110; Boxer manuscript 1975:190-191, 193; Chirino
1904:262-271; Loarca 1903:14g--151; Plasencia 1903b:190-195; San Antonio 1990:313,
316, 336-338; S;aaanchez 161n87). The Spanish sources suggest that both elite and
non-elite kin groups would hold these life-crisis ritual feasts if they possessed the
necessary resources, but for chiefs and other elites these events took on signifi
cant political meaning (see below) . Other noncalendrical ritual feasts are con
cerned with more ad hoc elite endeavors, such as the construction of a chiefly
residence or war canoe, and the initiation of maritime raiding or trading expedi
tions. As discussed in more detail below, calendrical rituals associated with plant
ing and harvesting agricultural crops also involved ceremonial feasting and were
intimately tied to chiefly tribute-mobilization systems.
Specialist priests, known as babaylans in Visayan and katulunan in Tagalog and
often elite females, performed the religious rites associated with curing or vari
ous rites of passage, as well as the ritual killing of the sacrificial animals (prefer
ably pigs or water buffalo) to be later consumed by the guests at the ritual feast.
An extended discussion of the cosmological underpinnings of these sacrificial
rites are beyond the scope of the present work (see Gibson 1986; Scott 199477-93,
233-241; and De Raedt 1989 for more detailed analysis of this aspect of feasting).
However, we note here that historical sources and ethnographic accounts sug
gest that animal sacrifice and consumption by the feast's participants was aimed
at warding off afflicting spirits who were seen as attempting to prevent the pas
sage of individuals through critical life stages (i. e., birth, marriage, death), who
were causing illness in an individual, or who were likely to prevent the successful
completion of community endeavors such as warfare, large-scale maritime trad
ing expeditions, agricultural production, or even the construction of a chiefly res
idence. There is some indication that these potentially malevolent spirits were
actually ancestors, who could bring misfortune on their descendants if the
proper ancestral sacrifices were not made at critical points in the life-cycles of the
living, and particularly if individuals had violated certain social proscriptions
such as engaging in "incestuous" marriages (e.g., Cole 1913:111-120; Gibson
1986:173-176; Scott 1994:237-238) .
273
Laura Leejunker
The sacrifice had to be performed for the kin group to regain the benevolence
of these ancestors or to ensure that malevolent spirits will not cross the threshold
of a house to cause misfortune for the inhabitants in their various endeavors. An
imals such as pigs were viewed as intermediaries between the human and spirit
worlds, who exchange their lives to bring "vitality" to the individuals sponsoring
the sacrifice. Anyone who participates in consuming the ritual animals' flesh is
also imbued with "vitality" and to some degree shares the supernatural protec
tion afforded by carrying out the sacrificial rites. Thus, the sacrificial rites were a
reaffirmation of social connectedness, whether these social relations involved a
kin-like alliance between social equals or asymmetrical patron-client ties.
R E C I P R O C I T Y A N D E X C H A N G E R E L AT I O N S I N F E A S T I N G
All of the feast participants, including chiefs and nobles as well as commoners
and slaves attached to the sponsor, were obliged to make contributions to the
feast (e.g., carabao or domestic water buffalo, pigs, fowl, rice, wine) and/ or offer
ings to be used in the accompanying sacrificial ritual and in payment to religious
specialists performing it. These contributions included gold, cotton, metal
weapons, other prestige goods, rice, pigs, chick.ens, and other "high-quality" sub
sistence items (Bobadilla 1990:334; Chirino 1904:262-271; Cole 1913:111-120,
1956:94-117; Boxer manuscript 1975:201; San Antonio 1990:314) . It is clear from
both the contact-period Spanish sources and early ethnographic accounts that
contributions from subordinates were exacted as a form of tribute or enforced
labor, whereas the prestations from members of the nobility took place in the
context of alliance-building reciprocal "gift" exchanges.
Those feasts marking critical points on the agricultural calendar, in particular,
were sufficiently regular to have been the key context in which tribute collection,
communal agricultural labor, and amassing of chiefly surplus took place (see
Cole 1913:111-120) . It is probably not a coincidence that the largest and most pro
tracted of the Bagabo feasts (the Gin Em feast) occurs just after the annual fall
rice harvest, when the granaries of subordinate chiefs and village commoners are
full of available rice for large tribute exactions. There are parallels with the
Makahiki festival and other chiefly agricultural rites of contact-period Hawaii
(Earle 1977:225-226; Peebles and Kus 1977:425; also see Kirch, Chapter 6 ) .
Ethnographic descriptions of these ritual events among the Tausug (the core
cultural group of the Sulu sultanate), Bukidnon, and Bagabo of Mindanao indi
cate that a significant component of these events is the reinforcement of elite
alliances (Claver 1985:74-75; Cole 1913:rn-112; Kiefer 1972:26, 97) . Alternately spon
sored in tb,eir own districts by local chiefs and periodically by regional chieftains,
these feasts appear to have been the primary opportunity for reciprocal gift ex
changes between elites (involving exchange of imported porcelains, valuable
274
R I T U A L F E A STI N G S Y ST E M S I N P R E H I S PA N I C P H I L I P P I N E C H I E F D O M S
metal gongs, gold jewelry and/ o r other prestige goods) . Elite prestations occur
ring at any single feast represented links in intertwined chains of ongoing recip
rocal exchange partnerships between kin-related and allied nobility, reflecting the
unique state of specific social relationships at any one point in time. Analogies
can be made with the alliance-based systems of prestige-good exchange recorded
for other areas of island Southeast Asia (e.g., Beatty 1991; Volkman 1985) and
Oceania (e.g., Gregory 1982; Rosman and Rubel 1978; Strathern 1971; also see
Weissner, Chapter 4) . Although a portion of the individual prestations are gener
ally returned and new debt created through the host's prestige-enhancing feast
ing of all the attending celebrants, the host also acquires the obligation to make
ceremonial contributions at future feasts sponsored by his exchange partners, en
suring the continuity of the exchange system. The social values transacted by
these prestations include political cohesion through the circulation of resources
and marriage partners, as well as the social prestige and political influence ema
nating from a wide network of exchange partners and sponsorship of lavish
feasts (Beatty 1991:230) .
Despite obligatory contributions from subordinates (exacted as a form of trib
ute) and from allied political leaders (framed in terms of alliance-building recip
rocal gift exchanges), the Spanish chroniclers suggest that the feast sponsor bore
the expensive burden of financing the feast (Boxer manuscript 1975:215; Chirino
1904:262-271; also see Claver 1985:27; De Raedt 1989:239; Prill-Brett 1989:3, 8 ) . Feast
financing presumably depended heavily on stored resources accumulated over
the long term by socially and politically prominent kin groups through their ex
tensive marriage ties and exchange networks.
S O C I A L P R E S T I G E O R " M E R I T " A S A T R A N S A C T E D VA L U E
The prestige-enhancing aspects of these feasts in the Philippines are emphasized
by Biernatzki's (1985) and Claver's ( 1985) ethnographic and historical work on the
Bukidnon, a traditional chiefly society of northern Mindanao (also see Cole
l956:94-n7) . At ceremonial feasts marking the succession of a new chief or other
politically significant events, the sponsoring chief typically invited as many chiefs
from neighboring districts and regions as possible to participate in the lavish feast
(Biernatzki 1985:36-37) . The sponsoring of feasts was key to chiefly political
strategies, since "the ease with which animals are procured, excellence in oratory,
and the ability to feast guests and people sumptuously [were] qualities par excel
lence of a 'first-class' datu [chief]" (Claver 1985:86) .
In his study of the feasting system amongst the Sagada Igorots and other
groups of interior northern Luzon, Voss (198T131) suggests that the prestige
gained from sponsorship of such feasts came from the creation of social debt
amongst the people attending the feast. In expending huge amounts of his
275
Laura LeeJunker
wealth in feeding the attendees at the feast, many of whom did not have the re
sources to reciprocate in kind, a feast's host expanded the number of people in an
asymmetrical relationship to him , as well as reinforced the "debt" of existing ties
of patronage. Voss notes that the prestige of an individual and of his immediate
kin group is gauged in terms of these accumulating debts or obligations that re
quire the debtors to subsequently provide agricultural labor or other services, as
well as return prestations, when called on by the feast's sponsor. Spanish sources
suggest that sixteenth-century datus often invested their surplus wealth in aiding
other men to sponsor feasts associated with marriage rites or death rites, thus
tightening their grip on these individuals as core supporters in the datu's alliance
network (Alcina 196ob:76-77, 180-198) .
Not only did the sponsoring chief or kin group increase their prestige in the re
gion by providing the enormous supplies of food and status goods disbursed at
such a ritual event, but the occasion provided an opportunity for all the attending
datus (chiefs) and male kin-group leaders to attempt to improve their position
within a ranked political hierarchy of chiefly authority. Biernatzki (1985:36-37) de
scribes this overt negotiation of power relations in the traditional "boasting con
test" accompanying ceremonial feasts among the ethnographically known
Bukidnon chiefdoms of northern Mindanao. In these feasts, each datu in turn
climbed a ladder to a high ceremonial platform laid out with a lavish meal, at each
rung of the ladder reciting the genealogical history that supported his inherited
claim to chieftainship and elite status. However, he would also present an even
more protracted recital of his personal exploits in trading, raiding, and other
"wealth"-producing activities, attempting to out-boast other attending chiefs and
win a place at the highest-ranking ceremonial table. These public recitations and
symbolic movement of chiefs up the status ladder allowed the community to com
pare chiefly prowess and to rank the multiple district chiefs in terms of appropriate
levels of deference. At least one contact-period Spanish text refers to similar public
recitations of status-validating genealogies and heroic accounts of trading and raid
ing exploits as part of feasting ritual in the sixteenth century (e.g., Bobadilla
1990:332). Oral presentations of status-enhancing genealogies and personal achieve
ments, allowing public assessment of the relative status, wealth, and power of at
tending chiefs, was associated with ceremonial feasting in other areas of Southeast
Asia and in Polynesia (Goldman 1970:522--s36; Schnitger 1964; Volkman 1989 ).
Animals (most commonly carabao, pigs, chickens, and dogs), rice and other
plant foods, and alcohol beverages were offered for sacrifice in the ritual compo
nent of Philippine ceremonies. Both the quality and quantities of sacrificial offer
ings reflected the social rank, wealth, and political power base of the sponsoring
chief (Biernatzki 1985:36-37; Chirino 1904:269-270; Claver 1985:74; San Antonio
1990:335-336). As in other complex societies of Southeast Asia, the maintenance of
276
the social status quo o r social mobility depended upon successful performance
over a lifetime in ritual feasting events (as well as abilities in warfare, trading, and
wealth acquisition; Beatty 1991; Kirsch 1973:26--27; Leach 1954:163). The carabao,
as a draft animal with a slow reproductive rate, and the pig, appear to have been
the most valued animals for both sacrifice and exchange in Philippine societies
(Barton 1949:74-75; Biernatzki 1985:43; Claver 1985:74-75, 86; Dasmarinas 1958:429;
Dozier 1966:84, 149, 194; Hart 1969:80, 88; Mendoza 1903:150; Prill-Brett 1989:1;
Reid 1988:32-33; Scott 1984:196; Voss 1987:128).
Food delicacies were served to high-ranking guests on the sponsoring chief's
collection of valuable imported mainland Asian porcelains and decorated earth
enware (Pigafetta 1975:59; also see Biernatzki 1985:33; Cole 1913:88, 92). This was
one of a number of occasions when chiefly household heirlooms were publicly
displayed, and the lavishness of meal "presentation" on finely made plates and in
delicate bowls (preferably imported porcelains) appears to have been a significant
factor in determining the amount of social prestige accruing to a feast's sponsor
(e.g., Alcina l96ob:133-136; Pigafetta 1975:59).
The importance of gaining prestige or "merit" through lavish feast sponsor
ship is echoed in ethnographic and historic accounts of other complex societies
in the southern Philippines, such as the Manuvu, Tagbanua, and Magindanao
(see Claver 1985:74; Cole l913:1n-120; Manuel 1973:196-197, 229-230, 266--267; C .
Warren 1977:255-257). Prestige o r "merit" translated into the expansion o f politi
cal alliance networks that were at the core of political structure in Philippine so
cieties, as examined in more detail below. The significance of ritual feasting in
legitimating the political order has been a common emphasis in analyses of polit
ical economy in other complex societies (see Kirch, Chapter 6; Schmandt
Besserat, Chapter 14).
F O O D A P P O RT I O N M E N T A N D S O C I A L RA N K I N G
Although the amount and quality of food conferred prestige on the feast's spon
sor, apportionment of the food reflected the social position of the feast's partici
pants and their relationship to the sponsor. As noted by Voss (1987:128), "it is the
system and the protocol of meat circulation which maintains and reproduces the
structure of the relationship between community members." Foodstuffs were di
vided for consumption among participants in accordance with their kinship ties
with the feast sponsor, their relative rank in the local and regional sociopolitical
hierarchy; their history of exchange relations with the sponsoring social unit, and
their perceived role in a chief's immediate alliance-building strategies (Biernatzki
1985:36; Pigafetta 1975:65-66; Plasencia l903a:174; San Antonio 1990:313). As de
scribed by Voss for the Sagada Igorot, an ethnographically known group of the
northern Luzon interior:
277
Laura Leejunker
Meat is distributed according to strict protocol. The most senior old men, who con
duct the ceremony, get the head and internal organs; other elderly relatives and dap
ay mates (men's council or group of political authorities) get the belly and side fat;
close relatives are given parts of the hams; the ribs go to the middle distant relatives;
while the neck and back go to distant relatives. One close relative of a man who was
giving such a ritual told us that "it is better to get a small piece of the appropriate
status meat than a kilo from somewhere else.'' One could say that the symbolic rep
resentation of the social order is embodied in the pig and in these rituals. (1987:129)
The portions allocated to specific individuals and kin groups were carefully scru
tinized and any perceived improprieties in meat distribution were openly criti
cized by the feast attendees. As recounted by Voss:
One man I interviewed talked sarcastically about "those capitalistic old men who
capitalize on their age and prestige by showing up at all kinds of feast to demand
more than their share of meat." In such instances, bitter arguments can result be
tween these old men-who feel they are getting their proper share-and those who
feel they are hogging too much of the pig. (198p30)
Spanish sources suggest that the chief's immediate relatives and other elites with
whom he was strongly allied were commonly served the choicest meat dishes
those consisting of carabao or domestic pig rather than smaller game animals
and, preferentially, the meatiest animal parts (e.g. , the forelimbs and, secondarily,
the vertebrae and ribs; Pigafetta 1975:65-66; San Antonio 1990:313; Santa Ines
1648:78; also see Biernatzki 1985:36-37; Barton 1949:74-75; Prill-Brett 1989). Pig and
carabao skulls were also preferred body parts in some societies (Plasencia
1903b:191), probably for reasons other than meat quality. Gibson's (1986:158)
analysis of the symbolic aspects of animal partitioning among the Buid, a con
temporary upland tribal group of Mindoro, suggests why the pig head might
have been preferred over even meatier animal parts. The head of the pig is
viewed by the Buid as the locus of the animal soul or vital spirit. Thus, by pos
sessing this portion of the animal, the consumers are able to transfer the "vital
ity" or spiritual power of the aillmal to themselves to ward off the weakening
attacks of predatory spirits (Gibson 1986:157).
Food apportionment as the primary public measure of social status differenti
ation has been described in significant detail for complex societies engaging in rit
ual feasting on the Indonesian islands of Nias (Beatty 1991) and Sulawesi
(Volkman 1985). On Nias, slaughtered pigs were traditionally divided into care
fully weighed meat portions, which were distributed according to recognized
rank in the social hierarchy, the history of prior exchange relations between the
host and guest, and kinship and co-residential ties between the host and guest
(Beatty 1991:224-225). Failure to receive the expected meat portion, in recognition
278
of rank and prior prestige payments, was considered a public affront that was fre
quently resolved through violent confrontation (Beatty 1991:225). In the Toraja
highlands of Sulawesi, public meat divisions at feasts required specialists who
could memorize the complicated "cutting histories" of hundreds of individu
als-using criteria of age, genealogy, wealth, achievement, and temperament, in
conjunction with the individual's past history and anticipated future of sponsor
ing and participating in feasts-to determine appropriate meat quantities and
cuts (Volkman 1985:96-rn3). Perceived slights in meat apportionment could
quickly escalate into a "meat fight" (Volkman 1985:rno-rn1) in which wounded
parties made inflamm atory speeches around piles of carabao meat, frequently
flinging meat and carabao excrement at the offending parties. Similarly, presenta
tion of food delicacies at Polynesian feasts followed strict rules of allotment by
social rank (Goldman 1970:5m---s o2, 504, 508; Thomas 1990:97). In the Philippines,
ritualized food distributions were occasions when social tensions over power dif
ferentials were overtly played out by individuals and larger factions. Those who
played the feasting game well, particularly those who could engineer convincing
public confrontations about their position in social and political hierarchies, were
rewarded with upward mobility.
C O M P ET I T I V E F E A STS O R " C H A L L E N G E F E ASTS"

A significant issue i s whether the Philippine feasts reported in contact-period
Spanish records and early ethnographic accounts represented what Beatty
(1991:232-233) refers to as "challenge feasts" in Nias society. A "competitive" or
"challenge" feast can be defined as one in which a primary goal was to achieve
political domination through an ever-escalating cycle of feasting "one-upmanship"
and public displays of generosity /hostility towards rival chiefs. The classic exam
ple of overtly competitive feasting is the so-called rivalry potlatch of Native
Americans such as the Kwakiutl of the Northwest Coast. The primary objective
in sponsoring a feast was to attempt to surpass the abilities of social peers and po
litical rivals to amass, display, distribute and, in some cases, destroy property. The
sponsor's performance resulted in immediate social antagonisms, social valida
tion or humiliation, and political realignment for the participants (see Codere
1950; Drucker 1967).
However, recent anthropological analyses have questioned the emphasis on
the socially combative nature of Northwest Coast feasting events and the over
riding image of "warring with property" (e.g., Kan 1986; Miller 1984; Walens
1981). Instead, status differences are viewed as emerging subtly and gradually
over the long-term process of prestige-good exchange relationships. This inter
pretive debate has recently entered analyses of the significance of feasting events
for chiefly status rivalry in sociopolitically complex Southeast Asian societies. Nu-
279
Laura Leejunker
merous authors cite the massive pig slaughters and ostentatious gifts of gold or
namentation as evidence of the strongly competitive ethos of Nias feasts
(Marschall 1976; Suzuki 1959; Schnitger 1964). However, other ethnographers
claim that overtly antagonistic "challenge" feasts were historically rare in Nias so
ciety, that most feasts had a highly reciprocal ethos in which enhanced status was
transitory, and that social "merit" and political legitimacy were gained only
slowly over the course of an individual's lifetime of ceremonial exchanges
(Beatty 1991; Slamet-Velsink 1995:130).
Some of this contention about the "competitive" nature of feasting events in
Southeast Asia and elsewhere derives from a lack of clarity in distinguishing feast
ing phenomena in which the social "merit" transacted is of a transitory and re
ciprocal nature, and feasting phenomena that attempt to accumulate permanent
wealth and to transform temporary status differentials into long-term and even
inheritable political power. Feasts that confer values of social "merit" but which
are not overtly "competitive," are characterized by cycles of balanced reciprocity,
in which surplus accumulation and status enhancement for any single individual
or kin group are transitory and eventually negated through the necessity of re
turning prestations to their partners in a feasting cycle. In "competitive" feasts,
there is an escalation of labor mobilization and surplus needed to finance future
feasts and the aim of translating feasting success into long-term political power
and economic profit (e.g. , Friedman 1979).
Hayden (1994:25, 64) has suggested that competitive feasts aimed at creating
economic inequalities and negotiating political power differentials are primarily a
structural feature of what he refers to as "transegalitarian" or "Big Man" societies
such as those in New Guinea and other areas of Melanesia characterized by in
stable (i.e . , wholly achievement-based) criteria of social ranking and political suc
cession. Hayden and others (e.g. , Clark and Blake 1994; Friedman and Rowlands
1978; Rosman and Rubel 1978) argue that competitive feasting is one of a number
of evolving strategies among "big-men" or "aggrandizers" for transforming tran
sitory, achievement-based political authority and differential wealth into the in
heritable political power and wealth, which are key to the emergence of
chiefdom-level societies. Hayden suggests that the need for competitive feasts is
largely obviated in chiefdom-level societies, since the competition for political
succession and economic surplus has been resolved in the emergence of inherit
able chieftainship and enforceable tribute systems maintaining permanent labor
pools and stored wealth (1994:64). In other words, political authority and the right
to economic surplus is no longer a matter to be negotiated in contexts of social
interaction such as competitive feasting, but rather a largely inalienable, ideolog
ically reinforced "fact" of ascribed chieftainship. According to this view, "feasts"
in chiefdom-level societies should primarily consist of ritual events exclusively
280
controlled by chiefly leaders and they should b e characterized by (1) a more uni
form scale and regular periodicity, (2) a strong schism between elite and com
moner modes of participation, and (3) a primary emphasis on ideological
reinforcement of existing political power hierarchies, ritual potency of the chief,
and a sense of community solidarity and economic well-being within the polity.
Hayden concludes:
Some forms of competitive feasts probably continued to be used among the lower
level elites in chiefdoms as criteria for promotion to positions of power and to in
crease the value of elite children, but they do not appear to be broadly based feasts
with obligatory interest payments geared to attract supporters such as found in
transegalitarian communities. (1994:64)
However, there is ethnohistoric evidence to support the idea that feasting in

many Philippine societies was overtly competitive, with chiefs attempting to up
the ante of displayed wealth and lavish food giveaways with each successive
round of ceremonial sponsorship (Anon. 199ob:no-n1; Boxer manuscript
1975:215; Chirino 1904:262-271; Colin 1906:75; also see Keesing 1962:165, 190; Scott
1982:192, 196). This competitive element is exemplified by the ethnographic ac
counts of feasting in the Bukidnon chiefdoms of central Mindanao. Public recita
tions extolling the hereditary superiority and grand achievements of the feast's
chiefly host were accompanied by attempts to sicken the guests by providing
prodigious amounts of food (particularly meat) that had to be consumed on the
spot (Biernatzki 1985; Claver 1985; Cole 1956). As Claver (198574, 86) notes, Bukid
non chiefs were continually striving to become the region's most socially es
teemed and politically influential datu by "providing food for the most people
and for the longest time" (1985:74). One Bukidnon datu is claimed to have
achieved the almost insurmountable feat of slaughtering more than 70 buffalo
and 180 chickens for a multiweek feast. That rival chiefs were overtly challenged
by chiefly sponsorship of a particularly elaborate feast, and that this social rivalry
could evolve quickly into social hostilities among elite participants, is evidenced
by a number of Spanish accounts of "treacherous feasts," in which intense emo
tions of rivalry between the feast's sponsors and invited guests erupted into
armed battle (e.g., Anon. 199oa79; Correa 1990:258-259; Pigafetta 1975=147).
At first glance, the Philippine case would seem to contradict Hayden's general
model of ritual feasting in chiefdoms. However, the political structures of Philip
pine polities and other complex societies in island Southeast Asia differ markedly
from the chiefdoms usually included in anthropological study of ceremonial
feasting. This divergent political structure appears to be connected to unique as
pects of social organization, ecology, population dynamics, and labor use in these
small-scale Southeast Asian island societies. We will now turn to a discussion of
281
Laura Leejunker
why ritual feasting is important not only in reinforcing, but also creating and re
forming political and social hierarchies in Philippine chiefdoms.
T H E P O L I T I C S O F F E A S T I N G I N S E G M E N TA R Y
O R "A L L I A N C E - S T R U C T U R E D " C H I E F D O M S
Southeast Asian complex societies have been variously described as "segmentary
polities," "galactic polities," or "theater states" (Geertz 1980; Kiefer 1972; Tambiah
1976; Winzeler 1981). These societies were typically decentralized, had weakly in
tegrated political hierarchies, and were particularly vulnerable to short-term cy
cles of alternating coalescence and disintegration. Political authority traditionally
relied, not on control over fixed territorial bases or well-defined unilineal descent
groups, but on cultivating ties of personal loyalty and expanding one's power base
through frequent and elaborate gift-giving and ceremonialism (Wolters 1982:6-15).
Recent work on comparative political structures by Blanton (Blanton et al.
1996) and Feinman (1995) has suggested that these ubiquitous features of South
east Asia polities are not extraordinary in state-level and chiefly societies, but in
stead represent an extreme emphasis on "network" strategies of political power
relations and political economy, a strategy for maintaining rulership that is prac
ticed to a lesser or greater degree in many complex societies. Blanton and Fein
man have contrasted "corporate-based" and "network-based" or "exclusionary"
power strategies as differing, but not mutually exclusive, ways of achieving polit
ical dominance in complex societies. They suggest that these strategies are part
of the political dynamics in all complex societies, but political actors in a particu
lar society and historical context may emphasize one mode of control more than
the other.
In the exclusionary or network power strategy, political actors try to create
personal networks of political dominance through the strategic distribution of
portable wealth items and symbolic capital (such as ritual potency and religious
knowledge). Alliances are typically created and maintained outside local groups
through "prestational events" (Blanton et al. 1996:4), involving exchanges of mar
riage partners, prestige goods, food, esoteric knowledge, and labor. Often these
prestational events occur in the context of ritual feasting. Individuals who can
most successfully translate the exchanges of foodstuffs, prestige goods, labor, and
other resources into patronage over a large network of allies and followers, gain
political preeminence in the community: However, as noted by Blanton and Fein
man, in societies where a network strategy is the predominant mode in the polit
ical economy, leadership is generally highly conflictive, unstable, and prone to
relatively short cycles of expansion and collapse.
Several factors may have been significant in creating the highly fragmented po
litical landscape, the relatively weak integration between vertically allied leaders,
282
R I T U A L F E A STI N G S Y S T E M S I N P R E H I S PA N I C P H I L I P P I N E C H I E F D O M S
and the highly personalized nature o f political power relations characteristic of

Southeast Asian polities. A number of historians and archaeologists have pointed
to the fragmented geography and diverse environments of Southeast Asia as pro
moting diffuse centers of political power and frustrating attempts at large-scale
political integration (Andaya 1992:405; Geertz 1973=331-338; Reid 1992:460-463;
Winzeler 1981:462) . In addition, most of precolonial Southeast Asia (with the pos
sible exception of Bali, Java, and parts of Vietnam) had exceedingly low popula
tion densities relative to land and resources (Reid l988:n-18; 1992:460-463) .
2
Anthony Reid estimates an overall average of 5.5 persons / krn for the region in
A.O. 1600, less than a fifth of that of India and China and roughly half that of Eu
rope (Reid 1988:n-18, 1992:460-463) . The Philippines and Borneo were the least
densely populated parts of Southeast Asia at the time of European contact (see
Junker 1999 for a more detailed discussion of population dynamics) .
Relatively low population levels, combined with an economic emphasis on
swidden cropping rather than intensive permanent agriculture, an abundance of
unoccupied fertile land, and a seemingly inexhaustible supply of wood and bam
boo for easily rebuilding settlements, meant that many island Southeast Asian
populations were inherently mobile and not particularly concerned with control
of land as a political and economic commodity (Hall 1992:187; Reid 198p57;
Winzeler 1981:462) . An obvious conclusion is that shortages of labor relative to
land engendered a political system in which a ruler's power base was measured in
terms of the size of the labor force bound to him through extensive alliance net
works, rather than fixed geographic territories. Thus, competition between polit
ical leaders (and between rulers and would-be rulers) focused on commanding
labor rather than commandeering land. This may explain the enormously strong
emphasis in island Southeast Asian complex societies on alliance-building activi
ties such as gifting of prestige goods, the creation of extensive marriage ties, ritu
alized feasting, and religious pageantry aimed at social cohesion.
The strongly competitive nature of these critical alliance-building activities is
undoubtedly related to the frailty of hereditary rights to political leadership. In
most ethnographically and historically known complex societies of Southeast
Asia, kinship is generally reckoned bilaterally, corporate descent groups are lack
ing, postmarital residence is bilocal or neolocal, and rank and wealth are inher
ited along both the maternal and paternal lines (Hall l985:no-rn; Reid 1988:147;
Winzeler 1976:628 ) . In the Philippines, as in many Southeast Asian complex soci
eties (e.g., Andaya 1992:409, 419; Reid 1988:152) , polygamy was the cultural norm
amongst rulers and other nobility (Saleeby 1905; Scott 1994) , producing multiple
heirs and further exacerbating conflicts over inheritance of the chieftainship.
Adding to the chaos of chiefly succession was a pronounced ideology of warrior
prestige that allowed men who were successful raiders and slave-takers to accu-
283
Laura LeeJunker
mulate the material base for chieftainship (Junker 1999). In many island South
east Asian societies, ambitious warriors were frequently able to garner significant
wealth and establish independent power bases through theiT close association
with elite patrons (Hall 1992:260) and ultimately threaten the latter's hegemony
(Reid 1988 :167).
In Philippine chiefdoms cognatic descent rules, and the widespread practice of
polygamy and demographic factors created an emphasis on coalition-building
rather than territorially based political power. Lasting political control was prob
lematic and the achievement-based dimension of political authority continued to
be important. A chief's social status and sphere of political authority were not
fixed through inheritance, but were instead continually renegotiated through
material exchanges that took place within the arena of ritual feasting. The cre
ation of social debt among a wide range of participants in a particularly lavish
feast enlarged the sponsor's alliance network, which could be called upon for
raiding, trading, and other wealth-generating activities. Although most of the
historic and ethnographic descriptions of ritual feasts do not provide enough de
tail to determine precisely which segment of society was most involved in the
competitive aspects of feasting, it is likely that competitive one-upmanship was
most fierce among lower-ranking chiefs and ambitious warriors rather than the
chiefly paramounts.
A R C H A E O LO G I C A L E V I D E N C E F O R C O M P E T I T I V E F E A S T I N G
I N T H E P R E H I S PA N I C P H I L I P P I N E S
The preceding discussion on the politics of feasting in traditional Philippine
chiefdoms has largely focused on structural dynamics and not the long-term evo
lution of these feasting systems, which often requires the use of archaeological
evidence. The presence of some type of ceremonial feasting system has been in
ferred for a number of complex societies on the basis of archaeological evidence
alone, in the absence of any ethnographic or historic sources. Evidence for differ
ential access to certain ritual foods or high-quality subsistence commodities, and
particularly their concentration in elite habitation contexts or in association with
ceremonial structures, is one approach in identifying ritual feasting activities in
the archaeological record (e.g. , Crabtree 1990; ]. Fox 1996; Kim 1994; Marcus and
Flannery l996:n5-n6; Pohl 1994; Welch and Scarry 1995). In addition, archaeolo
gists have inferred ritual feasting from the nature of ceramic assemblages found
at settlements. Some archaeologists have noted larger-than-normal cooking and
serving vessels at prominent regional centers, suggesting large-scale food or
drink preparation (e.g. , Blitz 1993; Dietler 1989, 1990). Archaeologists have also
suggested that ceremonial feasting can be recognized by the presence of ceramic
assemblages that consist of elaborate and aesthetically superior serving bowls,
284
cups, and plates, assemblages that vary significantly from mundane household
food preparation and serving assemblages (e.g. , Clark and Blake 1994; Welch and
Scarry 1995). Representations of what appear to be ceremonial events painted on
pottery, molded on metal artifacts, painted in murals, or carved into stone stelae
have also been interpreted by archaeologists as depictions of politically charged
ritual feasting (e.g. , Gero 1992; Higham 1996:133, 151-158). Despite the richness of
ethnographic and historic references to ritual feasting as part of traditional polit
ical and social dynamics in premodern Southeast Asia and China, there has been
limited archaeological research on this phenomenon (e.g., De Veyra 1986;
Higham 1996; Junker, Mudar, and Schwaller 1994; Kim 1994; Mudar 1997).
In terms of material evidence for ritual feasting in the Philippines, access to
feasting foods and feasting paraphernalia was highly correlated with a house
hold's social rank, with the most elaborate feasts sponsored by and participated
in by high-ranking chiefs. Furthermore, these sacrificial ritual feasts are reported
to have taken place within or directly adjacent to the sponsoring chief's residence
(Bobadilla 1990:334; Morga 1903:303-304; Pigafetta 1975:65; Plasencia 1903b:186;
Santa Ines 1990:78; also see Barton 1949:73-75; Cole 1913:65-67, 111-120; Manuel
1973:246-255) . This is true of even mortuary rites, since burials were traditionally
placed under contemporaneously occupied residences in many contact-period
Philippine complex societies (Junker 1993b). In other regions of island Southeast
Asia, early historic accounts suggest that ceremonial feasts took place at special
locales with "megalithic" monuments (including menhirs, dolmens, and circular
arrangements of stones; Slamet-Velsink 1995:113). Although the archaeological
record of megaliths on Borneo, the Malay Peninsula, and many island locales in
between is extensive, their association with ritual feasting in precolonial complex
societies has not been examined and, in fact, their cultural associations and
chronology are poorly known.
At ritual feasts in the Philippines, guests were served on the sponsoring elite's
household collection of finely made earthenware and imported porcelain plates
and bowls (Alcina 196ob:129-130; Pigafetta 1975:65), a ritually and socially signifi
cant ceramic assemblage that was distinct from "everyday" domestic wares.
Ethnographic and ethnoarchaeological studies of domestic activities and trash
disposal patterns in traditional lowland Philippine households (De la Torre and
Mudar 1982; Hart 1958; Nurge 1965:20-21) indicate that faunal debris, other sub
sistence remains, and broken serving dishes were routinely swept to the edges of
the house yard or under the pile houses to form concentrated middens. Over
time, archaeologically recognizable midden deposits and trash concentration
areas were formed that are spatially contiguous with associated domestic struc
tures. Thus, elite households that regularly sponsored large-scale feasting events
would be expected to contain significantly larger numbers of porcelain and high-
285
Laura Leejunker
quality earthenware serving dishes, as well as remnants of feasting foods, in con

trast to households that rarely sponsored feasts or sponsored these events on a
smaller scale.
At the outset, we should note that very little archaeological research in the
Philippines has been devoted to the specific issue of ritual feasting, and to recon
structing chiefly political economies in general. Twelfth- to sixteenth-century
Tanjay and fourteenth- to sixteenth-century Cebu are the only prehispanic
chiefly centers where archaeological research has been carried out at the house
hold level and where evidence relevant to evolution of ritual feasting has been
recorded. Therefore, much of what we have to say here about the material evi
dence is sketchy at best and in need of corroboration by archaeological investiga
tions at a wider range of prehispanic sites.
A N I M A L S A C R I F I C E A N D M E AT D I S T R I B U T I O N
I N P H I L I P P I N E F EASTI N G
Excavations of pile houses at the coastal chiefly center of Tanjay have yielded ar
chaeological evidence for distinct elite and non-elite residential zones, marked by
differences in house size and densities of "prestige goods" (foreign porcelains,
fancy earthenware, and metal goods) . Although these household wealth differen
tials are found in occupation phases dated back to the twelfth to fourteenth cen
turies A.D. , they became even more pronounced and complexly graded at the site
in the fifteenth and sixteenth centuries, as foreign porcelains and other exotics
flowed into the settlement in increasingly higher volumes and local production
of prestige goods expanded (see Junker 1993a and Junker, Mudar, and Schwaller
1994). Excavations at Tanjay have yielded a large sample of animal bone from
middens and trash pits within the two distinct residential sectors, which have
been analyzed for status-related differences in daily meat consumption and differ
ential access to ritual foods (Junker, Mudar, and Schwaller 1994; Mudar 1997). Ar
chaeological faunal assemblages have also been analyzed from several excavation
locales at Cebu, spanning the fourteenth and sixteenth centuries (Mudar 1997;
Nishimura 1992), although the complexities of the site make it much more diffi
cult to separate out factors of socioeconomic status, ecological change, and
Spanish impact on faunal use than at Tanjay (Mudar 1997:96).
At Tanjay, the faunal assemblages recovered from both the early second mil
lennium A.D. Santiago phase and the fifteenth- and sixteenth-century Osmena
phase contained a number of demonstrably domesticated species, including
Bubalus bubalus (carabao, or water buffalo), Gallus gallus (chicken), Sus scrofa (pig),
and Canis familiaris (dog). Wild tropical forest fauna included Philippine spotted
deer (Cervus alfredi), monkey (Macaca fascularius), civet cat (Viverra sp.), fruit bat,
and various reptiles. Marine fauna (fish, turtles, and rays) complete the house-
286
R I T U A L F E A S T I N G S Y S T E M S I N P R E H I S PA N I C P H I L I P P I N E C H I E F D O M S
hold subsistence assemblages (see Junker, Mudar, and Schwaller 1994 and Mudar
1997 for quantitative data). A similar range of species characterized the Cebu fau
nal assemblages (Mudar 1997). Quantitative comparison of the faunal assem
blages from the two residential zones at Tanjay indicates that the inhabitants of
the large, stockaded house-compounds with relatively high densities of "prestige
goods" derived a significantly greater proportion of their subsistence from do
mestic species (see Fig. 10.2). The inhabitants of the small, unstockaded house
compounds included only two species of small- to medium-sized domestic
animals in their diet (pig and chicken) but yielded a wider range of wild species
such as monkey, civet cat, and turtle. The elite households in both cultural phases
had a significantly higher percentage of water buffalo in their diets, as measured
by the Minimum Number of Individuals (MNI), raw bone weights, and percent
age by weight of the total faunal assemblage. Similarly, large mammals, including
water buffalo and pig, were found in significantly higher densities at what
Nishimura (1992) has identified (on the basis of foreign trade goods) as the highest
status residential locale in sixteenth-century Cebu. Historic sources cited earlier
suggest that water buffalo, pig, and other large mammals were preferred feasting
foods that were not consumed on a regular daily basis in the contact-period
Philippines.
There is also some evidence to suggest that inhabitants of the elite residential
zones at Tanjay and Cebu had greater access to meatier and culturally preferred
animal parts. This pattern may reflect ritual-associated redistribution practices
that accorded the best meat cuts to chiefs and their kinsmen at ceremonial feasts.
Ethnohistoric sources detailed earlier suggest that relatively meaty limb bones
were highly prized in meat distributions in the Philippines, as were the skulls of
water buffalo and pig for their symbolic value during associated ritual activities.
1 00
80
w
• ELITE ZON E
Cl
j! 60
z
w
0
• NON-ELITE ZONE
a:
w 40
...
20
Domestic animals Water buffalo and pig High-value body parts
Figure 10.2. Comparisons of the percentages by weight of domestic (vs. wild) animals,
water buffalo and pig (vs. smaller animals), and high-value parts (long bones and crania
vs. rib bones and neck bones) in household middens of elite and non-elite residences
dated to the fifteenth and sixteenth centuries at Tanjay.
287
Laura Leejunker
Whereas both residential areas at fifteenth- and sixteenth-century Tanjay showed

a preference for long bones and skulls compared to "low-value" body parts, the
presumed elite households had a significantly higher percentage by weight of the
prized long-bone meat cuts (see Fig. rn.2). Mudar (1997) found a strikingly similar
percentage of meaty long bones in elite habitation zones at fifteenth- and
sixteenth-century Cebu (64 percent by weight of the medium and large mammal
bones). Surprisingly, the percentage by weight of water buffalo and pig skull frag
ments in the fifteenth- and sixteenth-century habitation levels at Tanjay was
higher for the non-elite residential zone. Thus, whereas the meaty long-bone cuts
were preferentially distributed in elite households at Tanjay, ceremonial display
of animal heads was more frequently taking place in the non-elite sector during
this period.
At Tanjay, we see changes over time in household access to high-quality pro
tein and particularly ceremonial feasting foods such as water buffalo and pig.
Comparison of fauna! assemblages from Santiago phase (ca. noo-1400) and Os
mena phase (ca. 1400-1600) occupation levels suggests an overall increase in the
consumption of water buffalo and domestic pig within the settlement as a
whole in the fifteenth and sixteenth centuries. Mudar (1997) and Nishimura
(1992) note a similar increase in the use of pig and water buffalo in the diet in the
late fifteenth to early sixteenth centuries at Cebu. A possible interpretation is
that domestic pig herds and water buffalo stocks were increasing as stored forms
of wealth in Philippine lowland societies such as Tanjay and Cebu. They were
then being consumed in the context of inflationary competitive feasts sponsored
by chiefs and other "wealthy" individuals, a feasting system that required esca
lating inputs of sacrificial meat. At Tanjay, it is in the "non-elite" Osmena Park
residential zone that we see a particularly striking increase in the consumption
of water buffalo, and it is also in this non-elite sector that we see increasing
numbers of water buffalo and pig skulls, presumably for ceremonial status dis
play. One possible interpretation of this patterning is that participation in the
feasting system became more widespread among non-elite as well as elite spon
sors in fifteenth- and sixteenth-century Tanjay, with escalating competition
amongst individuals outside the hereditary elite for wealth, status, and political
power.
T H E D I S T R I B U T I O N O F R I C E A S A S TAT U S F O O D
Early Spanish sources indicate that root crops such as yams and taro were the sta
ple foods of the bulk of the population, except in Philippine regions such as the
Luzon cordillera, the upper Cotabato River plain, the Tagalog area around
Manila, and the Bicol peninsula where artificial terracing and the construction of
substantial-sized irrigation systems made wet rice production possible on a large
288
R I T U A L F E A ST I N G S Y ST E M S I N P R E H I S PA N I C P H I L I P P I N E C H I E F D O M S
scale (Scott 1994:181). The Spanish writer Francisco Alcina (196oa:88--93, 133-136;
196ob:121) emphasizes that rice was considered a high status food in comparison
to tubers in most areas of the Philippines lacking broad floodplains, and it was an
important prestige food to be served at ceremonial feasts.
Recent paleoethnobotanical work by Mary Gunn (1997) on plant material from
fifteenth- and sixteenth-century habitation contexts at Tanjay is consistent with
historical references to rice as a high-status food and particularly a ritual feasting
food. In a study of plant macrofossils from hearths and middens, she found that
rice was significantly more prevalent in the presumed elite habitation zone in
comparison to the non-elite residential zone. However, the prevalence of rice rel
ative to root crops as dietary components cannot be directly assessed, since root
crops are virtually absent from the both the plant macrofossil and pollen records.
In addition, few identifiable plant remains were recovered from pre-fifteenth
century habitation contexts. Therefore, we cannot presently say whether rice
consumption increased over time at chiefly centers such as Tanjay as a function
of expanding ritual feasting activity. Unfortunately, with the exception of palyno
logical investigations of early rice origins, there have been few archaeological
projects in the Philippines in which plant material has been collected and ana
lyzed.
TH E S E RV I N G ASS E M B LAG E I N P H I L I P P I N E F EASTI N G

Early Spanish accounts indicate that Philippine elites presented high-status
meats and other food delicacies on their impressive array of Chinese, Siamese,
and Annamese trade porcelains (Fig. 10 .3), along with some locally made dec
orated and pedestaled earthenware. Possession of these ritual feasting wares
was essential to a kin group's ability to participate in the feasting cycle. In ad
dition, competitive aspects of the ceremonial feasting system would have en
couraged potential sponsors of feasts to seek new avenues of production and
trade for obtaining an increasingly impressive array of food-serving vessels. In
the cases of Iron Age Europe and Formative period Mesoamerica, archaeolo
gists have suggested that competition for access to foreign trade vessels was
tied to expanding demands for elaborate feasting and drinking paraphernalia
(and possibly exotic foods and drinks; e.g. , Clark and Blake 1994; Dietler 1989,
1990) .
In his detailed analysis of imported porcelains from the Cebu excavations,
Masao Nishimura (1992) showed that the bulk of the imported porcelains com
ing into the Cebu chiefdom's maritime trade port in the fifteenth and sixteenth
centuries were forms that we might label as "plates" and "bowls," that is, food
serving dishes (Fig. 10.4). Nishimura suggests that Philippine elite of this period
are focusing their trade demands somewhat narrowly on those porcelain wares,
289
Laura Leejunker
F��!.
JARS AN D
JARLETS
\ ,
''t::�-�-:.1/
Celadon Bowl
Sung to Ming
Period
Blue-on White Dish

Covered Blue-on-White Jar
Vietnamese
Sawankhalok
Fifteenth century
Mid-fourteenth to early
sixteenth century
Stem cup
Sung to Ming SERVI N G PI ECES
Period
Large Brown Ware Jar

OTH ER Sung to Early Ming
FORMS
Lotus Jar
Sawankhalok (Thai)
Mid-fourteenth to
early sixteenth century
Vase Water dropper

Sawankhalok (Thai) Sawankhalok {Thai)
Mid-fourteenth to Box
early sixteenth century Yuan Period
Figure ro.3. Typical Chinese porcelain "serving" assemblages found at Philippine sites
dated to the eleventh through sixteenth centuries.
which functioned in displaying food and drink at ritual feasts. Consistent with
Nishimura's study of the Cebu porcelain forms, fifteenth- and sixteenth-century
porcelains recovered from the smaller chiefly center of Tanjay show a strong
bias towards "bowl" and "plate" forms (Fig. 10.4), suggesting that household
porcelains may have been used primarily in serving ritual foods at ceremonial
events. Burial assemblages with Late Ming porcelain also tend to concentrate
on a highly standardized range of plates and bowls, which may have contained
ritual foods, and are less likely to yield the more diverse porcelain jars, ewers,
figurines, and boxes of the Tang, Sung, Yuan, and Early Ming period cemeteries
(e.g. , R. Fox 1959, 1964, 1967; Hutterer 1973; Legaspi 1974; Nishimura 1988; Per
alta and Salazar 1974) . Quantitative comparisons of the frequency of porcelain
forms from the largest of these cemeteries, the Calatagan burial sites (R. Fox
1959 ), confirm the qualitative observations of archaeologists working at other
burial sites that cups, jars, covered boxes, and figurines are very rare as im
ported grave accompaniments (Fig. I0-4)- However, porcelain bowls are more
290
1 2th to 1 4th Century Sites
00 - - - - - - - - - - - - - - - - - - - -
• CEBU
settle ment
• TAN.JAY
settlement
D SANTA ANA
burial site
Plates Bowls Cups Jars other
1 6th and 1 6th Century Sites
00 - - - - - - - - - - - - - - - - - - - -
• CEBU
g settle ment
"' 60
'E
"'
u
a 40
• TAN.JAY
c..
settle ment
l!O D CALATAGAN
burial Site
0
Plates Bowls Cups Jars 0th•
Figure ro.4. Percentages of porcelain forms by weight in excavations of Tanjay and Cebu,
and percentage of porcelain forms by quantity (whole vessels) at cemeteries, dated to the
twelfth to fourteenth centuries and fifteenth to sixteenth centuries A.D.
common than the porcelain plates that dominate habitation sites. This pattern
may represent slight differences in the content of ritual feasting and mortuary
assemblages, or may simply reflect the higher breakage rates of plates as they
are used in consumption rituals (see DeBoer and Lathrap 1979 for a discussion
of the archaeological implications of differential breakage rates in household
assemblages).
At Tanjay, where we have a relatively long occupation sequence, there is ar
chaeological evidence that the demand for porcelain as feasting paraphernalia
was expanding over time. The density of foreign porcelains in habitation de
posits increased more than twofold in the fifteenth and sixteenth centuries over
the preceding twelfth- to fourteenth-century levels of occupation (Junk.er 1994).
As shown in Figure 10.4, porcelain imports derived from trash pits and other do-
291
Laura LeeJunker
mestic contexts at Tanjay are less restricted in their range of functional types in
this earlier phase of foreign trade, with a comparatively high proportion of non
serving pieces such as jars, ewers, figurines, and boxes. Unfortunately, Tanjay is
the only habitation site where we have quantitative information on imported
porcelain assemblages prior to the fifteenth century. However, the large
eleventh- to fourteenth-century cemetery of Santa Ana (near modern Manila)
yielded even larger volumes of porcelains that would not fit the category of
serving pieces (including jars of various sizes and forms, figurines, and lidded
boxes; Locsin and Locsin 1967) . The diversity of foreign porcelain assemblages
at both settlement and burial sites supports the observation that local prefer
ences for feasting wares had not strongly affected the import market yet in this
period.
At Tanjay, statistical analyses show that though porcelain was heavily concen
trated in the elite habitation zone with large houses in the twelfth- to fourteenth
century Santiago phase, it was more diffusely scattered in both elite and non-elite
habitation zones by the fifteenth and sixteenth centuries (Fig. 10.5). By the later
phase, foreign porcelains may have been available to a significantly larger portion
of the Tanjay population, including those outside the chiefly class who could at-
A. Foreign porcelain densities C. Decorated e a rthenware de nsities

On kg/m3) On kg/m3)
400 1 00
350
80
300
250 60
200
1 50 40
1 00 20
=L:��=:j��
5
� o ...k:'.��=---'.'._��_,,
1 1 th-1 4th centuries 1 5th-1 6th centuries 1 1 th-1 4th centuries 1 5th-1 6th centuries
D Ellte zone � Non-ellte zone D Ellte zone � Non-ellte zone
B. Southeast Asian/Chinese porcelains

(ratio)
Figure 10.5. Comparison of ceramic as
semblages in midden areas of elite vs. non
0.12
0.1
elite habitation zones at eleventh- to
0.08
fourteenth-century and fifeenth- to six
0.06 teenth-century Tanjay in terms of: (a)
0.04 overall densities (by weight) of foreign
0.02 porcelain, (b) ratio of Southeast Asian (An
o ...lL��__/ namese and Siamese) porcelain to Chinese
1 1 th-1 4th centuries 1 5th-1 6th centuries
porcelain, and (c) densities (by weight) of
D Ellte zone � Non-ellte zone decorated earthenware.
292
tempt t o improve their social rank by acquiring the porcelains necessary fo r criti
cal alliance exchanges and for sponsoring feasting events. The non-elite and elite
porcelain assemblages have almost identical proportions of bowl and plate forms
versus nonserving vessels. However, at both Tanjay and Cebu, areas of the site
with lower-ranking households (as measured by other criteria such as house size
and access to other prestige goods) tended to have a high percentage of low qual
ity foreign porcelains (e.g. , poorly made Sawankholok wares or Siamese and An
namese copies of Chinese forms) and/ or locally made "fancy" earthenware
mimicking Chinese styles (Fig. rn.5) . Thus, non-elites sponsoring this type of
feasting event emulated the choices of elites, when possible, in acquiring a presti
gious porcelain assemblage. Because they did not generally have the resources to
acquire the quantity and quality of porcelains available to elites, they substituted
aesthetically inferior Siamese or Annamese porcelains and locally made deco
rated earthenware in their ritual food displays.
The long-accepted explanation for the beginnings of Chinese mass production
of poorer-quality porcelains for Philippine export and the rise of competing
mainland Southeast Asian export porcelain manufacturers in the fifteenth cen
tury focuses on changing Chinese production modes and trade policies, rather
than transformations in Philippine trade demands. Robert Fox (1964, 1967) sug
gested that the advent of direct Chinese bulk shipping (rather than reliance on in
termediary Malay traders), the loss of revenue to expanding Siamese and
Annamese kilns , and the resulting establishment of large-scale export-focused
kiln sites in southern China, led to the loss of many delicately made jarlet forms
and the proliferation of compact, mass-produced types coming into Philippine
chiefly centers, particularly plate and bowls. An alternative interpretation ties this
shift in foreign trade volume and assemblage content to the expansion of a com
petitive feasting system in Philippine chiefdoms, increasing the demand for
prestige-enhancing foreign serving assemblages. The wider distribution of these
Late Ming mass-produced bowls and plates within the households of populations
of varying social rank at Cebu and Tanjay may indicate an expanded scale and
broadening participation in the competitive feasting system, hence creating a
massive demand for foreign serving vessels of varying quality and status value.
A significant issue in the evolution of ritual feasting systems in the Philippines
is whether the elaborately pedestaled and footed earthenware bowls and plates
found at Philippine sites traditionally dated to the "Metal Age" (ca. 1000 B.C .-A.D.
800; Fig. 10.6) might have functioned in feasting activities prior to the foreign
porcelain trade. There are several problems with the archaeological evidence that
make it difficult to assess whether decorated earthenware pottery was used in
these kind of commensal rituals during the prehistoric period. First, the vast ma
jority of Metal Age sites recorded by archaeologists are described as burial locales
293
Laura LeeJunker
pedestaled
plates
t �
CilJ bowls
'>lr'
jar
Figure ro.6. Characteristic pedestaled and decorated earthenware "serving assemblages"

from Philippine sites of Iron Age and later date (from Solheim 1 9 64).
with no clear habitation features (Solheim 1964; Tenazas 1974; Dizon 1996) in
which jars, lidded pots, funerary boxes, and anthropomorphic forms dominate.
Although there is a general impression amongst archaeologists who have exca
vated presumed Metal Age settlements that decorated plates and bowls may be
more common than in mortuary contexts (Dizon 1987; Junker 1993a; Mascunana
1986; Solheim 1964), none of these sites yielded adequate samples of pottery
sherds to quantitatively assess differences in household versus burial assemblages.
Secondly, there are serious problems with site chronologies prior to the historic
period. Many archaeologists have assumed that earthenware bowls, plates, and
jars with elaborately incised, impressed, stamped, and applique decoration are al
most exclusively "pre-porcelain" or "Metal Age" in date (Fox 1964, 1967; Solheim
1964), a chronological assignment that has been little tested through secure dat
ing methods. The assumption is that maritime trade for the aesthetically superior
foreign porcelains virtually replaced the local fine earthenware industry as pres
tige goods.
294
However, a recent restudy o f supposed "Metal Age" earthenware by Gunn and

Graves (1996) has demonstrated that stylistic analysis and seriation methods used
to assign consistently early dates to decorated earthenware complexes such as
the Kalanay wares may have been methodologically flawed. Excavations at Tan
jay, Cebu, and other locales Uunker 1993c; Nishimura 1992; Spoehr 1973) suggest
that indigenous production of slipped and decorated earthenware, particularly
forms associated with food serving, persisted and even expanded throughout the
period of foreign porcelain trade. Like the rising popularity of less costly foreign
imports, the steady and even increased local production of decorated earthen
ware vessels in the fifteenth and sixteenth centuries may be related to wider par
ticipation in, and sponsorship of, status feasts by lower-ranking individuals who
had limited access to the more valued foreign porcelains as serving pieces. Be
cause of the chronological uncertainties and problems of context with most of
the Philippine decorated earthenware, it is not currently possible to suggest, on
the basis of pottery assemblages, that ritualized feasting was a component of
chiefly political economies prior to the late first-millennium advent of foreign
trade.
T H E EVO L U T I O N O F T H E F EA S T I N G SYS T E M
The archaeological evidence from the Philippines suggests that overtly compet
itive feasting may have become increasingly key to status competition and polit
ical alliance building in the fifteenth- and sixteenth-century chiefdoms. Faunal
assemblages at chiefly centers such as Cebu and Tanjay show a differential
distribution of water buffalo and pig, known historically as important feasting
foods, to what are identified as probable elite households by the early second
millennium A.D. In addition, meaty body parts (particularly limb bones) of large
mammals are found in higher frequencies in these elite habitation contexts
where ceremonial meat divisions likely took place. However, the fifteenth
and sixteenth centuries see both an increase in the use of these high-status feast
ing foods within the chiefly centers as a whole and increased availability of
these prestige-enhancing foods to inhabitants of non-elite residential zones.
Serving assemblages also show an increasing lavishness of material display and
wider distribution of feasting paraphernalia in the two centuries prior to Span
ish contact.
These trends may be interpreted in terms of escalating status competition and
expanding scale of social participation in the feasting system. Unlike many chief
doms where feasting systems have been studied, prehispanic Philippine chief
doms had a political structure consisting of unstable alliance networks that were
maintained through continual material exchanges. In terms of political economy,
Philippine chiefdoms emphasized "wealth finance" (circulation of prestige
295
Laura LeeJunker
goods) rather than "staple finance" (investment in agricultural intensification and

tribute mobilization; D' Altroy and Earle 1985; Earle 1997). In a chiefly society
where political power was traditionally decentralized, where chiefly coalitions
lacked the stable core of territorial lineage groups, and where chiefs vied for con
trol of scarce labor, the circulation of prestige goods at ritual feasts were key to
the expansion of political patronage networks. I argue that, even as paramount
chieftaincies of significant scale emerged, ceremonial feasts became increasingly
important arenas for political and social competition, particularly amongst local
chiefs, lesser elites, and upwardly mobile commoners who were filling the lower
tiers of increasingly complex political hierarchies.
The expansion of ritual feasting systems appears to be linked to the develop
ment of foreign trade and the greater possibilities it offered for alliance-building
wealth circulation. Jonathan Friedman has developed an evolutionary scenario
that ties initial growth of a competitive feasting system to foreign trade wealth,
and eventually led to the evolution of societies with more permanent hereditary
stratification. Using the Kachin of Burma as an ethnographic example, Friedman
(1979) noted that foreign contacts with an adjacent more complex society (the
Shan kingdoms) would provide certain Kachin lineages advantage over initially
equal-ranked lineages in sponsoring ritual feasts. This is due to the inflow of
high-status Shan wives and lowland prestige goods to Kachin lineages strategi
cally located for interaction and trade. Through their greater control of foreign
wealth, Kachin chiefs could hold more elaborately financed feasts, as well as at
tract further lucrative marriage ties, thus creating asymmetrical alliances with
other lineages through their greater control of wealth. In Friedman's model, eco
nomic advantage and social prestige could then be translated into supernaturally
reinforced superiority of the wealthy lineage, since success in sponsoring ritual
feasts is conferred through successfully placating powerful ancestral spirits. Even
tually the enhanced ritual potency of this elite lineage might engender a quasi
historical linkage to these ancestral deities, creating an ideological justification
for a hereditary elite class.
An influx of exotic forms of feasting foods and/ or feasting paraphernalia can
also result in spiraling inputs in the feasting system by already competing elites
and, in some cases, emulation of elite status behavior by non-elites using now
devalued local goods. In an analysis of ceremonial feasting among the Chin, a
Tibeto-Burman people of the Indian-Burmese border, Lehman (1989) suggests
that the availability of external sources of wealth can set off an inflationary cycle
in an already socially competitive feasting system. Once foreign serving vessels
and exchangeable goods enter the feasting system as prestige symbols of excep
tionally high value, other feast-givers must obtain similar exotic goods and/ or in-
296
R I T U A L F E ASTI N G SYST E M S I N P R E H I S PA N I C P H I L I P P I N E C H I E F D O M S
flate their distributions o f now-devalued locally manufactured goods t o maintain

their relative ranking in the feasting system. In addition, the feasting system often
expands the diversity of social groups participating in status display, as the avail
ability of exotic prestige goods creates a two-tiered system of value in which
local luxury goods are more widely available for lower-ranking groups to engage
in elite-emulating ceremonial feasts.
In the Philippines, foreign porcelain trade reaches its peak in the few centuries
before Spanish contact, at a time when many Philippine chiefdoms grew in scale
and manifested more complexly tiered systems of social stratification. The for
eign porcelains coming into the Philippines in increasing volume in the Ming pe
riod have a narrower range of highly standardized forms compared to earlier
periods, emphasizing the plates, bowls, and other serving pieces associated with
ritual food presentations. We may suggest that chiefs who competed favorably
for control of foreign trade through strategic coastal locations, through spon
sored "trade missions" to the Chinese court, and through the development of at
tractive port facilities gained an advantage in the sponsoring elaborate feasts and
cirC:ulating the prestige goods necessary to consolidating political power. Chiefs
with restricted access to foreign feasting assemblages and local "men of renown"
who hoped to parlay military prowess and charisma into political sway probably
augmented their small porcelain assemblages with large volumes of "second
rate" decorated earthenware that allowed them to emulate the more elaborate
feasts of the trade-controlling paramounts. The presence of poorer-quality for
eign porcelains, decorated earthenware, and pig and water buffalo skulls outside
the wealthiest residential zone at fifteenth- and sixteenth-century Tanj ay may re
flect high-intensity status competition amongst lower-tier leaders to build their
own patronage and alliance networks.
The development of competitive feasting systems is likely to be tied to
changes in other aspects of chiefly political economies as Philippine chiefdoms
became more complex in the centuries before European contact. An escalating
scale, widening social participation, and expanding competitive focus in feasting
should be evident in changing subsistence production choices (e.g., a greater em
phasis on raising large pig herds) and in changing strategies for surplus mobiliza
tion (e.g. , agricultural intensification, wider tribute networks). Faunal
assemblages from household middens suggest that residents of chiefly centers
such as Tanj ay and Cebu increasingly concentrated their economic efforts on
raising large pig herds and water buffalo, which were the primary status foods in
ritual feasting. Geomorphological evidence for accelerating alluviation around
the coastal center of Tanjay (e.g., Schwab 1983) and decreasing numbers of trop
ical forest species in faunal assemblages at both Tanjay and Cebu (Mudar 1997)
297
Laura Leejunker
are indirect evidence for expanded land clearance for agriculture in the fifteenth
and sixteenth centuries. We noted that intensified rice production may be partic
ularly associated with expanding surplus in elite households, since rice is an im
portant ritual food distributed in competitive feasting events. However,
archaeological and historical evidence suggest that Philippine chiefs never made
large-scale investments in irrigation systems or other forms of intensification typ
ical of Polynesian chiefdoms and mainland Southeast Asian polities emphasizing
"staple finance" of rulership (Junker 1999) .
In the Marquesas, Thomas (1990:87-108) has suggested that a spiraling system
of competitive feasting is associated with increased interpolity raiding and war
fare, as chiefs vied for exotic resources and especially foreign captives to use in
status-enhancing sacrificial rites. In the Philippines, feasting is similarly linked
with slave-raiding and head-taking, since human sacrifice was a common compo
nent of feasts associated with chiefly mortuary rites, chiefly succession, annual
agricultural ceremonies (such as the Gin Em feast amongst the Bagabo), and per
haps other ceremonial occasions (Chirino 1904:303-305; Cole 1913; Keesing
1962:189; Junker 1993b; Scott 1991:51). Warren (1985) has suggested that the
tremendous escalation of maritime slave-raiding by the Ilanun (specialized sea
raiders supported by the still-independent Sulu sultanate) in the seventeenth
through nineteenth centuries was related to an expanding demand for "foreign"
captives to trade for ritual sacrifices in the Philippines and Borneo. Although es
calating competition for status in feasting may have precipitated increased raiding
to obtain sacrificial victims and other expendable resources, chiefly feasts also
served to mediate social conflict between competing groups and create at least
temporary social solidarity. This is most overtly manifested in "peace pact" or
'blood oath" ceremonies carried out at some ritual feasts, marking the cessation
of hostilities and newly formed alliances between formerly warring social units
or polities (e.g. , Keesing 1962:274; Kiefer 1972; Loarca 1903:160-163; Pigafetta
1975:56, 77, 79; Scott 1982:190).
In archaeologically known cases, such as Iron Age Europe, the Mississippian
chiefdoms, and Formative period Mesoamerica, there often appears to be a
strong correspondence between archaeological indicators for intensive elite
feasting activities and material evidence for escalating interpolity conflict (e.g. ,
Blitz 1993; Dietler 1989, 1990; Marcus and Flannery 1996:120). Both historic
sources and archaeological evidence point to an upsurge in the scale and inten
sity of interpolity conflict among Philippine chiefdoms in the two centuries
just prior to Spanish contact (Junker 1997, 1999) . Although Chinese accounts of
the eleventh through fourteenth centuries describe Philippine coastal ports
that are nonfortified and readily accessible by traders, the early Spaniards em-
298
R I T U A L F E A S T I N G SYST E M S I N P R E H I S PA N I C P H I L I P P I N E C H I E F D O M S
phasize the widespread use of both home fortresses (generally wooden stockade
and-ditch complexes) and interior refuge locales. Historic sources also point to
the rapid adoption of foreign military technologies (particularly Chinese
style iron cannons, swivel guns, and bronze helmets) and to the development
of a professional warrior class (maharlika in Tagalog-speaking chiefdoms)
with distinctive emblems some time after the fourteenth century (also see
Scott 1994) .
Archaeological investigations of coastal trading settlements and cemeteries
support ethnohistoric analysis in suggesting expanding interpolity conflict in the
fifteenth and sixteenth centuries. Archaeological remnants of fortification at
coastal centers and refuge locales are almost invariably late in date (Dizon and
Santiago 1994; Junker 1997). Analysis of skeletal remains from burials shows a
striking increase in the incidence of violent death and the use of decapitated
heads as burial accompaniments in the fifteenth and sixteenth centuries (Junker
1993b, 1997). Male burials in fifteenth- and sixteenth-century deposits at Tanjay,
Calatagan, and other sites (Fox 1959; Junker, Gunn, and Santos 1996) begin to ex
hibit standardized "emblems" of warrior status, including filed and gold-pegged
teeth, animal-tooth jewelry, trophy heads, and bronze and iron weaponry. Finally,
evidence for expanded metals production at coastal chiefly centers such as Cebu
and Tanjay (Junker 1999), and the appearance of foreign artillery at Manila and
Sulu (Scott 1994) suggest that control of metal weaponry becomes increasingly
important at this time. The relationship between military power and ideological
power is always complex in chiefdom-level societies (Earle 1997). In the Philip
pines, competitive aspects of ceremonial feasting may have precipitated raiding
as a strategy for obtaining status-enhancing resources, whereas socially integra
tive aspects of ceremonial feasting served to diffuse social tension and mediate
disputes between factions.
Figure 10.7 is a schematic presentation of inferred changes in political struc
ture, social stratification, agricultural production, foreign trade volumes, systems
of specialist production, the intensity of warfare, and the scale and frequency of
ritual feasting in Philippine chiefdoms in the two millennia prior to European
contact. Unfortunately, our understanding of processes of sociopolitical evolu
tion in the Philippines is based on limited ethnohistoric analysis and archaeologi
cal study of very few of these prehispanic polities. What is clear is that the ritual
feasting system, and particularly the social exchanges involved in its operation,
was increasingly critical to Philippine chiefly political economies, which empha
sized alliance-building through the circulation of prestige goods. Any study of
long-term political processes in the prehispanic Philippines must necessarily
focus on the evolution of ritual feasting systems.
299
Laura Leejunker
Phase P o l itical Social A g r i c u l t u ral F o re i g n S p e c i a l i ze d Wa rfare/ Ritual

S t r u ct u re S t rata I n t e n s i fi cat i o n Trade P r o d u c t i on S l ave r a i d s F e a st i n g
CHIEFS
AD. 1500
3-level
NOBILITY forest
hierarchy ilcreased
LATE
PORCELAIN
Ah COMMONERS
clearance
PERIOD
SLAVES
paramourt I
AD. 1000 chief, local
EARLY
�:r.l:::S
heacmen
chiefS, viUage
PORCELAIN I
burials
PERIOD I I
I
I TANG
I I
AD. 500
I
I frst
2-level porcelails
hierarchy I
simplelstatus
�::�ces in
A
local chiefS,
I
I EARLY
viiage IRON
AD. O heacinen
I
PGE
I TRADE
IRON AGE I
I possi:>le
I
I fancy
I
and iron
I earthenwar
?
? ? prod.Jc:tion
500 8.C. by
specialists
Figure 10.7. Schematic model of the relationship between changing sociopolitical organi
zation, agricultural intensification, increasing foreign trade, specialized production, in
creased warfare, and expanded ritual feasting in Philippine chiefdoms over the two
millennia prior to European contact.
A C K N OW L E D G M E N TS
The archaeological fieldwork in the Philippines that provided much of the empirical sup
port for my analysis of ritual feasting was sponsored over a number of field seasons by the
Fulbright Foundation, the Social Science Research Council, the Wenner-Gren Foundation,
Vanderbilt University Research Council, the Mellon Foundation, and the National Geo
graphic Society. I am grateful to Karl Hutterer, Mary Gunn, Karen Mudar, Marla
Schwaller, Masao Nishimura, Eusebio Dizon, Angel Bautista, Wilfredo Ronquillo, and
countless others for assisting in or facilitating my investigations of the animal and plant re
mains and ceramics at Tanjay. My interpretations of the Southeast Asian historic and ar
chaeological evidence, as well as the more general analysis of the role of ritual feasting in
chiefly political economies, benefited enormously from discussions with Brian Hayden,
Michael Dietler, Polly Weissner, Patrick Kirch, and other contributors to this volume. I
also appreciate the comments of historian colleagues at Brigham Young University (in
cluding Susan Rugh, Lee Butler, Tom Alexander, Jenny Pulsipher, Michael Murphy, David
Wright, Ignacio Garcia, David Montgomery, and many others) when I presented a version
of this paper at the faculty seminar.
300
REFERENCES
Alcina, F. I.
r96oa Historia de las Islas e Indios de las Bisayas, Book 3, Part 3. In The Munoz
[r688] Text of Alcina's History of the Bisayan Islands, edited by Paul Lietz. Chicago:
University of Chicago, Philippine Studies Program.
r96ob Historia de las Islas e Indios de las Bisayas, Book 3, Part 4. In The Munoz
[r688] Text of Alcina's History of the Bisayan Islands, edited by Paul Lietz. Chicago:
University of Chicago, Philippine Studies Program.
Andaya, B.
1992 Political Development between the Sixteenth and Eighteenth Centuries. In
The Cambridge History of Southeast Asia, Volume 1: From Early Times to c. 1800, ed
ited by N. Tarling, pp. 402-459. Cambridge: Cambridge University Press.
Anonymous
r99oa Relation of the Voyage to Manila. In Documentary Sources of Philippine History
[1570] Vol. .z, edited by G. Zaide, pp. 59-85. Manila: National Bookstore.
r99ob Relation of the Conquest of Manila and Other Islands. In Documentary
[1572] Sources of Philippine History, Vol. .z, edited by G. Zaide, pp. 93-rr5. Manila: Na-
tional Bookstore.
Barton, R. F.
1949 The Kalingas. Chicago: University of Chicago Press.
Beatty, A.
r99r Ovasa: Feasts of Merit in Nias. Bijdragen: Tot de Taal-, Land- en Volkenkunde
47:216--235.
Biernatzki, W E.
1985 Bukidnon Datuship in the Upper Pulangi River Valley. In Bukidnon Politics and
Religion, edited by A. Guzman and E. Pacheco, pp. 15-49. Institute for Philip
pine Culture, Papers No. rr. Quezon City: Ateneo de Manila Press.
Blanton, R., G. Feinman, S. Kowalewski, and P. Peregrine
1996 A Dual-Processual Theory for the Evolution of Mesoamerican Civilization.
Current Anthropology 37 (r): r-14.
Blitz, ].
American Antiquity 58:8Q-96.
Bobadilla, D. de
1990 Relation of the Philippine Islands. In Documentary Sources of Philippine History
[1640] Vol. 4, edited by G. Zaide, pp. 329-343. Manila: National Bookstore.
Boxer manuscript
1975 The Manners, Customs, and Beliefs of the Philippine Inhabitants of Long
[1590] Ago. In The Philippines at the Spanish Contact, edited and translated by F. L. Jo
cano, pp. 188-235. Manila: MCS Publishing.
Chirino, P.
1904 Relacion de las Islas Filipinas. In The Philippines, 1493-1898, Vol. 1.2, edited
[1604] by E. Blair and]. Robertson. Cleveland: Arthur H. Clark.
301
Laura Leejunker
Clark, ]. E., and M. Blake

1994 The Power of Prestige: Competitive Generosity and the Emergence of Rank
Societies in Lowland Mesoamerica. In Facticmal Competition in the New World,
edited by E. Brumfiel and]. Fox, pp. 17-30. Cambridge: Cambridge University
Press.
Claver, F.
1985 Dinawit Ogil: High Datu of Namnam. In Bukidnon Politics and Religion, ed
ited by A. Guzman and E. Pacheco, pp. 51-114. Institute for Philippine Cul
ture, Papers No. n. Quezon City: Ateneo de Manila Press.
Codere, H.
1950 Fighting with Property. American Ethnological Society, Monograph No. 18,
Washington, D.C.
Col\!, F. C.
1913 Wild Tribes of the Davao District. Publication No. 162, Anthropology Series 13,
No. r. Chicago: Field Museum of Natural History.
1956 The Bukidnon of Mindanao. Chicago: Chicago Natural History Museum Press.
Colin, F.
1906 Labor Evangelica, Ministerios Apostolicos de los Obreros de la Compania de
[1660] Jesus, Fundacion y Progressos de su Provincia en las Islas Filipinas. In The
Philippines, 1493-1898, Vol. 40, edited by E. Blair and ]. Robertson, pp. 37-48.
Cleveland: Arthur H. Clark.
1975 Native Races and Their Customs: Of the Origin of the Nations and Peoples
[1660] Who Inhabit These Islands. In The Philippines at the Spanish Contact, edi�ed
.
and translated by F. LandaJocano, pp. 147-187. Manila: MCS Press.
Correa, G.
1990 Lendas da India. In Documentary Sources of Philippine History, Vol. 1, edited by
[1563] G. Zaide, pp. 251-264. Manila: National Bookstore.
Crabtree, P.
1990 Zooarchaeology and Complex Societies: Some Uses of Fauna! Analysis for
the Study of Trade, Social Status, and Ethnicity. In Advances in Archaeological
Method and Theory, Vol. .2, edited by M. Schiffer, pp. 155-205. Tucson: Univer
D'Altroy, T., and T. Earle
1985 State Finance, Wealth Finance, and Storage in the Inka Political Economy:
Current Anthropology 26:187-206.
Dasmarinas, G. P.
1958 The Manners, Customs, and Beliefs of the Philippine Inhabitants of Long
[1590] Ago. Translated by C. Quirino and M. Garcia. Philippine journal of Science
87:389-445·
De la Torre, A., and K. Mudar
1982 The Becino Site: An Exercise in Ethno ¥chaeology: In Houses Built on Scattered
Poles: Prehistory and Ecology in Negros Oriental, Philippines, edited by K. But
terer and W. Macdonald, pp. n7-146. Cebu City, Philippines: University of
San Carlos Press.
302
R I T U A L F E A S T I N G SYST E M S I N P R E H I S PA N I C P H I L I P P I N E C H I E F D O M S
DeBoer, W , and D . Lathrap

1979 The Making and Breaking of Shipibo-Conibo Cerainics. In Ethnoarchaeology:
Implications of Ethnography for Archaeology, edited by C. Kramer, pp. 102-138.
New York: Columbia University Press.
DeMarais, E., L. J. Castillo, and T. K. Earle
1996 Ideology, Materialization, and Power Strategies. Current Anthropology 37:1s-3i.
De Raedt, J.
1989 Kalinga Sacrifice. University of the Philippines Cordillera Studies Center,
Baguio, Philippines.
De Veyra, E.
1986 Pigs and Rituals on Bohol Island, Philippines. In Southeast Asian Archaeology
1 9 86, edited by I. Glover. Oxford: BAR Publications in Anthropology.
Dietler, M.
1989 Greeks, Etruscans, and Thirsty Barbarians: Early Iron Age Interaction in the
Rhone Basin of France. In Centre and Periphery: Comparative Studies in Archae
ology, edited by T. Champion, pp. 127-141. London: Unwin Hyman.
1990 Driven by Drink: The Role of Drinking in the Political Economy and the Case
of Early Iron Age France. journal of Anthropological Archaeology 9:352-406.
Dizon, E.
1987 An Iron Age in the Philippines? A Critical Examination. Ph.D. dissertation,
University of Pennsylvania.
1996 The Anthropomorphic Pottery from Ayub Cave, Pinol, Maitum, South Cota
bato, Mindanao, Philippines. Bulletin of the Indo-Pacific Prehistory Association
14:186--196.
Dizon, E., and R. Santiago
1994 Preliminary Report on the Archaeological Explorations in Batan, Sabtang, and
Ivuhos Islands, Batanes Province, Northern Philippines. Unpublished report
submitted to the Philippine National Museum, Division of Archaeology.
Dozier, E. P.
1966 Mountain Arbiters: The Changing Life of a Philippine Hill People. Tucson: Univer
Drucker, P.
1967 The Potlatch. In Tribal and Peasant Economies, edited by G. Dalton. New York:
Natural History Press.
Earle, T. K.
1977 A Reappraisal of Redistribution: Complex Hawaiian Chiefdoms. In Exchange
Systems in Prehistory, edited by T. K. Earle and J. Ericson, pp. 213-229. New
York: Academic Press.
1997 How Chiefs Come to Power: The Political Economy in Prehistory. Stanford: Stan
ford University Press.
Feinman, G.
1995 The Emergence of Inequality: Focus on Strategies and Processes. In Founda
tions of Social Inequality, edited by T. D. Price and G. Feinman, pp. 255-279.
New York: Plenum.
303
Laura Leejunker
Fox, ]. G.
1996 Playing with Power: Ballcourts and Political Ritual in Southern Mesoamer
ica. Current Anthropology 37 (3): 483-510.
Fox, R.
1959 The Calatagan Excavations: Two Fifteenth Century Burial Sites in Batangas,
Philippines. Philippine Studies 7:325-390.
1964 Chinese Pottery in the Philippines. In Chinese Participation in Philippine Cul
ture and Economy, edited by S. Liao, pp. 96-n5. Manila: Liao.
1967 The Archaeological Record of Chinese Influences in the Philippines. Philip
pine Studies 15 (1): 41--62.
Friedman, ].
1979 System, Structure, and Contradiction in the Evolution of "Asiatic" Social Forma
tions. Copenhagen: National Museum Press.
Friedman, ]. , and M. Rowlands
1978 Notes towards an Epigenetic Model of the Evolution of "Civilization." In
The Evolution of Social Systems, edited by ]. Friedman and M. Rowlands,
pp. 201-276. Pittsburgh: University of Pittsburgh Press.
Geertz, C .
1973 The Interpretation of Culture. New York: Basic Books.
1980 Negara: The Theatre State in Nineteeth-Century Bali. Princeton: Princeton Uni
versity Press.
Gero, ].
wegian Archaeological Review 25 (1): 15-30.
Gibson, T.
1986 Sacrifice and Sharing in the Philippine Highlands. London School of Economics
Monographs in Social Anthropology No. 37. London: The Athlone Press.
Goldman, I.
1970 Ancient Polynesia Society. Chicago: University of Chicago Press.
Gregory, C. A.
1982 Gifts and Commodities. London: Academic Press.
Gunn, M. M.
1997 The Development of Social Networks: Subsistence Production and Ex
change Between the Sixth and Sixteenth Centuries A.O. in the Tanjay Region,
Negros Oriental, Philippines. Ph.D. dissertation, University of Hawaii.
Hall, K.
1985 Maritime Trade and State Development in Early Southeast Asia. Honolulu: Uni
versity of Hawaii Press.
1992 Economic History of Early Southeast Asia. In The Cambridge History of South
east Asia. Volume 1. From Early Times to c. 1800, edited by N. Tarling, pp. 183-275.
Hart, D. V.
1958 The Cebuano Filipino Dwelling in Caticugan: Its Construction and Cultural Aspects.
304
Southeast Asian Studies Cultural Reports Series, No. 7 . New Haven: Yale
University Press.
1969 Bisayan Filipino and Malayan Humoral Pathologies: Folk Medicine and Ethnohis
tory in Southeast Asia. Ithaca: Cornell University Press.
Hayden, B.
Foundations of Social Inequality, edited by T. D. Price and G. Feinman,
pp. 15-86. New York: Plenum.
Higham, C .
1996 The Bronze Age of Southeast Asia. Cambridge: Cambridge University Press.
Hutterer, K. L.
1973 An Archaeological Picture of Prehispanic Cebuano Community. Cebu City, Philip
pines: University of San Carlos Press.
1977 Prehistoric Trade and the Evolution of Philippine Societies: A Reconsidera
tion. In Economic Exchange and Social Interaction in Southeast Asia: Perspectives
From Prehistory, edited by K. L. Hutterer, pp. 177-196. Michigan Papers on
South and Southeast Asia No. 13. Ann Arbor: University of Michigan Center
for South and Southeast Asian Studies.
Jocano, F. L.
1975 Philippine Prehistory. Quezon City, Philippines: Garcia.
Junker, L. L.
r993a Archaeological Excavations at the Late First Millennium and Early Second
Millennium A.D. Settlement of Tanjay, Negros Oriental: Household Organi
zation, Chiefly Production and Social Ranking. Philippine Quarterly of Culture
and Society 21 (2): 146-225.
r993b Archaeological Excavations at the 12th-16th Century Settlement of Tanjay,
Negros Oriental: The Burial Evidence for Social Status Symboling, Head
taking, and Inter-Polity Raiding. Philippine Quarterly of Culture and Society 21
(r): 39-82.
r993c Craft Goods Specialization and Prestige Goods Exchange in Philippine Chief
doms of the Fifteenth and Sixteenth Centuries. Asian Perspectives 32 (r): r-35.
1994 Trade Competition, Conflict, and Political Transformations in Sixth- to
Sixteenth-Century Philippine Chiefdoms. Asian Perspectives 33 (2): 229-260.
1997 Slave-Raiding and Warfare in Philippine Maritime Trading Chiefdoms. Paper
presented at the Annual Meeting of the Society for American Archaeology,
Nashville, Tennessee.
1998 Integrating History and Archaeology in the Study of Contact Period Philip-
pine Chiefdoms. International]ournal of Historical Archaeology 2 (4): 291-320.
1999 Raiding, Trading, and Feasting: The Political Economy of Philippine Chiefdoms.
Honolulu: University of Hawaii Press.
Junker, L. L., M. Gunn, and M. ]. Santos
1996 The Tanjay Archaeological Project: A Preliminary Report on the 1994 and
1995 Field Seasons. Convergence 2 (2): 3o-68.
305
Laura LeeJunker
Junker, L. L., K. Mudar, and M. Schwaller

1994 Social Stratification, Household Wealth, and Competitive Feasting in 15th-16th
Century Philippine Chiefdoms. Research in Economic Anthropology 15:307-358.
Kan, S.
1986 The Nineteenth Century Tlingit Potlatch: A New Perspective. American Eth
nologist 13:191-212.
Keesing, F.
1962 The Ethnohistory of Northern Luzon. Palo Alto: Stanford University Press.
Kiefer, T.
1972 The Tausug: Violence and Law in a Philippine Moslem Society. New York: Holt,
Rinehart and Winston.
Kim, S.
1994 Burials, Pigs, and Political Prestige in Neolithic China. Current Anthropology 35
(2): II9-l4I.
Kirsch, A. T.
1973 Feasting and Social Oscillation: A Working Paper on Religion and Society in
Upland Southeast Asia. Cornell University, Dept. of Asian Studies, Southeast
Asian Data Paper No. 92, Ithaca, New York.
Leach, E. R.
1965 Political Systems of Highland Burma. Boston: Beacon Press.
Legaspi, A.
1974 Bolinao: A Fourteenth-Fifteenth Century Burial Site. Manila: National Museum
of the Philippines Publications No. 7.
Lehman, F. K.
1989 Internal Inflationary Pressures in the Prestige Economy of the Feast of Merit
Complex: The Chin and Kachin Cases from Upper Burma. In Ritual, Power,
and Economy: Upland-Lowland Contrasts in Mainland Southeast Asia, edited by S.
Russell, pp. 89-IOI. DeKalb, Ill. : Northern Illinois University Center for
Southeast Asian Studies Occasional Paper No. 14.
Loarca, M. de.
1903 Relacion de las Islas Filipinas. In The Philippines, 1493-1898, Vol. 5, edited by
[1582] E. Blair and]. Robertson,pp. 32-187. Cleveland: Arthur H. Clark.
Locsin, L., and C. Locsin
1967 Oriental Ceramics Discovered in the Philippines. Tokyo: Rutland.
Manuel, E. A.
1973 Manuvu Social Organization. Quezon City, Philippines: University of the
Philippines Press.
Marcus, ]. , and K. V. Flannery
1996 Zapotec Civilization. New York: Thames and Hudson.
Marschall, W
1976 Der Berg des Herrn der Erde. Munich: Deutscher Taschenbuch Verlag.
Mascunana, R.
1986 The Bacong Artifacts in the Sillinran Anthropological Museum Collection: A
Morphological Analysis of Displayed Material Culture. In Artifacts from the
306
R I T U A L F E A STI N G S Y ST E M S I N P R E H I S PA N I C P H I L I P P I N E C H I E F D O M S
Visayan Communities: A Study of Extinct and Extant Culture, edited by R . Cadelina

and]. Perez, pp. l-200. Dumaguete City; Philippines: Silliman University Press.
Mendoza, ]. G.
1903 History of the Great Kingdom of China. In The Philippines, 1493-1898, Vol. 6,
[1586] edited and translated by E. Blair and ]. Robertson, pp. 85-153. Cleveland:
Arthur H. Clark.
Miller, ].
1984 Feasting with the Southern Tsimshian. In The Tsimshian: Images of the Past,
Views for the Present, edited by M. Seguin, pp. 37-39. Vancouver: University of
British Columbia.
Morga, A.
1903 Sucesos de las Islas Filipinas. In The Philippine Islands, 1493-1803, Vol. 16,
[1609] edited by E. Blair and]. Robertson. Cleveland: Arthur H. Clark.
Mudar, K.
1997 Patterns of Animal Utilization in the Holocene of the Philippines: A Com
parison of Faunal Samples from Four Archaeological Sites. Asian Perspectives
36 (1): 67-105.
Nishimura, M.
1988 Long Distance Trade and the Development of Complex Societies in the Prehis
tory of the Central Philippines-The Cebu Archaeological Project: Basic Con
cepts and First Results. Philippine Quarterly of Culture and Society 16:107-157.
1992 Long Distance Trade and the Development of Complex Societies in the Pre
history of the Central Philippines: The Cebu Central Settlement Case. Ph.D.
dissertation, University of Michigan.
Nurge, E.
1965 Life in a Leyte Village. Seattle: University of Washington Press.
Peebles, C . , and S. Kus
1977 Some Archaeological Correlates of Ranked Societies. American Antiquity
42:421-448.
Peralta, ]. , and L. A. Salazar.
1974 Pre-Spanish Manila: A Reconstruction of the Pre-History of Manila. Manila: Na
tional Historical Commission.
Perez, D.
1951 Relacion de los Indios Zambales de la Playa Honda: Su Sitio, Sus Costumbres.
[1680] In Un Heroe Dominicano Mantanes en Filipinas, edited by Honoria Muiioz.
Manila: Santander.
Pigafetta, A.
1975 First Voyage Around the World. In The Philippines at the Spanish Contact,
[1521] edited and translated by F. L. Jocano, pp. 44--8 0. Quezon City: Garcia Publica
tions.
Plasencia, ]. de
l903a Las Costumbres de los Indios Tagalogs de Filipinas. In The Philippine Islands,
[1589] 1493-1803, Vol. 7, edited by E. Blair and ]. Robertson, pp. 173-185. Cleveland:
Arthur H. Clark.
307
Laura Leejunker
Instrucci6n de las Costumbres que Antiguamente Terrian los Naturales

de las Pampanga en Sus Pleitos. In The Philippine Islands, 1493-1803, Vol. 7,
edited by E. Blair and ]. Robertson, pp. 185-196. Cleveland: Arthur H.
Clark.
Pohl, M.
1994 The Economics and Politics of Maya Meat Eating. In The Economic Anthropol
ogy of the State, edited by E. Brumfiel, pp. rw-148. Monographs in Economic
Anthropology, No. rr. Lanham: University Press of America.
Powers, W K. , and M. Powers
1984 Metaphysical Aspects of an Oglala Food System. In Food in the Social Order;
edited by M. Douglas, pp. 40-96. New York: Russell Sage Foundation.
Prill-Brett, ].
1989 The Bontok Chuno Feast in the Context of Modernization. University of the
Philippines, Cordillera Studies Center Working Paper No. 12, Baguio, Philip
pines.
Reid, A.
1983 "Closed" and "Open" Slave Systems in Pre-Colonial Southeast Asia. In Slav
ery, Bondage, and Dependency in Southeast Asia, edited by A. Reid, pp. r56--r 8r.
New York: St. Martin's Press.
1988 Southeast Asia in the Age of Commerce 1450-1680, Vol 1: The Lands Below the
Winds. New Haven: Yale University Press.
1992 Economic and Social Change, c. 1400-1800. In The Cambridge History of South
east Asia. Volume 1. From Early Times to c. 1800, edited by N. Tarling. Cam
1993 Introduction: A Time and a Place. In Southeast Asia in the Early Modern Era,
edited by A. Reid, pp. 1-22. Ithaca: Cornell University Press.
Rosman, A., and P. Rubel
1978 Exchange as Structure, or Why Doesn't Everyone Eat His Own Pigs? Re
search in Economic Anthropology 1:rn5-130.
Saleeby, N. M.
1905 Studies in Moro History, Law, and Religion. Bureau of Printing, U.S. Depart
ment of the Interior, Ethnological Survey Publications, Vol. 4, Part 1, Manila.
San Antonio, F.
1990 Cronicas de la Apost6lica Provincia de San Gregorio de Religiosos
[1738] Descalzados de N.S.P.S. Francisco de las Philipinas, China, Jap6n, etc. In Doc
umentary Sources of Philippine History, Vol. 5, edited and translated by G. Zaide,
pp. 299-34r. Manila: National Bookstore.
Sanchez, M.
1617 Bocabulario de la Lengua Bisaya. Manila: National Archives.
Santa Ines, F. de
1990 Cronica de la Provincia de San Gregorio Magno de China, Jap6n, etc. In
[1676] Documentary Sources of Philippine History, Vol. 5, edited and translated by G.
Zaide, pp. 67-92. Manila: National Bookstore.
308
Schnitger, F. M .
1964 Forgotten Kingdoms in Sumatra. Leiden: Brill.
Schwab, A. M.
1983 The Geomorphology and Archaeological Geology of the Bais Anthropolog
ical Project, Negros Oriental, Philippines. M.A. thesis, Department of Geol
ogy; University of Michigan.
Scott, W H.
1982 The Discovery of the Igorots: Spanish Contacts with the Pagans of Northern Luzon.
Quezon City; Philippines: New Day Publications.
1984 Prehispanic Source Material for the Study of Philippine History. Quezon City;
Philippines: New Day Publications.
1991 Slavery in the Spanish Philippines. Manila: De la Salle University Press.
1994 Barangay: Sixteenth Century Philippine Culture and Society. Quezon City: Ate
neo de Manila Press.
Slamet-Velsink, I.
1995 Emerging Hierarchies: Processes of Stratification and Early State Formation in the
Indonesian Archipelago: Prehistory and the Ethnographic Present. Leiden: Konin
kluk Instituut Voor Taal-, Land- en Volkenkunde Press.
Solheim, W G.
1964 The Archaeology of the Central Philippines: A Study Chiefly of the Iron Age and Its
Relationships. Manila: Philippine Bureau of Printing.
Spoehr, A.
1973 Zamboanga and Sulu: An Archaeological Approach to Ethnic Diversity. Pittsburgh:
University of Pittsburgh, Ethnology Monograph No. r.
Strathern, M.
1971 The Rope of Maka. Cambridge: Cambridge University Press.
Suzuki, P.
1959 The Religious System and Culture of Nias, Indonesia. The Hague: Exelsior Press.
Tambiah, S.
1976 World Conqueror and World Renouncer: A Study in Religion and Polity in Thailand
Against an Historical Background. Cambridge: Cambridge University Press.
Terrazas, R.
1974 A Progress Report on the Magsuhot Excavations in Bacong, Negros Oriental,
Summer 1974- Philippine Quarterly of Culture and Society 2:133-155.
Thomas, N.
1990 Marquesan Societies: Inequality and Political Transformation in Eastern Polynesia.
Oxford: Clarendon Press.
Volkman, T.
1985 Feasts of Honor: Ritual and Change in the Toraja Highlands. Urbana, Ill . : Univer
sity of Illinois Press, Illinois Studies in Anthropology; No. 16.
Voss, ].
1987 The Politics of Pork and Rituals of Rice: Redistributive Feasting and Com
modity Circulation in Northern Luzon, the Philippines. In Beyond the New
309
Laura LeeJunker
Economic Anthropology, edited by J. Clammer, pp. l21-14r. New York: St. Mar
tin's Press.
Walens, S.
1981 Feasting with Cannibals: An Essay in Kwakiutl Cosmology. Princeton: Princeton
University Press.
Warren, C .
1977 Palawan. In Insular Southeast Asia: Ethnographic Section 4 : The Philippines, ed
ited by F. Lehar, pp. 229-290. New Haven: HRAF Publications.
Warren, ]. F.
1985 The Sulu Zone, 1768-1898: The Dynamics of External Trade, Slavery, and Ethnicity
in the Transformation of a Southeast Asian Maritime State. 2nd edition. Quezon
City, Philippines: New Day Publications.
Welch, P., and M. Scarry
1995 Status-Related Variation in Foodways in the Moundville Chiefdom. American
Antiquity 60 (3): 397-419.
Winzeler, R.
1976 Ecology, Culture, Social Organization, and State Formation in Southeast
Asia. Current Anthropology 17:623--640.
The Study of the Southeast Asian State. In The Study of the State, edited by H.
Claessen and P. Skalnik, pp. 455-467. The Hague: Mouton.
Wolters, 0. W
1982 History, Culture, and Region in Southeast Asian Perspectives. Singapore: Institute
of Southeast Asian Studies.
31 0
11
FEAST I NG AND THE E M ERGENCE
O F P LAT FOR M M O U ND CERE M ON I A L ISM
IN EASTERN NORTH A M ER I CA
Vernon James Knight
Freestanding platform mounds are a prominent feature of certain prehistoric

sites in the southeastern United States belonging to the period roughly between
rno B.c. and 700 A.O. These Woodland period mounds are the earliest examples of
the platform type found in the cultural chronology of the southeastern region.
In the interest of exploring the variability within this category, my discussion is
focused upon a subset of these early mounds that I believe has relevance to the
topic of this volume.
It has been known for some time that, despite similarities of form, earlier
Woodland period platform mounds of the Southeast exhibit fundamental differ
ences from the later and far better known late prehistoric platform mounds of the
same region. Generally, the late prehistoric Mississippian earthworks, dating to ap-
31 1
VernonJames Knight
proximately A.O. 1000-1550, are typified by the occurrence of summit buildings of

various kinds. These occur in the context of chiefdom-type societies. As to their
ancestry; their immediate prototype, it appears, is found in the slightly earlier plat
form mounds of the Coles Creek culture of the Lower Mississippi Valley (ca. A.O.
700-1200). Excavated examples of Coles Creek platform mounds at the Green
house and Morgan sites in Louisiana possess large, circular buildings on their sum
mits, which are arguably elite residences (Ford 1951; Fuller and Fuller 1987). By
contrast, those Woodland platform mounds that predate A.O. 700 seldom exhibit
summit buildings. The majority belonging to this earlier period of transegalitarian
societies were not lived on, and seem to be associated with altogether different
kinds of summit activity; at least partly of a ritual nature (Knight 1990; Mainfort
and Walling 1992; Jefferies 1994; Lindauer and Blitz 1997; Kohler 1997).
Despite this chronological shift in summit use, the mounds themselves show
important commonalities in their formal characteristics through time. These
commonalities include rounded quadrilateral contours, multistage construction,
and intentional use of contrasting fills. Specifically for the Mississippian case I
have argued that, in attempting to understand mounds-as-artifacts, summit use
ought to be decoupled from mound building per se, and that mound-building ac
tivity is best considered in itself as a repetitive ritual process. Some years ago I de
veloped an interpretive model using ethnographic analogy, drawing from the
languages, mythology, religious beliefs, and ritual practices of historic southeast
ern Indians, wherein Mississippian mounds were conceived by their builders as
earth icons, the "navels of the earth" as the Chickasaw called them, periodically
reburied in a purificatory process of world renewal (Knight 1981, 1986, 1991) . If
this model has merit, the continuity of formal characteristics I have men
tioned suggests that the analogy might be extended with equal plausibility to pre
Mississippian platform mounds. Fundamentally, from that point of view, the
entire history of platform mound building in the Southeast may be seen as a conserva
tive, long-term complex of world-renewal ritual. I will return to this point further on.
The known distribution of pre-Mississippian and pre-Coles Creek freestand
ing platform mounds is shown in Figure 11.1. This map is accompanied by a table
(Table II.I) giving the site names and basic references. Although there is uncer
tainty about the dating of some of these examples, particularly those excavated
early in the twentieth century; other potential claimants are excluded for short
age of evidence. Despite this uncertainty there is no longer any question that the
platform mound building tradition in the Southeast extends back in time to the
Middle Woodland period and that such constructions were very much a part of
southeastern Hopewellian and post-Hopewellian landscapes.
There is one seeming "hot spot" in the distribution: Northwest Florida and the
adj acent Chattahoochee River Valley on the northern Gulf Coastal Plain was an
312
F E A S T I N G A N D P L AT F O R M M O U N D C E R E M O N I A L I S M I N E A S T E R N N O RT H A M E R I C A
early center o f such activity. Otherwise, the remainder are rather widely spread
out. The settings differ as well, expressing something of a duality. Some sites,
such as Marksville, Florence, and Pinson, possess large mounds within essentially
vacant ceremonial centers, sometimes enclosed by earthen embankments. Else
where, in contrast, platform mounds lie at the margins of permanently occupied
settlements. It is intriguing that among the latter are some of the earliest well
defined nucleated villages, as contrasted to seasonal base camps, in the Southeast.
It is the examples tangent to permanent villages rather than those at vacant cen
ters that have special importance to the goals of the present paper.
As Richard Jefferies noted in his recent summary of this phenomenon, "The
characteristics of the Woodland platform mound surfaces suggest not only that
activities differed from those proposed for Mississippian platform mounds, but
that a great deal of variability existed among the [Woodland] mounds them
selves" (1994:83). This variability is indeed worthy of close attention (Knight
1990:172). For example, excavated summit surfaces at the Ingomar and Pinson
sites (Rafferty 1990; Mainfort and Walling 1989) are in character quite unlike
those reported elsewhere. Data are still scarce, however, and the temptation to
overgeneralize is better resisted.
Nonetheless, despite this variability there is at least one identifiable pattern of
mound-summit configuration that, at least to me, stands out from the rest and
therefore commands attention. A good way to introduce this pattern is by quot
ing from William Sears's first encounter with it. Of his excavation of Mound B at
the Kolomoki site in southwest Georgia in 1952, the veteran southeastern archae
ologist writes as follows.
This was the most unusual mound it has ever been the writer's misfortune to en
counter. As a result of a trench cut into it in the first season's work, it was hypothe
sized that this small structure, fifty feet in diameter and not over five feet high,
represented the remains of a collapsed earth covered lodge. It would be difficult to
have been more in error. . . . It rapidly became obvious that Mound B consisted of a
collection of post holes. Very large posts, twenty-four to thirty inches in diameter,
were erected successively in this small area . . . . Later posts often cut through the re
mains of earlier ones, so that in only a few instances were complete outlines pro
duced. (Sears 1956:10-n)
Sears goes on to describe post-insertion ramps leading down into the postholes
from one side, and he provides a sketch showing how these post-insertion ramps
must have been used. It is rather obvious that his bewilderment resulted from ex
pectations derived from a literature on Mississippian mounds that almost invari
ably possessed clear structure patterns; in finding multiple postholes Sears was
accustomed to looking for intelligible patterns that would belong to definite
313
Vernon]ames Knight
Figure 11.1. Truncated mounds in the eastern United States, 100 B.c .-700 A.O.
buildings. Here there were none; just a surface resembling Swiss cheese from the
repetitive emplacement of large posts.
The McKeithen site in northern Florida is closely related to Kolomoki in its
material culture. It was excavated by Jerald Milanich and his colleagues and thor
oughly reported in 1984. Like Mound B at Kolomoki, Mound A at McKeithen
also possesses a platform mound surface prominently marked by randomly situ
ated, large postholes having overlapping post-insertion ramps (Figure 11.2). Asso
ciated with this surface were other pits showing a complex intrusion sequence
indicative of frequent reuse, numerous small fire pits, a central hearth, two de
posits of red ochre, and two refuse deposits containing hundreds of potsherds
and pieces of deer bone (Milanich et al. 1984:94-ro5). McKeithen Mound A is situ
ated at one side of a permanent village that, like the village area at Kolomoki,
takes the form of a ring-shaped midden and additional mounds. At McKeithen
the other mounds consisted of a conical burial mound and a second platform
mound supporting a small building, the latter perhaps the earliest clearly docu
mented domiciliary use of a platform mound in the eastern United States.
314
F E A S T I N G A N D P LAT F O R M M O U N D C E R E M O N I A L I S M I N E A S T E R N N O RT H A M E R I C A
TABLE 1 1 . 1
Truncated Mounds in the Eastern Un ited States, 1 00 B.c.-700 A.D.
No. Truncated
Map No. Site Name Mounds References
Marksville 3* Fowke 1928; Vescelius 1957; Toth 1974

2 Troyville Walker 1936
3 Leist 1* Phillips 1970
4 Ingomar Rafferty 1987, 1990
5 Johnston 2 Kwas and Mainfort 1986
6 Pinson 5 Mainfort 1988; Mainfort and Walling 1983
7 Florence Boudreaux and Johnson 1998
8 Forkland Shogren 1989
9 Graveline Greenwell 1984
10 Mitchell Sears n.d.
11 Durant Bend Nance 1976
12 Walling Knight l990
13 Pierce 2* Moore 1 902
14 Waddells Mill Pond Anonymous 1974b; Brose 1979
15 Kolomoki 5 Sears 195 1 a, 195 lb, 1953, 1 956
16 Mandeville Kellar et al. 1962
17 Shorter Kurjack 1975
18 Annawakee Creek Dickens 1975
19 Ginther Shetrone 1925; McMichael 1964
20 Marietta Works Squier and Davis 1 848; McMichael 1964
21 Garden Creek Keel 1972, 1976
22 Cold Springs Jefferies 1994
23 Swift Creek Kelly and Smith 1975; Jefferies 1 994
24 Block-Sterns Jones, Penton, and Tesar 1 998
25 Hall l* Moore 1902
26 Yent Moore 1902
27 Aucilla River Moore 1902
28 Crystal River 2 Bullen 1953
29 McKeithen 2 Milanich et al. 1984
30 Fortson Williams 1992; Williams and Harris 1998
31 Murphy Island l* Moore 1 896
32 Tick Island l* Moore 1 894
33 Fort Center 1 0* Sears 1982
*One or more mounds circular or elliptical in plan.
In these particulars of site structure and configuration of Mound A, the Mc

Keithen site is strongly reminiscent of the situation at the Walling site in the Ten
nessee Valley of northern Alabama. The platform mound at Walling (Site Ma050)
was excavated by the author and reported in 1990. At Walling there is once again
a Middle Woodland village marked by a ring-shaped midden, with a low platform
31 5
VernonJames Knight
UIUDIB
Hearth in mound
� Hearth on mound
- Charred post
J2';Ll
Red ocher
® Fire pit & charred wood
0 Fire pit
D Removed charred post
c::J Pit in mound
6 � t 1•• N
... . �
Figure II.2. McK.eithen Mound A, plan of features. From Milanich et al. 1 9 8 4 ; reprinted
with the permission of the University Press of Florida.
mound standing tangent to it. On the opposite side of the village site from the
platform mound, lying at a distance, are two burial mounds, one of which has
been excavated and establishes a connection between the habitation area and the
Copena mortuary complex (Webb and Dejarnette 1942).
On the second stage of the Walling platform mound (Fig. 11.3), labeled Fill No.
2, were found three massive postholes with diameters of approximately one
meter, all provided with post-insertion ramps. One of these monumental posts
had, in addition, a separate post-extraction ramp. On the same mound summit
were dozens of additional, randomly scattered postholes of various sizes and fill
characteristics. Posts frequently intruded into other posts and pit features, show
ing intermittent reuse of the surface. Some of the larger postholes had peculiar
funnel-shaped profiles, with wide pits at the top transitioning into a vertical-sided
posthole at the bottom. Such features are plausibly interpreted as post-extraction
pits dug around the base of previously implanted large posts to facilitate their re
moval. Together with these posts were ash-laden surface hearths, small pits, and
concentrations of surface midden yielding potsherds, stone tools, and abundant
animal bone.
Large, funnel-shaped postholes and posts with post-insertion ramps are dupli-
316
F E A S T I N G A N D P L ATF O R M M O U N D C E R E M O N I A L I S M I N E A ST E R N N O RT H A M E R I C A
A - Middens N
B - Surface Hearths L-1
0 5
C - Large Post Features m
Figure 11.3. Walling Platform Mound, plan of features associated with Fill No. 2. After
Knight 1 99 0.
cated at Mound A of the Cold Springs site in northern Georgia, excavated by

Suzanne Fish and reported by Richard Jefferies (1994). Mound A is a low platform
mound at the northern end of a village of Swift Creek affiliation. Fish and Jef
feries estimated that this village possessed approximately eight off-mound
dwellings (Fish and Jefferies 1986) . A smaller, second mound located to the south
west of Mound A had, at least in part, a mortuary function as indicated by a cre
mation containing a fragmentary copper ear spool. In Mound A, the multistage
platform mound at Cold Springs, a portion of the surface of Floor 3 was exposed
(Fig. 11.4), revealing over mo postholes, many quite large and set as much as one
meter deep. The arrangement of posts was once again seemingly random.
Among the large postholes, Jefferies describes and illustrates obvious post-inser
tion ramps, and other large posts appear to have post-extraction pits as found at
Walling (Fig. 11.5). Other features of this mound stage included small pits and sur
face areas mottled with charcoal. Although Jefferies concludes that the Floor 3
surface "remained exposed for a long time and was the scene of much activity;"
he also concluded from the paucity of artifacts that this surface was not used for
ordinary domestic activity (1994:80).
Yet another, similar situation is Mound No. 2 at the Garden Creek site in west
ern North Carolina, reported by Bennie Keel in 1972 (see also Keel 1976). This is
317
VernonJames Knight
.aii. Motdcd Clay

IW with Charcoal
Q Pit Features
e Postmolds
North
®
•• A
o� ... . . �m
••
•
•
•
. �
� t:'\
{I �
� ®
Figure ri.4. Cold Springs, plan of features associated with Floor 3 . From Jefferies 1 994 (in
Ocmulgee Archaeology: 1 936-1 986, edited by David]. Hally [Athens: University of Georgia
Press]).
another low mound situated on the margin of a Woodland village site (desig
nated Hw 8). At least the first two mound stages at Garden Creek No. 2 are at
tributed to the Middle Woodland Connestee phase. Of most interest in this
thoroughly excavated case is the well-preserved summit of the primary mound
(Fig. n.6). This surface is riddled with hundreds of postholes of various sizes,
many intruding into others and indicating a protracted period of use. Despite
Keel's assertion that these postholes must indicate the presence of a structure, he
was unable to demonstrate any definite structure outline. Indeed, to judge from
his map, postholes appear to be haphazardly distributed across the surface. Keel
classifies three features as "large post holes." The largest of these, with a diame
ter of 88 centimeters and a depth of 149 centimeters, has a characteristic funnel
shaped orifice, in common with certain large posts at Walling and Cold Springs,
that in my view is probably a post-extraction pit. Other features of the primary
mound surface were a large, central burned area covered with ash, surface
hearths, ceremonial deposits of stones, small refuse-filled pits, and a deposit of
dark midden soil placed on one side of the mound which yielded over five hun
dred potsherds. Animal bone was not well preserved and no special analysis was
made of the primary mound fauna, although general field observations indicate
deer as the main species encountered in these deposits (Keel 1976:78-85, 149) .
These five mounds-Kolomoki Mound B, McKeithen Mound A, the platform
318
F E ASTI N G A N D P LAT F O R M M O U N D C E R E M O N I A L I S M I N E A ST E R N N O RT H A M E R I C A
�i
FEATIJRI! 108
u�FEATURI! 1 29
"'-N
0 30 .,.
=-=
!':if.I POSTMOU>
\.... PIT OU'ILINI!
l'EATURI! 121
FEATIJRI! 131 Fl!A1lJRB 110
Figure 1r.5. Cold Springs, cross-sections of large postholes. From Jefferies 1 994 (in Ocmul
gee Archaeology: 1 936-1 9 86, edited by David]. Hally [Athens: University of Georgia Press]).
mound at Walling, Cold Springs Mound A, and Garden Creek Mound No. 2-are
sufficient to illustrate the existence of a unitary phenomenon. The primary sig
nature of this phenomenon is the presence of dense, irregular scatters of post
holes of various sizes, sorrie of which are large and are equipped with
post-insertion ramps, extraction ramps, or extraction pits. The abundance of
these postholes, plus the evidence of intrusion sequences, points to repeated
reuse of these surfaces over a substantial period of time. Conspicuously absent
are posthole alignments or configurations that would indicate the presence of
roofed architecture. In general, this evidence is perhaps best denominated as scaf
folding behavior, repeated cycles of erection and subsequent dismantling of scaf
folds, racks, and isolated poles on platform mound summits. Moreover, this
activity seems to manifest itself at at least two different scales or modes. On the
one hand, we have evidence of scaffolds or racks employing vertical posts of or
dinary diameter, but on the other we have postholes that qualify as monumental.
The latter were footings for large, upright log poles, presumably of a height com-
319
\\
Hw 2
PRIMARY MOUND
D IUINEO AltEA
0 • C:HAtc°"'t FfUfD
00
I I
0 0 g 00 0 0
o @J � Gi° �°o
o
0 0
0 0
0 0
0
.
Do 0
0
Figure n.6. Garden Creek Mound No. 2, plan of features associated with the primary
mound. From Cherokee Archaeology by Bennie C. Keel, published 1 976 by the University of
Tennessee Press.
F E A S T I N G A N D P LAT F O R M M O U N D C E R E M O N I A L I S M I N E A S T E R N N O RTH A M E R I C A
mensurate to their girth, requiring special procedures and facilities to put up and
take down. Uprights of this sort were meant to impress.
Associated features on the mound summits include small pits, surface hearths,
and middens. Hearths and dense midden deposits, in particular, are strongly in
dicative of food preparation and consumption. Exotic raw materials, products of
long-distance exchange, and special-use pottery are also found in a number of
these contexts, pointing to some sort of ceremonial use. In this connection, the
Walling platform mound and Garden Creek Mound No. 2 both exhibit specific
Hopewellian-related artifacts in their respective mound assemblages. Table 11.2
summarizes the associations so far discussed (see also Knight 1990:170-172) .
The fact that these mounds all stand within the boundaries of small perma
nent settlements is significant. Access does not seem to be restricted in any im
portant way, and whatever activities occurred on them probably would have been
in full view of villagers residing only a few meters away. (A possible exception is
presented by evidence for a curving, pine post screening wall found on the south
west side of McKeithen Mound A [Milanich et al. 1984:97-98], but this screen
seems ill-placed to impede either view or access. Summit activities would still
have been perfectly visible from the vantage point of the village's central plaza.)
Owen Lindauer and John Blitz (1997) believe that public access is one of the chief
variables that distinguish such Woodland platform mounds from their later, Mis
sissippian counterparts. To use their terminology, whereas Woodland platform
mounds were principally communal, integrative facilities, Mississippian mounds
emphasized social dijfrrentiation.
In summary, minor variations aside, these mounds reveal a patterned combi
nation of behaviors. As I interpret them, these involve intermittent, repetitive ac
tivity involving manipulation of exotic artifacts, caching of goods in small pits,
food preparation and consumption, frequent scaffolding of objects unknown,
and monumental display of poles. All of this occurred on low, earthen stages, pe
riodically renewed and connected symbolically with world renewal.
Up to now I have sidestepped any acknowledgment of previous models of the
uses of these mounds. Jerald Milanich and his colleagues argued that McKeithen
Mound A was a charnel structure, on which the remains of the honored dead
were reduced to bones, later to be collected, bundled, and removed for final in
terment in the nearby burial mound, Mound C . In this interpretation the pits on
the summit of Mound A were for maceration. Milanich conjectures that the ad
jacent large posts held aloft pottery sculptures that functioned as charnel
"guardians" (Milanich et al. 1984:99-IOO) . What makes this argument particularly
compelling is the adduced complementarity of three contemporaneous mounds
in the McKeithen village. Mound C, the burial mound, contained numerous bun
dle burials in various states of decomposition; these remains clearly require a
321
TABLE 1 1 .2
Associations of Certain Woodla nd Platform Mounds
Dense Post-Insertion/ Post- Exotics or Mound at

Posthole Large Extraction Extraction Small Surface Special Edge of
Scatters Postholes Ramps Pits Pits Hearths Middens Ceramics Village
Kolomoki Mound B x x x x
McKeithen Mound A x x x x x x x
Walling Mound Ma050 x x x x x x x x x
Cold Springs Mound A x x x x x x
Garden Creek No. 2 x x x x x x x x
F E A S T I N G A N D P L AT F O R M M O U N D C E R E M O N I A L I S M I N E A S T E R N N O RT H A M E R I C A
charnel structure, which is supplied by Mound A , accounting fo r the full mortu

ary program. Richard Jefferies, in discussing the Cold Springs earthworks,
pointed to similarities between the monumental postholes in Cold Springs
Mound A and McKeithen Mound A. Noting that at Cold Springs, adjacent
Mound B contained at least one human cremation, he cautiously invoked Mi
lanich's McKeithen model as an analog (Jefferies 1994:78-80). Bennie Keel, in con
trast, interpreted the primary and secondary mound summits of Garden Creek
No. 2 as each supporting "some type of building" in the manner of later Missis
sippian platform mounds. This interpretation was based on the presence of an ir
regular burned area which he felt was a structure floor, in addition to the
ubiquity of postholes, which, as noted previously, failed to resolve into any defi
nite structure pattern (Keel 1976:78-86, 156) .
Turning again to the Walling platform mound, I have previously argued for a
nonmortuary and also nonresidential use. Postholes on the second mound stage
at Walling were numerous, as at Garden Creek No. 2, but were not necessarily
part of any building. As to any potential function as a charnel structure, neither
burials nor human bone scraps were found anywhere in association with the
Walling platform mound (except for obvious Mississippian intrusive burials,
which is also the case for Garden Creek No. 2), despite reasonable bone preserva
tion. This observation is in fact true of all five of the examples discussed; in no
case are burials or scraps of human bone reported in association with the Middle
Woodland platforms, even for McKeithen Mound A which has been so strongly
asserted to be a charnel platform. Certainly there are other Woodland platform
mound sites that have yielded human remains, the most prominent example of
which, perhaps, is Kolomoki Mound D, which is assuredly a mortuary monu
ment. But in general, mortuary use is not a prominent feature among the free
standing examples.
For Walling, I suggested that "the faunal and botanical data, taken together
with other signs of 'kitchen' activity such as midden accumulations, broken pots,
and surface hearths, ultimately support a scenario of communal food prepara
tion and consumption-L e., feasting-taking place at intervals on the summit"
of the second mound stage. I also highlighted the presence of exotic chert deb
itage on the mound summit, along with evidence for the use of copper-covered
artifacts, claiming that one aspect of mound use was the manufacture of goods
for ritual display and exchange. This evidence, I concluded, "agree[s] with a
broader picture of gift-giving and gift receiving in the context of feasting as a pri
mary social mechanism in Hopewellian exchange" (Knight 1990:160-161) .
It remains to discuss any direct evidence at these mounds for communal feast
ing in the most telling form: that of faunal and ethnobotanical data. Unfortu
nately, published reports for Kolomoki Mound B, Cold Springs Mound A, and
323
Vernon]ames Knight
Garden Creek No. 2 do not provide us with detailed data of this sort. However,
for McKeithen Mound A, we are fortunate to have an analysis of the faunal re
mains coming primarily from the two midden deposits on the edge of the plat
form (carried out by Arlene Fradkin and reported in Milanich et al. 1984:102).
Bone was only moderately well preserved. Much of the bone, however, was iden
tifiable and this sample was restricted largely to that of a single species, the white
tailed deer, the only exceptions being tiny amounts of duck, fish, and rabbit. As
to specific skeletal elements of the deer bone, "the elements are overwhelmingly
from the haunches of deer . . . Whoever was eating deer on the Mound A plat
form mound was receiving almost solely the haunches. Presumably the deer
were butchered elsewhere" (Milanich et al. 1984:102) .
For the Walling platform mound, an ethnobotanical analysis of flotation sam
ples was made by Margaret Scarry and two separate analyses of the faunal re
mains were made by John Worth and Gweneth Duncan. Of the botanical
remains Scarry (1990) says that the assemblage is an unusually diverse one, de
spite the small size of the samples. Some cultivated plants are present in small
amounts including maize (stalk and cupule fragments), cucurbits, sunflower,
chenopod, maygrass, and little barley. Besides these, Scarry reports a surprising
diversity of seeds from gathered wild greens and fruits, including plants that are
generally uncommon in archaeological deposits in this region. Nutshell remains,
ordinarily the most abundant component of botanical assemblages for this pe
riod, are relatively scarce. Scarry proposed that this unusual species composition
is the result of seasonality of use.
Feasts (or offerings) prepared in the summer might contain a high proportion of
early garden products and fruit, while a fall or winter feast might include nuts and
fall-ripening grains. The Walling assemblage as a whole contains few nut remains, a
varied assortment of seeds from summer-ripening fruits, and comparatively abun
dant grains of maygrass-an early summer crop. If season of deposit is a factor,
then most remains are derived from summer activities. (Scarry 1990:127)
Thus, although the plant food assemblage is relatively unusual, much of this pe
culiarity is plausibly accounted for by the season of use. Moreover, the assem
blage does not particularly bespeak of feasting; among the foodstuffs nothing is
so overrepresented as to suggest large-scale consumption, nor indeed do fruits
and greens seem especially well suited for bulk accumulation.
The faunal remains from the Walling platform mound tell a somewhat differ
ent story. The species list is remarkably lacking in diversity, as the faunal assem
blage is overwhelmed by bones of the white-tailed deer. Using formulas relating
bone weight to total biomass, Worth discovered that deer accounted for over 90
percent of the total meat weight in contexts related to the second mound stage
324
F E ASTI N G A N D P LAT F O R M M O U N D C E R E M O N I A L I S M I N E A S T E R N N O RTH A M E R I C A
summit, distantly followed b y remains o f beaver. In contrast t o faunal remains re

covered from the premound midden, those from the mound summit showed tio
evidence of gnawing by rodents or dogs. About one-third of the deer bone was,
however, burned; most of these sped.mens came from the vicinity of a surface
hearth. A study of bone elements revealed that heavy meat-bearing body parts
were overrepresented, mirroring the findings at McKeithen Mound A. Finally;
the ageable sped.mens revealed an unusually high frequency of adult deer. Worth
concludes that "the high proportion of deer in the total sample, and the possibil
ity that the meat-bearing portions of older de er may dominate the sample, sug
gests that the Walling site may in fact reflect Hopewellian ceremonial activity;
including the redistribution of meat through ceremonial feasting" (Worth
199o:uiJ). Duncan's (1990) analysis of separate faunal samples, primarily from the
upper mound stages at Walling, concurs with Worth's conclusion. It thus appears
that deer meat was being processed elsewhere and mobilized to the mound area
for roasting and consumption. It is unfortunate that the data are not available to
test the obvious question of whether these findings are generalizable to the other
sites herein reviewed.
Most of the postholes contributing to what I have called scaffolding behavior on
the Walling platform mound and elsewhere are not associated with surface
hearths. Such scaffolds, then, are mostly not racks for drying or roasting meat
like the ones depicted in the De Bry engravings of the sixteenth-century Timu
cua (Lorant 1946: 83; Fig. 11.7). However, it may not be too much of a leap to sug
gest that they might have served as racks for the collection and perhaps
conspicuous display of dried meat in the days prior to small-scale feasts. We have
here an alternative model to those which would invoke either mortuary practices
on the one hand, or mounds-as-substructures on the other, as the central use-pat
terns of these Woodland mound summits.
I agree with Lindauer and Blitz (1997) that these uses of Woodland platform
mounds were essentially communal. Access was apparently unrestricted within
small village communities. In our data the individual is evidently submerged,
suggesting that the main locus of interaction was between kin groups or com
munities serving as hosts. It is perhaps important for comparative purposes to re
view our evidence for the scale of this activity.
One point of reference is the size of the attached villages. Six or seven con
temporaneous houses have been estimated for the village size at Walling (Knight
1990:8), and eight for Cold Springs (Jefferies 1994: 80), suggesting permanent vil
lage populations perhaps in the range of 30-60 inhabitants for these sites. Kohler
(in Milanich et al. 1984: 89), using rates of sherd deposition, estimated the maxi
mum population of McKeithen at slightly over 100. Kolomoki is by far the largest
of the communities reviewed here. Based on the size of the Kolomoki phase
325
Vernonfames Knight
Figure 1r.7. Timucuan method of smoking meat on a scaffold, sixteenth-century Florida.

Engraving by Theodore De Bry, after a painting by Jacques le Moyne de Morgues. From
Lorant 1 94 6.
midden as reported by Sears (1956), a resident population in the low hundreds

would be a reasonable guess. On the question of whether such villages hosted
supracommunity festivals as well as gatherings at the level of the local commu
nity, current consensus weighs in favor of region-wide involvement (Walthall
1985; Knight 1990; Milanich et al. 1984; Smith 1986), particularly in view of the
strong evidence for external exchange at many of these sites. If so, the size of the
largest gatherings might considerably exceed the population of the host village.
There are other limiting factors on the potential scale of feasting at these sites.
One is the small absolute size of the facilities I have been describing, possessing
use-surfaces between 17-34 m in diameter. Notably; Kolomoki's Mound B, al
though it is at the largest site, has the smallest use-surface in the sample. A second
limiting factor is that systems of food production were only weakly developed in
these societies. Despite the use of some cultigens in a gardening complex, the
main source of carbohydrates was still wild gathered plants, while domesticated
animals other than dogs were absent. Venison, evidently the preferred food in the
326
F E A S T I N G A N D P L AT F O R M M O U N D C E R E M O N I A L I S M I N E A ST E R N N O RT H A M E R I C A
mound contexts, is not particularly suited to storage or bulk accumulation in ad

vance of a major feast, even when dried. As a hunted product, supplies would
have been limited at any given time. Moreover, if there is a correlation between
the size of social gatherings and the size of associated cooking vessels, it is my
impression that average vessel size in these mound deposits does not exceed that
from ordinary domestic contexts. Standard cooking vessels at the Walling mound
are small, short jars with orifice diameters of about 15-20 cm. I must say, how
ever, that the kinds of studies necessary to confirm such impressions have not
been done at any of the sites. All these factors considered, it is possible to envi
sion supracommunity festivals hosting no more than rno-200 participants at
most of these sites; perhaps they were smaller. Gatherings at Kolomoki, a larger
center, conceivably could have been at a somewhat grander scale, although that
site lies within a sparsely populated region and its scaffolding facilities, as shown
by Mound B, were quite small.
A principal function of feasting in these societies may have been the establish
ment and preservation of intervillage alliances. A security advantage would thus
be conferred to the community that could routinely host festivals; it is no acci
dent that the period in question in the Southeast is one in which evidence of
strife or warfare is depressed. However, if the nexus of feasting was indeed be
tween neighboring communities, it was not entirely reciprocal and symmetrical.
Most contemporary settlements within a given region lacked platform mounds,
suggesting that host groups who could muster the labor to erect one thereby
gained a measure of prestige that could be translated as a source of political in
fluence for individual village headmen.
As I have previously emphasized, for the Walling site, feasting seems to have
provided a context for the manufacture and giving of gifts, especially nonlocal,
finely crafted high-prestige items that would ultimately become concentrated in
burial mounds as mortuary offerings. Clearly some communities, and some indi
viduals, profited from such traffic more than did others.
In their level of sociopolitical organization these societies were transegalitar
ian, which is to say "those between chiefdoms and true egalitarian societies as
represented by generalized hunter I gatherers" (Chapter 2). In the Southeast, suc
cessful community leaders of the Woodland era have routinely been compared
to Big Men as described ethnographically in Oceania (e.g. , Milanich et al. 1984;
Smith 1986) . That such individuals enjoyed special advantages including access to
exotic craft goods and even retainer burials is clearly shown in the Mound D and
Mound E burials at Kolomoki (Sears 1953, 1956) . They also had privileged access
to special foods, as is shown by the individual centrally buried in the summit of
McKeithen Mound B. Isotopic analysis of skeletal remains has shown that of the
McKeithen individuals tested, this person alone routinely consumed maize (Mi-
327
VernonJames Knight
lanich 1997) . Maize, an introduced tropical cultigen, is rare in the Southeast at this
time level and may have been cultivated in small quantities in the vicinity of cer
emonial centers for special use (Scarry 1993) .
In sum, sites discussed in this paper exhibit at least some of the archaeological
signatures of feasting outlined by Hayden (Chapter 2). There is growing evidence
that communal consumption of food was an important social process among
Woodland societies of the southeastern United States. One of the things that
seems important here is that we find it connected in the present instance with a
specific kind of ritual activity, one having to do with platform mound building
and, as I perceive it, world-renewal ceremony. Recently Robert Hall (1997) has
made the claim that aspects of historic world-renewal symbolism in the Plains
and Eastern Woodlands can be linked to the material remains of Middle Wood
land cultures, and mound building in particular. Hall in fact suggests that the
whole complex Woodland burial mound tradition and interregional
Hopewellian exchange of goods is best thought of in terms of a combination of
elements of world-renewal and reincarnation ceremony as practiced historically.
This is certainly a provocative model that has the potential to explicate much of
what we know about the Middle Woodland cultural climax.
The thrust of the present paper is to suggest, however, that in parts of the
Woodland Southeast, burial-mound ceremonialism was segregated from a sec
ond kind of ceremonialism. This second kind centered on world renewal and
feasting, emphasizing community integration, very similar in type to the Green
Corn ceremonialism practiced throughout the Eastern Woodlands in the historic
era. Green Corn ceremonialism was, and in several cases still is, a major Native
American Indian ritual associated with the ripening of corn. It was (and is) per
formed within a consecrated ceremonial area that is a world symbol, renewed by
the sweeping or the addition of earth (Witthoft 1949) . It is ironic, therefore, to
consider how often historic Green Corn ceremonialism has been invoked-awk
wardly-as a model for Mississippian ritual practices, whereas in fact the historic
Green Corn practices, with their emphatically communal ethic, may be much
more closely akin to decidedly pre-Mississippian forms of ritual expression.
REFERENCES
Anonymous
1974 Early Swift Creek Indian Mound. Archives and History News 4:2. Tallahassee:
Florida Division of Archives, History, and Records Management.
Boudreaux, Edmond A., and Hunter B. Johnson
1998 Test Excavations at the Florence Mound in Northern Alabama. Report sub
mitted to the Alabama Historical Commission and the City of Florence, Al
abama. Tuscaloosa: University of Alabama Department of Anthropology.
328
F E A S T I N G A N D P LAT F O R M M O U N D C E R E M O N I A L I S M I N E A S T E R N N O RT H A M E R I C A
Brose, David S .
1979 An Interpretation of the Hopewellian Traits in Florida. In Hopewell Archaeol
ogy: The Chillicothe Conference, edited by David S. Brose and N' omi Greber,
pp. 141-149. Kent, Ohio: Kent State University Press.
Bullen, Ripley P.
1953 The Famous Crystal River Site. The Florida Anthropologist 6:9-37·
Dickens, Roy S.
1975 A Processual Approach to Mississippian Origins in the Georgia Piedmont.
Southeastern Archaeological Conference Bulletin 18:31-42.
Duncan, Gweneth A.
1990 A Supplemental Vertebrate Faunal Analysis of the Walling Site (Ma050). In
Excavation of the Truncated Mound at the Walling Site: Middle Woodland Culture
and Copena in the Tennessee Valley, by V. J. Knight Jr., pp. 144-152. Report of In
vestigations 56. Alabama Museum of Natural History, Division of Archaeol
ogy, University of Alabama, Tuscaloosa.
Ford, James A.
1951 Greenhouse: A Troyville-Coles Creek Period Mound Site in Avoyelles Parish,
Louisiana. Anthropological Papers 44, Part r. New York: American Museum
of Natural History.
Fowke, Gerard
1928 Archaeological Investigations-II. Forty-Fourth Annual Report of the Bureau of
American Ethnology, pp. 399-436. Washington, D.C.
Fuller, Richard S., and Dianne Silvia Fuller
1987 Excavations at Morgan: A Coles Creek Mound Complex in Coastal Louisiana. Bul
letin n. Cambridge: Lower Mississippi Survey, Peabody Museum, Harvard
University.
Greenwell, Dale
1984 The Mississippi Gulf Coast. In Perspectives on Gulf Coast Prehistory, edited by
Dave D. Davis, pp. 125-155. Gainesville: University of Florida Press.
Hall, Robert L.
1997 An Archaeology of the Soul: North American Indian Belief and Ritual. Urbana:
University of Illinois Press.
Jefferies, Richard W
1994 The Swift Creek Site and Woodland Platform Mounds in the Southeastern
United States. In Ocmulgee Archaeology: 1936--1986, edited by David ]. Hally,
pp. 71-83. Athens: University of Georgia Press.
Jones, B. Calvin, Daniel T. Penton, and Louis D. Tesar
1998 1973 and 1994 Excavations at the Block-Sterns Site, Leon County, Florida. In A
World Engraved: Archaeology of the Swift Creek Culture, edited by Mark Williams
and Daniel T. Elliot, pp. 222-246. Tuscaloosa: University of Alabama Press.
Keel, Bennie C .
1972 Woodland Phases of the Appalachian Summit. Ph.D. dissertation, Depart
ment of Anthropology, Washington State University, Pullman.
329
Vernon]ames Knight
1976 Cherokee Archaeology: A Study of the Appalachian Summit. Knoxville: University

of Tennessee Press.
Kellar, J. H., A. R. Kelly; and E. V. McMichael
1962 The Mandeville Site in Southwest Georgia. American Antiquity 28 :338-355.
Kelly; Arthur R. , and Betty A. Smith
1975 The Swift Creek Site, 9Bi3, Macon, Georgia. Ms. on file, National Park Ser
vice, Southeastern Archaeological Center, Tallahassee, Florida.
Knight, VernonJames, Jr.
1981 Mississippian Ritual. Unpublished Ph.D. dissertation, Department of An
thropology; University of Florida, Gainesville.
1986 The Institutional Organization of Mississippian Religion. American Antiquity
51:67y-687.
1989 Symbolism of Mississippian Mounds. In Powhatan's Mantle: Indians in the
Colonial Southeast, edited by P. H. Wood, G. A. Waselkov, and M. T. Hatley;
pp. 279-29!. Lincoln: University of Nebraska Press.
1990 Excavation of the Truncated Mound at the Walling Site: Middle Woodland Culture
and Copena in the Tennessee Valley. Report of Investigations 56. Alabama Mu
seum of Natural History, Division of Archaeology; University of Alabama,
Tuscaloosa.
Kohler, Timothy A.
1997 Public Architecture and Power in Pre-Columbian North America. Paper pre
sented at the symposium, "Power, Monuments, and Civilization," Nara,
Japan, December.
Kurjack, Edward B.
1975 Archaeological Investigations in the Walter F. George Basin. In Archaeological
Salvage in the Walter R George Basin of the Chattahoochee River in Alabama, edited
by D. L. Dejarnette, pp. 87-198. Tuscaloosa: University of Alabama Press.
Kwas, Mary L., and Robert C. Mainfort
1986 The Johnson Site: Precursor to Pinson Mounds? Tennessee Anthropologist
n:29-4L
Lindauer, Owen, and John H. Blitz
1997 Higher Ground: The Archaeology of North American Platform Mounds.
journal of Archaeological Research 5:169-207.
Lorant, Stefan, ed.
1946 The New World: The First Pictures of America. New York: Duell, Sloan, and Pearce.
Mainfort, Robert C . , Jr.
1988 Middle Woodland Ceremonialism at Pinson Mounds, Tennessee. American
Antiquity 5p58-173.
Mainfort, Robert C., Jr., and Richard Walling
1983 Excavations at Pinson Mounds: Ozier Mound. Midcontinentaljoumal of Ar
chaeology 17:n2-136.
McMichael, Edward V.
1964 Veracruz, the Crystal River Complex, and the Hopewellian Climax. In
330
F E A S T I N G A N D P L AT F O R M M O U N D C E R E M O N I A L I S M I N E A ST E R N N O RT H A M E R I C A
Hopewellian Studies, edited by Joseph R . Caldwell and Robert L . Hall,

pp. 124-132. Springfield: Illinois State Museum.
Milanich, ]. T.
1997 Preface. In Archaeology of Northern Florida, A.D .20D-900: The McKeithen Weeden
•
Island Culture, by]. T. Milanich, A. S. Cordell, V. J. Knight Jr. , T. A. Kohler, and

B. ]. Sigler-Lavelle. Gainesville: University Press of Florida.
Milanich, J. T. , A. S. Cordell, V. J. KnightJr. , T. A. Kohler, and B. J. Sigler-Lavelle
1984 McKeithen Weeden Island, the Culture of Northern Florida, A.D .20()-900. Orlando,
•
Florida: Academic Press.

Moore, Clarence B.
1894 Certain Sand Mounds on the St. Johns River, Florida. journal of the Academy of
Natural Sciences of Philadelphia 10:130--246.
1896 Certain River Mounds of Duvall County, Florida. journal of the Academy of
Natural Sciences of Philadelphia 10:449-516.
1902 Certain Aboriginal Remains of the Northwest Florida Coast. Part 2. journal
of the Academy of Natural Sciences of Philadelphia 12 (2): 125-335.
Nance, C. Roger
1976 The Archaeological Sequence at Durant Bend, Dallas County, Alabama. Special
Publication 2. Orange Beach: Alabama Archaeological Society.
Phillips, Philip
1970 Archaeological Survey in the Lower Yazoo Basin, Mississippi: 1949-1955· Papers of
the Peabody Museum of American Archaeology 60. Cambridge: Harvard
University.
Rafferty, Janet
1987 The Ingamar Mounds Site: Internal Structure and Chronology. Midcontinen
tal]ournal of Archaeology 12:147-173.
1990 Test Excavations at Ingamar Mounds, Mississippi. Southeastern Archaeology
9:93-102.
Scarry; C. Margaret
1990 Plant Remains from the Walling Truncated Mound: Evidence for Middle
Woodland Horticultural Activities. In Excavation of the Truncated Mound at the
Walling Site: Middle Woodland Culture and Copena in the Tennessee Valley, by V. ].
Knight Jr., pp. n5-129. Report of Investigations 56. Alabama Museum of Nat
ural History; Division of Archaeology, University of Alabama, Tuscaloosa.
1993 Variability in Mississippian Crop Production Strategies. In Foraging and Farm
ing in the Eastern Woodlands, edited by C . M. Scarry; pp. 78-90. Gainesville:
University Presses of Florida.
Sears, William H.
n.d. An Investigation of Prehistoric Processes on the Gulf Coastal Plain. Final re
port submitted to the National Science Foundation. On file, Florida Museum
of Natural History; Gainesville.
l95ra Excavations at Kolomoki, Season I. University of Georgia Series in Anthropol
ogy, No. 2. Athens.
331
Vernon]ames Knight
l951b Excavations at Kolomoki, Season II, Mound E. University of Georgia Series in

Anthropology, No. 3. Athens.
1953 Excavations at Kolomoki, Seasons III and IV; Mound D. University of Georgia Se
ries in Anthropology, No. 4. Athens: University of Georgia Press.
1956 Excavations at Kolomoki: Final Report. University of Georgia Series in Anthro
pology, No. 5. Athens: University of Georgia Press.
1982 Fort Center: An Archaeological Site in the Lake Okeechobee Basin. Gainesville : Uni
versity Presses of Florida.
Shetrone, Henry C.
1925 Exploration of the Ginther Mound. The Miessee Mound. Ohio Archaeological
and Historical Quarterly 34:154-168.
Shogren, Michael G.
1989 A Limited Testing Program at Four Mound Sites in Greene County, Alabama. De
Soto Working Paper rr. Alabama De Soto Commission. Alabama Museum of
Natural History, University of Alabama, Tuscaloosa.
Smith, Bruce D.
1986 Archaeology of the Southeastern United States: From Dalton to de Soto,
ro500-500 B.P. Advances in World Archaeology 5:1-92.
Squier, E. G., and E. H. Davis
1848 Ancient Monuments of the Mississippi Valley. Smithsonian Contributions to
Knowledge r. Washington, D.C . : Smithsonian Institution.
Toth, Alan
1974 Archaeology and Ceramics at the Marksville Site. Anthropological Papers 56. Ann
Arbor: University of Michigan Museum of Anthropology.
Vescelius, Gary S.
1957 Mound 2 at Marksville. American Antiquity 22:416-420.
Walker, Winslow M.
1936 The Troyville Mounds, Catahoula Parish, Louisiana. Bulletin 133. Washington,
D.C . : Bureau of American Ethnology.
Walthall, John A.
1985 Early Hopewellian Ceremonial Encampments in the South Appalachian
Highlands. In Structure and Process in Southeastern Archaeology, edited by R. S.
Dickens Jr. and H. T. Ward, pp. 243-262. Tuscaloosa: University of Alabama
Press.
Webb, William S., and David L. Dejarnette
1942 An Archaeological Survey of Pickwick Basin in the Adjacent Portions of the States of
Alabama, Mississippi, and Tennessee. Bulletin 129. Washington, D.C . : Bureau of
American Ethnology.
Williams, Mark
1992 Archaeological Investigations of Fortson Mound, Wilkes County, Georgia. Lamar
Institute Publication 3. Watkinsville, Georgia: Lamar Institute.
Williams, Mark, and Jennifer F. Harris
1998 Shrines of the Prehistoric South: Patterning in Middle Woodland Mound
332
F E ASTI N G A N D P LAT F O R M M O U N D C E R E M O N I A L I S M I N E A ST E R N N O RT H A M E R I C A
Distribution. In A World Engraved: Archaeology of the Swift Creek Culture, edited

by Mark Williams and Daniel T. Elliot, pp. 36-47. Tuscaloosa: University of
Alabama Press.
Witthoft, John
1949 Green Corn Ceremonialism in the Eastern Woodlands. Occasional Contributions
13. Ann Arbor: University of Michigan Museum of Anthropology.
Worth, John
1990 Vertebrate Faunal Analysis for the Walling Site (Ma050) , 1986 Excavations. In
Excavation of the Truncated Mound at the Walling Site: Middle Woodland Culture
and Copena in the Tennessee Valley, by V. J. Knight Jr., pp. 129-144· Report of In
vestigations 56. Tuscaloosa: Alabama Museum of Natural History, Division
of Archaeology, University of Alabama.
333
12
A CASE O F RIT UA L FEASTING
AT THE CAHOKIA SITE
Lucretia S. Kelly
The site of Cahokia is an early Mississippian (A.D. 1050-1350) center located in the
American Bottom region of the central Mississippi River floodplain, just east of
St. Louis on the northern edge of Mississippian development. The Mississippian
cultural tradition (A.D. 1000-1500) extends spatially over portions of the Midwest
and Southeast United States (Fig. 12.1) including societies that display distinctive
ceramic technology, platform mounds, intensification of agricultural crops, and
ranked sociopolitical systems. Cahokia is the largest Mississippian site, with over
rno mounds, and represents the most complex of all Mississippian communities
(Emerson 1995; Fowler 1974, 1975, 1978, 1989; Mehrer 1988, 1995; Milner 1990, 1998;
Pauketat 1991, 1994).
During the past two decades, much has been written about the nature and
334
THE
A M E RICAN
BOTTOM
ca. 1800
e Mound Group
M i i es
Kllom•l•n
Figure 12.r. The American Bottom region, with inset showing the extent of the Middle
Mississippian cultural tradition.
Lucretia S. Kelly
form(s) of political economy associated with Mississippian societies (see Muller

1997 for a discussion). Two dichotomous schools of thought about its structure
and function have recently been articulated and termed the "top-down" and
'bottom-up" perspectives (Pauketat l997b; Saitta 1994). In this chapter, I approach
Cahokia's early economy from an empirical perspective, emphasizing the analy
sis of fauna and other classes of material recovered from a large pit at Cahokia
designated sub-Mound 5r. The faunal assemblage, in particular, provides very
specific information on ritual feasting activity. I suggest that investigation of rit
ual feasting and its integration into the larger body of existing literature about so
ciopolitical mechanisms can provide finer resolution to questions concerning
Cahokia's political economy early in its florescence as a mound center.
C A H O K I A' S S E T T L E M E N T S YS T E M
A N D P O L I T I CA L E CO N O M Y
During the 1970s and l98os, a large number of sites in the American Bottom were
excavated as part of the FAl-270 highway project (Bareis and Porter 1984) . Results
of these excavations include a highly refined regional chronology (Bareis and
Porter 1984; Fig. 12.2), and a better understanding of changes in the regional set-
A.D. �--'P"-'e"'ri:..:.o.;::.d_�-----"P""'h_as'-'e--,
Sand Prairie
1 300
Moorehead
1 200
Stirling
1 1 00 - - - - - - - - - - - - - r= - - - - - - - - - - - - - ,,....=d
Lindhorst
Edelhardt I;;' Lindemann

1 000 - - - - - - - - - - - - - ....
rD ------------- �
z n
I- <( Merrell !§! George Reeves �
950 z c:: - - - - - - - - - - - - - =I: - - - - - - - - - - - - - --i
w c..
--i iil
(!) - 0..
a: V) iil
w !!! ;::;:
::::E V) Loyd � Range
w !!!
(5'
"'
g·
8 50 ::::E
Collinsville Dohack
8 00 i...�
.. ��.i...�
.. ��......
. ��-
...
Figure 12.2. Late prehistoric chronology for the American Bottom.
336
R I T U A L F E A S T I N G AT T H E C A H O K I A S I T E
tlement pattern between A.D. 800 and noo. During the Emergent Mississippian
(A.D. 800-1050) , regional population aggregated into large nucleated villages
(J. Kelly 1992) . Beginnings of social ranking appear at this time, and form the
roots of Cahokia's future complexity (J. Kelly 1990, 1992; Pauketat 1994) .
The initial Mississippian episode (the Lohmann phase, A.D. 1050-noo) is desig
nated by major and rather rapid changes in the regional landscape. Emergent
Mississippian populations that were aggregated in large nucleated villages either
disperse to small farmsteads and civic / ceremonial nodes, or coalesce into mound
centers (Emerson 1995; ]. Kelly 1992; Mehrer 1995; Pauketat 1991 ) . At Cahokia it
self, there is a five- to tenfold population increase (Pauketat and Lopinot 1997) ,
and status differentiation is evident (Fowler 1989; Pauketat 1994) . The creation of
the site's plan (Fowler 1973, 1989; ]. Kelly 1996, 1997) is centered upon Monks
Mound, the Grand Plaza (Holley, Dalan, and Smith 1993; Pauketat and Rees 1996 ) ,
and several flanking plazas (Fig. 12.3) . The Grand Plaza was demarcated as an im
portant area of the site by major earthmoving pubhc-works projects. Recent ar
chaeological investigations in this area have revealed that a meter of clayey
deposits overlying a ridge and swale topography were stripped and used for bor
row, possibly for the initial construction of Monks Mound (the site's largest
mound) to the north. This area was then reclaimed by filling in the swales and
raising the plaza to about its premodified level to form a clean and level surface
(Dalan 1997; Dalan et al. 1994) . This provided a large centrahzed space to accom
modate ritual activities attended by the Cahokia population (Dalan 1997) .
There is now fundamental agreement that the Mississippian settlement pat
tern in the northern American Bottom was hierarchical and that Cahokia was at
its center. Other smaller mound centers (with either multiple or single mounds)
and numerous nonmound communities that may have served several functions
complete a proposed three-tiered settlement hierarchy (Emerson 1995, Finney
1993, Mehrer 1995, Milner 1990) . Emerson (1995 ) beheves the lowest tier of non
mound communities can be differentiated from each other by how they may
have functioned in the overall regional settlement system. He has identified
nodal centers that served speciahzed pohtical, rehgious, and economic functions.
Members of the two schools of thought on the development of pohtical econ
omy at Cahokia disagree about how the small farmstead/ hamlet entities of the
third settlement tier were tied (if at all) to Cahokia or to the smaller mound cen
ters (e.g., about how much regional control Cahokia actually had). Advocates of
the top-down perspective (Emerson 1997; Pauketat 1994) view Cahokia's pohtical
economy from the theoretical stance of agency. They think that the three-tiered
settlement hierarchy was a manifestation and exphcit creation of the Cahokia
ruhng class (Emerson 1997) . Ehtes in the countryside were responsible for pro
duction and mobihzation of foodstuffs to support the central ehte at Cahokia. In
337
Lucretia S. Kelly
�� \
��r \ \ /
I
I
I
e NorthPlaza /
..
CENTRAL
CAHOKIA
0 km
..- - · ---�
ohmann/Stirling Phase
Config uration
Figure 12.3. Site plan of central Cahokia showing Monks Mound, the Grand Plaza, and
flanking plazas.
effect, the Cahokia elite were not leaving the procurement of their subsistence to
chance (Emerson 1997).
Those with the bottom-up perspective (Mehrer 1995; Milner 1990, 1998; Muller
1997) view Cahokia's economy from an environmental and functionalist stand
point. They believe the Cahokia settlement system to have been somewhat de-
338
R I T U A L F E ASTI N G AT T H E C A H O K I A S I T E
centralized and loosely structured. They argue that Cahokia elites would have
had only limited, direct influence over the hinterland communities, which are
viewed as autonomous, self-sufficient units. However, it is also argued that sur
plus produced by these farmsteads would have been sent to Cahokia as tribute,
but, perhaps on a periodic basis, and may have been a means by which leaders
augmented their positions and expanded and reinforced patron/ client relation
ships (Milner 1990, 1998) .
SOCIAL AND POLITICAL DIMENSIONS OF FOOD

Tribute or tributary systems are usually considered inherent to chiefdoms such as
those of the Mississippian Southeast (Anderson 1994; Carneiro 1981; DePratter
1991; Service 1962; Steponaitis 1978), but, tribute can encompass a variety of items
including food, prestige or exotic goods, and labor. In this paper I examine food
and its political and social dimensions within a tributary system. The term tribute,
however, means different things to different people. For instance:
At one end of the scale we have the complete and equitable reassignment of a �
lage's harvest back to its producers by a chief who is merely a temporary and be
nign custodian of it [a bottom-up perspective]. At the other end [a top-down
perspective] there is enforced appropriation of a part of a society's food supply by a
powerful ruler for his own benefit and that of a small ruling elite. (Carneiro
1981:59-60)
Because of the range of meanings, Muller (1997=14) argues that the term should
not be used, especially in reference to prestate societies, and instead prefers the
use of prestation. But, "tribute" is commonly used in the literature as a catchall
term for the general mobilization of goods and labor by the elite (Blitz 1993) . It
can encompass a number of provisioning strategies including gift-giving, ritual
feasting, or a collected tax. The term is useful as long as its meaning is defined.
Various models of tribute collection and redistribution have been outlined for
Mississippian societies (see Steponaitis 1978; Welch 1991) . However, they do not
offer much information about what mechanisms governed collections and redis
tribution (if any) . For instance, was tribute collected on a regular basis, was it a
type of usury tax (paid for use of land or hunting territories), was it collected for
the exclusive use of the elite, or did collected stores of produce serve as insurance
against hard times (risk management)? The way one thinks about tribute is usu
ally dependent upon the chiefdom model that one embraces.
However, by using empirical evidence in conjunction with theoretical models,
we may be able to identify some of the mechanisms involved in tributary systems
or food mobilization, and thus gain a better understanding of the construction of
social relations within a society (sensu Dietler 1996) . The most direct evidence
339
Lucretia S. Kelly
would be the food remains themselves, that is, faunal and floral materials. It is
only recently that zooarchaeological and archaeobotanical data have been used
to directly address such social issues (see Blitz 1993; Jackson and Scott 1995; Welch
and Scarry 1995).
Fortunately, large-scale archaeological excavations, undertaken as part of the
massive highway projects mentioned earlier, have generated an enormous
amount of zooarchaeological and archaeobotanical data for the American Bot
tom (Johannessen 1984, 1988, 1993; L. Kelly and Cross 1984; L. Kelly 1997b;
Lopinot 1991, 1992, 1994; Rindos and Johannessen 1991; Simon and Parker 1995).
This extraordinary biological database can be used to address the role played by
the higher-ranked segment of Cahokia's population in food provisioning and
possible tributary systems. Was the chiefly class just a temporary caretaker of a
village's harvest, holding it for the benefit of all society, or was it coercively ap
propriating food for the exclusive benefit of the elite? In other words, by examin
ing the way one segment of the population relates to another in the provisioning
of food, can we gain a better understanding of the articulation of social relations
among the various social segments? To examine alternative hypotheses about
tribute at Cahokia, one must take into account several factors regarding the na
ture of the archaeological evidence, and how food items might fit into a tributary
system.
FOOD AS TR I B UTE
For instance, consider fresh meat from game animals (e.g., deer). Unlike domes
tic animals in other areas of the world, deer may not have made good, regular
sources of tribute (Zeder 1996). Acquisition of hunted wild game is unpre
dictable, especially during certain times of the year. Fresh meat would also be dif
ficult to transport long distances and to stockpile for future use because of
spoilage problems (Jackson and Scott 1995) . On the other hand, dried meat and
skins probably did enter into a tributary system (Zeder 1996). We know these
were common tribute items ethnohistorically in the Southeast (Clayton, Knight,
and Moore 1993), but animal products in these forms would be almost impossible
to detect archaeologically:
It has been argued (Blitz 1993) that maize and other native grain crops may
have been difficult to control or redistribute by the elite unless or until they were
centrally stockpiled or stored in large quantities. Ethnohistorical information in
dicates that large granaries were controlled by the chiefs in the Southeast (Clay
ton, Knight, and Moore 1993; Barker 1992; Smith and Hally 1992) . Above-ground
granaries are difficult to identify archaeologically, but those that have been recog
nized at Cahokia are located in domestic or residential areas. They are inter
preted as having been communal resources rather than having been under chiefly
340
KEY:
II High
lilll M id
D Low
Figure 124 Portions of deer belonging to the high, mid, and low Food Utility Index
(FUI) categories. Reprinted from Cahokia: Domination and Ideology in the Mississippian
World, edited by T. R. Pauketat and T. E. Emerson, by permission of the University of
Nebraska Press, copyright University of Nebraska Press (1 997).
control (Collins 1990, 1997; Fritz and Johannessen 1996; Lopinot 1994). It has also
been documented ethnohistorically, however, that chiefs' granaries were some
times located in areas nearer the producers (Smith and Hally 1992) than the
chiefs' residences. 1b.erefore, interpreting control solely on location may be
problematical (Muller 1997) .
A R C H A E O LO G I C A L E V I D E N C E
For the American Bottom and Cahokia, differential access to animal resources
based on status has been investigated by the employment of two different types
of faunal studies: taxonomic diversity and body-part distribution of deer, the
largest vertebrate species recovered in any abundance from Cahokia (L. Kelly
1997b) . Deer body-part studies, particularly those made on assemblages from Ca
hokia, indicate a flow of meaty portions of deer into the site (L. Kelly 1979,
1997b). However, the studies of faunal assemblages from outlying communities
do not yet provide the data necessary to determine whether they were furnishing
deer meat to Cahokia. Faunal assemblages from these sites are small, and it is dif
ficult to distinguish faunal refuse resulting from daily subsistence from that re
sulting from butchering animals to be used as tribute.
Plant products such as shelled maize appear to have been a very important
component of tribute ethnohistorically (Clayton, Knight, and Moore 1993; Smith
and Hally 1992) and theoretically should be visible in the archaeological record.
341
Lucretia S. Kelly
However, archaeobotanical evidence for tribute from the American Bottom

tends to be equivocal (Lopinot 1997; Pauketat l997a). The data do not exhibit
strong differences among site types and locations, but rather they indicate that
the same varieties of crops were grown and eaten in all parts of the region
(Lopinot 1997).
There is evidence of agricultural intensification for a suite of native cultigens
(knotweed, chenopod, maygrass) and maize (Fritz and Johannessen 1996;
Lopinot 1994) in the Emergent Mississippian to Mississippian. And, there is some
indication of a decrease in maize kernel-to-cob ratios from the Emergent Missis
sippian to Mississippian (Lopinot 1994). This has been interpreted as movement
of shelled maize from outlying, producing communities to Cahokia (e.g. , trib
ute; Emerson 1995; Esarey and Pauketat 1992; Pauketat 1994). Fritz and Johan
nessen (1996), however, do not think there is convincing evidence that shelled
maize was being consumed in greater quantities by higher-ranked social groups.
In fact, -stable carbon isotope studies run on high-status, Lohmann-phase burials
in Mound 72 at Cahokia indicate high-ranking personages may not have eaten as
much corn as other Mississippians (Buikstra, Rose, and Milner 1994).
Because the database for the American Bottom area is so large, Fritz (personal
communication, 1998) dismisses the argument that preservational problems asso
ciated with archaeobotanical materials may obscure the fact that maize and/ or
cultivated native crops were part of a tributary system of foodstuffs being utilized
differentially along class lines. Fritz and Johannessen (1996) do not see evidence for
chiefly manipulation or centralized political control of agricultural surplus, espe
cially maize. There is evidence, however, that other kinds of plants-such as to
bacco and red cedar-may have had more restricted use by ritual specialists (Fritz
and Johannessen 1996; Lopinot 1994; Lopinot and Woods 1993).
The above discussion demonstrates that the zooarchaeological and the ar
chaeobotanical evidence from the American Bottom are somewhat equivocal
when used to examine broad questions concerning food tribute. If the focus is
narrowed, however, it may be possible to identify tribute or food-provisioning
mechanisms that are more visible in the archaeological record. This, in turn, may
help elucidate broader questions regarding the articulation of social relations. Re
cent analysis of material from the sub-Mound 51 pit is providing strong evidence
for large public events held during Cahokia's development that may be linked to
one such mechanism: ritual feasting.
T H E CA H O KI A EXA M P L E : S U B-MO U N D 5 1
An extraordinary assemblage of cultural material was recovered in the late 1960s
from a large, prehistoric borrow pit, revealed under Mound 51 when that mound
was removed by the owner for fill (Chmurny 1973). It is located 175 m to the south-
342
east o f Monks Mound, the site's largest mound, and i s on th e northeast edge of
the Grand Plaza (Fig. 12.3) . The dirt removed from this borrow pit was presumably
used in the construction of nearby mounds, possibly even in the early stages of
Monks Mound, and for the leveling of the Grand Plaza. After the pit ceased func
tioning as a borrow pit, it was left open for several months before being rapidly
filled (possibly within one to three years) by a sequence of at least seven distinct
depositional episodes or zones (Fig. 12.5; Chmurny 1973; Pauketat l997b). Mound
51 was subsequently built in two stages on this reclaimed surface (Chmurny 1973) .
Salvage excavations of the sub-mound pit were conducted between 1966 and
1970 (Bareis 1975). Because only a portion of the pit was excavated, its exact shape
and horizontal dimensions are unknown, but excavations did indicate a mini
mum north-south dimension of 53 m, a width of more than 20 m, and a depth of
3 m. Five 3-by-3 m units were excavated between 1966 and 1968 . The first unit, dug
in 1966, was not excavated by cultural zone but by arbitrary levels. The four addi
tional units, excavated in 1967 and 1968, were dug by cultural strata. One of the
four units was only partially excavated (Pauketat l997b).
E385
E388
Zones
01
02
F1
dark gray silty clay
F2
mottled dark gray

silty clay
G
H
• - burned thatch
f;g - unburned thatch
Figure 12.5. A profile of one of the sub-Mound 51 excavation units (S53 E3 85-388) show
ing the various zones.
343
Lucretia S. Kelly
The preservation of organic materials from this pit is unparalleled for this area
of the Midwest U.S., with a portion of the plant remains being preserved in an
uncarbonized state. Optimum preservational conditions resulted from an anaer
obic environment created by the rapid burial of the deposits, which were subse
quently sealed beneath the mound. William Chmurny analyzed a large sample of
the organic material recovered from the 1966 and 1967 excavation seasons for his
1973 dissertation. More recently, very fine-grained analyses of additional samples
1
were undertaken as part of Timothy Pauketat' s Early Cahokia Project.
Though it is important to interpret the data from this pit in its entirety (i. e.,
from all classes of material), in this paper I focus on the faunal remains. A full ac
counting of the various analyses performed by all the participants of this project
has not yet been made, but preliminary results have been presented. 2
FA U N A L A N A LY S E S : C A H O K I A A N D T H E A M E R I C A N B O T T O M
In order to interpret the faunal assemblage from sub-Mound 51, one must com
pare it to faunal patterns of general subsistence for the American Bottom during
Emergent Mississippian and early Mississippian periods. Faunal data from Amer
ican Bottom sites indicate that Emergent Mississippian populations were self
sufficient, relying on faunal resources in proximity to their settlements to meet
their animal-protein requirements (Kelly and Cross 1984) . They were apparently
directing their fauna! exploitation away from the procurement of large animals,
such as deer, to smaller animals such as marsh birds, medium-sized mammals,
and fish (L. Kelly 1990, 1997b). This localized exploitation pattern is also present
3
at Cahokia. However, there is a major difference in deer body-part distribution
for the Emergent Mississippian occupations at Cahokia. Elements representing
4
the higher-utility portions of the deer are represented in much higher propor
tions than they are at other Emergent Mississippian sites (L. Kelly 1997b ). The Ca
hokia percentages indicate that complete deer carcasses were not brought to the
site. This may be a reflection of Cahokia's distance (4-5 km) from the uplands ad
jacent to the American Bottom where deer would be most abundant. 5 Because of
this distance, I proposed that only the higher-utility portions of the deer were
transported to Cahokia (L. Kelly 1979) .
A rather dramatic shift in faunal composition is observed at Cahokia during
the initial Mississippian, Lohmann phase and is interpreted as resulting from po
litical and social changes. The relative quantity of mammals jumps from about IO
percent NISP (number of identified specimens) in the Emergent Mississippian,
Edelhardt phase to about 67 percent NISP in the Mississippian, Lohmann phase.
The main reason for this increase is the concoinitant increase in deer remains ( 6
percent NISP to 63 percent NISP; Fig. 12.6). Fish remains decrease from 77 per
cent to IO percent NISP and bird remains gradually increase (L. Kelly 1997b).
344
1 00
90
80
70
60
'16
N
I so
s
p
40
30
"ICT-11
+sub-Md. 51
zo 1 = non-elite
Z = elite
10
Emergent Mississippian ----- Mississippian ------
Figure 12.6. Change in percentage of NISP of deer at Cahokia.
Deer body-part representation does not change greatly, however. Deer were ap
parently still being procured at a distance from the site, but deer meat was seem
ingly making up a larger portion of the faunal diet. A possible explanation for this
increase in venison is discussed below.
Unfortunately, the bulk of the faunal data from both the Emergent Mississip
pian and Lohmann phases at Cahokia are from non-elite domestic contexts. No
Lohmann phase assemblages from purely elite contexts are yet known. There are
assemblages from the subsequent Stirling phase (A.D. 1100-1200), however, that
may represent elite subsistence refuse (L. Kelly 1997b). These indicate that deer
may have been distributed along class lines with the higher-status segment of the
population receiving more hindquarters; but it cannot be determined whether
this was also the case in the earlier Lohmann phase.
S U B - M O U N D 5 1 FA U N A L A S S E M B L A G E
Before analysis of the material recovered from the sub-Mound 51 pit was under
taken, it was postulated that it represented elite subsistence and craft-production
refuse (Pauketat et al. 1993) . However, my immediate impression of the sub
Mound 51 faunal assemblage was that it was very different from Mississippian as
semblages elsewhere in the American Bottom and from Cahokia itself, and that
impression was confirmed by subsequent study.
I have examined a faunal sample not studied by Chmurny that was recovered
from the 1967 and 1968 excavations. Only well-provenienced remains in my sam-
345
Lucretia S. Kelly
ple were identified. As noted above, part of the faunal debris recovered from the
1966 and 1967 excavation seasons was analyzed by William Chmurny for his 1973
dissertation. As far as can be determined, no overlap exists in the materials in
cluded in the present study and those of Chmurny.
In his dissertation, Chmurny reported the faunal material as a single unit,
rather than by depositional zones. He did this because different sample sizes were
recovered, and because he was addressing ecological questions rather than social
issues, so keeping the material separated by zone was not thought to be neces
sary. Some notes and tables not presented in his dissertation included informa
tion by zone, but he does not provide provenience data other than zone
designations nor is the excavation year given. I do not know what criteria he may
have used to determine zone designations for the 1966 material. Hence, I report
his findings by zone, but cannot verify the accuracy of that information.
For the present discussion, I present data from only one zone-D2-(Fig. 12.5)
a very rich provenience unit in the upper half of the pit. Faunal remains were re
covered from all zones of the sub-mound pit (Fig. 12.7) in varying amounts, but
the majority of sub-Mound 51 faunal remains (D2 NISP=8666) was recovered
from zone D2 (Fig. 12.7) . This zone also yielded an array of Lohmann phase ex
otics and craft goods, in addition to items that can best be termed "sumptuary,"
such as crystals, exotic arrowheads, axehead debitage, and sherds from engraved
and painted vessels (Pauketat 1997b). Unusual or unique faunal remains that can
be placed in this sumptuary category include a drilled alligator tooth (the most
zones 02 G H F
assemblage K C K C K c K c
----------------
NISP 1 553 2847
-------------
583
-------------
638 21 5 250
-------------
295 392
- -------------
----------------
# taxa 3
-------------
5 3
-------------
5 2 2
-------------
1 3
- - - -----------
% NISP deer 99.7 99.2 99.5 98.6 98.6 99.6 1 00.0 99.0
----------------
NISP 206 1 3 63
-------------
25 363 30 1 09
-------------
2 327
- - - -----------
# taxa 5 19 4 14 7 11 8
----------------
NISP 873 1 549
-------------
261 1 359 27 1 52
-------------
2 1 037
- - - - - ---------
# taxa 9 9 7 7 4 5 8
----------------
NISP -------------
1 ------------- ------------- - - - - ----------
# taxa 1
Not ident.
NISP 274 1 75 44 25
K = Kelly's Assemblage
C = Chmumy's Assemblage
Figure 12.7. Summary of faunal materials from sub-Mound 5 1 pit.
346
unusual faunal item in zone D2), a bone harpoon, a bone earspool, and marine
shell beads .and spoons. Bone harpoons and marine shell beads were also part of
the display of exotic goods recovered from an elite Lohmann phase burial in
Mound 72.
The deer bone stood out immediately from all other assemblages that I have
studied. It made up 99.7 percent (NISP=1253 / 1257) of the identified mammalian
remains in the assemblage I identified (Fig. 12.7) . Elements with low structural
densities-vertebrae, innominates, and scapulae-that are fragmentary or non
existent in many assemblages, were found relatively whole in sub-Mound 51.
About 72 percent of such elements recovered from zone D2 are half or more than
half complete. Many elements with unfused epiphyses were recovered in such
close association with their epiphyses that they can be readily refitted. There are
a number of deer bones from my zone D2 sample that appear to articulate, par
ticularly sections of thoracic and lumbar vertebrae. Chmurny (1973) noted simi
lar articulations in his assemblage. A number of bones from the forelimb are
whole or nearly so, including two radii, two humeri, and three ulnae. The femora
and tibiae do have broken shafts, but they have not been shattered. Long bone
fragments are not abundant. Chmurny (1973) suggests the breakage of the
hindlimb long bones in his sample is a product of butchering rather than of mar
row extraction, and I believe this to be true for my sample as well.
It should be noted that FUis (Food Utility Index) for the sub-Mound 51 deer are
figured on percentage of minimum numbers of skeletal elements (MNE) . Skele
tal element here is a complete element, and, where applicable, rights and lefts are
summed. I calculated deer FUis for Chmurny's data, even though I am not com
pletely confident about his zone designations. The MNE calculations may not be
entirely accurate either, because I used the data tables Chmurny constructed
rather than visual observations of the bones themselves. But even with these po
tential problems, the results are remarkable. The deer FUis calculated from
Chmurny's data were almost identical to those for the assemblages that I studied.
Each zone is similar regardless of the size of the sample. This indicates that the
cuts of meat or portions of deer utilized were consistent throughout the deposit.
The articulation and completeness of many bones probably reflect the manner in
which the meat was brought to the area, and the way it was prepared for consump
tion. Primary butchering debris, such as skull fragments and the lower limb bones of
metapodials, phalanges, and carpals, are almost nonexistent in the assemblage I ana
lyzed, except for several carpals. These appear to articulate with one of the radii and
may have been riders on a forelimb. Chmurny's assemblage does contain a few loose
teeth, phalanges, and metapodials, but they represent only 3 percent of his zone D2
MNEs. It seems reasonable to conclude that the meat was brought to the area in
large or bulk cuts after having been initially butchered elsewhere.
347
Lucretia S. Kelly
On the basis of large numbers of blowfly pupal remains and flesh-eating bee
tles recovered from the sub-mound pit (Elias and Pauketat 1997), it appears at
least some of the bones were deposited in the pit with raw, soft tissue still adher
ing to them. These insects feed almost exclusively on decaying flesh and are un
likely to colonize cooked meat. This suggests that the meat was cut from the
bones prior to cooking. Cutting meat from the bone prior to cooking would not
seem to be an efficient way to maximize the amount of meat prepared and may
suggest that meat was plentiful. Another indication that meat was in good supply
is that most of the bones were not processed for the recovery of marrow or bone
grease. The easiest method of cooking meat removed from the bone would be
boiling, possibly in stews. It is also possible that the meat being stripped from the
bones was dried for future use. In either case, the bones would be left relatively
intact when deposited.
The bird and fish components of the present D2 assemblage are also unique in
comparison to other Cahokia and American Bottom assemblages. In contrast to
the mammals, they differ in several ways from the assemblage Chmurny ana
lyzed. His assemblages were much larger and the birds were more varied (Fig.
12.7). Only five taxa of birds were identified from my zone D2 assemblage (Fig.
12.8). This is quite low when compared to other assemblages at Cahokia or in the
American Bottom-usually twice as many are identified-whether from elite or
non-elite contexts. The other unusual aspect is that over half my zone D2 bird re
mains represent swan, a bird not frequently recovered, but there are no wings
represented among the swan elements present. Chmurny's assemblage also con-
Swan Prairie Goose Merganser Duck Sandhill Turkey Other

Chicken Taxa Crane
Figure 12.8 . Zone DZ birds from sub-Mound 5 1 .
348
tained a large number o f swan elements, representing 2 1 percent o f his identifi

able bird NISP. But he does not report wing elements for this assemblage. Hence,
it appears that the swans were being processed, at least in part, for their wings.
The other avian taxa I identified from zone D2 include Canada goose, com
mon merganser, mallard, and prairie chicken. Prairie chicken remains at 28 per
cent make up a relatively large proportion of the total. Chmurny's assemblage
contained 22.5 percent prairie chicken (Fig. 12. 8). Prairie chicken remains recov
ered from Mississippian occupations at Cahokia have been previously identified
only from later Stirling phase, elite contexts. They were also noted in non-elite
residential contexts at Cahokia during the Emergent Mississippian, however. It is
possible these birds increased in value for some reason or acquired some sym
bolic meaning, making them attainable only by the elite segment of the popula
tion during the Mississippian.
Goose and mallard are represented by low percentages in my assemblage (15
percent combined). However, for Chmurny's D2 assemblage, ducks and geese
make up almost 50 percent of the identified NISP (Fig. 12. 8). His sample con
tained nine other taxa of duck besides mallard, and includes pintail, gadwall,
green- and blue-winged teals, shoveler, woodduck, redhead, ring-necked or lesser
scaup, and ruddy duck. Other taxa identified include coot, sandhill crane, turkey,
and perching birds.
It is difficult to evaluate the differences in sample sizes and number of taxa
present between the bird assemblages in the present analysis and those of
Chmurny's. They could be due to significant depositional variation within the
zones, or could be a result of postexcavation factors such as mixing of materials
from various zones that may be reflected in Chmurny's sample. Nevertheless,
swan and prairie chicken make up the majority of avian remains from both as
semblages.
Fish assemblages from American Bottom sites are typically large and quite var
ied, with as many as twenty-five or more taxa represented. In contrast, the sub
Mound 51 zone D2 assemblage I examined (Fig. 12.7) is small (NISP=873), has
relatively few taxa (nine), and is represented by large individuals. This in part may
be because I analyzed only material recovered from hand excavation. It is possible
that small fish bone was missed, creating a bias toward larger fish remains. I ex
amined bulk samples that were recently dry fine-screened in the paleoethno
botany lab at Washington University, however, and although some small fish
bones were observed, they were not abundant, and in fact they could represent
the stomach contents of some of the larger fish.
The assemblage I studied consisted mostly of buffalo sucker (79 percent), fol
lowed by gar (Fig. 12.9). The remaining taxa identified from zone D2 include
members of the catfish fainily, freshwater drum, and bass. Chmurny reports a
349
Lucretia S. Kelly
Suckers Catllsh Sunfish Gar Freshwater Sturgeon

Drum
Taxa
Figure 12.9. Zone DZ fish from sub-Mound 5 1 .
lower percentage of suckers than I found, but they still represent the most abun
dant family: He noted higher percentages of bullheads, gar, and freshwater drum,
and also reported sturgeon. More taxa of fish from the sub-Mound 51 deposits
represent large river fish than backwater slough varieties. The latter are generally
more common in fish assemblages from the American Bottom, including
Cahokia.
There were no reptile or amphibian remains in any of the sub-mound pit as
semblages I analyzed. The drilled alligator tooth was the only reptile remain
Chmurny reports for zone D2.
I N T E R P R E TAT I O N O F T H E S U B - M O U N D 5 1 A S S E M B L A G E
The fauna! remains recovered from the sub-Mound 51 pit, especially those from
zone D2, differ in significant ways from other Cahokia assemblages that are in
terpreted as household refuse, elite or non-elite. Most of us involved in the analy
sis of the sub-mound material now agree that the contents were the product of
large-scale Cahokian events (Fritz 1997; L. Kelly l997a; Pauketat l997b). The
sumptuary items of crystals, exotic arrowheads, axehead debitage, and sherds
from engraved and painted vessels, listed earlier, reinforce the interpretation that
elite personages were associated with the represented activities. But non-elite
participation is indicated as well by the presence of domestic cooking vessels
(Pauketat 1997b) and the domestic variety of subsistence plants (Fritz 1997) . What
is believed to be represented here are public ritual and feasting activities attended
by all segments of the community:
350
The faunal refuse from sub-Mound 51 has many o f th e signatures proposed by

others as representing feasting. There is a relatively low taxonomic diversity; it
contains high-yielding meat species from all classes represented; bulk cuts of
meat can be identified, as well as bulk cooking, probably by boiling; little
butchering debris is present (Jackson and Scott 1995). Bones are not completely
broken or disarticulated, and bones are present in great quantities in a single de
posit (Hayden 1996) that was laid down in a short period of time.
The contents of this pit are perceived to be, in large part, the result of feasting
preparation rather than its aftermath. The material discarded in the pit may be
derived from a staging area where the different ingredients for a ritual feast were
being gathered, prepared, or manufactured. For example, the faunal remains in
dicate that deer meat was being stripped from the bones. The wings of swans
were being disarticulated, removed, and perhaps curated for future use. Large
fish were being filleted. The manufacture of quartz crystal arrowheads, basalt axe
heads, and exotic chert tools also appears to be taking place in the area as evi
denced from the recovered quartz crystal, basalt, and chert debitage. In addition,
there is evidence that paints and pigments were being mixed and painted onto ce
ramic vessels after they had been fired (Pauketat 1997b). A large number of un
carbonized squash seeds may be the result of the preparation or cleaning out of
the squash fruits. Fritz reports that squash seeds in some of the bulk samples she
examined were "stuck together and covered with material as if they represented
the contents of fruits that had been scooped out and discarded, the way most of
us treat j ack-o-lantern seeds" (1997:5) .
Chiefly involvement appears to be evident in the acquisition of some of the
animals represented. The consistent body-part representation for deer in this pit
indicates that the portion of the deer brought to the area was prescribed. This
may suggest that hunters attached to the hosting group, or members of the host
ing group themselves, procured the meat (see DuPratz in Swanton 1911) . This
suggestion may be further strengthened by the fact that basically only one kind
of mammal is present in the deposit. A variety of mammals and a more random
sample of deer parts would be expected if commoners brought in portions of
their hunt.
Seeman (1979 ), in his study of Ohio Hopewell ritual feasting, hypothesizes that
meat, particularly venison, was a vital resource in maize farming societies in the
Eastern Woodlands. As such, it may have been ritually regulated. Applying See
man's argument to the American Bottom, one might conclude that, as a result of
crop intensification and increased population, the demand for tillable and habit
able floodplain land may have diminished prime habitat areas for deer and other
wildlife, or at least pushed them farther away from Cahokia. 6 This may have in
creased the value of deer meat in particular, and enhanced its procurement and
351
Lucretia S. KeUy
consumption in ritual contexts. Helms believes hunting can be equated with

long-distance trade and skilled craft activities. Hunting is a "strongly ritualized
activity involving acquisitional and transformative dealings with outside powers
(e.g., supernatural masters of animals) and since success in hunting frequently is
taken as evidence of supernatural approval and support," it is often associated
with the elite (Helms 1992:189). Chiefly ritualization of deer hunting could possi
bly account, in part, for the large increase in deer remains at Cahokia at the be
ginning of the Mississippian.
Historically; ritual was also involved with the serving of deer. For instance, the
Menomini (Douglas 1976) used red cedar to render parts of the deer safe to eat.
Without the ritual precaution, the deer would remain supernaturally charged.
High frequencies of red cedar branchlets along with moderate amounts of to
bacco seeds were recovered from zone D2 of the sub-Mound 51 pit (Fritz 1997).
There are also numerous ethnographic references to tobacco being used for pu
rification in ritual contexts.
It is probable that the birds represented in this pit were eaten, but it is more
likely that they were primarily used for their feathers. Birds were imbued with
symbolism, being representations of the Upper World in Southeastern Indian
cosmology (Hudson 1976; Jackson and Scott 1995; Swanton 1946) . Skins and feath
ers of birds, such as mallards, cranes, herons, owls, hawks, were used for orna
mental clothing and headdresses, as well as used in ceremony (Swanton 1946).
Swan feathers seem to have been particularly important in the ritual aspect of
feasting, as ethnohistorically documented. As an example, during the Winnebago
Thunderbird feast, the host scatters swan feathers in the lodge and wears them
on his head (Radin 1990) . Swan feathers have also been mentioned as important
fu Southeast Indian culture, and in many instances they were associated with per
sons of rank (DuPratz 1972; Swanton 1946) . For instance, in an account by Adair
of a Chickasaw green corn ceremony (DePratter 1991: 62), he describes the rekin
dling of the sacred fire. The priest selected special wood chips and used a fire drill
to start the fire. The flame was fanned by a swan's wing. DePratter believes when
chiefdoms were still in existence, it would have been the chiefs who would have
rekindled the fire because of their descent from the sun.
Unlike the mammals, a variety of bird species is represented in the sub-Mound
51 pit, especially in Chmurny's assemblage. This may indicate that the types of
birds furnished for the feast were not prescribed, different segments of the popu
lation may have procured them, or a variety of symbolic meanings were attached
to a wide range of bird species.
Fish are not generally considered to be animals of as high value as hunted ani
mals (Kent 1989), but this would vary regionally (for instance, fish were highly
352
valued in Northwest Coast societies). Rees (1997) found fish were rated highly
enough to be given as tribute ethnohistorically in the central Mississippi river val
ley region of the interior Southeast. Larger fish (Scott and Jackson 1995) and
those difficult to capture may have had increased value. In the sub-Mound 51 pit,
the remains of larger-sized river fish that are more difficult to catch were recov
ered in much higher quantities than the smaller, more easily obtainable backwa
ter slough varieties.
Seeman (1979) claims meat was the principal food consumed at mortuary
feasts in Hopewell times. Unfortunately, ethnohistoric accounts of feasts in the
Southeast generally do not state what foods were served. It would seem logical
that different foods might be served at different feasts depending on the season of
the year the feast was held (see DuPratz 1972 for description of monthly feasts
held among the Natchez). Deer would have been the primary source of meat at
the feast(s) represented by the sub-Mound 51 faunal assemblage. The season of
the year represented by the faunal remains in this pit can only be suggested at this
time. Tooth eruption cannot be used, because no skulls or mandibles were recov
ered, and epiphyseal fusion cannot give the fine-grained information needed
(Purdue 1983) . The availability of deer was at its peak in the fall and early winter
(September-December; Rue 1997; B. Smith 1975). Swans and other waterfowl
would be at their peak abundance during the spring migrations in March and fall
migrations in October-November (Bent 1962) . The fall swan migration and
prime deer-hunting season would have coincided. The bones were deposited dur
ing warm weather because large numbers of insects were also recovered. Based
on this information, it is most likely that the feasting activity represented by the
faunal remains from zone D2 of sub-Mound 51 took place in the early fall before
the weather turned cold. Late summer and fall would correspond to the time
Swanton (1946:256) indicates the greatest feasts and ceremonies were held in the
Southeast. These would include feasts of the first fruits and would be a time
when the maximum amount of food would be available.
The extraordinary preservation of materials from the sub-Mound 51 pit has
allowed a rare opportunity to recover many ingredients resulting from a large
public gathering at Cahokia. The contents and context of this pit provide one of
the most unequivocal data sets for ritual feasting recovered from a Mississippian
site. The faunal remains indicate that food played a very important part in the
event(s) represented, and that many of the animals were used because of the
symbolic meaning they carried. This may also be true for the other material
classes recovered. 7 The pit contents seem to reflect events orchestrated by hosts
of elevated social status to convey particular messages to the attendees, includ
ing commoners.
353
Lucretia S. Kelly
F EAST I N G : A M EC H A N I S M O F T R I B U T E
A N D S O C I A L I N T E G R AT I O N
Ritual and feasting have a long history in the Eastern Woodlands, dating back to
at least the Middle Woodland period (rno B.c .-300 A.D. ; Blitz 1993; Seeman 1979;
Styles and Purdue 1991) . Blitz (1993) has very effectively demonstrated how both
ritual and feasting were manifested in Mississippian chiefdoms to the south of
Cahokia, but feasts have not been emphasized in sociopolitical development at
Cahokia (Emerson 1997; ]. Kelly 1996). The sub-Mound 51 material allows us to
focus our attention on the role large, public ritual feasting events may have played
in Cahokia's history.
Ritual feasting can be considered a mechanism by which food was mobilized to
the main center (DePratter 1991), and it adds a new dimension to discussions of trib
ute. It must be understood, however, that feasts were not the only mechanism in
volved in food tribute, and the functions of feasts go far beyond food mobilization.
It has been shown that the feast represented by the sub-Mound 51 material was
a public event with both commoners and higher-ranked individuals in atten
dance. This would indicate that the chiefs as hosts were (re)distributing food to
the populace, rather than being the sole beneficiary of it. It is also evident that at
least some of the foods (e.g., deer) were furnished or prescribed by the chiefly
class. Whether the distribution or "sharing" of the food was equitable cannot be
determined. Because chiefs were hosting the event(s), based on the location of
the pit, and may have been determining what was being served, it could not be
termed a purely "potluck'' affair (Muller 1997) . Therefore, this form of tribute
collection would seem to fall somewhere between the two ends of the tributary
scale-equitable reassignment, and enforced appropriation of food-and, thus,
between the perspectives expressed by proponents of the top-down and bottom
up schools of Mississippian political economy. Obviously; both schools of
thought could incorporate feasts into their formulations. There were most likely
a multitude of different feasting events, big and small, taking place throughout
year in the Cahokia region. For this discussion, only the large public-type feast
represented by the empirical evidence from sub-Mound 51 is considered.
Feasts of this type would have solved several of the problems previously men
tioned in regards to food being part of a tributary system. They would have af
forded a reason to transship large quantities of food to the center in a short
period of time. This would solve the spoilage problem for meat. If grain was not
to be stored for long periods of time, granaries may not have been built, at least
ones that could be detected archaeologically. To mobilize the labor force for the
collection and preparation of the amount of food necessary to feed a large public
gathering, superior managerial ability as dictated by tradition would be needed
by the hosts. The feast would therefore serve as a means for the hosting individ-
354
R I T U A L F EA S T I N G AT T H E C A H O K I A S I T E
ual o r group t o gain (or maintain) prestige by displaying managerial skills and
economic superiority. Thus, large, public ritual feasts such as that represented by
sub-Mound 51, served functions besides food mobilization that define broader so
cial articulations within the developing Cahokian society.
Feasts would also be a venue whereby individuals or groups sanctified their po
sitions within the population at large by establishing their connections to gods or
ancestors (Dietler 1996; Friedman 1984) . Social change could be achieved through
the ritual and ideological realm by consensus, a feature of kin-based societies
such as Cahokia (Blitz 1993). The top and bottom segments of society, as well as
those residing at the mound centers and at outlying communities, were linked by
kinship ties, a point often forgotten when political economy is discussed. As
noted by Knight (1990:19), commoners and nobles were tied together by kinship
in webs of rights and obligations. Kinship rights and obligations could be, at least
partially; operationalized through ritual feasts.
The spatial layout of Mississippian Cahokia reflects the way in which the soci
ety was organized socially; politically; and ideologically. It is an elaboration or em
bellishment of principles that extend back to the Emergent Mississippian, when
the beginnings of social ranking can be seen (J. Kelly 1990). The early plan for Ca
hokia utilized the cardinal directions and incorporated the principles of central
ity, quadrilateralism, and dualism that were so important to the Mississippian
worldview (J. Kelly 1996) . The construction of Monks Mound and the central rit
ual spaces (plazas) flanked by smaller mounds is closely linked with the ideology
of the dominant social group. It is argued that leading kinsmen, to sanctify their
elevated social and ideological position, co-opted a long-standing sacred symbol,
the mound, by placing their structures on top of them (Knight 1986; Steponaitis
1986:386) . The public plazas next to these mounds served as focal points or arenas
for public gatherings including chief-sponsored ritual and feasting activity (Blitz
1993; Smith and Williams 1994). The deep-rooted sacredness of these areas can be
observed in the maintenance of these areas over many generations.
The labor needed for mound building could have been effectively mobilized
through feasts. The mounds at Cahokia were enlarged on a continuous and
seemingly periodic basis (Knight 1986; Pauketat 1993), and therefore may not have
required a large, centrally coordinated labor force (Dietler 1996; Lindauer and
Blitz 1997; Muller 1997) . Mound enlargement was probably accomplished in con
nection with rites of renewal and intensification and chiefly succession (Blitz
1993; Knight 1986). Labor for large-scale projects, such as the leveling of the
Grand Plaza, may also have been coordinated through feasts. The resulting mon
ument or sacred space, beyond having immense cultural significance, would be
"a conspicuous advertisement of the scale of labor capable of being mobilized"
(Dietler 1996:105).
355
Lucretia S. Kelly
Through ritual and competitive feasting (Hayden 1990, 1996), ranks of social
inequity within Mississippian Cahokia may have been formalized. One lineage of
the American Bottom population may have been able, over time, to accumulate
symbolic capital and social credit (Dietler 1996) substantial enough for it to affect
the behavior of the rest of the population (Drennan 1993). Blitz argues that ritual
acts can support the status quo or can serve as a force for social change. "Individ
uals do not merely react to the social idiom of ritual but continually create, alter,
reinterpret, or manipulate its information content with dramatic results" (Blitz
1993:23). This would have been a noncoercive but very effective way of establish
ing who was in charge and why, and a means by which all could participate and
benefit. In this latter respect, ritual feasts would have a socially integrating effect
by promoting community solidarity and achieving a more homogenous form of
Cahokian Mississippian.
CO N C L U S I O N
The sub-Mound 51 material provides the best evidence to date for the occurrence
of large ritual feasts at Cahokia. Quantities of deer bone representing uniform
bulk cuts of meat in conjunction with sumptuary items of crystals, engraved and
painted pottery sherds, finely made chert projectiles, and axehead debris com
posed part of the debris recovered from the pit located within the sacred precinct
of the site. These materials indicate elite involvement, if not actual hosting, of a
major feasting event. But, the sheer volume of debris along with domestic cook
ing vessels and common plants indicate the attendance of the general populace.
The ritual activity that accompanied the feast included the use of swans, tobacco,
red cedar, and squash.
The specific event cannot be determined, but it appears to have occurred in the
late summer or early fall. Historically, within the southeastern United States,
feasts and ceremonies connected with food harvesting and world renewal such as
the widespread Green Corn Ceremony (Swanton 1946; Witthoft 1949) were held
at this time of the year. The sub-Mound 51 material represents only one type of
communal ceremonial event. It is presumed that many types of large community
based and smaller household rituals and feasts would have taken place at Cahokia
and the surrounding region. It is important that other examples are identified ar
chaeologically. Different types should have different archaeological signatures
(see Jackson and Scott 1995).
The roles of large ritual feasts in early Cahokian society may not be able to be
specifically ascertained based on current data and methods, but several possibili
ties can be offered based on our knowledge of how feasts functioned ethnohis
torically in the southeastern United States and ethnographically in other parts of
the world. It can be hypothesized that at Cahokia, large feasting events, such as
356
that represented b y the sub-Mound SI material, were a means t o integrate the Ca
hokia community or region. Evidence suggests they were more than a "potluck"
gathering of the general populace (sensu Milner and Muller) . Large public feast
ing events likely provided articulation and coordination between the center and
outlying communities and between the various ranked segments of the popula
tion. It appears Cahokia and its hinterland were not totally autonomous nor was
the hinterland directly dominated by Cahokia at this time.
It can be further postulated that highly ranked individuals may have been at
tempting to solidify their more favorable and influential roles within a prescribed
social order through the hosting of large, conspicuous ritual feasts. Whether con
scious negotiation of a new Cahokian political order (sensu Pauketat I997b) was
taking place is difficult to determine. However over time, the more prominent
Cahokian lineages that had been able to accrue a large amount of social credit
and capital (sensu Dietler I996) from past ritual, feasting events may have been
able to advantageously mobilize the growing Lohmann phase population for
mound building, plaza leveling, and the provisioning of large quantities of food,
thus effecting changes and elevating the Cahokian sociopolitical system to a
more complex level. A continued effort to identify and study remains from other
feasting events should allow this interpretation of political economy and social
relations at Cahokia to be refined.
Discussions of ritual activity for the American Bottom region have recently be
come more frequent (Emerson I997; Kelly I996), however, because evidence for
large ritual feasts has not been previously identified at Cahokia, the role of such
socioreligious activity has not been a major component of the sociopolitical mod
els proposed for the site. The sub-Mound SI material now provides evidence for a
mechanism, ritual feasting, that helps explain some of the dynamics associated
with the sociopolitical system of early Cahokia. It draws our attention away from
the extremes previously characterized as the top-down and bottom-up models, to
a more central position and one more in line with the community-based manner
in which many Native American societies functioned historically. Rituals and
feasts have been and still are an integral part of Native American society.
AC K N OW L E D G M E N TS
I would like t o thank Tim Pauketat, University of Illinois, fo r including me in his NSF
funded Early Cahokia Project. The opportunity to study a faunal assemblage such as tbat
from sub-Mound 51 is a zooarchaeologist's dream. I would like to thank Karli White and
Terry Martin of tbe Illinois State Museum for giving me access to the comparative osteol
ogy collections in order to identify some of the specimens. I would like to acknowledge
tbe Cahokia Mounds Museum Society for providing support for part of tbe sub-Mound 51
faunal analysis. John Kelly graciously gave of his time to draft two of tbe figures. Several
357
Lucretia S. Kelly
people, including Fiona Marshall, Gayle Fritz, Patty Jo Watson, David Browman, and John
Kelly; all of Washington University; read and gave constructive criticisms of various drafts
of this paper. I would also like to thank the anonymous reviews for their helpful sugges
tions. I am, however, solely responsible for any mistakes or inaccuracies.
N OT E S
I . Support for the sub-Mound 51 portion of Tim Pauketat's Early Cahokia Project was
provided by National Science Foundation (SBR-9305404). Support was also provided to
the author for further faunal analysis of the sub-Mound 51 remains by the Cahokia
Mounds Museum Society.
2. Preliminary analytical results were presented at the 54th Southeastern Conference in
Baton Rouge, LA, Nov. 5-8, 1997 in a symposium entitled "New Evidence of Early
Cahokian Provisions and Rituals" and a co-authored journal article is forthcoming.
3. The study of body-part distribution must be undertaken cautiously. Many factors,
cultural and noncultural, can play roles in what body parts form an assemblage (Klein
1989; Lyman 1984; Marshall and Pilgrim 1991). Noncultural taphonomic factors such as
weathering, trampling, and animal gnawing do not appear to have significantly
affected the composition of the Cahokia assemblages, however. The consistent
pattern at Cahokia throughout its occupational history is a high proportion of those
parts of the deer that have low structural density (i.e., are thin and fragile such as
scapulae) and a very low proportion of some bones (e.g. , metapodials) with high
structural density; a pattern that is not density-mediated. Therefore, a cultural
explanation is more likely for most of the assemblages' compositions.
4. As a convenient way of illustrating the relationship between food utility and deer
body part distribution, I have employed Purdue's (Purdue, Styles, and Masulis 1989)
adjusted Food Utility Index (FUI) categories of high, mid, and low. These are based on
Metcalfe and Jones's (1988) continuous FUI variable but emphasize the extremes of
the FUI spectrum where human behavior may be less variable. Figure 12.4 illustrates
which portions of the deer are considered to belong to the low, mid, and high FUI
categories.
5. Deer would not have been overly abundant in the floodplain near Cahokia because of
lack of preferred habitat. The area around Cahokia was mostly wetlands and prairie
(Lopinot 1991; White et al. 1984). The highest density of deer would have occurred in
the bluff edge zone and the adjacent uplands that contained an extensive oak/hickory
woodland zone (Halls 1984; B. Smith 1975).
6. It could be argued, however, that the increasing land under crop cultivation would
increase edge habitats that might attract more deer.
7. Because several types of squashes are represented by many seeds (>3,000) throughout
the sub-Mound 51 deposit, Fritz (1997) believes squash may have been a ritually
significant plant. This evidence lends support to an earlier interpretation (Emerson
1982; Prentice 1986) that squash (cucurbits) may have had particular significance in
fertility-related rituals. This earlier interpretation is based on a Mississippian, Stirling
phase figurine recovered from the BBB Motor site. This figurine depicts a woman
358
kneeling o n the back o f a serpent with a hoe embedded in its back. The tail o f the
serpent splits into squash vines and fruits identified as Cucurbita argyrosperma (Fritz
1994) that climb the woman's back.
REFERENCES
Anderson, D. G.
1994 The Savannah River Chiefdoms: Political Change in the Late Prehistoric Southeast.
Tuscaloosa: University of Alabama Press.
Bareis, C. J.
1975 Report of 1971 University of Illinois-Urbana Excavations at the Cahokia Site.
In Cahokia Archaeology: Field Reports, pp. 9-11. Papers in Anthropology No. 3.
Springfield: Illinois State Museum.
Bareis, C . J. , and]. W Porter, eds.
1984 American Bottom Archaeology. Urbana: University of Illinois Press.
Barker, A. W
1992 Powhatan's Pursestrings: On the Meaning of Surplus in a Seventeenth Cen
tury Algonkian Chiefdom. In Lords of the Southeast: Social Inequity and the Na
tive Elites of Southeastern North America, edited by A. W Barker and T. R.
Pauketat, pp. 61-80. Archaeological Papers No 5. Washington, D.C . : Ameri
can Anthropological Association.
Barker, A. W, and T. R. Pauketat, eds.
1992 Lords of the Southeast: Social Inequity and the Native Elites of Southeastern North
America. Archaeological Papers Papers No 5. Washington, D.C . : American
Anthropological Association.
Bent, A.
1962 Life Histories of North American Wild Fowl, Part II. New York: Dover
Publications.
Blitz, J.
1993 Ancient Chiefdoms of the Tombigbee. Tuscaloosa: University of Alabama Press.
Buikstra, J. E., J. C. Rose, and G. R. Milner
1994 A Carbon Isotope Perspective on Dietary Variation in Late Prehistoric West
ern Illinois. In Agricultural Origins and Development in the Midcontinent, edited
by W Green, pp. 155-170. Report No. 19. Iowa City: Office of the State
Archaeologist.
Carniero, R. L.
1981 The Chiefdom: Precursor to the State. In The Transition to Statehood in the
New World, edited by G. D. Jones and R. R. Kautz, pp. 37-79. Cambridge:
Chmurny; W W
1973 The Ecology of the Middle Mississippian Occupation of the American Bot
tom. Unpublished Ph.D. dissertation, Department of Anthropology; Univer
sity of Illinois, Champaign-Urbana.
359
Lucretia S. Kelly
Clayton, L. A., V. J. Knightjr., and E. C. Moore, eds.

1993 The De Soto Chronicles: The Expedition of Hernando De Soto to North America in
153!)-1543. Tuscaloosa: University of Alabama Press.
Collins, ]. M.
1990 The Archaeology of the Cahokia Mounds JCT-II: Site Structure. Illinois Cultural
Resources Study rn. Springfield: Illinois Historic Preservation Agency.
1997 Cahokia Settlement and Social Structures as Viewed from the ICT-II. In Ca
hokia: Domination and Ideology in the Mississippian World, edited by T. R.
Pauk:etat and T. E. Emerson, pp. 124-140. Lincoln: University of Nebraska
Press.
Dalan, R. A.
1997 The Construction of Mississippian Cahokia. In Cahokia: Domination and Ide
ology in the Mississippian World, edited by T. R. Pauk:etat and T. E. Emerson,
pp. 89-102. Lincoln: University of Nebraska Press.
Dalan, R. A., W. Watters Jr., G. R. Holley, and W. I. Woods
1994 Sixth Annual Cahokia Mounds Field School: Understanding Mound
Construction. Office of Contract Archaeology, Southern Illinois University
at Edwardsville. Submitted to Illinois Historic Preservation Agency,
Springfield.
DePratter, C. B.
1991 Late Prehistoric and Early Historic Chiefdoms in the Southeastern United States.
New York: Garland Press.
Dietler, M.
Perspective, edited by P. Wiessner and W. Schiefenhi:ivel, pp. 87-I26. Oxford:
Berghahn Books.
Douglas, ]. G.
1976 Collins: A Late Woodland Ceremonial Complex in the Woodfordian North
east. Unpublished Ph.D. dissertation, Department of Anthropology, Univer
sity of Illinois, Champaign-Urbana.
Drennan, R.
1983 Ritual and Development at the Early Village Level. In The Cloud People: The
Divergent Evolution of the Zapotec and Mixtec Civilizations, edited by K. V. Flan
nery and]. Marcus, pp. 46-so. New York: Academic Press.
Du Pratz, L. P.
1972 The History of Louisiana or of the Western Parts of Virginia and Carolina: Con
[1754] taining a Description of the Countries that Lie on both Sides of the River Missis
sippi: With an Account of the Settlements, Inhabitants, Soil, Climate, and Products.
Reprint. Baton Rouge: Claitor's Publishing Division.
Elias, S., and T. R. Pauk:etat
1997 The Paleoentomology of Cahokia's Sub-mound 51 Pit. Paper presented at
the 54th Southeastern Archaeological Conference, Baton, Rouge, LA.
360
Emerson, T. E.
1982 Mississippian Stone Images in Illinois. Circular No. 6. Urbana: Illinois Archaeo
logical Survey.
1995 Settlement, Symbolism, and Hegemony in the Cahokian Countryside. Un
published Ph.D. dissertation, Department of Anthropology, University of
Wisconsin, Madison.
1997 Cahokia and the Archaeology of Power. Tuscaloosa: University of Alabama
Press.
Esarey, D., and T. R. Pauketat
1992 The Lohmann Site: An Early Mississippian Center in the American Bottom. Ur
bana: University of Illinois Press.
Finney, F. A.
1993 Spatially Isolated Structures in the Cahokia Locality: Short-term Residences
or Special Purposes Shelters? In Highways to the Past: Essays in Honor of
Charles ]. Bareis, edited by T. E. Emerson, A. Fortier, and D. McElrath,
pp. 381-392. Illinois Archaeology, Volume 5 (1 and 2). Urbana: Illinois Archae
ological Society.
Fowler, M. L.
1973 The Cahokia Site. In Explorations into Cahokia Archaeology, edited by M. L.
Fowler, pp. l-30. Illinois Archaeological Survey Bulletin 7, Urbana.
1974 Cahokia: Ancient Capital of the Midwest. Addison-Wesley Module in Anthro
pology, No. 48.
1975 A Pre-Columbian Urban Center on the Mississippi. Scientific American 233 (2):
92-ror.
1978 Cahokia and the American Bottom: Settlement Archaeology. In Mississippian
Settlement Patterns, edited by B. D. Smith, pp. 455-478. New York: Academic
Press.
1989 The Cahokia Atlas: A Historical Atlas of Cahokia Archaeology. Studies in Illinois
Archaeology No. 6. Springfield: Illinois Historic Preservation Agency.
Friedman, ].
Fritz, G. J.
1994 Precolumbian Cucurbita argyrosperma ssp. argyrosperma (Cucurbitaceae) in the
Eastern Woodlands of North America. Economic Botany 48 (3): 280-292.
1997 Special Plants from Early Cahokia: Deposits from Sub-Mound 5r. Paper pre
sented at the 54th Southeastern Archaeological Conference, Baton Rouge,
LA.
Fritz, G. ]. , and S. Johannessen
1996 Social Differentiation in the American Bottom: Late Prehistoric Plant Re
mains from Household, Communal, and Ceremonial Contexts. Paper pre
sented at the 61st Annual Meeting of the Society for American Archaeology,
New Orleans.
361
Lucretia S. Kelly
Halls, L. K., ed.

1984 White-tailed Deer Ecology and Management. Harrisburg, Pa. : Stackpole Books.
Hayden, B.
tion. Journal of Anthropological Archaeology 9:31-69.
An Interdisciplinary Perspective, edited by P. Wiessner and W. Schiefenhavel,
Helms, M. W.
1992 Political Lords and Political Ideology in Southeastern Chiefdoms: Comments
and Observations. In Lords of the Southeast: Social Inequity and the Native Elites
of Southeastern North America, edited by A. W. Barker and T. R. Pauketat,
pp. 185-194. Archaeological Papers of the American Anthropological Associa
tion No. 5.
Holley, G. R. , R. A. Dalan, and P. A. Smith
1993 Investigations in the Cahokia Site Grand Plaza. American Antiquity 58 :306-318.
Hudson, C.
1976 The Southeastern Indians. Knoxville: University of Tennessee Press.
Jackson, H. E., and S. Scott
1995 The Faunal Record of the Southeastern Elite: The Implications of Economy,
Social Relations, and Ideology. Southeastern Archaeology l4 (2): 103-n9.
Johannessen, S.
1984 Paleoethnobotany. In American Bottom Archaeology, edited by C. J. Bareis and
]. W. Porter, pp. 197-214. Urbana: University of Illinois Press.
1988 Plant Remains and Culture Change: Are Paleoethnobotanical Data Better
than We Think? In Current Paleoethnobotany, edited by C. A. Hastorf and V. S.
Popper, pp. 145-166. Chicago: University of Chicago Press.
1993 Food, Dishes, and Society in the Mississippi Valley. In Foraging and Farming in
the Eastern Woodlands, edited by C . M. Scarry, pp. 182-205. Gainesville: Uni
versity Press of Florida.
Kelly, ]. E.
1990 The Emergence of Mississippian Culture in the American Bottom Region. In
The Mississippian Emergence, edited by B. D. Smith, pp. n3-152. Washington,
1992 The Impact of Maize on the Development of Nucleated Settlements: An
American Bottom Example. In Late Prehistoric Agriculture: Observations from
the Midwest, edited by W. I. Woods, pp. 167-197. Studies in Illinois Archaeol
ogy No. 8. Springfield: Illinois Historic Preservation Agency.
1996 Redefining Cahokia: Principles and Elements of Community Organization.
The Wisconsin Archeologist 77 (3 /4): 97-n9.
1997 Stirling Phase Sociopolitical Activity at East St. Louis and Cahokia. In Ca
hokia: Domination and Ideology in the Mississippian World, edited by T. R.
Pauketat and T. E. Emerson, pp. 141-166. Lincoln: University of Nebraska
Press.
362
Kelly, L. s .
1979 Animal Resource Exploitation by Early Cahokia Populations on the Merrell Tract.
Circular 4, Illinois Archaeological Survey, Department of Anthropology, Uni
versity of Illinois, Urbana.
1990 Range Phase Faunal Analysis. In The Range Site 2: The Emergent Mississippian
Dohack and Range Phase Occupations, by J. E. Kelly, S. J. Ozuk, and J. A.
Williams. American Bottom Archaeology, FAI-270 Site Report 20. Urbana:
University of Illinois Press.
l997a Lohmann Phase Faunal Provisioning at the Cahokia Site. Paper presented at
the 54th Southeastern Archaeological Conference, Baton Rouge, LA.
l997b Patterns of Faunal Exploitation at Cahokia. In Cahokia: Domination and Ideol
ogy in the Mississippian World, edited by T. R. Pauketat and T. E. Emerson,
pp. 6!}---88 . Lincoln: University of Nebraska Press.
Kelly, L. S., and P. G. Cross
1984 Zooarchaeology. In American Bottom Archaeology, edited by C. J. Bareis and
J. W. Porter, pp. 215-232. Urbana: University of Illinois Press.
Kent, S.
1989 Cross-cultural Perceptions of Farmers as Hunters and the Value of Meat. In
Farmers as Hunters: The Implications of Sedentism, edited by S. Kent, pp. l-17.
Klein, R. G.
1989 Why Does Skeletal Part Representation Differ between Smaller and Larger
Bovids at Klasies River Mouth and Other Archaeological Sites? journal of Ar
chaeological Science 6:363-38I.
Knight, V. J. , Jr.
1986 The Institutional Organization of Mississippian Religion. American Antiquity
51 (4): 67y--687.
1990 Social Organization and the Evolution of Hierarchy in Southeastern Chief
doms. Journal of Anthropological Research 40 (1): l-23.
Lindauer, 0. , and). H. Blitz
1997 Higher Ground: The Archaeology of North American Platform Mounds.
journal of Archaeological Research 5 (2): 16!}---207.
Lopinot, N. H.
1991 Archaeobotanical Remains. In The Archaeology of the Cahokia Mounds ICT-II:
Biological Remains, by N. H. Lopinot, L. S. Kelly, G. R. Milner, and R. Paine.
Illinois Cultural Resources Study 13. Springfield: Illinois Historic Preservation
Agency.
1992 Spatial and Temporal Variability in Mississippian Subsistence: The Archaeo
botanical Record. In Late Prehistoric Agriculture Observation from the Midwest,
edited by W. I. Woods, pp. 44-94. Studies in Illinois Archaeology No. 8 .
Springfield: Illinois Historic Preservation Agency.
1994 A New Crop of Data on the Cahokian Polity. In Agricultural Origins and De
velopment in the Midcontinent, edited by W. Green, pp. 127-154. Report No. 19,
Office of the State Archaeologist, University of Iowa, Iowa City.
363
Lucretia S. Kelly
1997 Cahokian Food Production Reconsidered. In Cahokia: Domination and Ideol

ogy in the Mississippian World, edited by T. R. Pauketat and T. E. Emerson,
pp. 52--68. Lincoln: University of Nebraska Press.
Lopinot, N. H., and W I. Woods
1993 Wood Overexploitation and the Collapse of Cahokia. In Foraging and Farming
in the Eastern Woodlands, edited by C. M. Scarry, pp. 206-23r. Gainesville: Uni
versity Press of Florida.
Lyman, R. L.
1984 Bone Density and Differential Survivorship of Fossil Classes. journal of An
thropological Archaeology 3:259-299.
Marshall, F. , and T. Pilgrim
1991 Meat versus Within-bone Nutrients: Another Look at the Meaning of Body
Part Representation in Archaeological Sites. journal of Archaeological Science
18:149-163.
Mehrer, M. W
1988 The Settlement Pattern and Social Power of Cahokia's Hinterland House
holds. Unpublished Ph.D. dissertation, Department of Anthropology, Uni
versity of Illinois, Champaign-Urbana.
1995 Cahokia's Countryside: Household Archaeology, Settlement Patterns, and Social
Power. DeKalb: Northern Illinois University Press.
Metcalfe, D. , and K T. Jones
1988 A Reconsideration of Animal Body Part Indices. American Antiquity 53 (3):
486--504.
Milner, G. R.
1990 The Late Prehistoric Cahokia Cultural System of the Mississippi River Val
ley: Foundations, Florescence, and Fragmentation. journal of World Prehistory
4 (1): l-43.
1998 The Cahokia Chiefdom: The ArchaeoloK)I of a Mississippian Society. Washington,
Muller, ].
1997 Mississippian Political Economy. New York: Plenum.
Pauketat, T. R.
1991 The Dynamics of Pre-State Political Centralization in the North American
Midcontinent. Unpublished Ph.D. dissertation, Department of Anthropol
ogy, University of Michigan, Ann Arbor.
1993 Temples for Cahokia's Lords: Preston Holder's 1955-1956 Excavations of the Kunne
mann Mound. Memoirs No. 26. Ann Arbor: University of Michigan Museum
of Anthropology.
1994 The Ascent of Chiefs: Cahokia and Mississippian Politics in Native North America.
Tuscaloosa: University of Alabama Press.
l997a Cahokian Political Economy: In Cahokia: Domination and Ideology in the Mis
sissippian World, edited by T. R. Pauketat and T. E. Emerson, pp. 30--5 r. Lin
coln: University of Nebraska Press.
364
l997b New Evidence o f Early Cahokian Provisions and Rituals. Paper presented at
the 54th Southeastern Archaeological Conference, Baton Rouge, LA.
Pauketat, T. R., G. ]. Fritz, L. S. Kelly, and N. H. Lopinot
1993 Early Cahokia: A New Research Project in the American Bottom. Paper pre
sented at the 1993 Midwest Archaeological Conference, Milwaukee, WI.
Pauketat, T. R., and N. H. Lopinot
1997 Cahokian Population Dynamics. In Cahokia: Domination and Ideology in the
Mississippian World, edited by T. R. Pauketat and T. E. Emerson, pp. ro3-123.
Lincoln: University of Nebraska Press.
Pauketat, T. R., and M. A. Rees
1996 Early Cahokia Project 1994 Excavations at Mound 49, Cahokia (n-S-34-2),
Early Cahokia Proj ect Papers, No. 2. Submitted to the Illinois Historic
Preservation Agency, Springfield.
Prentice, G.
1986 An Analysis of the Symbolism Expressed by the Birger Figurine. American An
tiquity 51:239-266.
Purdue, ]. R.
1983 Epiphyseal Closure in White-tailed Deer. journal of Wildlife Management 47
(4): 1207-1213.
Purdue, J. R. , B. W Styles, and M. C. Masulis
1989 Faunal Remains and White-tail Deer Exploitation from a Late Woodland Up
land Encampment: The Boschert Site (23SC609), St. Charles County, Mis
souri. Midcontinentaljournal of Archaeology I4 (2): 146-163.
Radin, P.
1990 The Winnebago Tribe. Lincoln: University of Nebraska Press.
Rees, M. A.
1997 Coercion, Tribute, and Chiefly Authority: The Regional Development of
Mississippian Political Culture. Southeastern Archaeology 16 (2): n3-133.
Rindos, D., and S. Johannessen
1991 Human-Plant Interactions and Culture Change in the American Bottom. In
Cahokia and the Hinterlands: Middle Mississippian Cultures of the Midwest, ed
ited by T. E. Emerson and R. B. Lewis, pp. 35-45. Urbana: University of Illi
nois Press.
Rue, L. L.
1997 The Deer of North America. New York: Lyons and Burford.
Saitta, D. J.
1994 Agency, Class, and Archaeological Interpretation. journal of Anthropological
Archaeology 13:201-227.
Seeman, M. F.
1979 Feasting with the Dead: Ohio Hopewell Charnel House Ritual as a Context
for Redistribution. In Hopewell Archaeology: The Chillicothe Conference, edited
by D. S. Brose and N. Greber, pp. 39-46. Kent, Ohio: Kent State University
Press.
365
Lucretia S. Kelly
Service, E. R.
1962 Primitive Social Organization: An Evolutionary Perspective. New York: Random
House.
Simon, M., and K. Parker
1995 Detours and Divergences on the Pathway of Prehistoric Plant Exploitation.
Paper presented at the 6oth Annual Meeting of the Society for American Ar
chaeology, Minneapolis, Minnesota.
Smith, B. D.
1975 Middle Mississippi Exploitation of Animal Populations. Anthropology Papers
No. 57. Ann Arbor: University of Michigan Museum of Anthropology.
Smith, M. T. , and D. ]. Hally
1992 Chiefly Behavior: Evidence from Sixteenth Century Spanish Accounts. In
Lords of the Southeast: Social Inequity and the Native Elites of Southeastern North
America, edited by A. W Barker and T. R. Pauketat, pp. 99-IIO. Archaeologi
cal Papers No 5. Washington, D.C . : American Anthropological Association.
Smith, M. T., and M. Williams
1994 Mississippian Mound Refuse Disposal Patterns and Implications for Archaeo
logical Research. Southeastern Archaeology 13 (1): 27-35.
Steponaitis, V P.
1978 Location Theory and Complex Chiefdoms: A Mississippian Example. In Mis
sissippian Settlement Patterns, edited by B. D. Smith, pp. 417-453. New York:
Academic Press.
1986 Prehistoric Archaeology in the Southeastern United States, 1970-1985. Annual
Reviews of Anthropology 15:363-404.
Styles, B. W, and]. R. Purdue
1991 Ritual and Secular Use of Fauna by Middle Woodland Peoples in Western Illi
nois. In Beamers, Bobwhites, and Bluepoints: Tributes to the Career of Paul W. Par
malee, edited by ]. R. Purdue, W E. Klippel, and B. W Styles, pp. 421-436.
Illinois State Museum Scientific Papers, Vol. XXIII and The University of
Tennessee, Department of Anthropology Report of Investigations No. 52.
Springfield: Illinois State Museum.
Swanton, ]. R.
l9II Indian Tribes of the Lower Mississippi Valley and Adjacent Coast of the Gulf of
Mexico. Bulletin 43. Washington, D.C . : Bureau of American Ethnology.
1946 Indians of the Southeastern United States. Bulletin 137· Washington, D.C . : Bu
reau of American Ethnology.
Welch, P. D.
1991 Moundville's Economy. Tuscaloosa: University of Alabama Press.
Welch, P. D., and C. M. Scarry
1995 Status Related Variation in Foodways in the Moundville Chiefdom. American
Antiquity 60 (3): 397-419.
White, W P., S. Johannessen, P. G. Cross, and L. S. Kelly
1984 Environmental Setting. In American Bottom Archaeology, edited by C. ]. Bareis
and]. W Porter, pp. 15-33. Urbana: University of Illinois Press.
366
R I T U A L F EA S T I N G AT T H E C A H O K I A S I T E
Witthoft, ).
1949 Green Corn Ceremonialism in the Eastern Woodlands. Occasional Contributions
No. 13. Ann Arbor: University of Michigan Museum of Anthropology.
Zeder, M.
1996 Zooarchaeological Approaches to Complexity: A View from the Old World.
Paper presented at the 53rd Annual Meeting of the Southeastern Archaeolog
ical Conference, Birmingham, Alabama.
367
FEASTING O N THE P ERI P HERY
T H E P R O D U CT I O N O F R I T U A L F EA S T I N G A N D V I L LA G E
F E S T I VA L S AT T H E C E R E N S I T E , E L S A LVA D O R
Linda A. Brown
And they [th e s ixteenth-ce ntury Yucatec Maya] often spend on one banq uet what they have
earned by tra d i n g and barg a i n i n g many days. And they h ave two ways of celebrating these
feasts: the first, which is that of the nobles and of the principal people, obl iges each one of the
i nvited guests to g ive a n other similar feast. And to each g u est they give a roasted fowl, b read
and d r i n k of cacao i n a b u n d a nce; and at the end of the repast, they were accustomed to give a
manta [cloth] to each to wea r, a n d a l ittle sta n d a n d vessel, as beautifu l as possible. And if one
of the g uests s h o u l d d i e, h i s household or h i s relations a re obliged to repay the i nvitation. The
seco n d way of giving feasts was used among kinfolk when they ma rry their c h i l d ren or cele
brate the memory of the deeds of their a n cestors, and this does n ot oblige the g u ests to give
a feast in return, except if a h u n d red person s h ave i nvited an I n d i a n to a feast, he a lso i nvites
them a l l when he gives a banq uet or ma rries his c h i l d ren. They h ave strong fri e n d s h i p a n d
they remember for a long t i m e these i nvitations, a lthough they a re fa r apart from one a n other.
Bishop Diego de Landa (Tozzer 1 94 1 :92)
In the Maya region, feasting was a pivotal component of rituals for the elite and
non-elite alike. As noted by Bishop Diego de Landa, elite Maya rulers engaged in
a form of competitive feasting and gift-giving with strict understandings for re
payment. Social obligations were so embedded in sixteenth-century Maya elite
consumption rituals that the debts acquired during feasts did not end with death
but would be inherited by surviving kin (Tozzer 1941:92).
But in addition to the competitive feasting of the elite, Landa mentioned feast
ing among the commoners, or "kinfolk," associated with ancestor veneration
and life-cycle celebrations. Many of these feasts occurred in conjunction with rit
ual activities, such as dancing with animal headdresses, bloodletting, animal sac
rifice, carving new idols, and displays of ideologically charged icons, which took
368
R I T U A L F E A S T I N G A N D V I L L A G E F E ST I VA L S AT T H E C E R � N S I TE, EL S A LVA D O R
place during public religious performances at community festivals (e.g. , Tozzer

1941).
In this chapter I explore this second type of feast, that of the ''kinfolk," focus
ing on the archaeological correlates of ritual feasting at the site of Cereo, El Sal
vador. The Cereo site, located in the Zapotitan Valley, was a flourishing Middle
Classic period agricultural community located on the southern Maya periphery
(Sheets 1992a; Fig. 13-1). Around A.D. 590 a volcanic vent, located only 600 meters
from the site, opened up beneath the nearby Rio Sucio and buried the commu
nity under 6 meters of ash (Sheets 1992a) . The suddenness of the eruption pre
cipitated a catastrophic abandonment of the community leaving virtually
complete artifact assemblages in their context of use or storage, in addition to
preserving fragile earthen architecture and organic artifacts. The unique mode of
abandonment and subsequent extraordinary preservation of the site provides ar
chaeologists with a rare glimpse of rural village life, including material remains
that can be interpreted as functioning in the production of community festivals
and ritual feasting. The archaeological signature suggesting participation in vil
lage feasting at Ceren adds important new criteria to the archaeological identifi
cation of feasts, one of the major themes of this volume.
Pacific Ocean
0 50 kilometers
Figure 13-1. Map of western El Salvador.
369
Linda A. Brown
Material correlates of feasting inferred from the Ceren data include many of
the expected categories as proposed by Hayden (1995,1998, Chapter 2) including:
(1) a specialized permanent facility, known as Structure ro, that was used for food
storage, food preparation during feasts, and as a storage house for ritual para
phernalia between events; (2) a cleared area around the exterior of the building
interpreted as a gathering space for food consumption and ritual performances;
(3) an artifact assemblage indicative of large-scale food preparation and storage;
and (4) ritual items that included a deer-skull headdress with other likely compo
nents of a ceremonial dance costume.
In this chapter, my objective is to examine the archaeological signature of rit
ual feasting. First the archaeological evidence of feasting at Ceren is examined
and data suggesting that one household may have been linked with feasting are
presented. Then I turn to ethnographic accounts to begin to generate possible
expectations for the archaeological recognition of ritual feasting in Maya rural
village communities. Finally, building on the growing body of research into an
cestor veneration and the built environment in the Maya area (e.g., McAnany
1995), I propose that we may be able to recognize rural lineage feasts from dis
tinct types of remains in the archaeological record.
A few terms should be defined at the outset. In using the termfeast, I refer to
the communal consumption of food and drink in a context that differs from that
of daily household food consumption practices. By the term ritual feasting, I
mean that social network in which communal food and alcohol consumption co
occurs with a series of requisite ritual performances. Finally, in using the termfes
tival, I refer to the period of time, usually several days, which is set apart for open
public celebrations, including ritual feasting, ceremonial performances, and
other forms of entertainment.
THE CEREN SITE

THE E NVI RO N M E N T
The Ceren site is located in the Zapotitan Valley of west-central El Salvador
at an elevation of 450 meters (1,500 feet) above sea level. The site is situated
along the western terrace of the Rio Sucio, the main river of the valley.
Today the Zapotitan Valley, an intermontane basin that covers an area of ap
proximately 182 square kilometers, is extremely fertile and productive agricul
tural land. But the Zapotitan Valley was not always an advantageous
environment.
Around A.O. 260, the southern Maya periphery underwent a regional disaster
when the massive eruption of the Ilopango volcano, in central El Salvador,
spewed tephra over millions of square kilometers (Hart and Steen-Mcintyre
1983) . As a result of this eruption, the Zapotitan Valley was virtually depopulated
370
R I T U A L F E ASTI N G A N D V I L L A G E F E ST I VA L S AT T H E C E R E N S I TE, E L S A LVA D O R
and evidence suggests that it took two centuries before the soil recovered and
people resettled the valley (Sheets 1983:287) . Based on survey data (Black 1983),
we know that by the sixth century people had migrated back into the valley and
the area was emerging as a "complex stratified society, with hierarchical settle
ment system" complete with occupational specialists controlling obsidian indus
tries at San Andres, the primary regional center (Sheets 1983:290). It was during
this reoccupation that the agricultural village of Ceren was founded.
Attempting to reconstruct population estimates for the Ceren community is
particularly challenging as most structures remain buried under 6 meters of vol
canic ash. Currently; ten structures have been excavated and an additional seven
structures have been identified by test pits (Fig. 13.2). Data gathered from ground
penetrating radar studies suggest that at least eighteen more structures, yet to be
verified in test excavations, may remain buried (Conyers 1995). Based on these
data, Sheets (personal communication, 1998) estimates that the ancient Ceren
community consisted of at least 150 individuals.
Although estimating the size of the population is problematic, we do know
that Ceren residents settled on top of a very thin, yet extremely productive soil.
Residents exploited and grew numerous plant species including maize, beans,
squash, manioc, maguey; cacao, and chili, as well as a number of medicinal
and / or ceremonial plants (Lentz et al. 1996) . Additionally community members
kept and tended ducks, raised domesticated dogs, and exploited wild fauna such
as white-tailed deer, peccary; and freshwater turtles (Brown, in press).
Household 3 Joya de Ceren, El Salvador
0 10 m eters
ICEICll:>CO
$
Plaza
CJSio
Str. 1 4 � Str. 1 3
o
Household
!..m �
·.:: .
2 <:::; � ---T' Str. 2
Str. 7 _
,
�
Str. 1 2
Str. 1 8 . �.
,__
&I
fa llow milpa
) .. m1lpa
'Q
Str. 9 ' ' · "' . tfPJ /'·,,
Str. 1 7 '"'1/
Figure 13.2. Site map for Cerfo, El Salvador.
371
Linda A. Brown
E X C AVAT I O N S AT T H E C E R E N S I T E
Payson Sheets (principal investigator) has conducted archaeological investiga
tions at the site since 1978, with a hiatus between 1980 to 1989 during the height of
the El Salvador civil war (Sheets 1992a) . Excavations resumed in 1989, and in the
following years portions of four household clusters and associated extramural
areas, three ceremonial structures, a civic building, a midden, and various agri
cultural areas were excavated.
Of the household clusters excavated to date, Household 1 is the best known
(Beaudry-Corbett, Simmons, and Tucker, in press) . It consists of three separate
buildings-a kitchen (Structure 11), a storeroom (Structure 6), a domicile (Struc
ture 1)-as well as a covered open work area (Structure 5), cleared extramural ac
tivity areas around the compound, and agricultural zones.
Structure IO is located immediately to the east of Household 1. This building
was excavated during the 1992 and 1993 Ceren field seasons under the supervision
of Andrea Gerstle (1992, 1993). The following section on Structure IO, except
where otherwise noted, is a condensed version of Brown and Gerstle (in press)
and two preliminary reports (Gerstle 1992, 1993) .
ST R U CT U R E 1 0
The archaeological evidence suggests that Structure IO was used as a headquar
ters for the production of community festivals that included ritual feasting. Struc
ture IO is a thatched-roof wattle-and-daub building that was oriented
approximately 23 degrees east of north (Fig. 13.3). The superstructure was con
structed on a square clay platform 3.7 meters on a side and has two rooms, an east
(front) room and a west (back) room. The only entranceway into the building
was through a wooden pole door that faced west toward the Household 1 com
pound. Unlike other buildings excavated at the site, walled corridors were
erected outside and along the north and east sides of the superstructure and the
corridors were covered with a thatch roof. Moreover, Structure IO does not fol
low the dominant 30 degrees east of north alignment used in domestic buildings
at the site.
Structure IO was divided into several functionally distinct activity areas. The
north exterior corridor was utilized for food preparation as indicated by the pres
ence of two hearths, a metate and mano (grinding and hand stone), bone and
antler corn huskers with empty ears of corn discarded just outside the entrance
way, food-serving vessels, and a number of large cooking and storage vessels, one
of which was found resting on one of the hearths.
The east corridor was utilized primarily for ceramic vessel storage. The south
ern portion of the corridor remains unexcavated under several fallen walls; how
ever, the excavated northern portion was packed with at least seventeen medium
372
R I T U A L F E A S T I N G A N D V I L L A G,E F E S T I VA L S AT T H E C E R E N S I TE, E L S A LVA D O R
....... - ""'. ,.?

\ mN �
.
\
\.
I
.______,
0 1 meters
0 structure walls/columns
Figure 13.3. Plan view of Structure 10 at Ceren, Bl Salvador.
to large storage jars, serving vessels, and several painted gourds (Beaubien and
Lundberg 1993; Beaudry-Corbett 1993) .
A half-height wall forms part of the exterior wall of the east corridor directly
abutting the food-preparation area (Fig. 13.4) . This low wall is interpreted as a
food-dispensing area and likely functioned as a pass-through for food and drink to
festival participants gathered outside of the building (Gerstle 1993). This interpre
tation is supported by the observation that the ground to the north and east of
Structure IO was highly compacted, flattened, and smoothed, suggesting an area
of high foot traffic (Simmons and Villalobos 1993). Additionally, this area was rel
atively free of artifacts and plants suggesting a well-maintained area that was reg
ularly swept clean. Presumably, this cleared hard packed surface was where
participants gathered for ceremonial celebrations that included outdoor feasting.
Moving inside Structure IO, the east room is notable for both the wall treat
ment and the kinds of artifacts present. It was the sole painted room in the build
ing, with the eastern face of the dividing wall, cornices, and door pilasters
painted red and the lowermost portion of the pilasters accented with white. This
373
Linda A. Brown
Figure 13-4· Artist's reconstruction of Structure 1 0 (roof over walled corridors not de
picted). Modified from an illustration by Karen Kievit.
room was used for the storage of ceremonial and unique artifacts. Ritual items
included a deer-skull headdress, which was painted red, in association with other
components of a possible dance costume-including a matched set of bone
beads, bone ornaments, and deer scapulae. The headdress-formed from the cra
nium minus the mandible of an adult white-tailed deer stag (Odocoileus virgini
anus), was in storage on a high shelf. It was recovered with bits of twine still
wrapped around the antler bases. Presumably the twine was used to secure the
headdress on the wearer.
A large jar with an appliqued caiman head effigy and another large jar contain
ing squash seeds were recovered in situ on the east room floor. The caiman effigy
jar was full of achiote seeds, used by various contemporary Maya as food color,
and by the Lacandon Maya to produce a red paint symbolic of human blood (Coe
1994; McGee 1990; Tozzer 1907) . Additionally, twenty ears of corn were stored in
an elevated context in this room. Apparently, the corn was stored shucked as no
remnant of husks remained.
In contrast with the special-use items stored in the east room, the west room
was used for the storage of utilitarian vessels, including a large jar full of beans.
The pattern of the storage of ceremonial artifacts in the east room is consistent
with practices of contemporary Maya in Zinacantan who place household altars
and associated ceremonial paraphernalia along the east wall of their homes, the
wall toward the direction of the rising sun (Vogt 1976).
374
R I T U A L F E A S T I N G A N D V I L L A G E F E S T I VA L S AT T H E C E R E N S I TE, E L S A LVA D O R
A few other notable patterns in the artifact assemblage should b e mentioned

here. Structure IO contained relatively few serving vessels, with only five ceramic
serving vessels and two painted gourds, a number less than some of the domestic
structures at the site (Beaudry-Corbett 1993) . Instead of serving vessels, Structure
IO contained the highest number of large utilitarian food storage jars of any
Cereo building excavated to date (Beaudry-Corbett 1992, 1993).
STRU CTU R E 1 0 A N D D E E R C E R E M O N I A L I S M
In addition to feasting, festivals at Cereo may have involved the display of white
tailed deer artifacts. The presence of the deer headdress in storage suggests that
the white-tailed deer stag was associated with ceremonial activities at Structure
IO and the headdress may have been part of a dance costume used at village festi
vals. Mary Pohl (1981) argued that the white-tailed deer stag was a prominent pre
Columbian deity who played a significant role in ritual. Specifically; Pohl argued,
the white-tailed deer was associated with the cuch ceremony that linked commu
nity leaders with agricultural fertility; the sun, rain, economic prosperity; the
cyclic nature of time, death, renewal, and rebirth.
L I N KA G E S B ET W E E N H O U S E H O L D 1
A N D F E A S T I N G AT S T R U C T U R E 1 0
One of the questions posed in this volume (see Clarke, Chapter 5, and Hayden,
Chapter 2) concerns how archaeologists might identify households that regularly
sponsored feasts from remains found in the archaeological record. At Cereo, the
archaeological evidence suggests that Household 1 was involved in the processing
of consumable goods for public feasting at Structure IO (Beaudry-Corbett, in
press; Beaudry-Corbett, Simmons, and Tucker, 1 99 8; Brown and Gerstle, in
press). This relationship is inferred from numerous lines of material remains,
which are reviewed below.
Physical Proximity and Building Modifications

Evidence linking Household 1 with Structure IO is implied from building proxim
ity (Beaudry-Corbett, Simmons, and Tucker, in press) and the orientation of
structure entranceways, with inferred foot-traffic patterns (Simmons and Villalo
bos 1993). Structure IO is located only s meters to the east of Household 1. The
only access into Structure IO is through a wooden pole door that faces west, di
rectly toward Household I. Meanwhile, the entrance to the Household 1 store
room faces east toward Structure IO (Beaudry-Corbett, Simmons, and Tucker, in
press) . This is notable because all other entranceways into Cereo domestic store
rooms open to the north, toward the household domicile. Presumably this would
have facilitated the flow of people and goods between Structure IO and House
hold i's storeroom.
375
Linda A. Brown
Groundstone
Only one metate was recovered inside of Structure IO. A single metate hardly
seems sufficient for the amount of grinding necessary for the scale of food prepa
ration as inferred from the size of the large pot resting on the nearby hearth. In
contrast, a total of five complete metates and five curated metate fragments were
recovered from Household 1 (Fig. 13.5). Four metates were in use positions,
mounted on forked sticks or on the kitchen floor, whereas one was in temporary
storage resting a pair of forked sticks that presumably it would be mounted on
during use (Beaudry and Tucker 1989; Mobley Tanaka 1990) . The number of
complete metates in use or temporary storage at Household 1 suggests that
women were grinding more maize than would be necessary for household con
sumption alone. It is conceivable that a surplus of maize could have been pro
duced for household life-cycle celebrations. However, the evidence of large-scale
food preparation in Structure IO would suggest that the Household 1 compound
was used for labor-intensive food grinding associated with public feasts.
This interpretation is further supported by use-wear analyses. Only one
metate, located on the floor of the kitchen (Structure 11), showed evidence of
0 5 m
.....
.....
.....
&
� ....
....
....
....
A metate
@ earthen floor/collapsed wal l
Figure 13.5. Household 1, Structure 10, and Structure 12 showing positions of complete
metates in use contexts or temporary storage.
376
R I T U A L F E A S T I N G A N D V I L L A G E F E STIVALS AT T H E C E R E N S I TE, E L S A LVA D O R
heavy use-wear suggesting that it was the main metate used daily by this house
hold (Sheets 1990, 1992b ). The other four complete metates exhibited minimal
wear, suggestive of short-term or periodic use, a material expectation for metates
used occasionally for feasts. Interestingly, not all metates had matching manos
(hand stones), as only three manos were found (Sheets 1992b).
Fauna/ Remains and Blood Residue Analysis

In addition to grinding, Household 1 appears to have been the location of meat
processing. Although a sample of obsidian blades from all site contexts was tested
for organic residues, thus far only tools from Household 1 indicate the presence of
animal protein. Two obsidian blades in Household 1, from the kitchen and store
room respectively, tested positive to artiodactyla (deer or peccary) antiserum by
irnmunoelectrophoresis whereas two blades discarded outside of Household 1
tested positive to Canidae (domesticated dog, fox, coyote, wolf; Newman 1993) .
The presence of deer and dog faunal remains recovered from storage and discard
contexts at this household would suggest that these were the species being
processed here (Brown in press). A duck was recovered in situ tied to a pole wall in
side of the Household 1 storeroom, once again a deviation from other Ceren con
texts as this is the only household to have a live animal tethered inside. It should be
kept in mind that currently only ten structures have been fully excavated at the
site; thus the sample size is quite small. However, given the available data, this pat
tern suggests that this household was the locus of more meat processing com
pared with other households excavated thus far. It is conceivable that Household 1
members processed meat for feasting activities at Structure IO.
Frequency and Type of Ceramic Vessels

The ceramic assemblage further supports the interpretation of a relationship be
tween Household 1 members and feasting at Structure IO. Household 1 had more
large jars without handles than did other Ceren households, a pattern interpreted
as related to a household's need for long-term food storage (Beaudry-Corbett, in
press) . Furthermore, Household 1 had more utilitarian bowls with handles than
other households suggesting a greater need for transferring and transporting a
higher volume of consumable goods (Beaudry-Corbett, in press). Beaudry
Corbett (in press) argued that Household i's greater need for long-term food stor
age, as well as for transferring and transporting a higher volume of consumable
goods than other households, could reflect a role in feasting at Structure IO.
Ceramic Compositional Analyses

In addition to the amount and types of vessels recovered from Household 1, com
positional analyses suggest further linkages between this household and Struc-
377
Linda A. Brown
ture IO. Beaudry-Corbett (in press) identified two imported red wares in the
Cereo assemblage. One originated from within the Zapotitan Valley, close to the
main civic-ceremonial center, San Andres; the other was produced close to the
site of Chalchuapa, approximately 40 kilometers from Cereo. Notably all im
ported red ware identified at Ceren to date was found exclusively in either
Household 1 or the two ceremonial buildings, Structures IO and 12. Interestingly;
although the ceramics originating from Chalchuapa consisted of both utilitarian
jars and food serving vessels, red wares produced within the Zapotitan Valley
close to the site San Andres, the civic-ceremonial core, consisted exclusively of
polychrome food-serving vessels (Beaudry-Corbett, in press) .
S U M MARY
Based on the archaeological evidence, a number of inferences concerning the
structure of festivals at Cereo can be proposed. Cereo festivals included ritual
feasting in the context of deer ceremonialism. The limited space inside Structure
IO, in addition to the well-trodden and well-maintained exterior ground surface,
suggest that festivals were public events and food consumption occurred in open
outdoor areas.
The Household 1 compound may have been used periodically for food prepa
ration prior to and during feasts at Structure IO. The restricted distribution of ob
sidian blades testing positive to animal blood residue suggests that meat was
processed here. Additionally; the number of functional metates recovered in
Household 1 suggests that this household was the locus of more food grinding
than would be necessary solely for household consumption. Women living in
Household 1 likely assisted with labor-intensive grinding for feasts prepared at
Structure IO. However, the presence of worn metates, without matching manos,
suggests that some women living in other households brought their own manos
to this compound to assist in grinding (Sheets 1997 personal communication) .
Thus, responsibility for food preparation for feasts seems to have involved a small
group of women living in and beyond the immediate household.
The use of household space for the preparation of ritual feasts associated
with nearby ceremonial buildings has parallels among the contemporary Maya.
In the Guatemalan village of Santiago Atitlan, co.fradia members stage festivals
that include feasting. The co.fradia building, a specialized ceremonial structure
that houses the corporately owned sacra of the sodality and serves as the site
for festivals, is always located within the domestic household compound of the
highest-ranking male and female co.fradia members. Prior to a festival, a small
work party of women gather at the household of the highest-ranking mem
bers to begin preparing food for ceremonial feasts (Fig. 13 .6). Metates used peri
odically for festivals are moved from storage locations, washed, and then set up
378
R I T U A L F E A S T I N G A N D V I L LAG E F E S T I VA L S AT T H E C E R £ N S I TE, E L S A LVA D O R
Figure 13.6. Women's work group using household space t o grind com fo r a ritual feast at
a nearby cofradia building in Santiago Atitlan, Guatemala. (Photo by the author)
in use contexts, generally on the ground under an outdoor roofed area close to
the kitchen. While a few women tend to the food cooking over hearths, the
majority of the women are involved in various labor-intensive food preparation
activities such as grinding corn or preparing tamales. Thus, for several days
prior to a festival and for the duration of the festivities, the household com
pound is converted into a semipublic work area for the organized female labor
needed to produce feasts associated with festivals at the nearby ceremonial
building. It is conceivable that the Household 1 compound at Ceren may have
functioned in a similar way in the time immediately before and during festivals
at Structure IO.
Returning to Ceren, festival foods included corn, beans, and squash, as indi
cated by empty corn cobs discarded outside of Structure IO, the husked corn
cobs stored in an elevated context inside the building, and the large jar of beans
379
Linda A. Brown
and squash seeds stored in Structure IO. Corn, beans, and squash also are found
in domestic contexts at the site and probably were consumed daily. But in addi
tion to these everyday provisions, some more unique food items are present, in
cluding deer, dog, and a large number of achiote seeds, which probably do
represent special festival foods. Early Colonial accounts note that meat com
monly was consumed in the context of public ritual feasting (e.g. , de Palacio
1985; Tozzer 1941). The faunal evidence from the Household l compound sug
gests that meat consumption at Ceren likely followed a similar pattern.
Estimated vessel capacities for cooking and serving vessels make it possible
to approximate the size of feasts at Structure IO . An extremely large cooking
pot with a capacity of 56,429 cc was recovered in situ, resting directly on one of
the Structure IO hearths (Beaudry-Corbett 1993) . Currently, vessel capacities of
approximately one third of all Ceren ceramic individual serving vessels have
been directly measured (Stacy Barber, personal communication, 1998), allow
ing for preliminary inferences of the maximum number of individuals this pot
could serve. Assuming that they were completely filled, Ceren individual ce
ramic serving vessels could hold a mean of 8 85 cc, or approximately 3.5 cups of
food. Thus, at maximum capacity, the large cooking pot could have served up
to 64 individual servings or approximately one half the 150 individuals that
Sheets (personal communication, 1998) estimates may have lived in the village.
Another similar extremely large cooking vessel was recovered beside the
hearth in the kitchen of Household l and if it were used simultaneously it
would raise the number that could be served to 128 individuals. Although pre
liminary; estimates of Ceren festival crowd size based on maximum capacities
of cooking vessels, then, would suggest that some feasts at Structure IO were
open to the entire community.
Feasts at Ceren were frequent enough for hosts to construct, and continuously
maintain, a specialized building solely dedicated to feasting activity and the stor
age of ritual paraphernalia. The association of two permanent building types, rit
ual architecture in close proximity to a residential compound, suggests that the
relationship between Household l members with activities at Structure IO was in
stitutionalized. Thi� interpretation is further supported by a series of building
modifications to both Structure IO and Household l suggesting some time-depth
to existence of these buildings.
In addition to linking Household l with feasting at Structure IO, interpreta
tions based on ceramic compositional analyses also link Household l members
with social networks beyond the immediate community (Beaudry-Corbett, in
press). The restricted distribution of imported ceramics within the village sug
gests that Household l members, in their role as producers of ritual feasting and
380
R I T U A L F E A S T I N G A N D V I L LAG E F E S T I VA L S AT T H E C E R � N S I TE, E L S A LVA D O R
festivals, had greater access t o extravillage social relations, apparently a social net
work not available to all community members (Beaudry-Corbett, in press) .
R U R A L F E A S T I N G I N T H E M AYA A R E A : E T H N O G R A P H I C
S O U R C E S A N D A R C H A E O LO G I C A L I N F E R E N C E S
Ethnohistoric sources note that at the time of Contact, Maya lineages maintained
permanent lineage houses or small temples (Fox and Cook 1996). Although line
age houses no longer exist in the Maya area, there is evidence that suggests
that the modern Maya fiesta system (specifically cofradias), a syncretistic
Catholic / Maya institution involved in the production of village festivals includ
ing ritual feasting, subsumed some of their functions (Cook 1981) . An examina
tion of the material correlates associated with contemporary Maya village
festivals Inight elucidate features potentially useful for the archaeological identifi
cation of rural feasting spaces.
S P E C I A L I Z E D F E A S T I N G S PA C E S
In general, two types of physical spaces were noted by ethnographers: those
marked by permanent dedicated ceremonial buildings, and open-air feasting
spaces usually in association with permanent structures.
Permanent Dedicated Ceremonial Buildings

Ethnographic references frequently stated that rural ceremonial facilities were
constructed siinilarly to domestic buildings. Thompson noted of Cehach Maya
villages that "specialized temple buildings seem to be absent; instead, as among
the Lacandon, a hut of normal residential type, although perhaps a little larger,
serves the religious needs of the village" (1970 :76) . Likewise, among the Xpichil
and Xyatil Maya of Quintana Roo, Shattuck (1933) noted that ceremonial huts
were constructed of siinilar material and size as the fainily hut.
The Lacandon Maya (McGee 1990) have a sacred hut that is shared by several re
lated fainilies. Interestingly, these structures are associated with accompanying sa
cred cook huts: "The sacred hut has its own fire and its own utensils, which are
exclusively used for the celebration of religious observations. To bring any food
into the domestic hut renders it unfit to be offered to the gods" (Tozzer 1907:91---92).
Ceremonial kitchen huts were utilized by the ritual sponsor's wife, who prepared
tamales and food offerings in these locations (McGee 1990). The use of separate
kitchens, artifacts, and features for the preparation of ceremonial food may reflect
the concept of ritual purity (Thompson and Thompson 1955; Tozzer 1907).
Among the Chorti Maya of Guatemala, Wisdom (1940 :384-386) noted the
presence of two types of specialized permanent facilities utilized in the produc-
381
Linda A. Brown
tion of festivals: ceremonial houses and cofradia houses. Both the ceremonial
house and the cofradia house functioned as storage areas for corporately owned
ritual paraphernalia, as well as for food preparation and dispensing during com
munal feasting. Ceremonial houses were common features of rural settlements
whereas cofradia houses were associated with the more populated towns.
Ceremonial houses, constructed and maintained by the communal labor of
members of the community, were said to be communally owned by all members
of the aldea (rural settlement; Wisdom 1940:281). These rural ceremonial struc
tures were the loci for ritual feasting and rites for rainmaking, agricultural fertil
ity, and Day of the Dead celebrations.
Most of the larger aldeas have a ceremonial house, the construction of which is like
that of any other house, consisting usually of a sleeping-house and kitchen. The for
mer is actually the ceremonial house, while the latter is used by the women for
cooking food during ceremonies. (Wisdom 1940:384)
[After the ceremony] the Indians then return to the ceremonial house, where feast
ing, drinking, and dancing go on through the night. A large group of women pre
pare tamales, atol [maize gruel], chilate, tortillas, meats, and coffee, using the sacred
water for this purpose. (Wisdom 1940:438-439)
Cofradia houses were located in more populated settlements. In addition to serving

as the locus of festivals, the cofradia house was the temporary residence of the may
ordomo (native religious official) and his family during his year of service in office.
The cofradia is found only in the pueblos and is used in connection with festivals.
The mayordomos live in it during their period of service at the church and use it as a
storage house for the festival paraphernalia. On festival days the Indians bring their
contributions of maize to this house and the women grind it and cook it into foods,
especially chilate and tortillas, which are often given gratis to those who attend the
festival. The house is usually built at the edge of the pueblo, near some important
trail which leads to large outlying aldeas. It is built like any other house but it is
much larger. (Wisdom 1940:385)
Open-Air Feasting Spaces Associated with Ceremonial Structures

Although descriptions of buildings used for feasting were scant in the literature,
the use of outdoor space in association with feasting was even less frequently re
ported by ethnographers. However, Wisdom (1940) noted that the cleared space
outside of the ceremonial house was used for several types of activities related to
the production of festivals.
Indians from all over the munidpio [township] gather in the open space surrounding
the house, bringing large quantities of chicken tamales, cacao, chilate, atol, fruits, and
chicha [fermented sugarcane drink]. Each family contributes a peso or two until
about one hundred are collected. The money and food are given to the leader as
382
R I T U A L F E A S T I N G A N D V I L L A G E F E S T I VA L S AT T H E C E R t N S I TE, E L S A LVA D O R
payment fo r performing the ceremony. Each family brings additional food which is
cooked by the women in the ceremonial house kitchen and eaten by everyone.
The land upon which any ceremonial house . . . sits is considered a sacred spot
and is not to be used for any secular purpose. The ceremonial house has a large
space around it, kept clear of vegetation by communal aldea labor, where the Indi
ans congregate during ceremonies. (Wisdom 1940:426)
Meanwhile, at the larger cofradia house the open exterior space also was used for
food preparation and consumption:
A group of fifty women, appointed by the mayordomos, are busy at the cofradia
making chilate, atol, coffee, and tortillas in the large ovens and fireplaces in the cofra
dia courtyard. Long benchlike tables are covered with gourd bowls filled with the
chilate, which all the Indians drink free of charge. (Wisdom 1940:449-450)
It [the cofradia house] is built like any other house but it is much larger and has a
large courtyard, since on important festivals many hundreds of Indians congregate
around it. In the courtyard stands a clay oven, used for boiling chilate served to Indi
ans on festival days. (Wisdom 1940:385)
S O C I A L G R O U P S T H AT S P O N S O R F E A S T S
The Archaeological Signature o f Lineage Spaces
As seen in this volume, it is possible to recognize the material correlates of vari
ous types of feasts from remains in the archaeological record. Moreover, those
households that regularly engaged in feasts are expected to have an artifact as
semblage indicative of this role. Among the contemporary Maya, rural house
holds occupied by lineage heads (Hayden and Cannon 1984:178), as well as
cofradia leaders, have more large fiesta pots and metates than other households,
reflecting participation in lineage or sodality feasts during fiestas. The question
arises as to whether it would be possible to identify the type of social group (lin
eage versus sodality) that sponsored feasts from remains found in the archaeolog
ical record. In the Maya area, it may be possible to identify lineage-sponsored
feasts when they occur in a specialized context, such as a lineage house or shrine,
by combining the expectations of feasting with the material correlates proposed
for demarcating sacred lineage space (McAnany 1995). McAnany argued that one
of the ways in which lineage groups demarcated sacred space had a strong mate
rial component that left a distinct archaeological signature.
[Ancestors] were venerated by name because particular resource rights and obliga
tions were inherited through them by their descendants. This is precisely why an
cestors "slept" within the construction mass of residential compounds-to insure
the chain of continuity in resources transmitted between generations . . . Burial of
ancestors often marks the termination of the use of an older structure and the com
mencement of the construction of a new one. Over time, these places of the ancestors
often become sacralized locales at which ritual structures such as temples or shrines com-
383
Linda A. Brown
pletely replace domestic structures. I have cited one such sequence from K' axob, but the
pattern is ubiquitous throughout the lowlands of Mesoamerica. (McAnany 1995:161;
emphasis added)
Although many archaeological household compounds in the Maya area are likely
to have burials under domestic platforms, those representing the practice of de
marcating sacred space for lineage ancestors showed a fundamental change in
building function (e.g. , a domestic space that was transformed into a specialized
ritual locale; McAnany 1995). This pattern is dramatically illustrated in the main
acropolis at Copan, where elite members of the founding lineage were buried
under floors of their domestic compounds. These underwent impressive build
ing renovations until eventually the entire domestic compound was literally en
tombed within a large temple known as Structure 16 (Stuart 1997).
The transformation of domestic space into a sacred lineage house is not
unique to the Maya area. For example, Kirch (Chapter 6) noted that ritual feast
ing in the village of Tikopia, Polynesia, also occurred at lineage and clan temples
that historically functioned as residences for ancestors. After years of domestic
use and the subsequent burial of ancestors within them, some residential build
ings were transformed into "holy houses" used for ritual feasting.
Assuming that such a change in building function could be inferred from re
mains in the archaeological record, then domestic structures displaying a similar
life history, in conjunction with the material correlates of feasting, may enable us
to infer ritual feasting revolving around a lineage organization.
Social Groups and Feasting at Structure 7 0

Returning to Ceren, can we infer anything about the social unit that sponsored
feasts at Structure IO? As suggested earlier, the Household 1 compound appears
to have been used periodically for large-scale food preparation. Similarly, the ce
ramic assemblage and architectural orientation suggest linkages between House
hold 1 members and Structure IO. That this linkage may reflect a lineage activity
is suggested by the particular history of building modifications at Structure IO.
As mentioned earlier, at the time of the eruption Structure IO had undergone
at least three building modifications. With the exception of building alignment, a
point to which I will return below, the original structure and architectural fea
tures followed the same building plan as domiciles at the site: a two-room super
structure constructed on a elevated square platform and a centrally placed door
in the interior wall that allowed access into the back room where the "sleeping"
bench was located. At some point, Structure IO underwent a series of renova
tions. A low clay floor was constructed around the east and north sides of the su
perstructure, walls were erected restricting access, and this area was roofed. A
kitchen was placed in the north corridor. A new doorway was built in the north
384
R I T U A L F E ASTI N G A N D V I L L A G E F E ST I VA L S AT T H E C E R E N S I TE, E L S A LVA D O R
wall of the corridor t o be later moved t o the west wall opening toward House
hold 1. The sleeping bench was obliterated and buried during a renovation in
which the entire floor level of the back room was raised.
The conservation ethic at Cereo prevents excavating into the floors of struc
tures, thus precluding an important source of data from subfloor burials and
caches to fully test this hypothesis. However, the above-ground architectural pat
tern suggests that Structure IO was a domestic residence that, at some point in its
life history, underwent modifications to become a ceremonial locale that perhaps
was a rural lineage house. If this interpretation is correct, then the deviation in
the Structure IO building alignment would suggest that original occupants of
Structure IO held a distinctive status in the community.
CO N C LU S I O N S
A number of features potentially useful for the archaeological identification of
spaces used for Maya village feasting can be inferred from ethnographic, ethno
historic, and archaeological evidence from Cereo. Although rural Maya ceremo
nial houses often have a similar architectural plan and scale as domestic
compounds, they should have more hearths or other cooking features than ex
pected for domestic use. Those areas used for large-scale food preparation might
have features that facilitate the distribution of food serving, such as the low serv
ice wall. If ceremonial houses are the same size as domestic compounds, then
they should be associated with open exterior spaces that are used by groups gath
ered for feasting and ritual performance. Areas used for ceremonial gatherings
may be regularly maintained and kept clear of trash and debris. Moreover,
cleared ceremonial gathering areas are expected to have a hard-packed ground
surface indicative of an area of heavy foot traffic.
Households or groups involved in sponsoring festivals may reside in buildings
located in close proximity to the specialized ceremonial structures. The architec
ture may be designed to facilitate the movement of people and goods between
the living quarters of the hosts and the specialized ceremonial structure and as
sociated cooking features. Areas used for the large-scale food processing should
have a higher frequency of food-preparation artifacts, especially those heavy; dif
ficult-to-transport items such as metates, than expected in domestic household as
semblages. Moreover, manos and metates used only periodically for preparation of
feasting foods may show evidence of lighter use-wear when compared with those
used in daily household food preparation.
Ethnohistoric, ethnographic, and archaeological sources suggest that Maya
feasts often occurred in the context of lineage celebrations that honored the an
cestors. Combining the material correlates of lineage ceremonial buildings with
those associated with feasting may allow us to infer lineage-sponsored feasts from
385
Linda A. Brown
the archaeological record, specifically when these feasts occur in specialized line
age houses.
So what were the advantages to those involved in sponsoring ritual feasting
and festivals at Cereo? A range of benefits, from functional equilibrium to more
competitive interest-theory models, have been proposed for participation in the
contemporary Maya fiesta system (e.g., Camara 1952; Cancian 1965; Carrasco
1961; Chance and Taylor 1985; Collier 1979; Farris 1984; Hayden and Gargett 1990;
Nash 1958; Vogt 1969; Wolf 1959) . Interestingly; ritual feasting among contempo
rary agriculturalists in Asia suggests that both models may have heuristic value
(Ahern 1973; Freedman 1958, 1970) . For example, in some Asian villages perma
nent ceremonial lineage halls are used for ritual activities including feasts. In ad
dition to this utilitarian function, ceremonial halls take on a symbolic role and
stand as visible icons of the economic strength of a lineage as the "hall requires
for its building and its maintenance and for the upkeep of its rites that it be en
dowed; hence a hall is a mark of riches" (Freedman 1970: 168). As lineages con
stantly assess themselves by comparing their halls with those owned by other
lineages in the village, the material trappings of feasts as well as the actual line
age-feasting event have strong boundary-marking potential (Ahern 197p12) .
Both promote within-group solidarity as well as being a means to compete with
other lineages and engage in self-aggrandizing behaviors (Ahern 1973). A similar
observation was made by Clarke (see Chapter 5) concerning the feasting prac
tices of the rural Akha in Northern Thailand.
Whether feasting at Cereo followed a similar trajectory is unknown. However,
if the interpretation that Household 1 members sponsored feasts at Structure IO
is correct, then the archaeological evidence does suggest a possible benefit.
Rather than the accumulation of a disproportionate amount of material goods,
inferences based on the ceramic assemblage suggest that Household 1 partici
pated in a geographically broader social network that did other households exca
vated to date (Beaudry-Corbett, in press). Scholars have argued that extended
social networks created by consumption rituals are risk-reducing strategies that,
during times of crisis, can be accessed to provide critical support and information
to participants (Douglas and Isherwood 1996:xxii) . The sponsorship of feasting at
Structure IO may have provided a similar advantage to Household 1 members.
In a broader perspective, feasting was likely an activity for forming alliances in
the Zapotitan Valley. The restricted distribution of polychrome serving vessels
produced close to San Andres, which were found only in Household 1 and the
two ceremonial buildings to the east, suggests that Household 1 members may
have been more closely linked to the civic-ceremonial center than other house
holds excavated to date (Beaudry-Corbett, in press). Feasting may have been a so
cial arena used in alliance formation linking the civic-ceremonial center to more
rural settlements throughout the valley. If so, then perhaps the gifting of food-
386
R I T U A L F E A S T I N G A N D V I L L A G E F E S T I VA L S AT T H E C E R t N S I T E , E L S A LVA D O R
serving vessels, vessels that function in both utilitarian and symbolic realms of
consumption rituals, were tangible reminders of the social relations and obliga
tions embedded in participation in those networks.
AC K N OW L E D G M E NTS
I would like to thank Brian Hayden and Michael Dietler for inviting me to participate in
the SAA symposium and subsequent volume. Thanks go to Payson Sheets who gener
ously allowed me use of the Ceren data and to Marilyn Beaudry-Corbett for use of her ce
ramic analyses. I also thank Payson Sheets, Barbara Voorhies, and Harriet "Rae" Beaubien
for valuable comments on earlier drafts.
REFERENCES
Ahern, E. M.
1973 The Cult of the Dead in a Chinese Village. Stanford: Stanford University Press.
Beaubien, H. F., and H. Lundberg
1993 Artifact Conservation during the 1993 Field Season. In Cerro Project 1993 Pre
liminary Report, edited by P. Sheets and S. Simmons, pp. 164-172. Boulder:
University of Colorado.
Beaudry; M., and D. Tucker
1989 Household l Area Excavations. In A Preliminary Report: Cerro Project 1989, ed
ited by P. Sheets and B. McKee, pp. 29-40. Boulder: University of Colorado.
Beaudry-Corbett, M. P.
1992 Ceramic Analysis, Joya de Ceren: 1992 Season. In 1992 Investigations at the
Ceren Site, El Salvador: A Preliminary Report, edited by P. Sheets and
K. Kievit, pp. 82--95· Boulder: University of Colorado.
1993 Ceramic Analysis, Joya de Ceren: 1993 Season. In Cerro Project 1993 Prelimi
nary Report, edited by P. Sheets and S. Simmons, pp. l38-15r. Boulder: Uni
versity of Colorado.
in press Ceramics. In Before the Volcano Erupted: The Ancient Cerro Village in Central
America, edited by P. D. Sheets. Austin: University of Texas Press.
Beaudry-Corbett, M. P., S. E. Simmons, and D. B. Tucker
in press Household r. In Before the Volcano Erupted: The Ancient Cerro Village in Cen
tral America, edited by P. D. Sheets. Austin: University of Texas Press.
Black, K. D.
1983 The Zapotitan Valley Archaeological Survey. In Archaeology and Volcanism in
Central America: The Zapotitdn Valley of El Salvador, edited by P. Sheets,
pp. 62--97· Austin: University of Texas Press.
Brown, L. A.
in press Household and Village Animal Use at Ceren. In Before the Volcano Erupted:
The Ancient Cerro Village in Central America, edited by P. D. Sheets. Austin: Uni
versity of Texas Press.
Brown, L. A., and A. Gerstle
in press Structure ro: The Production of Village Festivals. In Before the Volcano Erupted:
The Ancient Cerro Village in Central America, edited by P. D. Sheets. Austin:
University of Texas Press.
387
Linda A. Brown
Camara, F.
I952 Religious and Political Organization. In Heritage of the Conquest, edited by
S. Tax, pp. 142-I?J. Free Press, Glencoe, Illinois.
Cancian, F.
I965 Economics and Prestige in a Maya Community. Stanford: Stanford University
Press.
Carlsen, R.
I997 The Warfor the Heart and Soul of a Highland Maya Town. Austin: University of
Texas Press.
Carrasco, P.
I96I The Civil-Religious Hierarchy in Mesoamerican Communities: Pre-Spanish
Background and Colonial Development. American Anthropologist 63:483-497.
Chance, ]. , and W. Taylor
I985 Cofradias and Cargo: An Historical Perspective on the Mesoamerican Civil
Religious Hierarchy. American Ethnologist I2:I-26.
Coe, S. D.
I994 America's First Cuisines. Austin: University of Texas Press.
Collier, J. F.
I979 Stratification and Dispute Handling in Two Highland Chiapas Communities.
American Ethnologist 6:3oy-328.
Conyers, L.
I995 Paleogeography of the Ceren Site, El Salvador. Ph.D. dissertation, University
of Colorado. Ann Arbor: University Microfilms.
Cook, G. W.
I98I Supernaturalism, Cosmos and Cosmogony in Quichean Expressive Culture.
Ph.D. dissertation, State University of New York. Ann Arbor: University
Microfilms.
de Palacio, D. G.
1985 Letter to the King of Spain: Being a Description of the Ancient Provinces of Guaza
capan, Izalco, Cuscatlan, and Chiquimula, in the Audiencia of Guatemala, with an
Account of the Languages, Customs, and Religion of Their Aboriginal Inhabitants,
and a Description of the Ruins of Copan. Translated by Ephraim Squier. Culver
City, Cal.: Labyrinthos.
Douglas, M., and B. Isherwood
1996 The World of Goods: Towards an Anthropology of Consumption. New York:
Routledge.
Farris, N. M.
I984 Maya Society under Colonial Rule. Princeton: Princeton University Press.
Fox, ]. , and G. Cook
1996 Constructing Maya Communities: Ethnography for Archaeology. Current An
thropology 37 (5): 8n-830.
Freedman, M.
I958 Lineage Organization in Southeastern China. London: Athlone.
388
R I T U A L F E A ST I N G A N D V I L L A G E F E ST I VA L S AT T H E C E R E N S I TE, E L S A LVA D O R
1970 Ritual Aspects of Chinese Kinship and Marriage. In Family and Kinship in Chinese
Society, edited by M. Freedman, pp. 163-187. Stanford: Stanford University Press.
Gerstle, A. I.
1992 1992 Excavations at Structure IO, Joya de Ceren. In 1992 Investigations at the
Ceren Site, El Salvador: A Preliminary Report, edited by P. Sheets and K Kievit,
pp. 30-54. Boulder: University of Colorado.
1993 1993 Excavations at Structure IO, Joya de Ceren. In Ceren Project 1993 Prelimi
nary Report, edited by P. Sheets and S. Simmons, pp. 46-90. Boulder: Univer
sity of Colorado.
Hart, W J. E., and V. Steen-Mcintyre
1983 Tierra Blanc ]oven Tephra from the A.D. 260 Eruption of Ilopango Caldera.
In Archaeology and Volcanism in Central America: The Zapotitan Valley of El Sal
vador, edited by P. Sheets. Austin: University of Texas Press.
Hayden, B.
Foundations of Social Inequality, edited by T. D. Price and G. Feinman,
pp. 15-86. Fundamental Issues in Archaeology, G. Feinman and T. D. Price,
general editors. New York: Plenum.
1998 Fabulous Feasts: A Prolegomenon to the Archaeological and Theoretical Im
portance of Feasting. Paper presented at the Society for American Archaeol
ogy Annual Meeting, Seattle.
Hayden, B., and A. Cannon
1984 The Structure of Material Systems: Ethnoarchaeology in the Maya Highlands. SAA
Papers No. 3. Washington, D. C . : Society for American Archaeology.
Hayden, B., and R. Gargett
1990 Big Man, Big Heart? A Mesoamerican View of the Emergence of Complex
Society. Ancient Mesoamerica q-20.
Lentz, D. L., M. P. Beaudry-Corbett, M. L. Reyna de Aquilar, and L. Kaplan
1996 Foodstuffs, Forests, Fields, and Shelter: A Paleoethnobotanical Analysis of
Vessel Content From the Ceren Site, El Salvador. Latin American Antiquity 7
(3): 247-262.
McAnany, P.
1995 Living with the Ancestors: Kinship to Kingship in Ancient Maya Society. Austin:
University of Texas Press.
McGee, R. J.
1990 Life, Ritual, and Religion among the Lacandon Maya. Belmont, Cal. : Wadsworth
Publishing Company.
Mobley Tanaka, ].
1990 1990 Excavations in Operation l, Ceren El Salvador. In 1990 Investigations at
the Ceren Site, El Salvador, edited by P. Sheets and B. McKee, pp. 36-47. Boul
der: University of Colorado.
Nash, M.
1958 Political Relations in Guatemala. Social and Economic Studies ?:65-75.
389
Linda A. Brown
Newman, M.
1993 Organic Residues on Obsidian Blades. In Geren Project 1993 Preliminary Report,
edited by P. Sheets and S. Simmons, pp. 182-184. Boulder: University of
Colorado.
Pohl, M.
1981 Ritual Continuity and Transformation in Mesoamerica: Reconstructing the
Ancient Maya Cuch Ceremony. American Antiquity 46 (3): 513-529.
Sheets, P. D.
1990 Chipped Stone, Ceren, 1990 Season. In 1990 Investigations at the Geren Site, ffi
Salvador: A Preliminary Report, edited by P. Sheets and B. McKee, pp. 176-181.
Boulder: University of Colorado.
l992a The Geren Site: A Prehistoric Village Buried in Volcanic Ash in Central America.
Case Studies in Archaeology Series. Fort Worth, Tex. : Harcourt Brace &
Company.
l992b Chipped Stone and Ground Stone Artifacts fromJoya de Ceren, 1991--92. In 1992
Investigations at the Geren Site, ffi Salvador: A Preliminary Report, edited by P. Sheets
and K Kievit, pp. 86-104. Boulder: University of Colorado.
Sheets, P. D. , ed.
1983 Archaeology and Volcanism in Central America: The Zapotitan Valley of El Sal
vador. Austin: University of Texas Press.
Simmons, S. and S. Villalobos
1993 Landscape Archaeology in Operation 8 between Household l and the Struc
ture ro Patio. In Geren Project 1993 Preliminary Report, edited by P. Sheets and
S. Simmons, pp. 31-45. Boulder: University of Colorado.
Stuart, G. E.
1997 The Royal Crypts of Copan. National Geographic 192:68--93.
Thompson, D. , and]. E. S. Thompson
1955 A Noble's Residence and its Dependencies at Mayapan. Carnegie Institution of
Washington, Current Reports 2 (25): 225-25r.
Thompson, ]. E. S.
1970 Maya History and Religion. Norman: University of Oklahoma Press.
Tozzer, A. M.
1907 A Comparative Study of the Mayas and the Lacandones. New York: Macmillan.
1941 Landa's relaciOn de las cosas de Yucatan. Translated by Alfred Tozzer. Cambridge:
Peabody Museum of Archaeology and Ethnology; Harvard University.
Vogt, E. Z.
1969 Zinacantan: A Maya Community in the Highlands of Chiapas. Cambridge: Har
vard University Press.
1976 Tortillas for the Gods: A Symbolic Analysis of Zinacanteco Rituals. Cambridge:
Harvard University Press.
Wisdom, C.
1940 The Chorti Indians of Guatemala. Chicago: University of Chicago Press.
Wolf, E.
1959 Sons of Shaking Earth. Chicago: University of Chicago Press.
390
FEASTING IN THE ANCIENT NEAR EAST
In the ancient Near East, the feast was a leitmotif in art (Collon 1992; Pinnock
1994) and cuneiform texts (Bottero 1994). Images carved in stone illustrate ban
quets of 4,500 years ago and economic tablets record the movement of goods
they occasioned. In this paper, I draw information from both art and texts to as
sess the role of feasting in Sumer in the third millennium B.C. I argue that the
feast was a significant factor in the Mesopotamian redistribution economy. I also
contend that the economic function of feasting was maxinllze d by the fact that
festivals were religious and sodopolitical events.
T H E M E S O P O TA M I A N C I T Y S TAT E S
Before starting my discussion on the feast, I summarize briefly the sodopolitical
and economic conditions of Mesopotamia in the third millennium B.C. Situated
391
in present-day southern Iraq, the country of Sumer was a low alluvial plain,
made of silt and lacking the most elementary raw materials such as stone, metal,
or even timber wood. As it is today, it was cursed in summer by a scorching heat
and would have been a barren desert if not watered by two rivers, the Tigris and
Euphrates. It is along the banks of the Euphrates, the less unpredictable of the
two rivers, that the Sumerians first settled. They transformed the dry silt by irri
gation and drained swamps into rich fertile soils. They could thus engage in
farming. They specialized in the cultivation of dates and barley, as well as the
raising of sheep and cattle.
The population was clustered where irrigation was possible along the river and
the canals. The first settlements were villages, but around 3500 B.C. the first cities,
such as Uruk and Girsu, appeared in the south. The process spread to the north
and along the Tigris with the rise of Nippur, Kish, Shuruppak, and Khafaje. Each
city with its surrounding villages was separated from the next by long stretches
of swamps and desert. The hinterland was a desert where nomadic tribes were
probably roaming. This geographic situation, in which each settlement was con
fined to relative isolation, favored the development of small political units called
city-states. In the third millennium, individual city-states may have had popula
tions of 5,000 to rn,ooo people.
The political leadership of the city-states was first in the hands of high priests
of the powerful Mesopotamian temple. However, by 29 00 B.C., a king ruled from
a royal palace, built next to the sacred temple area. The king's main function was
to build and keep in good repair the temples of the gods, to plan, oversee and
maintain the irrigation system and the defensive fortification wall surrounding
the city He was also in charge of raising and supplying the army, including costly
metal armaments. Finally, we may assume that the king and his queen played an
important ceremonial role, in particular, during the monthly festivals, which, I
will argue, generated the capital to support the royal endeavors.
F EASTI N G I N ART
The famous "Standard of Ur," which features the most detailed image of a Near
Eastern feast, illustrates the Sumerian king hosting a banquet (Moorey
1982:98-rn2; Fig. 14.r). The panel, inlaid with shell, limestone, and lapis lazuli,
shows the ruler as the largest figure of the top register, and sporting a kilt with
multiple tiers of fringes. The six guests facing him are identical in size, wear the
same garment with a single fringe, sit on similar wooden stools, make the same
gesture, and hold similar cups. A harpist and a singer entertain the elegant as
sembly while attentive servants are busy helping. Below, nineteen offering bear
ers, organized in five delegations led by ushers, stand in a long line stretching
over two registers. Unlike the stereotyped banquet guests, the gift bearers have
392
c:::::::>
=
Figure 14.r. Standard of Ur. After Denise Schmandt-Besserat, Before Writing, p. 1 73 , fig. 10 4 (Austin: University of Texas Press, 1 992).
individual features and assume different positions. Sumerians with shaven heads
and wearing fringed kilts lead the procession. Some of them pull or push a cow,
one escorts a small flock of goats and rams while another carries bundles of fish.
Foreigners follow wearing skirts overlapping in front. They are depicted with
curly hair, bearded or clean-shaven, and occasionally with a headdress. One of
the men brings a kid in his arms, another leads a team of asses, still others carry
heavy loads of goods on their backs or shoulders. The procession advances to
wards the right indicating that it is proceeding toward the banquet hall in order to
parade the goods before the king. The occasion for the feast is no doubt a mili
tary victory since the opposite side of the standard shows the king inspecting
prisoners in the aftermath of a battle. The inlaid panel therefore discloses that a
Mesopotamian victory feast was attended by males (military leaders?), drinking
and music was involved, and it included a procession of local offering bearers fol
lowed by the display of booty or tribute from foreign countries. The fact that
similar inlaid standards excavated at Mari, Syria, also combined banqueting with
an exhibition of goods (Parrot l953:figs. 64-66 and 70-72) confirms that the two
scenes belonged together as two episodes of a victory feast. It is noteworthy that
more space and attention is devoted to the display of merchandise suggesting
that the parade of goods was deemed the more significant of the two.
About twenty carved stone plaques excavated in the Sumerian city of Nippur
(Fig. 14.2) in sites of the Diyala Valley (Fig. 14.3) and Elam, depict a more popular
feast where music and competitive sports, such as wrestling, were taking place
(Boese 1971:Pl. X:2; Fig. 14-4). Humans and sometimes deities wearing horned
headdresses (Boese l971:Pl. XVIII:8-n; Fig. 14.2) are shown banqueting on the
upper register. As on the standard, the guests sit on wooden stools and usually
hold cups. Here, however, females are present. Servants and musicians are of ei
ther sex (Boese l971:Pl. 1:2; III:2; XVII:1). More importantly, the king faces a
woman of equal size and therefore of equal status-most likely the queen (Boese
1971:Pl. I:r; III:3; etc.; Fig. 14-3). On the lower registers, individuals are pictured
leading bulls, goats, and kids on the hoof, hauling jars, carrying on their heads
piles of goods or trays loaded with merchandise, while still others are holding
rings of precious metal or bundles of fish (Boese 1971:Pl. 1:2; III:1; V:2; IX:1; XVI:2,
3; XVII:r) . It is conceivable that the chariot with a team of four onagers, some
times depicted on the last register, was also a prized offering (Pallis 1926:155). The
carved plaques therefore concur with the standard for suggesting that feasting in
the ancient Near East meant not only banqueting but also gift giving.
The glyptic artists competed in virtuosity to depict feasts on the minuscule
field of cylinder seals (Collon 1992; Fig. 14-5). Here food was served. Meat
haunches and bowls filled with prepared dishes were displayed on high chests
(Amiet 198o:Pl. 9o:n82-n88). Beer, kept in tall jars, was sipped with metal straws
394
Figure 4.2 (top). Perforated plaque from Nippur. After Johannes Boese, Altmesopotamische
Weihplatten, PL XVIII: ! (Berlin: Walter de Gruyter, 1 97 1). Figure 4.3 (bottom). Perforated
plaque from Khafaje. After Johannes Boese, Altmesopotamische Weihplatten, PL IX: l
(Berlin: Walter de Gruyter, 1 97 1).
Figure 14-4- Perforated plaque from Khafaje. After Johannes Boese, Altmesopotamische
Weihplatten, Pl. X:Z (Berlin: Walter de Gruyter, 1 97 1).
(Amiet 198o:Pl. 9o:n83, n86-n87). As on the perforated plaques, the banquet

guests were human or sometimes divine (Amiet l98o:Pl. 92:1218-1221; Fig. 14.6).
They could include both sexes (Amiet l98o:Pl. 9o:n81-n84) or be exclusively
males or females (Amiet 198o:Pl. 9o:n92). Dancing took place to the tune of
harpists, flutists, and clappers (Amiet l98o:Pl. 9o:n93-n94). Offering bearers are
shown bringing fish, a kid, or a large jar (Amiet l98o:Pl. 9o:n87, l190-n91; Pl.
92:1218).
In sum, feasting was a popular motif of Mesopotamian art. It is noteworthy that
the monuments focused only on the economic aspect of the feast and never pic
tured the religious rituals that certainly took place during the festivals (Postgate
Figure 14-5· Seal impressions from Ur. After Pierre Amiet, La Glyptique Mesopotamienne Ar
chafque, PL 9 0: 1 1 90-1 1 9 1 (Paris: Editions du Centre National de la Recherche Scien
tifique, 1 9 80).
396
F EASTI N G I N TH E A N C I ENT N EAR EAST
Figure 14 . 6 . Seal impression from unknown site. After Pierre Amiet, La Glyptique Meso
potamienne Archafque, Pl. 92 : 1 2 1 8 (Paris: Editions du Centre National de la Recherche
Scientifique, 1 9 80).
1992:123-124). Instead they glorify banqueting and the parade of goods. This sug
gests that, in the ancient Near East, the importance of festivals was collecting as
well as consuming victuals. In fact, seen through art, the delivery of offerings may
be regarded as the most significant aspect of feasting. This seems very similar to the
chiefly feasting systems described by Kirch (Chapter 6) and Junker (Chapter ro).
F EAST I N G I N T H E T E XTS
Feasts were also a frequent topic o f mythological (Vanstiphout 1992:9-21), royal
(Postgate 1992:146-147), and economic texts, such as those excavated at Girsu
(Rosengarten l96oa:249-302). The cuneiform tablets complement the informa
tion provided by art with the advantage that the texts spell out the purpose of the
celebrations, their location, duration, and frequency. Moreover, the economic
tablets record the movement of goods that took place during the festivals and the
individuals involved.
The texts, as the art monuments, refer to multiple types of feasts. Some
marked special events such as the king's investiture, military victories, alliances
with neighboring countries (Bottero l994:n) and the inauguration of a temple or
a palace (Sauren 1970). However, Mesopotamian life revolved around monthly
festivals celebrating the multiple gods of the pantheon (Cohen 1996:14). For ex
ample, twice a year the people of Girsu honored the goddess Nanshe by a feast
lasting up to seven days (Cohem993:45). The two feasts celebrating the god
Ningirsu were four days long (Rosengarten l96oa:262) and that of his consort,
the goddess Bau, was two days (Rosengarten l96ob:39).
The tablets elucidate various features of the banquet images. For example,
they make it clear why deities were among the banqueters. According to the
397
texts, the feasts marked the days when the deities descended from heaven to be
worshipped. Since mankind was created to provide for their upkeep, the gods
and goddesses expected not only rituals, but also a banquet and lavish offerings
(Oppenheim 1964:187-193) . At her feasts, Nanshe was given silver crowns, vessels
of stone and of precious metal, and goods including animals on the hoof (Cohen
1993:46) . Other gods received meat, fish, flour, beer, oil, special dairy dishes, and
dates in quantities according to their rank in the pantheon (Rosengarten
196ob:40) .
The texts also clarify the prominent place given to the queen in the stone re
liefs by revealing her role as the intermediary between people and gods. The
household servants, herders, and farmers brought her offerings of small cattle,
bread, and beer, which she was in charge of sacrificing to the pantheon on their
behalf (Cohen 1993:37) . She also organized pilgrimages to attend the festivities
honoring Nanshe in the various cities of the state (Cohen 1993:44-45) .
We gather from the texts that the culmination of the religious festivals was a
sacrificial meal shared by the gods, the royal family; the priesthood, and the citi
zenry (Pallis 1926:173) . As pictured in art, the events involved drinking, toasting,
and merry making. The names of the two yearly Nanshe and Ningirsu festivals,
"the feast of barley eating" (She-ku) and "the feast of malt eating" (Bulug-ku),
leave no doubt that the feasts were the occasion for massive consumption. The
fabulous quantities of food involved are disclosed in an inscription of Ur-Nanshe,
king of Lagash, ca. 2500 B.c., who boasted of the 70 guru (equivalent to 30,800
kiloliters) of barley consumed at the inauguration of the temple of Ningirsu
(Cooper 1986:29) . In turn, the feast of Ur Nanshe presages the extravagant ban
quet offered by the Assyrian king Assurnasirpal II for the inauguration of his
palace of Nimrud in 879 B.C. The 69,574 guests were then entertained at the ex
pense of 300 oxen, 1,ooo calves, 15,000 sheep, 1,000 lambs, 500 deer, 500 gazelles,
·
10,000 eggs, 10,000 loaves of bread, 10,000 jars of beer, 10,000 skins of wine,
10,000 measures of chick pea, and so on (Finet 1992:31-35) .
That collecting goods was an essential part of the feast comes out loud and
clear in the tablets as it does in art. The stone carvings illustrating endless proces
sions of worshippers, bringing a kid in their arms, pushing animals on the hoof,
hauling jars, or carrying trays and loads of goods are echoed by the innumerable
lists of offerings compiled by the palace administration. The tablets stipulate sev
eral types of gifts. Some, specially earmarked for the banquet (Nig-ku-da, liter
ally "with" the meal), were offered by the priesthood, high officials, and palace
employees. They consisted of fish, meat, a certain milk dish or cheese, particular
types of breads, and beer. Second, goods including animals, bread, and beer were
brought to the palace by household servants, the palace farmers, and herders to
be sacrificed to the gods by the queen (Rosengarten 196ob:44) . Finally; the most
398
F E A S T I N G I N T H E A N C I E N T N E A R EAST
significant contributions were the so-called offerings for the feast provided by
professional guilds, such as that of the fishermen, who brought baskets of vari
ous kinds of fish and turtles as well as jars of fish oil (Rosengarten l96oa:255-256).
Individuals also contributed offerings for the feast. Servants of the princes, high
administrators, and foremen brought one animal. For example, a certain Ur
Lama is recorded as delivering a kid for "the feast of barley eating" and the same
for "the feast of malt eating" (Rosengarten l96oa:256). However-noblesse
oblige-the priest of Nanshe and his wife, are listed as each offering three animals
or respectively, one sheep and two lambs, and one sheep, one kid, and one lamb
(Rosengarten l96ob:41). The other high priests and their wives gave exactly the
same. For the feast of Bau, the contribution of the Priest of Nanshe and his col
leagues changed to one sheep, one kid, strong beer, black beer, loaves of bread,
bunches of vegetables, dates, and roasted barley (Rosengarten l96ob:39 ). As illus
trated in art, the most frequent offerings reported in the texts were animals on
the hoof, fish, and other victuals. But the contributions were not limited to edi
bles. Some texts from Umma and Ur list considerable deliveries of bundles of
reeds, and firewood (Limet 1970:67).
Although the texts reveal in amazingly specific details the monthly feast offer
ings a certain Ur-Lama paid some four thousand years ago, they fail to disclose
the big picture. For instance, the records are not complete enough to evaluate the
amount of revenues generated by banquet gifts, compared to those derived from
other palace resources, such as landholdings. It is also unclear what percentage of
the offerings was consumed at the feast and how much was levied by the palace.
The most significant information revealed by the economic texts is that the
same offerings were repeated over and over again. Month after month, the sea
fishermen delivered the same amounts of their catch-namely some 120 fish and
IO kg per fisherman; the administrators and foremen brought a kid; and the
priests invariably contributed three animals for Nanshe and two with other foods
and drinks for Bau. This strongly suggests that the "offerings for the feast" were
strictly regulated, mandatory contributions (Rosengarten l96ob:49). In our own
vocabulary they were monthly taxes.
T H E F EAST AS A R E L I G I O U S EVENT
The texts make it clear that the monthly festivals established a calendar stipulat
ing the days of the month when goods were to be delivered at the palace or the
temple. The feast, however, did far more than providing a rhythm for giving. The
fact that the festivals were religious made giving an awesome obligation.
The religion of the Sumerians was greatly influenced by their land. The deities
of their pantheon were natural powers on whose mercy they depended for their
daily sustenance. Major gods were Enki, the god of fresh water, Enlil, that of the
399
wind, and Utu, the sun. Inanna was the goddess of the storehouse and primarily
a fertility deity. Each city stood under the special protection of a patron god and
was responsible for catering to every possible need of that deity-who, like hu
mans, required food, drink, and shelter.
The temple perpetrated the idea that mankind had been created in order to
provide the pantheon with its upkeep (Lambert 199p97-198). Each household
was therefore expected to produce more goods than what it required for its own
needs and surrender the surplus to the gods as "gifts." The matter was urgent be
cause the deities were capable of doing evil as well as good. They were loving
gods, when satisfied with the care they received from their city, but wrathful
when displeased. They could in a whim generate favorable rains, produce plenti
ful harvests, multiply the animals of the herds, and bring prosperity to families.
Just as well, they could send famine, turn fields and orchards into barren waste
land, plague flocks or people with diseases, and unleash devastating storms. Each
individual, therefore, had the awesome responsibility of fulfilling the expecta
tions of the powerful deities. Each household was to avoid irritating the pan
theon by failing to attend the rituals and to provide quality goods in enough
quantities. Each individual held in his hands the welfare and the doom of the city.
Giving plentifully was not a matter of choice: it was a stern obligation.
T H E F EAST AS A S O C I O PO L I T I CA L EVE N T
The feast was not only religious, it was an important social and political event.
There can be no doubt that Sumerian festivals, like feasts all over the world, at
tracted crowds from far and near for the pageantry; music, and drama. City
dwellers and villagers flocked to view the king and his court in their ceremonial
paraphernalia and to hear the high priests utter sacred words while performing
sacred gestures. More importantly; the presence of the gods at the sacrificial meal
brought the entire community together. Finally; as depicted in art, after the ritu
als were over, it was time for merry making. Music, dance, and competitive
games gave everyone a chance to display elegance, beauty, grace, strength, and
dexterity in front of all. The feast was a place to see and be seen.
The royal family was the focal point of _the feast. We may assume that they
stood out from the crowds by wearing lavish ceremonial regalia. Most impor
tantly; the king and the queen played a key role in the ritual. As shown on the
Standard of Ur, the king presided over the procession of offerings. As represented
on the plaques, the royal couple hosted banquets. Finally; the texts spell out the
active part played by the queen in collecting offerings and making sacrifices. We
may infer that in all these various circumstances the king and the queen occupied
positions of honor segregated from the rest of the community. For instance, the
standard of Ur presents the king seated on one side, facing the guests. The festi-
400
F E A STI N G I N TH E A N C I E N T N E A R EAST
vals thus singled out the royal family, highlighting the importance of monarchy.
The feast reinforced the king and the queen's prestige and authority. By the same
token, the actual presence of the king to review the parade of gifts, and that of
the queen to receive the offerings, further increased the pressure for giving.
The procession of worshippers bringing gifts, glorified in art, certainly was a
main attraction of the feast. We would like to know whether it was compulsory
or voluntary to show publicly the offerings presented to the gods. Was each head
of household obliged to parade his contribution in view of the king and the en
tire community? Or, was the procession a special honor reserved for the bearers
of unusually priced products? In any case, one can well imagine that the public
exhibition of offerings created an extraordinary social pressure-because in
Sumer, as elsewhere, the human virtue most admired was generosity, and the
vice most despised, stinginess. Those who brought the most lavish gifts were
praised and admired, whereas those who supplied little were ridiculed and chas
tised. The competitiveness of giving is well illustrated by a wisdom composition
featuring the heated dialogue between Enten and Emesh, personifying Winter
and Summer, arguing the worth of their respective banquet gifts. The text
amounts to a battle of words between the two protagonists boasting about the
value of their respective offerings. Enten claimed that the lambs, kids, cattle,
honey, and wine he gave were far superior to the bountiful harvests brought by
Emesh, and vice versa (Vanstiphout 1992:14-15). At the feast, citizens sought ap
proval from the gods, but also prestige in the community.
Texts and art are mute concerning the consequences of failing to attend the
feast and to bring offerings. Martial coercion is not discussed or depicted, except
in a series of early fourth-millennium seals showing the priest-king presiding over
the beating of prisoners (Schmandt-Besserat 1992:180-181). It is conceivable that
in the third millennium B.c., ideological coercion was put to work for people to
bring the most lavish gifts in the greatest quantity. Failing to partake was proba
bly deemed such a profound insult to the gods and the community that, as a rule,
it did not occur. Otherwise, society deals efficiently with delinquents with a
gamut of punishment ranging from ridicule to ostracism. The festivals were the
time when fame and shame were created.
CO N C LU S I O N
The combined information derived from the art and economic texts of the ancient
Near East suggests that, in the third millennium B.C., the religious, social, and politi
cal aspects of feasting were manipulated to amass palace wealth. (1) The feasts hon
ored deities and therefore giving was obligatory. Failing to provide for the gods
spelled divine rage, famine, epidemics, and doom for the community. (2) The pro
cession of offerings was a social event when, in the midst of pageantry, before the en-
401
tire citizenry; worshippers rivaled each other in presenting the most generous gift. (3)
The queen, acting as intermediary between the people and the pantheon, gave the
palace economic control. Finally; the king was the central figure of the ceremonies
and the ultimate recipient of the offerings. The delivery of "gifts to the gods" set in
motion the wheels of the Mesopotamian redistribution economy. The contributions
forced each household to produce a surplus. Of course, all the monthly deliveries of
animals and merchandise were not consumed at the feast, but increased the palace
herds and granaries until their redistribution (Rosengarten 196oa:44). In
Mesopotamia the feast was a fulcrum of the state redistribution economy.
ACKNOWLEDGMENT
Research assistance by Ines Rivera is gratefully acknowledged.
REFERENCES
Amiet, Pierre
1980 La glyptique Mesopotamienne archafque. Paris: Editions du Centre National de
la Recherche Scientifique.
Boese, Johannes
1971 Altmesopotamische Weihplatten. Berlin: Walter de Gruyter.
Bottero, Jean
1994 Boisson, banquet, et vie sociale en Mesopotamie. In Drinking in Ancient Soci
eties, edited by Lucio Milano, pp. 3-13. Padova: Sargon srl.
Cohen, Mark E.
1996 The Sun, the Moon, and the City of Ur. In Religion and Politics in the Ancient
Near East, edited by Adele Berlin, pp. 7-20. Bethesda: University Press of
Maryland.
1993 The Cultic Calendars of the Ancient Near East. Bethesda: CDL Press.
Collon, Dominique
1992 Banquets in the Art of the Ancient Near East. In Banquets d'Orient, edited by
R. Gyselen, pp. 23-30. Res Orientales Vol. IV, Groupe pour l'Etude de la Civi
lisation du Mayen-Orient: Bures S /Y.
Cooper, Jerrold S.
1986 Sumerian and Akkadian Royal Inscriptions, Vol. r . New Haven: The American
Oriental Society.
Finet, Andre
1992 Le Banquet de Kalah offert par le Roi d'Assyrie Ashurnasirpal II (883-859) . In
Banquets d'Orient, edited by R. Gyselen, pp. 31-43. Res Orientales Vol. Iv,
Groupe pour l'Etude de la Civilisation du Mayen-Orient: Bures S / Y.
Lambert, W G .
1993 Donations of Food and Drink to the Gods in Ancient Mesopotamia. In Ritual
and Sacrifice in the Ancient Near East, edited by]. Quaegebeur, pp. 191-201. Leu
ven: Uitgeverij Peeters en Departement Orientalistiek.
402
F E A S T I N G I N T H E A N C I E N T N E A R EAST
Limet, H.
r970 L'organisation de quelques retes mensuelles a 1' epoque neo-Sumenenne. In
Actes de la XVIIe Rencontre Assyriologique Internationale, edited by Andre Pinet,
pp. 59-74· Ham-sur-Heure: Comite Belge de recherches en Mesopotamie.
Moorey, P. R. S.
r982 Ur 'of the Chaldees': A Revised and Updated Version of Sir Leonard Woolley's Exca
vations at Ur. Ithaca: Cornell University Press.
Oppenheim, A. Leo
r964 Ancient Mesopotamia. Chicago: University of Chicago Press.
Pallis, Svend Aage
r926 The Babylonian Akitu Festival. Copenhagen: Bianco Lunos Bogtrykkeri.
Parrot, Andre
r953 Mari. Collection des Ides Photographiques, vol. 7. Neuchatel: Edition des
Ides et Calendes.
Pinnock, Frances
r994 Considerations · on the 'Banquet Theme,' in the Figurative art of
Mesopotamia and Syria. In Drinking in Ancient Societies, edited by Lucio Mi
lano, pp. r5-26. Padova: Sargon srl.
Postgate, ]. N.
r992 Early Mesopotamia. London: Routledge.
Rosengarten, Yvonne
r96oa Le Concept Summen de Consommation dans la Vie Economique et Religieuse. Paris:
Editions E. de Boccard.
r96ob Le Regime des Ojfrandes dans la Societe Summenne. Paris: Editions E. de Boc
card.
Sauren, H.
r970 Les fetes neo-sumeriennes et leur periodicite. In Actes de la XVIIe Rencontre As
syriologique Internationale, edited by Andre Pinet, pp. n-29. Ham-sur-Heure:
Comite Belge de recherches en Mesopotamie.
Schmandt-Besserat, Denise
r992 Before Writing. Austin: University of Texas Press.
Vanstiphout, H. L. ].
r992 The Banquet Scene in the Mesopotamian Debate Poems. In Banquets
d'Orient, edited by R. Gyselen, pp. 9-2!. Res Orientales Vol. Iv; Groupe pour
l'Etude de la Civilisation du Moyen-Orient: Bures S/Y.
403
GAR BAGE AND THE M ODERN A M ERICAN FEAST
I n the last several decades, home enterta i n i n g i n America has evolved d ra m atica lly
from a matter of rules a n d reg imens into a very personal, freewh eel i n g affa ir.
Martha Stewart (1 982: 12)
The Crystal Ballroom in Portland, Oregon, operated between 1914 and 1968,
hosting a diverse cross-section of American music and dance, from Dixieland
jazz, waltzes, and "old time" traditional (square) dancing to soul music and psy
chedelic light shows. Historian Tim Hills's (1997) monograph on the ballroom
documents that in the late 1950s and early 1960s, the operators augmented their
revenues by leasing out the hall for Gypsy celebrations, including Feasts of the
Dead. Attempts by Hills in the 1990s to interview participants at these rnid
twentieth-century feasts were stymied by the Gypsy community's secretive and
exclusionary nature, and their inherent mobility (Timothy Hills 1998, personal
communication). The only reports on these Gypsy feasts are recollections by the
operators of the Ballroom, who described the "vibrant attire," bouquets of flow-
404
G A R B A G E A N D T H E M O D E R N A M E R I C A N F E AST
ers, tall candles, unfamiliar music, and exotic dance steps. What was most clearly
remembered, however, was the massive cleanup afterwards, with gallons of
spilled beer and pounds of roast lamb, pork, chicken bones, and waste scraps that
had been casually discarded on the floor. One operator went to the trouble of
taping cardboard over the entire dance floor prior to the Gypsy feasts to facilitate
the subsequent cleanup.
In discussing feasting in industrialized societies, and the United States in par
ticular, the Gypsy celebrations at the Crystal Ballroom are important anecdotes
because they illustrate the difficulty in characterizing an ''.American" feasting be
havior. The broad ranges of ethnic groups and economic classes that make up
the United States participate in an equally daunting array of feasting activities,
and it is difficult, even for those who live within the culture, to understand the
full range of extant feasting behaviors. The Gypsy feasts also illustrate the prob
lem of ethnographic-respondent recall and reporting, especially when the feast
organizers and participants are unwilling to be interviewed. This bias in inform
ant interviews, which has been discussed at length elsewhere (e.g. , Ritenbaugh
and Harrison 1984; Sechrest 1979; Webb et al. 1966), underscores the importance
of the archaeological examination of feasting behavior, since the study of mate
rial objects is not subject to the same biases as traditional ethnographic tech
niques (Rathj e 1995; Rathje and Murphy 1992:53-78).
In this chapter, we espouse a materials-based approach to the study of the
American feast. We argue that the social and economic processes that have led to
the widespread availability of specialty foods, the mass-production of low-cost
commercial goods, and the shrinking size and increased financial independence
of households in urban and suburban settings, have democratized, downsized,
and fragmented the American feast. We suggest that many of the functions of
American feasts have fallen away with their formal attributes, leaving "parties"
that function primarily as solidarity feasts.
TYPES O F A M E R I CA N F EASTS
In attempting t o discriminate between the many types of feasts in the United
States, it is first appropriate to segregate those feasts held in and by the household
from those given by larger social groups in nonhousehold settings. Although one
might expect that this distinction has something to do with the number of feast
participants or the size of the feast, this is not always the case. Some household
gatherings can include hundreds of guests, whereas some business feasts in
clude only a handful of people. Sometimes the only distinguishing feature of a
household-level feast is its setting in a residence.
Inspection of a range of twentieth-century ''.American" cookbooks and enter
taining guides suggests a variety of idealized "emic" types of household celebra-
405
tions. Biddle and Blom's (1937), The Book of Table Setting, for example, details
those times when "significant" table settings are appropriate, including family
birthdays, Washington's Birthday, Lincoln's Birthday, Easter, Election Day, Valen
tine's Day, Independence Day, New Year's Day, St. Patrick's Day, April Fools Day,
Memorial Day, Columbus Day, and wedding anniversaries. Berolzheimer's
(1963:55-59) The American Woman's Cookbook identifies menus for "special occa
sions," including (in order), a St. Patrick's Day Luncheon, three varieties of
Thanksgiving dinner, two varieties of Christmas dinner, four wedding menus, af
ternoon tea, afternoon or evening refreshments, a children's party, a school re
ception, a bridge party supper, a men's card party, and a cocktail party. Likewise,
Meta Given's (1952:1573, 1576-1579) Modem Encyclopedia of Cooking, identifies two
types of wedding breakfast, a wedding tea, a wedding dinner or supper, three
types of Thanksgiving dinner, two types of Christmas dinner, two types of New
Year's Day dinner, three menus for showers, two types of Easter menu, two types
of Fourth of July menu, a Washington's Birthday luncheon, a Christmas lunch
eon, a 'Hallowe' en" party, a birthday luncheon or dinner party, an after-skating,
skiing or football "get-together," and an after-school Valentine's Day party. Other
holidays not mentioned in these mid-twentieth-century guides include Martin
Luther King's Day, Veterans Day, and Presidents' Day. The generic nature of
most twentieth-century cookbook and entertaining guides is reflected in the fact
that no non-Christian religious occasions are mentioned.
Cocktail parties are perhaps the most American of feasts, but as a type they are
hard to define. Martha Stewart's (1982:30-61) menus for cocktail parties range
from small gatherings of 6-8 people to a '1arge extenuated bash" of 200. Susan
Roane's (1988:133) business guide, How to Work a Room: A Guide to Successfa.lly Man
aging the Mingling, suggests that there are three types of cocktail parties: social,
business, and fundraisers. Social cocktail parties often have a theme, based on a
holiday or event, such as an engagement, housewarming, or holiday. Business
cocktail parties include no-host receptions at conferences and business socials, in
cluding company picnics and the infamous year-end office party.
Sally Quinn (1997), who provides a unique insight into feasting behaviors in
Washington D.C . , describes a variety of functions, including (1) political parties,
where information and ideas can be exchanged in an informal setting, (2) eco
nomic parties, which are designed to meet new contacts and reconnect with old
ones, and (3) parties that are given for a person or to celebrate an event, such as a
book signing.
Another set of American feasts, for which there are no guides, are those re
lated to institutionalized settings, including campus fraternity and sorority par
ties, and payday and "leave" binges at logging camps, military posts, and other
work-related settlements. Alcohol-dominated parties, related to pathological
406
drinking behavior, also include those conducted by youths in areas hidden from
parental and local authorities. "Raves" and other youth parties are poorly docu
mented and only partly understood. Although Staski (1984) has detailed the po
tential for historical archaeology to address alcohol-consumption behavior, rarely
have archaeologists considered their bottles of liquor, ale, beer, and wine as a
possible reflection of feasting behavior.
Other American feasting types include dances and balls, bar mitzvahs, and fu
nerals. Tourist entertainment feasts, such as Hawaiian luaus or "medieval ban
quets," are feast types specifically designed for tourist consumption (literally) .
FIVE TUCSON F EAST I N G CASES

Though idealized American feasting types can be identified in the menus and en
tertaining guides that have flourished in the late twentieth century, they cannot
alone predict how feasts in America are actually held, including how feasting ma
terials are purchased, prepared, used, consumed, and wasted. In this section, we
examine feasting refuse in modern Tucson, Arizona, discussing some of the vari
ability in the material remains of these feasts and what they generally reveal
about feasting in the United States. Our focus is narrowed to the household-level
feast and limited to the distinctions in feasting that are visible archaeologically
and that can be studied in a material sense. In our discussion, below, we treat
American feasting in a more general context, extrapolating the patterns seen in
the Tucson households to other social groups.
Over the years we have discerned many differences between the reported per
ceptions and understanding of a particular behavior and the material realities
that represent the consequences of that behavior. Through taking an archaeolog
ical, rather than ethnographic or sociological approach to American feasting, we
hope to provide information that is of practical value to archaeologists and that
will contrast with our own, imperfect, understanding of feasting in industrialized
societies (see Rathje 1995; Rathje and Murphy 1992:53-78).
Household garbage in Tucson, Arizona, is picked up twice a week. Since 1972,
sanitation workers in Tucson have systematically gathered samples of household
garbage left at the curbside and delivered them to student volunteers and research
archaeologists of the University of Arizona's Garbage Project. The unit of cul
tural deposition for this study, therefore, is the individual household pickup, repre
senting all the refuse placed at the curb for collection on a particular pickup day.
Because we focus on the household pickup, our analysis is limited to those behav
iors that occur within the household, and excludes those feasts that take place in
restaurants, meeting halls, ballrooms, and elsewhere outside the household.
Residues from two different time periods were examined for evidence of feast
ing behavior: the Fall 1991 sample and the combined Spring and Fall 1994 sample.
407
Each pickup sampled during these two periods was painstakingly pored over by
University of Arizona student researchers, documenting each specific item in the
garbage. The goal of the present analysis is to identify the material remains of a
small sample of Tucson feasts, and to detail those material associations for the
types and quantities of foods, beverages, tobacco, and other material items used
and discarded, the types of special serving vessels, the presence of conspicuous
waste of food, and any evidence for the number and composition of people pres
ent. Based on the material remains, we also attempt to make some inferences on
the likely organizers and contributors to the feast, the likely purpose of the feast,
aspects of material redistribution, and other important elements of feasting
behavior. Although we rely almost entirely on the material record to guide our
interpretations, we acknowledge that we are participants in the culture we
are studying and shamelessly employ our unique knowledge and experiences,
and also an understanding of the American feasting calendar, to help in our
interpretations.
The Fall 1991 sample comprised 65 pickups collected at various times between
September 27 and December 5. Project archaeologists recorded more than 8,ooo
individual items. Four of the sample pickups were collected on December 2, the
Monday after the November 28 Thanksgiving holiday. Two of these pickups con
tained good evidence for Thanksgiving feasting behavior. The first feasting type
we discuss, then, is perhaps the most traditional of all the secular holidays of the
United States. Both pickups were collected from a census tract of Iniddle-class,
predominantly Anglo-American families on the east side of Tucson.
CAS E O N E : T H A N KSG IVI N G

A picture of the Thanksgiving feast associated with Sample C45-120291 (which we
refer to as Case One) emerges from the close examination of the material items.
Noting that any of these material objects could be found in any household
pickup, it is the association of items, the high frequency of items, the diversity of
items, and the date of deposition, that provides the best evidence for feasting be
havior. Artifacts believed to originate from the Thanksgiving feast are listed in
Table 15.1. In the practice of American historical archaeology; we have included
the particulars of brand names and container sizes, along with frequency and
type distinctions. This provides a clear material-culture context for the artifacts
and an associational framework for the comparative study of feasting behavior in
historical archaeology. Other students of material culture interested in identify
ing feasting behavior also may find it useful.
As discussed in the introductory chapter, one distinguishing attribute of feast
ing behavior is the preparation and serving of special foods. Not surprisingly,
408
TABLE 1 5 . 1
Artifacts Related t o a Tha n ksgivi ng Feast i n S a m p l e (45- 1 2029 1
Type Material Brand Size Description No.
Food plastic Sager Crest turkey wrapper

Food fauna! turkey bones 17
Food botanical 48 g potato peel
Food plastic preprepared potato wrapper 2
Food paper Kingston 7 oz. macaroni and cheese wrapper 1
Food ferrous can Cablecar 20 oz. peach can
Food ferrous can Kingston IO oz. sweet pea can
Food plastic Rambow 48 oz. bread wrapper
Food paper I ferrous Pillsbury 8 oz. biscuit mix box
Food faunal/botanical 350 g "slop"
Snack food paper Fritos 3 oz. corn tortilla chip wrapper
Snack food plastic Saguaro 1 oz. potato chip wrapper
Snack food plastic Tootsie Roll 5 lb candy wrapper
Snack food paper Brachs chocolate candy wrapper
Snack food paper Lifesavers hard candy wrapper
Snack food paper Topps 1 oz. bubble gum wrapper
Snack food paper Starburst fruit chew candy wrapper 2
Snack food paper Kit Kat 2 oz. chocolate candy wrapper
Food preparation/ serving aluminum foil sheets 7
Food preparation/ serving textile kitchen towel
Food preparation/ serving textile I plastic table cloth
Food preparation/ serving plastic trays 4
Food preparation/ serving paper plates 38
Food preparation/ serving plastic cups 3
Food preparation/ serving paper paper towel roll 2
Food preparation/ serving plastic Bounty paper towel wrapper
Nonalcoholic beverages aluminum Pepsi 12 fl. oz. soda can
Alcoholic beverages aluminum Old Milwaukee 12 fl. oz. "light" lager beer can 3
Alcoholic beverages aluminum Hamms 12 fl. oz. '1ight" lager beer can 3
Alcoholic beverages aluminum Budweiser 12 fl. oz. '1ight" lager beer can
Alcoholic beverages paper Old Milwaukee 12-pack lager beer label
Beverage related plastic six-pack ring
Other plastic Crystal Ice ice bag
Tobacco textile I paper Marlboro filter tip 149
Tobacco textile I paper Marlboro Lights filter tip 18
Tobacco textile /paper Kool Lights filter tip 10
Tobacco textile I paper Wmston filter tip 12
Tobacco textile I paper Viceroy filter tip
Tobacco paper I plastic Doral Lights pack
Tobacco paper Winston carton
roasted turkey is one of the more archaeologically visible correlates of Thanks

giving feasting in the United States. For example, Dickens and Bowen (1980:51--s3)
used the relative frequency of turkey bones from a 1910 refuse deposit, associated
with a "working class" neighborhood in Atlanta, Georgia, as one means to infer
early-twentieth-century Thanksgiving and Christmas holiday behavior.
As shown in Table 15.1, the most obvious food items related to the Case One
feast include a plastic "Sager Crest" turkey wrapper and 17 turkey bones. The
bones found in this single pickup represent over one-quarter (27 percent) of the
poultry bones (n=64) and 16 percent of the total bones (n=109) documented dur
ing the Fall analysis. As in the Dickens and Bowen (1980) study; the frequency of
turkey items in Case One provides a strong indication of holiday feasting behav
ior. Other food items that may be related to the feast (Table 15.1) include various
fresh and packaged food items, including 350 grams of mixed and unidentifiable
food debris (classified as "slop" by the Garbage Project researchers). Snack foods
are well represented, including both chips and candy.
There are several materials in the Case One pickup related to the preparation
and serving of the Thanksgiving feast. These include aluminum foil sheets, plas
tic trays, and 38 paper plates, suggesting that at least part of the feast was served
on disposable dinnerware. Evidence for beverage use at the feast is dominated by
alcoholic beverages, with at least three types of low- to moderate-cost American
lager beers. Tobacco consumption is represented by 190 filter tips from five differ
ent brands of cigarettes. These filters represent over 14 percent of all the cigarette
butts recorded in the Fall 1991 sample of pickups. The quantity and diversity of
cigarettes recorded in the Case One pickup provide good evidence that this
Thanksgiving feast included an extended-family or suprahousehold gathering,
with perhaps as many as six different smokers.
The image that emerges from Case One is of an extended family gathering
with a traditional turkey dinner, notable alcohol consumption, the use of dispos
able paper plates, and a smoke-filled environment. Although some foods were
apparently prepared from scratch, such as the turkey, many were purchased
preprepared from the grocery store. There is no evidence for the use of expen
sive or exotic materials and much of the serving materials are cheap disposables.
CASE TWO: THAN KSGIVI N G

Sample D45-120291 (Case Two) also contained good evidence for a Thanksgiving
feast, but one that differs somewhat from that documented in Case One. For the
sake of brevity; we have not listed in detail the materials related to feasting, but
merely describe their general characteristics and how they compare with Case
One. Case Two contained plastic wraps from a "Honeysuckle" brand turkey and
a four-pound "Foster Farms" duck, but only a single bone (obviously, there were
41 0
G A R B A G E A N D TH E M O D E R N A M E R I C A N F E AST
leftovers). Other fresh and packaged feast-related food items include oysters, po
tatoes (peels), broccoli, corn-on-the-cob, rolls, and pumpkin bread. Materials re
lated to cooking and serving include an aluminum pie tin, an aluminum plate, a
ferrous metal bowl, eight plastic trays, five plastic dishes, a paper cup, a plastic
cup, one ferrous metal fork, and two plastic spoons.
As with Case One, there is evidence for consumption of nonalcoholic bever
ages, including coffee and soda, but the majority of beverage materials are re
lated to the consumption of alcohol, including 40-ounce glass bottles of
inexpensive malt liquor and quart bottles of inexpensive lager beer. Tobacco
products in Case Two are also represented, although in much lower numbers
(n=43 filters and two packs) and varieties (n=3) than Case One. This suggests to
us that there were relatively few smokers at the feast, and possibly a smaller
group of feasters, perhaps only a single household.
The material discards from the two Tucson Thanksgiving feasts are quite dif
ferent, even though they were held in the same year in the same neighborhood.
Case Two contains fewer packaged items, more fresh foods, and probably was
held for fewer people than the Case One feast. One common element in both,
however, is the consumption of relatively large amounts of alcoholic beverages.
Even in the 1990s, when the health effects of alcohol and the dangers of drunk
driving have been well publicized, alcohol appears necessary to the success of
many American feasts (see Quinn 1997:115).
C A S E T H R E E : H A L LO W E E N
The largest quantity of alcohol-related items found in a sample pickup from the
Fall of 1991 is also closely associated with a typically American holiday. Four sam
ple pick.ups were collected on November 5, which was the Tuesday pick.up after a
previous Thursday Halloween. Of the four, three contained evidence of partici
pation in the Halloween celebration. Two of these contained a relatively diverse
range of candy wrappers and some Halloween paraphernalia, including bags and
balloons. In contrast, Sample D45-110591 (Case Three) contained no candy wrap
pers or food debris of any kind. Instead, the sample contained 92 cans of low- to
average-cost lager beer in 12- and 16-ounce sizes, comprising three brands and five
varieties, including two varieties of "light" beer. These represent 66 percent of all
the beer artifacts recorded in the entire Fall 1991 sample. Tobacco smoking was
represented by 37 filter tips and an empty pack, all representing a single brand of
cigarettes. The remainder of the sample consisted of 9 paper grocery sacks.
According to Martha Stewart (1982:33) the ideal home bar for "entertaining"
has 21 different types of hard liquor and liqueurs, 6 bottles of white wine, 3 bot
tles of red wine, a 6-pack of '1ight" beer, a 6-pack of "dark" beer, and a 6-pack of
imported beer. It is likely that the home bars associated with the Case One to
41 1
Case Three samples fall short of this ideal. The dominance of inexpensive lager
beers in the samples belies the fact that they are derived from a middle-income,
Anglo-American neighborhood. In this case, as with many instances of American
feasting behavior, the cultural norms dictated by entertainment guides, cook
books, and advertising differ considerably from the actual behavior of American
households.
C A S E F O U R : P R E S I D E N T S ' D AY
Turning to the 1994 sample, we find two other pickups associated with high alco
hol consumption, each of which, however, represent very different expressions of
feasting behavior. A total of 63 pickups was collected between February 15 and
December 1, 1994, representing nearly 11,000 recorded items. The artifacts associ
ated with feasting in Sample C7-022394 (Case Four) are listed in Table 15.2. As
with the Case Three Halloween feast, many artifacts are associated with alco
holic beverages (n=33 cans and 3 cartons), representing 33 percent of all the
recorded beer items in the 1994 sample. These containers, however, represent
only a single variety of moderate-cost lager. As shown in Table 15.2, tobacco con
sumption is also well represented by 126 filter tips and 11 packs, together repre
senting two brands and four varieties of cigarettes. As listed in Table 15.2, a trace
TABLE 1 5.2
Artifacts Related to a Presidents' Day Feast in Sample Cl-022394
Food corrugated
cardboard pizza boxes 2
Food preparation/ serving paper Dixie plates 7
Food preparation/ serving plastic Dart plastic fork

Alcoholic beverages paper Budweiser 12-pack lager beer can 3
Alcoholic beverages aluminum Budweiser 12 fl. oz. lager beer can 33
Nonalcoholic beverages glass Seagrams 1 liter ginger ale bottle
Nonalcoholic beverages plastic Coca Cola 20 fl. oz. cola bottle
Tobacco textile I paper Marlboro filter tip 92
Tobacco textile I paper Marlboro light filter tip 23
Tobacco textile I paper Camel light filter tip 10
Tobacco textile I paper Camel light
menthol filter tip
Tobacco paper I plastic Marlboro pack 6
Tobacco paper I plastic Marlboro light pack 2

Tobacco paper I plastic Camel Light pack 3
Illicit drugs trace marijuana
41 2
of marijuana was also recorded. The pizza boxes provide the best evidence for
foods consumed at the feast, although fast-food pack.aging found in the pick.up
(not listed) also could be related to the feasting activities. Serving items recorded
in the sample include paper and plastic dinnerware.
Because there was only a single brand of beer, we initially thought that Case
Four was associated with pathological alcohol consumption. A consideration of
the date of the pick.up, though, along with the amount and variety of cigarette
butts, suggests that it is not. The sample was collected Wednesday, February 23,
1994, which closely followed the Presidents' Day holiday, on Monday, February
21. If these materials do represent pathological drinking, then this behavior is cor
related with, and perhaps accentuated by, the holiday weekend. Holidays, such as
paydays, are probably often an excuse for binge drinking. Many entertainment
guides stress that there should be a defined purpose for a party. We hypothesize
that in many American households, the excuse for feasting is that there is a sur
plus of cash, a day or two off work, and a thirst for ethyl alcohol.
C A S E F I V E : M I D W E E K PA R T Y
Sample A7-031194 (Case Five) was collected o n March 11, 1994, which was a Friday
pick.up not associated with any common United States holiday. The contents of
Case Five are listed in Table 15.3. Alcoholic beverage artifacts are again dominated
by beer containers, although a much greater variety of beers are present (seven
brands), including more expensive imported Mexican lagers and specialty "craft
brewed" ales. Other alternative "malt beverages" are also present, and perhaps
most interesting from a material-culture perspective, all of the containers are
glass. The beer containers in this pickup represent 29 percent of all the beer items
recorded in the 1994 sample. As with the other cases, nonalcoholic beverages are
few in number.
Food items in Case Five are dominated by a variety of snack foods, including
nuts, popcorn, and chips. The two plastic deli cups and three plastic containers
may have held take-out party foods or snacks. Notably, there were no paper
plates, cups, or silverware. The 60 grams of fresh lime waste are undoubtedly as
sociated with the serving of the Mexican beer. There were no cigarette butts or
other tobacco packaging.
Obviously, the materials recorded in Case Five contrast greatly with the Case
Four Presidents' Day feast. A variety of higher-cost, imported beers are repre
sented, and there is a selection of snacks and specialty foods. Smoking items are
conspicuously absent. Based entirely on the types of artifacts present, it is very
likely that the goals of the Case Five feast were considerably different from that
of Case Four, especially in portraying status through beverage selection (a variety
of more expensive imported and craft beers), food selection (snack. foods), a feast-
41 3
TABLE 1 5.3
Artifacts Related to a Midweek Pa rty i n Sample A?-03 1 1 94
Food paper Wonder Roast chicken wrapper

Food fauna! 504 g chicken bones
Food botanical 60 g lime
Snack food plastic Laura Scudder popcorn wrapper 2
Snack food paper Golden Valley 3 oz. popcorn wrapper
Snack food botanical 18 g popcorn waste
Snack food plastic Eagle 6 oz. potato chip wrapper
Snack food plastic Reina 6 oz. tortilla chip wrapper
Snack food plastic Penguina hard candy wrapper
Alcoholic beverages glass Dos Equis 12 fl. oz. lager beer bottle 11
Alcoholic beverages glass Corona 12 fl. oz. lager beer bottle 8
Alcoholic beverages glass Amstel Light 12 fl. oz. lager beer bottle 4
Alcoholic beverages glass Bohemia 12 fl. oz. lager beer bottle 2
Alcoholic beverages glass Columbus 12 fl. oz. lager beer bottle
Alcoholic beverages glass Bridgeport 12 fl. oz. ale bottle
Alcoholic beverages glass McTarnahan's 12 fl. oz. ale bottle
Alcoholic beverages corrugated
cardboard Corona 12-pack lager beer box
Alcoholic beverages paper Dos Equis 6-pack lager beer label
Alcoholic beverages paper Winterfun 6-pack beer label
Alcoholic beverages glass Zima 12 fl. oz. malt beverage bottle 5
Alcoholic beverages paper Zima 6-pack malt beverage label
Nonalcoholic beverages aluminum Pepsi Cola 12 fl. oz. cola can 2
Nonalcoholic beverages paper coffee filters 2
Nonalcoholic beverages botanical coffee grounds
ing "etiquette" or a lifestyle that does not include tobacco consumption, and food
service that does not employ disposable dinnerware.
DISCUSSION
This brief foray into feasting in modern Tucson, Arizona, provides some primary
material data to discuss feasting behavior in the United States. Our first impres
sion is that there are many feasts. The five cases of feasting, noted here, represent
4 percent of the 128 pickups for the combined Fall 1991 and Spring/ Fall 1994 sam
ples, and we have merely scratched the surface of the data, pulling out the most
compelling examples. Even at 4 percent, it can be roughly estimated that feasts in
Tucson are occurring at a rate of one per household every roo days or so.
Second, certain types of foods, and most types of alcohol and tobacco prod
ucts, are consumed at United States feasts in high frequencies compared with reg-
41 4
ular, everyday consumption of these items. Archaeologically, the material

residues of feasting in Tucson skew the total frequencies of residues related to
food and beverage use. Certain specialty items, such as fresh, whole turkey, are
consumed predominantly during year-end holiday seasons. Because of the
unique associations of materials and the diversity and high frequency of items,
the residues of many modern feasts are probably archaeologically observable.
T H E D E M O C R AT I Z AT I O N O F T H E A M E R I C A N F E A S T
Another obvious characteristic of the five cases documented here is in the flexi
bility of expression of the feasting behavior. The particular choices of foods, bev
erages, and table settings, is based somewhat on tradition and "emic"
prescriptions, but also clearly on choices dictated at the household level about
what is appropriate for the guests. This observed flexibility is echoed in the
twentieth-century entertaining guides. For example, Stewart notes that there are
no longer rigorous prescriptions for social habits and rituals, but that "the grow
ing body of experience in America-social as well as culinary-has fostered a
new openness and respect for diversity" (1982:12). Biddle and Blom advise 'l\.s we
may consider ourselves unhampered by the requisites of the formal dinner we
may let our motto be: try anything" (193p9) .
There are a few bastions of formal feasting etiquette that remain. For example,
The International Guild of Professional Butlers (199 S ), provides standards for
service techniques over the Internet, including topics on "the Banquet," "the For
mal Dinner," and "Entertaining." Quinn (1997:153-154) recommends that persons
of "obvious rank or stature," such as the President of the United States, the sec
retary of state, a state's governor, or a city's mayor, should be given the seat of
honor. However, in almost the same breath, she discourages seating guests
"below the salt" and espouses the use of round tables to balance status differ
ences and to treat all guests as if they were "celebrities." In America, the status
differences of feasters appear to be kept well hidden.
It is also important to note that all of the items found in the Tucson refuse
samples could be purchased from any grocery store in any town or city in the
United States. In a sense, the United States has democratized feasting through
making delicacies available to almost everyone year-round and through the mass
production of feasting paraphernalia, including low-cost china, decorations, and
other elements of buffet and table settings. Only the most expensive, truly elite
goods, are out of reach of most Americans. Although it is within the economic
means for many households to pay $20 to $40 for a bottle of wine for a special oc
casion, there are only a few who will consume the most expensive products. No
tably, the archaeological evidence for use of these extremely high-end products is
exceedingly rare. They are simply not found in common grocery stores, or even
41 5
in many specialty stores, and since such a small segment of the population is
using them, they are archaeologically invisible. The notion of "saturation point"
or decreasing marginal utility for white ceramics, considered by Spencer-Wood
and Heberling (1987:60) is valuable in this regard. After attaining a certain level of
value in feast foods and table settings, little is gained by additional expenditures,
regardless of income and wealth. We would argue that in most American house
holds this threshold has already been achieved.
The demise of formal, prescribed feasting rituals, such as the English tea cere
mony and the formal dinner, is probably directly related to industrialization and
the advent of mass-produced consumer goods in late-nineteenth- and twentieth
century America. This in turn has resulted in the flexibility that is a characteristic
of the modern American feast. The American Woman's Cook Book identifies "the
problem of the formal meal," suggesting that "elaborate meals are not justified
from any point of view, social, physiological or economic, and that even the most
formal meal must follow the rules of health" (Berolzheimer 1963:54).
Biddle and Blom (1937) noted the mid-twentieth-century economic processes
that were changing the American feast: "Variety and change are not for the
wealthy, the favored few, but for Everywoman." They suggest that the increased
availability of mass-produced consumer goods resulted in "a change of glasses"
being equal to the price of a movie, and the cost of a centerpiece bowl being
equal to the cost of a pack of cigarettes. It is also clear that by the mid-twentieth
century, there was a reemphasis away from the home production of feasting
foods to the greater use of prepackaged substitutes. Feasting foods discussed in
Jacqueline Williams (1996) The Way We Ate: Pacific Northwest Cooking, 1843-1900,
are dominated by home-prepared cakes and sweet breads. In contrast, the 1954
Cooked to Taste cookbook of the Junior League of Portland provides a section en
titled "The Cook Is at the Party" that lists meals that can be made in the morning,
thereby releasing the ''hostess" to enjoy her party.
Part of the trend away from formality is reflected in the abundance of dispos
able dinnerware noted in our Tucson feasts. Although entertainment guides dis
courage the use of paper and plastic plates, cups, and other disposable china and
tableware, these are often used in lieu of ceramics, glass, and other durable items
for a variety of American feasts. As early as 1964, the National Industrial Confer
ence Board noted a downward trend in the United States production of ceramic
products due partly to substitute products, including pressed glass and plastic ma
terials. In the late twentieth century, paper and plastic disposable substitutes have
become much more dominant. Franklin Associates (1998) estimates that the con
tribution of paper and plastic plates and cups in the municipal solid waste stream
of the United States has risen from 0.3 percent by weight in l960 to 0.9 percent in
1996.
41 6
This flexibility in feasting behavior is a characteristic of the United States and

may be a characteristic of industrialized nations. Historical archaeologists have
begun to track the material implications of the greater availability and lower
costs of mass-produced ceramic tablewares and other consumer goods, and the
decreasing reliance on home-processed and prepared foods reflected in late
nineteenth-century and early-twentieth-century refuse deposits (e.g. , De Cunzo
1987:282; Purser 1992). The modern garbage data provides evidence for the
newest and most disposable stage in the evolution of the American feast.
While the United States has democratized feasting, it also has downsized it. By
downsizing, we mean both the size of the feasting group tends to be smaller and
the percentage of material resources poured into the feast is much less. We hy
pothesize that feasts in America usually involve one or two extended families, say
no more than 15 to 45 people. All of our examples, above, involved relatively low
numbers of participants, usually one household, with perhaps no more than 6 to
12 additional people. As some papers in this volume illustrate, in nonindustrial so
cieties, feasts often involve the best food and drink within the group's finances,
and sometimes even beyond the group's finances. In contrast, American feasts
have been streamlined, especially the larger ones, with standard menus, no sec
onds, and short serving times. We have routinized our fares. Even our largest and
most grand ''big man" feasts-the inauguration of the President of the United
States, the annual meeting banquets for the strongest companies, and retirement
dinners-cannot be termed food extravaganzas. Balls and socials, which were
very popular in the late nineteenth and early twentieth centuries, have largely
given way to cocktail parties and socializing in bars, taverns, and pubs. The roots
of this shift in American feasting are undoubtedly related to the late-nineteenth
and early-twentieth-century shift from a largely agrarian society to one that is
urban and suburban. Downsizing of the American family has led to the downsiz
ing of the American feast.
F U N C T I O N A N D F R A G M E N TAT I O N
Though it is possible that competitive feasting of the very wealthy in America
may be to enhance status, secure positions, or to make alliances with business
partners, the feasts of most Americans, as revealed in Tucson's garbage, are more
closely tied to entertainment. Consider how many traditional religious and secu
lar holidays in the United States have been corrupted, from honoring religious
and traditional values to overt materialism. For many Americans, the original
purposes of many of these feasting days, to celebrate social and economic mobil
ity or to venerate a public figure or deity, has become another reason for a party.
Drinking celebrations, such as St. Patrick's Day, or even Mardi Gras, are excellent
examples. As Sally Quinn (1997:21) advises, "there are a thousand reasons to have
41 7
a party or to entertain," but the one legitimate reason is to have a good time.
As noted above, there are some practical benefits of the feast in industrialized
societies. American feasts undoubtedly provide valuable settings for informal in
formation exchange, yet because the size of households is small and the links
with other relatives and friends are weak, they are probably qualitatively different
from nonindustrial societies.
Perhaps the best illustration of how we differ from feasting in nonindustrial so
cieties is how much feasting we do in restaurants (Fig. 15.1). Just try to get a din
ner reservation on Christmas, Thanksgiving, or Easter! This also illustrates how
the social value of feasting has been lost. By transporting them to restaurants, we
have, in a sense, both emas culated and secularized feasts. Compare hundreds of
All of the trimmings.

None of the hassle.
This Thanksgiving, make a pilgrimage to our table

fur a traditional fuast without the traditional clean-up.
Figure 15.r. A Marriott hotel advertisement for a

restaurant feast.
41 8
G A R B A G E A N D TH E M O D E R N A M E R I C A N F E AST
Enga (New Guinea) feast participants in and around a Kepele ancestral cult house
(Wiessner, Chapter 4) with 200 Americans "feasting" in a restaurant. There is co
hesion and integration in the United States because everyone knows the social
rules for restaurant dining and everyone gets in and out in an orderly manner, but
there is also tremendous luxury and individuality-anything can be ordered on
the menu.
Perhaps the most striking element of the United States feast in restaurants,
however, is fragmentation. There are numerous small social units that mingle in
the same feasting hall, but do not acknowledge one another, do not know each
other's background or where they are from. At the payment time, debts are not
marked up in the social and political order, but each little social fragment pre
sents a credit card and is gone with a full belly but no relationships left behind.
CO N C LU S I O N
We are brought back to the Gypsy feasts at the Crystal Ballroom. The complexity
of the American feasting system is daunting, and our generalities noted here can
not come close to characterizing the full range of feasts in the United States. We
acknowledge that there are many feasts that still preserve religious and other im
portant qualities, and like the Gypsy celebrations, they are difficult to approach
using traditional methods. The archaeological evidence for feasting in the United
States, however, is present in the abundant wastes discarded from feasting behav
ior. Through a materials-based approach, archaeologists and other anthropolo
gists can begin to document systematically the variability in feasting in the
United States and elsewhere, and begin to quantify and better specify our general
statements about American feasting behavior.
REFERENCES
Berolzheimer, Ruth, ed.
1963 The American Woman's Cook Book. Garden City; New York: Garden City
Publishing.
Biddle, Dorothy, and Dorothea Blom
1937 The Book of Table Setting. New York: Doubleday, Doran & Company.
De Cunzo, Lu Ann
1987 Adapting to Factory and City: Illustrations from the Industrialization and
Urbanization of Paterson, New Jersey. In Consumer Choice in Historical Arch
aeology, edited by Suzanne M. Spencer-Wood, pp. 261-295. New York:
Plenum.
Dickens, Roy S., Jr., and William R. Bowen
1980 Problems and Promises in Urban Historical Archaeology: The MARTA Pro
ject. Historical Archaeology 14:42-57.
419
Franklin Associates, Ltd.

1998 Characterization of Municipal Solid Waste in the United States: 1 997 Update.
Franklin Associates, Ltd., Prairie Village, Kansas. Submitted to the U.S. Envi
ronmental Protection Agency; Municipal and Industrial Solid Waste
Division, Office of Solid Waste. Report No. EPA530-R-98-007. Copies avail
able from the U.S. Environmental Protection Agency; Municipal and Indus
trial Solid Waste Division, Office of Solid Waste.
Given, Meta
1952 Meta Given's Modern Encyclopedia of Cooking. Chicago: J. G. Ferguson and As
sociates.
Hills, Tim
1997 The Many Lives of the Crystal Ballroom. Gresham, Ore.: Accuprint.
The International Guild of Professional Butlers
1998 The International Guild of Professional Butlers web page. Electronic docu
ment. http: / / www.butlersguild.com /201.htm
Junior League of Portland, Oregon
1954 Cooked to Taste. Portland, Ore.: The Junior League of Portland.
National Industrial Conference Board, Inc.
1964 Stone, Clay, and Glass Products. New York: National Industrial Conference
Board.
Purser, Margaret
1992 Consumption as Communication in Nineteenth-Century Paradise Valley;
Nevada. Historical Archaeology 26 (3): rn5-n6.
Quinn, Sally
1997 The Party: A Guide to Adventurous Entertaining. New York: Simon and Schuster.
Rathje, William L.
1995 Forever Separate Realities. In Expanding Archaeology, edited by James M.
Skibo, William H. Walker, and Axel E. Nielsen, pp. 36-43. Salt Lake City: Uni
versity of Utah Press.
Rathje, William L., and Cullen Murphy
1992 Rubbish! The Archaeology of Garbage. New York: HarperCollins.
Ritenbaugh, Cheryl K., and Gail G. Harrison
1984 Reactivity of Garbage Analysis. American Behavioral Scientist 28 (1): 51-70.
Roane, Susan
1988 How to Work a Room: A Guide to Successfully Managing the Mingling. New York:
Shapolsky.
Sechrest, L., ed.
1979 Unobtrusive Measures Today: New Directions far Methodology of Behavioral Sci
ence. San Francisco: Josey-Bass.
Spencer-Wood, Suzanne M., and Scott D. Heberling
1987 Consumer Choices in White Ceramics: A Comparison of Eleven Early Nine
teenth-Century Sites. In Consumer Choice in Historical Archaeology, edited by
Suzanne M. Spencer-Wood, pp. 55-84. New York: Plenum.
420
Staski, Edward
1984 Just What Can a 19th Century Bottle Tell Us? Historical Archaeology 18 (r):
38--5r.
Stewart, Martha
1982 Entertaining. New York: Clarkson N. Potter.
Webb, E. ]. , D. T. Campbell, R. D. Schwarts, and L. Sechrest
1966 Unobtrusive Measures: Nonreactive Research in the Social Sciences. Chicago:
Al dine.
Williams, Jacqueline B.
1996 The Way We Ate: Pacific Northwest Cooking, 1843-1900. Pullman , Wash. : Wash
ington State University Press.
421
Index
Page numbers followed by f refer to alligator tooth, 346, 350

figures Amazonian tribes, 6, 216; big drinking
fiesta (marriageability), 216-217, 218-223;
feasting vessels, 41, 218
achiote, 41, 374, 380 American feasts, 406-407; democratized,
adaptive behavior, 13-14, 25, 58-59 415-417; as entertainment, 417-419;
adaptive value, 28 food and drink frequencies, 414-415,
adultery, 222 417; household, 405-406, 407, 408-414
Aeatee cult, 134-135, 136 ancestor worship, 145, 149, 153, 368,
Africa, 6, 14, 68, 80, 83, 243, 247 383-384
age grades, 79, 82 archery contest, 220
aggrandizers, 30-31, 188-189, 191, 192-193,
197-198, 201, 207, 2II, 232 B
agrarian societies, 2, 16, 145, 240, 271, 392 Bagabo, 269, 274, 298
agricultural labor, II, 82, 91---92, 256; by Baganda, 83
women, 92, 95, 150, 225. See also labor bananas, 85, 95
mobilization and exploitation barley; 324, 392, 398
Ain's cult, 137 bats, 286
Akha: Blacksmith position, 146-147; beakers, 48
converts to Christianity, 158; elders, 147, beans, 95, 371, 379, 380
150, 152, 154; feast hosts, 31, 56F, 151, 156-157; bears, 154
feast role, 6, 26, 144, 146, 147; feast types, beavers, 325
151-156, 163; hunting and gathering and beef, 95; consumption, 89, 97
shifting cultivation, 1 4 5; lineage alliances beer: brewing and sales by women, 92;
and competition, 51, 144, 146, 148-151, 155, consumption, 81-82, 89, 96, 97, 220,
156, 158, 163; lineage feasts, 149, 152, 222-223, 243, 245, 249, 398, 410, 4n,
153-154, 156-158; Ritual Reciters, 146, 154/i 412, 413; dippers, 225, 226f; feasts, 79,
village feasts, 149, 154-155; Village 82, 247-248, 253; halls, 97---9 8 ;
Founder-Leader, 146, 154, 155; village household production, 85, 218, 250;
political unit, 145, 146-147; wealth, mugs, 218, 223, 225, 227; pots, 89, 96,
56f, 147 97, 98---99f, 223, 228; storage jars, 218;
alcohol, 10, 153, 174, 218, 276; agricultural straws, 96---97, 394; type served
resources for, 81-82, 218, 250; binge differentiation, 88, 243, 413
parties, 406-407, 413; symbolic value, 57, "begging for blessing," 158
72-73, 150. See also beer; cane liquor; Bemba, 81, 84, 95
whiskey; wine betel nuts, 155
alliances, 4, 17, 26, 30, 37, 57, 123, 190, 270, betrothal feasts, 173
271, 282, 327, 397; kinship-based, 51, 55, big men. See under Enga
201-202, 386 birds, 344, 348-349, 352
423
I N D EX
bloodletting, 368 ceramics: broken vessels, 228-229, 285, 291,

blowflies, 348 346; decorated, 277, 289, 294-295,
boar, 156 351, 378; import consumption, 8, 17,
Boas, Franz, 187, 188, 207 292-293; size and function, 223-229,
boasting, 6 284-285, 321, 377, 380; tradition, 48, 49,
body adornment. See feasts, clothing and 225-226. See also porcelains
adornment ceremonial gate rebuilding feast, 154
bone artifacts, 347, 372 Ceren (El Salvador) site, 9, 369-37!,
Bourdieu, P., 28, 76, 244, 252 384-385. See also Maya
breadfruit, 169, 174 Chagga, 85
bridewealth, 99, 125, 251 challenge feasts, 279-280
British Isles. See Bronze Age; Neolithic chenopod, 324
Bronze Age (British Isles), 53, 58 Chickasaw; 312, 352
Buid, 278 chickens, 95, 145, 153, 155, r6r, 253f 272,
Bukidnon, 274, 275, 276, 281 276, 281, 286, 287
burial: assemblages, 290-291; feast, 9; chiefs, 46, 58, 83-85, 87, ro2-ro3, 244. See
mound, 314, 32r. See also funeral also Cahokia site; Maya; Philippines;
feasts Sumer
Burkina Faso, Sr, 82 child-growth feasts, 191, 198-201
butchers' feasts, 153 chili, 37!
Chin, 296
c China, 53, 268-269, 289-292
cabbage, 145 Christmas feasts, 29
cacao, 37! cigarettes, 156, 4ro, 4rr, 412
Cahokia (Ill. ) site, 9, 334-336, 338f 351; circumcision. See initiation feasts
animals in feasting, 41, 346--350, civet cats, 286, 287
351-353; chiefs and elites, 337, 339, 345, clans, 148
355, 356; feasting, 57, 58, 342-350, 351, clay vaginal plug, 222, 225, 226f
352, 356--357 clitoridectomy, 217, 221, 222
caiman effigy, 374 clothing. See feasts, clothing and
calamity feasts, 37 adornment
Cameroon. See Koma; Meta Coast Salish, 200, 202, 203
candy; 155, 156, 4rr cocktail parties, 406, 417
cane liquor, 218, 222, 382 coconut oil, 169
canoes, 178, 223 cofradias, 381, 382, 383
carabao. See water buffalo Cole, Fay Cooper, 269
cassava, 95 Coles Creek culture, 312
cassowary, 124 collective work events (CWEs), 79-80, 247,
caterpillars, 89 254-256, 257-258; defined, 24r. See also
cattle, 79, 82, 95, 145, 162, 251, 253f work feasts
392, 393f 394; parading of, 96. See also commensality; 243, 255. See also politics
beef and feasts, commensal
celebratory feasts, 268 Comox, 202
424
I N DEX
competitive feasts, 37, 38, 51-53, 57--58, 79, domestication o f plants and animals, 41,
173, 190, 206--209, 279-282, 368, 401, 59, 286--2 88, 324, 371
417. See also ritual feasting Dorze, 79
"condensed meaning," 4, 71 ducks, 145, 324, 348f, 349, 371, 377
Conibo-Shipibo, 216, 217-218, 220, 221, dyadic friendship feasts, 54
230-233; ceramics, 223-229. See also
marriageability fiesta E
consumption patterns, 8, 68 earspool, 347
contractual debts, 31, 35, 45 economic advantage. See competitive
Cook, James, r68 feasts; labor mobilization and
Cook Islands, 247 exploitation; work feasts
cooperative relationships, 29, 37 economic success, 14, 15, 36, 37, 150, 163
copper, 317, 323 ecstatic rituals, l4
corn/maize, 95, 145, 253, 324, 327-328, egg ritual, 156
371, 379, 380; huskers, 372; storage, elites, 28, 174, 193, 195-196, 202-203, 268,
340-341, 342 272, 285-288, 312; women, 9r. See also
corvee, 244, 249 chiefs; diacritical feasts
cranes, 348f, 349 El Salvador, 9, 369
Crete, 48 emic type feasts, 37, 185, 4oy-406, 415
crystal, 346, 351 empowering feasts, 76--82, 88, 94, 96, ro3.
cuch festival, 375 See also work feasts
cucurbits, 324, 351, 371, 374, 379, 380 Enga, 14, 15, 17, n9, 121-122, 124, 137;
culturalist social theory, 2 big men, 77, Sr, rr9, 121, 126--127,
Cumancaya site, 223, 225, 226 133-136, 138; cults, 124, 127, 128,
cuneiform texts, 391, 397-399 129-133, 134-135, 137-139; feasting, sacred
curassows, 218 and secular, l2y-133; oral tradition,
curing feasts, 149, 153, r54f rrl)-125; polygyny, 81, 127; trade goods,
CWEs. See collective work events 124, 126, 138; value, exchange and
feasting, rr7-n9, 122-124, 125, 139,
D l40; warfare, l2I, 122, 123-124, 134,
dancing, 4, 96, n7, 174, 221, 368, 396; 136, 137
plazas, 175 entrepreneureal feasts, 67
dates, 392 equal-return feasts, 191, l9y-206
deer, 286, 353, 371, 377; body parts Ethiopia, 79, 245
remains, 324-325, 34r, 344-345, 34;r, exchange: reciprocal, 272, 274-275;
351, 353; ceremonialism, 3 75, 3 78; -skull relationships, 251, 274, 275, 282; value,
headdress, 370, 374; venison, 351, 380 251-252, 253. See also Enga; ritual
diacritical feasts, 5, 9, ro, 37, 57, 85-88; as exchange
political action, 93-94; and social
boundaries, 88-90 F
dogs, 145, 162, 286, 326, 371, 377; as food, feasts: archaeologically identifiable, 4, 5,
153, 169, 178, 380; sacrifices, 276 6-9, 17-r8, 36--42, 47-54, 5;r, 72, 90,
dolphins, 221, 225 rr6--117, 128-129, 133, 157-162;
425
I N DEX
architectonic elaboration, 9-IO, 39, 82, 3s3; consumption of animals, 49, S4,
41-42, S3, S7, S9, 89, n6, 131, 169, 276, SS, 87, ISI, rss, 160-162, 218, 220--22!,
28s, 372; in art, 391, 392-397; 277-279, 281, 286--2 88, 344-349, 3Sl-3S2,
classification, 4, 6, 30, 3S-42, 190-191; 3S3, 3S4, 377, 380; consumption of
clothing and adornment, 89, rs2, 174, vegetables, 126, rs3, 289, 379, 398;
218, 229, 3S2, 368, 370, 374, 394; as cooked by males, lSl; differentiation, S4,
culturally and evolutionary adaptive, s8; 88, 132, 243, 277, 278, 380; distribution,
defined, 3, 4, 28, 67, 169; as drama, 220; n6, 126, 179, 186, 196, 270, 271, 272,
duration, 96, rs2, 221, 281; and ecology, 277-278, 279; everyday; 89; human flesh as,
12-16, 26--27; ethnographic manifes 169, 174, l7S; portable containers, 223;
tations, 4, s-6, IO, 23-24, 38r-38s; preferences, 89, 287-288; preparation, 30,
fighting at, 221-222; form and content, 80-81, 84, 379; preparation containers
2s-26; and gender, ro-n, 42, 43, and facilities, 42, 47-49, SS, 88, lS7,
9 �3, 98--99, 124, 174, 178, 179, 379; rs8-r6o, 223, 22s-226, 3SO, 3Sl, 372,
household size, 47-49, SS, rs3, 194; 376-378, 416; preparation for elite, 86--8 8,
invitations to, 3S, 39, 80, 219-220; 174, 277, 278; served by males, lSI;
location, 8--9, IO, 47, rsr, rs6; practical servingvessels, 30, 39, 41, 47-49, 72, 86,
benefits, 29-3S, 79, 124, 189; as ritual, 88, 89, 90, 98, n6, lS7, lS8, 160, 223, 268,
3-4, s, 9, 13, s8, 6s, 67, 69-7S, 272, 3S3; 270--271, 277, 289-29S, 387, 4IO, 4n, 416;
roles, 6--7, 12-16, 24, 4s-46, s8, 6s, 3s7; special, 4, SS, 89, 127, lSO, lS2, 169, 270,
seasons for, 3s3; serving order, 88, 90, 277; storage, 80, 223, 340, 370, 373, 37S
196; size, 36, 38-39, 47, SS, S7, s8, 80-82, Food Utility Index (FUI), 347
n6, rs6, 219; social boundaries, 88--90; foraging societies, n7
and social relations, 4s, 6s, 67, 69, 272, Formative period (Mesoamerica), 289
297-299, 3S4, 387; social units at, 39, S3, FUI. See Food Utility Index
96--97, 99f; spatial differentiation, 9, IO, functional type feasts, 37-38
sef, 88, 91, rs2, 169; specialized funeral feasts, 14, 37, sof, lSO, lS2,
paraphernalia, 41, S3, S7, S9, 22s-228, rss-1s6; fauna! remains in graves, 162;
374; and surplus, 13, 24, 27, 30, 37, 4s, and political advancement, 20s; serving
s8-s9, 71-72, 73, 192; symbolic content, vessels, 90; for women, 82. See also
3S, 67, 8s, I03; texts on, 391, 397-399; as under Luo
widespread, 24-2s. See also alcohol; Fur, 247, 2s3
American feasts; collective work feasts;
food; ritual feasting G
"feasts of merit," 270, 271, 277 gambling, lS6
Female Spirit cult, 138-139 garbage, 6; disposal, 169, 180, 416;
festivals, 378, 382, 397; defined, 370 fires and middens, l6r, 162, 228, 28s,
Firth, Raymond, 170 29r. See also Tucson
fish, 89, 9s, r69, 178, r80, r92, 286, 324, geese, 348f, 349
344, 349-3SO, 3S2-3S3, 393f, 394, 399 Ghana, 87
Florida, 312 gift giving, 27, 30, 31, 3S, 4S, 73, 190,
food: abundance, 96, n7, 179, 243, 272, 398; 203-204, 270, 272, 274, 394
avoidances, 87, 89, 179; consumption, s8, ginger, r4s
426
I N D EX
goats, 95, 145, 155, 194, 393f, 394 Inca, 58

gold, 274, 275 indebtedness, 79, 189, 200-201, 368
Goldman, Irving, 169, 173 Indonesia, 278
Goody, Jack, 85-86 inebriation, 4, IO, 221; permissible, 91
Great Ceremonial Wars (Enga), 123-124, initiation feasts, 6, 9, 37, 171, 207. See also
134, 136 marriageability fiesta
greater-return feasts, 191, 196, 206--209 intertribal meetings, 207
Green Corn ceremonialism, 328, 352, 356 iron, 245, 249, 251
grooved ware (British Isles), 48 Iron Age (Europe), 289
group exogamy, 203 irrigation, 372
guilds, 399 Iteso, 243
Gypsy feasts, 404-405, 419
K
H Kamares ware, 48
Haida, 205 · Kamba, 247
Halloween, 4n-412 Kenya, 245, 247. See also Luo
Handy, E. S. C . , 173, 174 Kepele cults, 124, 129-133, 134f
harpoon, 347 kings, 83, 87, 178, 244. See also Hawaii,
harvest feasts, 96, 154-155, 174, 178, 269 royal feasting; Sumer
Hawaii: chiefdom, 169-170, 177; cult kitchens, 161
feasts, 178-179; diacritical mode, 86; knives, 221-222, 225, 226f
gender and social class, IO, 169, 178, Kofyar, 247
180; kings, 168, 177, 178; royal feasting, Koma, 79, 82
14, 177-178, 181, 274; temples, 179-180 Kwakiutl, 187, 201, 205, 207, 208-209
hearths: male and female, 161; size and
number, 47, 48F, 50-52f, 314, 372 L
Hill Tribes (Thai). See Akha labor mobilization and exploitation, 6, II,
Hmong, 50J, 52f 17, 29, 30, 37, 69, 92, 190, 192-193, 218,
hoes, iron, 250-251, 252-253f 241, 246, 249-256, 354-355; by chiefs, 84,
hogs, 174, 179· See also pigs 85, 244, 283; of women, II, 91-92,
Hopewell culture, 312, 321, 323, 351, 353 255-256. See also collective work events;
horticultural societies, 44-46, n7, 145, 218 work feasts
hospitality, 242, 243, 252. See also chiefs land use patterns, 117
human sacrifice, 269 lentils, 95
hunter-gatherer societies, 2, 42-46, 49-50, Liangzu burial offerings, 53
54, 59, 151, 351-352 limited-goal feasts, 39
lineage feasts, 30, 125, 370, 383-384;
structures, 53, 54f
icons, 368-369 Lords (Spirits) of the Earth feasts, 154
idols, 368 lu'au, 168
!gala, 88 Luo, 95, 99, IOO, I01-I02, 249; beef, 96;
Igbo, 84-85 beer, 89, 95, 96; beer pots, 89, 96, 97,
Ilanun, 298 98.fi beer straw, 96--97, 98.fi feast
427
I N DEX
location, 96; feast markers, 9Q--91; feast memorial feasts, 174

role, 6, 14, 17, 96, 100; fish, 89; funeral menopause feasts, 155
feast, 82, 96, 100; male status, 91i---97, 99f, Meta: hospitality; 83-84, 88-89; noble
100, 102-103; patron-role feast, 100-101; game animals, 87
power, 102-103; serving containers, 98; metal goods, 250, 268, 272, 299
women, 97, 98-99, 10of, 256 metate and mano, 372, 376-377, 378, 385
Mexico, 253
M middens, 314, 315. See also garbage
Maale, 245, 254 midweek party; 413-414
magic, 102 milk, 95
Maguez, 371 millet, 95, 145, 250
maize/ corn, 95, 145, 253, 324, 327-328, Minimum Number of Individuals
371, 379, 380; huskers, 372; storage, (MNI), 287
340-341, 342 minimum number of skeletal elements
Mambila, 79, 82 (MNE), 347
Mamprusi, 87 mining, 250
manatees, 218, 221 Mississippian sites, 7, 3n-312, 321,
Manga (Burkina Faso), 82 336-337. See also Cahokia site
manioc, 218, 371 MNE. See minimum number of skeletal
marijuana, 413 elements
marine fauna, 286 MNI. See Minimum Number of
Marquesas, 169; chiefs, 173, 174, 181, 298; Individuals
feasts, 173-174, 177, 178, 181 monkeys, 218 , 220-221, 286, 287
marriageability fiesta, 216-223, 229-230; Mossi, 82
ceramics, 223, 22s-229 moundbuilders. See platform mounds
marriage feasts, 93, 96, 150, 151, 155, 173, multi-centric economy; 250, 251
201-204; Akha, 156-158; Tikopia, I7I
marriage partners, 30, 37, 150, 190, 197, N
204; abduction of, 230, 23rf naming feasts, 149, 153, 197, 198-199
marsupials, 124, 126 narcotics, 150, 155, 156, 413
Marx, Karl, 28, 169 Neolithic (British Isles), 48, 58
Masa, l4 new homestead feast, 96, 153
mask feast, 85 New Year's festival, 154, 155
matrilocality; 230 Nias, 279, 280
Maya, 7, 34; chiefs, 368; feasts, 41-42, Nigeria. See Igala; Igbo; Kofyar; Mambila;
48, 57, 368, 369, 378, 380, 383-385; Tiv
specialized structure, 370, 372-378, NISP. See number of identified specimens
380, 381-382 Nootka, 187, 202
maygrass, 324 no-return feasts, 191, 192-195, 196
meat feasts, 247-248 Northwest Coast (North America), 6, 9,
medicinal plants, 222, 371 53, 18_5-186; borrowing, 199-201, 202;
megaliths, 285 feasts, 188, 190-191, 201-204, 210-211;
memorial ceremonies (kuure), 82 post-contact, 192, 209; slaves, 200, 206,
428
I N D EX
209; warfare, 200-201, 210--2II. See also and skulls, 131-132, 161, 278, 287; and
aggrandizers; potlatches; resource sites rank and status, 175, 278; sacrifices, 178,
number of identified specimens (NISP), 272, 274, 276; slaughtered for sale, 153; in
344, 345.f. 346, 349 Tee exchange, 122, 123, 129, 137
nuts, 324 pizza, 412, 413
Nyoro, 83 platform mounds, 9, 3II-312; building
activity, 312, 320--321, 325, 326fi charnel
0 structures, 221, 223; elite residences,
obsidian, 371, 377 312, 314; feasting, 323-328; sites,
Oliver, Douglas, 169 314-321; specialized feasting structures,
onagers, 394 42, 57, 355. See also Cahokia site
open societies, 173 platforms (Marquesas), 175, l76f
opium, 156 poi, 178
oratorical displays, 4, 6, 96, 100, 276 politics and feasts, 6, 12, 15; commensal,
66, 67, 68, 73-75, 93--94, 100, 127;
p contests, 16, 30, 84-85; egalitarian, 79,
pantheists, 145 lOI; and male power, II, 91; and power,
Papua New Guinea, 6, 26, II5, 254. See also 15, 17, 26, 30, 37, 66, 68, 70, 123-124,
Enga 155, 190, 268, 276, 282-283; and
patricians, 145 prestige, 1 4 , 15, 3 1-32, 7 8, 8 4 ; ritual and
patrilineages, 146 symbolism, 70--75; and segmentary
patron-client networks, 17, 67, 83 lineage system, 101-102; and status,
patron-role feasts, 82-85, 94, 100--101 14-15, 31-32, 77, 79, 84, 268, 277; for
peanuts, 145 support and position, 205-206, 283-284.
pearl shells, 138, 139 See also diacritical feasts; empowering
peccaries, 218, 221, 371 feasts; patron-role feasts; ritual feasting
penalty feasts, 30, 154 polygamy, 202, 204, 283
Pentlatch, 202 polygyny, 95; and household labor, 81, 85,
peppers, 145 99, 255; and male political power, II, 91,
Peru. See Amazonian tribes 99, 103
Peten Polychromes (Maya), 48 Polynesia, 6, 9, 42, 169. See also Hawaii;
Philippines, 6-7, 8, 17, 26; chiefdoms / Marquesas; Tikopia
chiefs, 57, 86, 267-273, 276, 281-282, 284, Pondo, 83, 87, 247-248
295-299, 3oofi feast characteristics, 272; porcelains, 268, 270--271, 274, 277, 285,
feasting vessels, 41, 268; maritime trade, 289-295, 297; foreign sources, 289
268, 269, 289, 294; Metal Age, 293-295; pork, 123, 124, 125, 126, 132, 133, 155, 169,
population density, 283. See also ritual 178
feasting potlatches, 9, 26, 206-209, 210; for
pigs, II9, 120, 124-125, 133-140, 145, 155, 161, alliances, 30, 188, 196; and chiefs,
174, 178, 179, 270, 286, 287; and 191-192, 196; as different from feasting,
Ain's cult, 137-138; deaths from disease, 186, 187; as investment, 206-207; and
149; for exchange, 277; feasts, 14, 17, 125; prestige and status, 187, 188, 2n; as
in Kepele cult, 129-133, 137; mandibles redistribution mechanism, 187-188
429
I N D EX
power. See elites; politics and feasts s

prairie chicken, 348f, 349 Sagada Igorots, 275-276, 277-278
praise singers, 96, roo Sahlins, Marshall, 169
Presidents' Day, 412-413 salmon, 194, 202
prestations, 4, 69, 282. See also gift giving "salvage ethnography," 173
promotional feasts, 196-198 . See also labor Samia, 6, 245, 249-252
mobilization and exploitation; Samoa, 247
marriageability fiesta; marriage Sears, William H., 313-314
partners seasonal feasts, 36, 37
puberty feast, 197, 198 sea turtles, 169
puddings, 169, 172, 175 Sebei, 247
punishment feasts, 54 secrecy, 222
purification feasts, 152, 154 segmentary lineage system, ro1-ro2, 282
self-interest, 6, 12, 13, 15, 16, 26, 31, 32, 58,
R 76, 189, 201, 2n; and cultural traditions,
rabbits, 324 33, 34
raiding, 103, 230, 283, 298 sequential monogamy, 202
rays, 286 sham marriage, 202
reciprocal: debts, 30, 35; feasts, 35, 36, sheep, 95, 392
73-74, 201-205, 242 shell beads, 347
red cedar, 352 Shipibo. See Conibo-Shipibo
red ochre, 314 silk, 268
refuse pits, 180, 314, 350. See also garbage singing, 4, 96, roo, n7, 174, 221, 392
religion and feasts, 6, 37, 273, 391, slaves, 85, 200, 206, 209, 256-257, 271, 298
397-398, 406; economic aspects, social classes, 5, 88, 91, 243; and gender,
396-397; and elites, 28, 155, 178-179; ro, n, 9r. See also chiefs; elites
political dimension, 78, 146-147, 155· social control, 69
See also Enga; Kepele cults; Tikopia social debt, 83, 272, 275-276
resource sites (hunting, fishing, gathering), social formation, 169
192-193, 194, 198, 199, 2II social relationships, 30, 36, 270
restaurant feasting, 418-419 social technology, 26
rice, 145, 153, 270, 274, 276, 288-289 social transactions, 4, 16-17, 27'j-277
ritual exchange, 268, 269, 270 social transformations, 13, 58, 256, 270,
ritual feasting, 268-269, 270, 284, 336, 279, 295-299, 321, 354
356-357, 368; archaeological evidence socioeconomic inequalities, 26, 147-151,
fo� 284-295, 342-353, 370, 378-385, 185, 241, 251, 270
386-387; defined, 272; politics of, solicitation feasts, 30, 37, 54, 190, 195
282-284, 296, 354; priests, 273 solidarity feasts, 38, 49-50, 52, 54---55 , 149,
ritual items, 30, 268 151, 190, 192-194
rivalry feasts, 209 sorghum, 95
roasting pits, 49 South Africa, 83, 87
rock concerts, 13, l4 sports, 220, 394
rubbish man, 33 squash. See cucurbits
430
I N DEX
sugar cane, 2r8 tribute feasts, 58, 194, 275, 339

Sumer, 42, 392; feasts and kings and Tsimshian, 205
queens, 392-396, 397, 398, 400; Tucson (Ariz.), 6, 10; feasting rate, 4!4;
redistribution economy, 402; religion, garbage, 407-4!4
397, 399-400; tribute feasts, 58, 400 tumeric dye, 171
sunflowers, 324 turkeys, 145, 349, 410, 415
survival, !4, 27, 58 turtles, 286, 287, 371, 399
swans, 348-349, 351, 353 Tvvana, 202, 204-205
sweet potato, 95, II9-I20, r21, 123, 131, twin birth ceremony, 96
2r8; for pigs, 137
swidden cropping, 283 u
symbolic capital, 252, 253 Uganda. See Baganda; Nyoro
University of Arizona Garbage Project,
T 407, 408
Ta Oi marriage feast hearths, 5rf Upper Paleolithic, 45, 47
tapirs, 218, 223
Tarahumara, 253 v
taro, 120, r72, 178, 288 value construction, 4, 6, 17, n7-n9
tattooing, r73 vegetables, 126, 145
Tausug, 274 victory feasts, 394, 397
taxes, 399 Vietnamese lineage house, 54f
tea, 145, 153 villages, 337. See also platform mounds
technological change, 38 volcanic eruption, 369, 370---371
Tee ceremonial exchange, 122-123, 125,
129, r34, 135, r36 w
Thailand, 6, 41. See also Akha wage labor, 247
Thanksgiving feasts, 408-4n Wainira, 254
Thomas, Nicholas, 173, I77 warfare, 103, 122, 173, r98, 200---2or, 271,
Tikopia: chiefs, 169, 170, 180; feast food 283, 299
and facilities, 172, 180---1 81; kinship, 170, water buffalo, !45, r55, r62, 270, 272, 276,
172, 176-177; ritual and secular feasts, 277, 281, 287; horns, 160; skulls, 278,
170, I71, 172 287
Timucuan, 326f wealth: accumulation, 30, 32, 56f, 95, 103,
Tiv, 92, 248-249 188, 207, 271, 283-284, 286, 401;
Tlingit, 205 destroyed, 53, 57, 59, 209; manipulation,
tobacco, 57, 342, 352 n9, 123, r25, r99, 202-203, 207-208. See
toucan beak knife blade, 221, 226f also Enga, big men
trade partners, So, 122, 268 wedding feasts. See marriage feasts
trading activity, 8, 9, 121-123, 124, whiskey; 153, 155, 156
250---251, 289, 299 White-Skirted Woman (Akha role), r55
transegalitarian societies, 44-46, 47, 50---53, wine, 398
58-59, 280, 312, 327. See also Akha women: as agricultural, culinary, and
"triangulation," 5, 59 serving labor, 92, 95, 98-99, 100.f. 150,
43 1
I NDEX
255-256, 379; beer consumption, 97; skills, 245; voluntary, 244, 257; women's,
elite, 91, 273; status, II, 92, 253f See also 92, 256
feasts, and gender workmen's feasts, 153
Woodland cultures, 7, 313. See also "Work of the Gods," 170
Cahokia site; platform mounds world-renewal ritual, 312, 328, 356
work exchange, 256; as CWEs, 241,
242-243; quality, 246 y
work feasts, II, 17, 30, 39, 55, 79-80, 85, yams: feasts, 85; mounding, 248-249, 288
190, 194-195, 247-256; and colonialism, Yao, 48f
249; as CWEs, 241-242, 243-246, 254,
255; and obligation, 243, 244, 254; and z
quality of work, 246; size, 243, 254; Zambia. See Bemba
432

Archaeological and Ethnographic Perspectives On Food, Politics, and Power Edited by Michael Diet/er and Brian Hayden

Uploaded by

Copyright:

Available Formats

Archaeological and Ethnographic Perspectives On Food, Politics, and Power Edited by Michael Diet/er and Brian Hayden

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Archaeological and Ethnographic Perspectives On Food, Politics, and Power Edited by Michael Diet/er and Brian Hayden

Uploaded by

Copyright:

Available Formats

ARCHAEOLOGICAL AND

ON FOOD, POLITICS, AND POWER

Edited by Michael Diet/er and Brian Hayden

SMITHSONIAN INSTITUTION PRESS

WASHINGTON AND LONDON

COPY EDITOR: Peter Donovan

Library of Congress Cataloging-in-Publication Data

British Library Cataloguing-in-Publication Data available

Manufactured in the United States of America

List of Illustrations vii

2.1. Schema of feast types 38

10.Z. Faunal remains in household middens of elite and non-elite residences

2.1. Archaeological Signatures of Feasts 40

Linda A. Brown, Department of Anthropology; University of Colorado, Boulder

Michael Diet/er and Brian Hayden

what Hayden calls "triangulation." The difficulties of distinguishing on a priori

Thus, Hayden (Chapter 2) examines feasting from a cross-cultural and ecolog­

women), (4) quantitative distinctions (i.e., differences in the amounts of food or

analysis of feasts because feasting frequently involves a gendered asymmetry in

a major role in promoting and defending individual or factional self-interests�

Everyt h i n g that is n ot g iven is lost

I am always a bit uncomfortable and self-conscious when people ask me what I

lier ethnographies, few anthropologists address the theoretical importance of

problems related to symbolical meanings. Many of the articles in this volume, in

difficult issue is how to meaningfully categorize the extensive variety of feasts

3. create cooperative alliances between social groups (including political sup­

compelled to accumulate large amounts of wealth in order to be able to dispense

itarian societies, is embodied by people who are systematically reviled as "lazy;"

Yan (1996:44ff) discusses classification in regard to the parallel phenomenon

RECIPROCAL COM PETITIVE

Figure 2.1. A schematic diagram of the most important divisions in

tions of material indicators, creating new distinctions between feasting

Food Rare or labor-intensive plant or animal species (especially

TABLE 2.1 continued

Ritualized items of etiquette Smoking or other narcotic paraphernalia

factors should be useful in trying to distinguish different types of feasting

nomic hospitality would b e displayed for potential or continuing allies in other

households or communities and that the ceramic tradition is relatively devel­

poor or politically isolated families could be disenfranchised or even enslaved or

occur as a rule within normal residences, without special paraphernalia other

preparation vessels, serving vessels, and perhaps even temporary architecture

In conclusion, it is worth reiterating that our theoretical and comparative un­

anthropology and ethnography have not been of great service to archaeologists

structural typologies without considering the intervening kinds of social prac­

enue of analysis, we need not only a greater range of empirical information

F EASTS AS RITUAL, RITUAL AS POLITICS

In the work of reproducing established relations-feasts, ceremonies, exchange 'of

misunderstandings, particularly in traversing linguistic frontiers in which the cul­

ready supply of agricultural produce. The institutional arrangements for mobi­

gins and significance of cuisine. According to Goody, the development of such

orate, cuisine and/ or consumption paraphernalia, depending upon the nature of

archaeologist to distinguish readily between "diacritical feasts" and these other

production, although the relative gendered contribution in this domain is by no

tion, reproduction, and transformation of gender identity; as well as in the gen­

fisch 1987) . Yet generosity in commensal hospitality is an essential practice in

symbolic capital. However, the suppression of warfare and raiding eliminated

Douglas, M . , and C . Isherwood

Middleton, ]. , and D. Tait, eds.

Bonding groups xxx xxx x xxx x x

Thus, Hayden (Chapter 2) examines feasting from a cross-cultural and ecolog

3. create cooperative alliances between social groups (including political sup

households or communities and that the ceramic tradition is relatively devel

In conclusion, it is worth reiterating that our theoretical and comparative un

structural typologies without considering the intervening kinds of social prac

misunderstandings, particularly in traversing linguistic frontiers in which the cul

ready supply of agricultural produce. The institutional arrangements for mobi

tion, reproduction, and transformation of gender identity; as well as in the gen

of "good name" was not insignificant in a society where approximately 25 per

�------;M;,:OTE HOUSE ��

In Akha society; socioeconomic inequity; competition, and aggrandizing be

-hunter-gather-type solidarity feasting and complex-society competitive feast

historic record of the long-term development of feasting behavior might be pos

serves a s a mechanism fo r assuring social solidarity, binding the community to

8 . Work-Party Feasts. Work-party feasts obtain labor for specific labor

less fornicators. Peccaries are classic Amazonian symbols of unrestrained sexual

em and archaeological contexts. These include rattle-based beer mugs (Ray