Factional competition

and political
development in the
New World

Edited by
E L I Z A B E T H M. B R U M F I E L
and
J O H N W. F O X

BIQUCAVCA L V n OdUSAUEZ

CAMBRIDGE
UNIVERSITY PRESS

132256
PUBLISHED B Y T H E PRESS SYNDICATE OF T H E U N I V E R S I T Y O F C A M B R I D G E
The Pitt Building, Trumpington Street, Cambridge, United Kingdom
C A M B R I D G E U N I V E R S I T Y PRESS
The Edinburgh Building, Cambridge C B 2 2RU, UK
40 West 20th Street, New York N Y lOOI I-4211, U S A
477 Williamstown Road, Port Melboume, VIC 3207, Australia
Ruiz de Alarcn 13, 28014 Madrid, Spain
Dock House, The Waterfront, Cape Town 8001, South Africa
http ://www. Cambridge. org
Cambridge University Press 1994
This book is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without
the written permission of Cambridge University Press.
First published 1994
First paperback edition 2003
A catalogue record for this book is available from the British Library
Library of Congress catalogiiing in publication data
Factional competition and political development in the New
Worid/edited by Elizabeth M. Brumfiel and John W. Fox.
p. cm. - (New directions in archaeology)
ISBN O 521 38400 I hardback
I. Indians- Polities and govemment. 2. Political anthropologyAmerica. 3. Social archaeology - America.
4. Indians - Antiquities.
5. America - Antiquities. I Brumfiel, Elizabeth M.
II. Fox, John W., 1947- . III. Series.
E59.P73F33 1993
306.2-dc20 92-32371 CIP
ISBN O 521 38400 I hardback
ISBN O 521 54584 6 paperback

je: *

.,

1
Factional competition and
political development in the
New World: an introduction
E L I Z A B E T H M. B R U M F I E L

This volume calis attention to the importance of factional competition as a forc of social transformation. It
arges that factional competition is implicated in developments as diverse as the spread of ceramic technology
and maize agriculture, the origins of permanently instituted leadership offices, the expansin and collapse of
States, and the European domination of indigenous New
World peoples. Although this volume focuses upon the
New World, its perspective is relevant to the social
histories of other reas of the world as well, because all
non-egalitarian societies, both ancient and modern, are
shaped by the dynamics of factional competition. Bringing an agent-centered perspective to the study of political development, this volume also contributes to a
general understanding of social stability and change. An
agent-centered perspective maximizes the amount of
data drawn into the analysis and thus permits the most
detailed and complete account of specific cases of political continuity and transformation.
Our studies of factional competition both complement
and critique the two prevailing approaches to prehistoric
social change: cultural ecology and Marxism. Cultural
ecology focuses upon the dynamic interactions of human
populations and their local environments. As a complement to this, the studies in this volume examine the
intemal dynamics of local populations, dynamics that
help to shape the strategy of resource exploitation.
Marxist theory focuses upon the dynamics of class
struggle: a model postulating solidarity within classes
and struggle between them. As a complement to class
struggle, the essays in this volume emphasize the importance of conflicts within classes and alliances between
them. As critiques of cultural ecology and (particularly

structural) Marxism, the studies in this volume demnstrate the necessity of replacing a theory of strict systemic or structural determination of human behavior with
a theory that integrales agent-centered and systemcentered analyses into a single framework. Most of the
studies in this volume employ versions of an agentcentered practice theory developed in the work of Barth
(1966), Giddens (1979), and Ortner (1984).
Practice theory is especially suited to the analysis of
factional competition. Factions are characterized by an
informal, leader-focused organization (Nicholas 1965,
Bujra 1973, Silverman 1977); thus, it seems reasonable to
initiate analysis by examining the goals and strategies
employed by faction leaders as individual social actors.
Furthermore, factions are groups whose single function
appears to be gaining access to limited physical and
social resources (Bailey 1969:52), and these goals are
best achieved through the application of a pragmatic,
advantage-seeking, maximizing strategy such as that
imputed to agents in practice theory. Finally, factions
are structurally and functionally similar groups that
compete for advantages within a larger social unit such
as a kin group, ethnic group, village or chiefdom. Practice theory anticipates conflict between individuis similarly positioned within society while cultural ecology
and Marxism do not. As argued below, this internal
competition supplies the dynamic for political development.
This introduction begins by defining factions and factional competition. Then, it discusses the relationship
between factional competition and "ecological" variables such as population, warfare, agricultural production, and long-distance trade. Third, it proposes how
factional competition and class struggle interact to
produce the social formations discussed in the case
studies that follow. Forth, it considers how factions
and factional competition can be identified in the
archaeological record. Finally, it examines the wider
theoretical implications of an agent-centered perspective.
What are factions?
Aside from a few brief discussions of factions prior to
1955 (Lasswell 1931, Linton 1936:229, French 1948,
Fenton 1955), anthropological interest in factions developed as part of an effort to expand the scope of social
anthropology beyond the description and analysis of
formal social structure. This was accomplished rst
through the recognition of the many informal, noncorporate groups present in contemporary non-Western
3

Elizabeth M. Brumfiel

societies (including action groups, diques, networks,

factions, and patron-client dyads). and second, through
the investigation of how social structure is generated by
individuis acting to maximize their self-interest given
their particular sets of cultural and material constraints
(Whitten and Whitten 1972; Cohn 1974:40-3; Vincent
1978). From the mid-1960s to the mid-1970s, the construction of modeis of informal groups and individualcentered social transactions was a flourishing, if somewhat insular, enterprise within social anthropology, and
factions became the object of considerable interest.
At issue was the definition of factions (Lewis and
Dhillon 1954; Firth 1957; Boissevain 1964; Nicholas
1965), whether factions served positive social functions
or were a form of social pathology (Siegel and Beals
1960; Schwartz 1969); the relationship between factions
and class conflict in peasant societies (Sandbrook 1972;
Alavi 1973; Gross 1973; Schryer 1977), and whether
factions were an obstruction or an mpetus to social
change (Bujra 1973; Silverman and Salisbury 1977).
However, this interest in factions abruptly collapsed in
the late 1970s when anthropologists turned away from
local-level, agent-centered studies to pursue the issues
raised by the world systems perspective. Although factions have continued to serve as a basic construct for
analyzing peasant polities (e.g., Hegland 1981; Greenberg 1989; Munson 1989), the theoretical debates surrounding factions have ceased.
There are two reasons for resuscitating what appears
to be a dead horse. First, in the sudden move away from
factions fifteen years ago, a number of theoretical issues
concerning factions were left underdeveloped or
unresolved. Second, while factional competition was
analyzed extensively in relation to contemporary
peasant polities, its usefulness for understanding the
polities of prehistoric societies has not been explored.
Dealing first with the most important of the unresolved
questions, we can ask, how should factions be dened?
During the 1960s, factions were defined in terms of
their characteristics as a group. Factions were said to be
politically oriented conflict groups whose membership
was recruited and maintained through the efforts of a
leader (Nicholas 1965; Bujra 1973; Silverman 1977). In
such groups, unity derives from ties between leaders and
followers; lateral ties among followers are poorly developed (Nicholas 1965:28-9; Bujra 1973:134). This was
said to account for the loosely structured, personalistic
character of factions (Firth 1957:292). Factions were
also said to be based upon calculations of self-interest
rather than moral commitment to the group (Bailey
1969:52), to be transitory groups with membership

recruited on many different bases (Nicholas 1965), and

to be lacking in corporate property, frequent meetings,
structural complexity, and rules governing succession to
leadership (Boissevain 1964; Bujra 1973). Explicitly or
implicitly, factions were contrasted with corporate
groups, the traditional focus of structural-functional
analysis in social anthropology.
These definitions are quite useful for investigating the
character of factions as a type of informal group, but
they divert attention from the most interesting dynamic
property of factions, namely the competitive relationships between them. I f the object of investigation is to
discover how factional competition acts as a forc of
social transformation, it is preferable to view factions in
terms of what they do rather than in terms of what they
are (Salisbury and Silverman 1977). Therefore, in this
volume, factions are defined as structurally and functionally similar groups which, by virtue of their similarity,
compete for resources and positions of power or prestige.
In this definition, factions are understood to be groups
engaged in political competition which are neither
classes or functionally differentiated interest groups.
The lack of structural and functional diTerentiation
between factions has been frequently noted. In a South
India village, Siegel and Beals (1960:396) found "few
indications of consisten! differences between them in
terms of policy or kinds of people who belonged." In
Boissevain's (1964:1276) view, a faction is "a loosely
ordered group in conflict with a similar group." Similarly, Sandbrook (1972:111) defines a faction as "a
segment of a clientage network organized to compete
with a unit or units of similar type."
Bujra (1973:136-8) provides an excellent theoretical
account of the structural similarity of factions. She
explains that faction leaders come from similar social
backgrounds because while "social distance restricts
competition, social contiguity engenders it . . . Conflicts
thus often begin between people who are more socially
alike than different." In addition, faction leaders tend to
come from the "dominant" sectors of society, since it is
these individuis who have the resources needed to
recruit large followings. Coming from the same privileged sector of society, faction leaders are likely to share
similar political goals, and these goals are not likely to
challenge the basic structure of society.
Bujra (1973:137) adds that leaders, wishing to enlarge
their followings, will seek supporters in all the different
sectors of society, claiming allegiance on many different
bases: past or future patronage, proximity of kinship, a
common religin or ethnic identity, etc. Therefore, the
individuis forming a faction lack an identity of interests

Introduction
that would engender common political goals beyond
winning advantages for their own faction. or are there
clear differences between the members of different factions that might result in policy differences between
them.
Thus, while factions compete for resources, their
structural similarity insures that they will hold similar
ideas about what the world is like and what it should be
like. Factional competition tends to be nonrevolutionary in intent. The objective of factional competition is to achieve a favorable allocation of existing
benefits; each faction hopes to gain more while its competitors gain less. Participants conceptualize factional
competition as a zero-sum game in which one party's
gain is another's loss. Thus, in factional competition,
debate generally centers upon the relative legitimacy of
each faction's claims rather than the merits of substantively different social programs.
Given the lack of structural differences between competing factions, it is at first difficult to see how they might
act as vehicles of social transformation. Because factional competition is non-revolutionary in intent, it has
often been regarded as non-revolutionary in consequence (Siegel and Beals 1960, Gross 1973, Sandbrook
1972). However, Salisbury and Silverman (1977:6-7)
observe that factionalism has an inherent dynamism
grounded in competitive strategizing: "Each confrontation [between factions] changes the terms on which the
next confrontation will take place . . . [T]he strategy of
one side . . . does not produce an exact or mirror-image
strategy . . . Relations are, in fact, systematically oblique
and groupings are systematically unalike. Factionalism,
in short, produces actions and reactions that do not
simply balance o u t . . . "
The issue of whether, as Silverman and Salisbury
suggest, factional competition could serve as a
mechanism of social transformation has not been
thoroughly explored. A major obstacle to gauging the
transformative power of factional competition has been
that, prior to this volume, factions have been studied in
contemporary communities under the dominance of a
state: Native American communities supervised by the
Burean of Indian AlTairs and peasant communities in
colonial or recently post-colonial nations. In such communities, the state constrains both the scale of conflict
and the degree of structural change resulting from factional competition. For this reason, previous studies
may easily have underestimated the transformative
power of factional competition. Almost certainly, factional competition was a more dynamic forc in pre-state
societies than it is in the modern world.

Another issue that has not been thoroughly explored

is the proper scale of analysis for the study of factional
competition. In the ethnographic literature, outside
forces have sometimes been seen as influencing factional
competition. For example, the rise of factions is often
attributed to the decline of traditional, power-holding
corporate groups under the impact of Western contact
(French 1948; Siegel and Beals 1960; Nicholas 1965;
Nagata 1977). And several observers have suggested that
the strength of factions and faction leaders varied
according to their access to outside sources of revenue
and influence, most often supplied by the state (Schwartz
1969; Sandbrook 1972; Bujra 1973; Gross 1973; Schryer
1977; Salisbury 1977). Nevertheless, factions have been
regarded as an aspect of "local level polities" (Schwartz
1968), impinged upon by regional or national polities
but not affecting them in reverse.
In contrast, several essays in this book investgate the
interplay of local and regional processes in tribal polities. Spencer (Chapter 3) arges convincingly that the
internal and external dimensions of tribal leadership are
intertwined. The formation of competing factions within
communities goes hand in hand with the development of
alliances between faction leaders in different communities. The net effect of this process is to turn an entire
regin into a single political "arena," a community
within which competing coalitions of faction leaders vie
for resources. In a similar fashion, it becomes extremely
difficult to diTerentiate between local, internal dimensions of political violence (the suppression of rebellion)
and regional, external dimensions of political violence
(the pursuit of warfare) once communities became
enmeshed in intervillage alliances that compete at the
regional level (see Anderson, Chapter 6, for Mississippian societies and Pohl and Pohl, Chapter 13, for the
Classic Maya).
In examining factional competition and political
development in prehistoric societies, the essays in this
volume raise a number of issues, most of which are new
to archaeology. These include: the opportunities and
constraints presented by different mdiums of competition including feasting (Clark and Blake, Chapter 2),
external alliances and trade (Spencer, Chapter 3), and
warfare (Redmond, Chapter 4; Helms, Chapter 5); the
sources of factional competition in the kinship and political structures of chiefdoms (Anderson, Chapter 6),
States (Byland and Pohl, Chapter 11; Pohl and Pohl,
Chapter 13; Fox, Chapter 14) and empires (van Zantwijk, Chapter 9); factional competition and ethnic
identity (Pollard, Chapter 7; Brumfiel, Chapter 8); and
factional competition and imperial expansin (Hicks,

Elizabeth M. Brumfiel

Chapter 10). The apparent absence of corporate groups

and factional competition is analyzed for the Valley of
Oaxaca (Kowaiewski, Chapter 12), cycles of factional
competition and political change are defined for the
southeastern United States, the Maya lowlands and the
Postclassic highlands (Anderson, Chapter 6; Pohl and
Pohl, Chapter 13; Fox, Chapter 14), and the shifting
composition of factions and factional conflict in preimperial, imperial, and colonial societies is examined for
the central Andes (D'Altroy, Chapter 15).
All the essays in this volume are concerned with the
relationships of factional competition to ecological conditions and class conflict. The next two sections of this
introduction explore these relationships.
Factional competition and cultural ecology
At first glance, the study of factional competition and
cultural ecology would seem to have little to offer each
other. Factional competition focuses attention upon the
inner dynamics of social systems while ecosystem theory
derives the dynamics of social change from the interaction of human populations with their environments
(HiU 1977:88; Binford 1983:221). The study of factional
competition invoives consideration of strategic decisin
making by self-promoting leaders while, in the view of
cultural ecologists, social change is unrelated to the
perceptions and motives of social actors (Hill 1977:66-7;
Price 1982:720). Despite these differences, cultural
ecology can only benefit from a more explicit consideration of factional polities. Such studies would reveal the
internal needs and resource requirements of complex
political institutions that affect their distribution in time
and space. And studies of factional polities would reveal
the importance of the traits that accompany complex
political institutions but appear to serve no critical ecological function. For these reasons, perhaps, a concern
with factional polities is already present in the work of
several ecosystems theorists (Flannery 1972; Webster
1975, 1976; Yoffee 1979; Spencer 1982).
Cultural ecologists generally assert that socio-political
hierarchies evolve because chiefly and state hierarchies
provide for a more effective relationship of a population
to its environment; under certain demographic and
environmental conditions, political hierarchies are
adaptive.
Ecosystem theorists often assume that incipiently
complex political institutions are at least sporadically
present in simpler societies, ready to be pressed into
service when they are favored by demographic and
environmental conditions. The timing and location of

incipiently complex institutions is said to be random;

they do not in themselves constitute a problem suitable
for research. Variation, as Price (1982:716) observes,
"arises constantly in all living systems and does not, in
terms of an evolutionary paradigm, require explanaron." But this is not entirely true. Just as sociopolitical complexity might be precluded by environmental
problems that have no managerial solution, complex
political institutions might be precluded by ecological
conditions that do not meet their own institutional
requirements.
An excellent example of this principie is supplied by
Clark and Blake (Chapter 2, summarizing Hayden 1990;
Hayden and Gargett 1990). All but the most ephemeral
forms of political leadership require a disposable surplus,
a "fund of power" (Sahhns 1968:89). But so long as
humans relied upon limited and fluctuating resources,
the competitive accumulation of surplus depleted communal resources and was not tolerated. Henee, despite
the managerial benefits that more powerful specialized
leadership might have conferred upon the population,
such leadership did not emerge until after subsistence
carne to be based upon rich and reliable food resources.
A second and even more interesting example concerns
long-distance trade. Long-distance trade has a tendency
to increase as political institutions become more
complex. Cultural ecologists, with their attention perennially fixed on population-environment interaction,
have supplied three accounts of long-distance trade. One
regards it as a means of procuring critical resources that
are not locally available (Rathje 1971; Johnson and
Earle 1987:245). A second regards it as a means of
gaining alliances and valuables (storable wealth) that
enhance subsistence security (Flannery 1968; Halstead
and O'Shea 1982). The third regards long-distance trade
as unimportant because it is most often concerned with
sumptuary, as opposed to subsistence, goods (Price
1977; Binford 1983:227-31; Sanders 1984).
However, all these accounts draw attention away from
the fact that valuables acquired from distant sources
supply considerable political control because of their
ability to attract followers, allies and patrons and to
maintain hierarchies of control (Schneider 1977; Earle
1978; Friedman and Rowlands, 1978; Helms 1979; Kristiansen 1981:257; Brumfiel and Earle 1987; Gosden
1989). Coalition building is an essential activity in creating and maintaining political power. But we cannot fully
appreciate the importance of this aspect of long-distance
trade in valuables until we stop looking for a directly
adaptive function for this institution or, unable to find
one, assess such trade as epiphenomenal.

Introduction
Although factional competition must be eonsidered in
ecological analyses of political complexity, ecological
variables are essential for understanding factional competition. As Hayden and Gargett (1990) suggest, factional competition will not exist so long as subsistence is
based upon limited and fluctuating resources associated
with generalized foraging. When factional competition is
present, the success of faction leaders is partly determined by local resource productivity and trade route
accessibility. Faction leaders will be most successful in
reas that are most productive, giving the prevailing
methods of resource exploitation. For example, under
conditions of low agricultural intensication, faction
leaders will do best in reas (like the American Bottoms
regin of the middle Mississippi River) where a large
following can gather to enjoy the benefits of factional
membership without incurring the costs of intensified
subsistence effort. But under conditions of higher agricultural intensification, leaders will do best in reas with
the greatest quantity of intensifiable resources (irrigable
land, etc.).
Pohl and Pohl (Chapter 13) suggest that ecological
variables affecting agricultural production also shape
the onset and intensity of factional competition. Rainfall
agriculture permits a more mobile commoner population, easily able to shift allegiance from one leader to
another. Leaders then compete to control segments of
this mobile population. More intensive agricultural
regimes tie farmers to the land, lessening competition
between political lites and permitting greater political
stability. Price (1984) has also noted this difference,
adding that leaders in rainfall agricultural regimes are
more likely to engage in conspicuous generosity in order
to attract followers.
Anderson (Chapter 6) suggests that extreme competition and violence are common during periods of
environmental instability or change. He believes that
European contact intensified factional competition
among native peoples in the southeastern United States,
first through the introduction of European diseases that
killed individuis who occupied strategic positions in the
regional alliance network, and second through the introduction of European trade goods that opened new possibilities for acquiring wealth items. Spencer (Chapter 3)
cites a case where disease in a Shavante village led to a
realignment of factions.
Geography, by affecting the shape of interaction networks, also shapes factional competition. Clark and
Blake (Chapter 2) arge that more open settlement
systems with greater possibilities for regional interaction
have greater potential for being dominated by a single.

advantageously situated authority. Conversely, both

Anderson (Chapter 6) and Byland and Pohl (Chapter 12)
suggest that, in patchy environments, the difiiculty of
maintaining C o m m u n i c a t i o n s between scattered communities prevents any one from dominating the others.
Under these conditions, factional competition tends to
persist in a more or less stable equilibrium.
The recognition that factional competition is shaped
by ecological variables does not imply that factional
competition is always, at the base, caused by subsistence
shortages. Cultural ecologists have frequently argued
that intercommunity warfare is the result of growing
populations competing for scarce subsistence resources
(Sanders and Price 1968; Carneiro 1970; Webster 1975;
Ferguson 1984; Johnson and Earle 1987). Alternative
views on the motivation and character of warfare appear
in this volume. For example, Redmond (Chapter 4) finds
that, in northern South America, tribal warfare is motivated by the desire for revenge rather than the desire for
resources. A careful consideration of Panamanian chiefs
leads Helms (Chapter 5) to conclude that their leadership of warfare was motivated by their desire for personal gain without the added spur of population
pressure. Spencer (Chapter 3) suggests that the elitist
character of warfare in ranked societies is revealed
archaeologically in the fact that lite centers are fortified
while smaller communities are not.
But if warfare arises from the desire for revenge, why
does it occur in some societies but not others (Johnson
and Earle 1987:124, 134)? And if warfare arises from the
selfish motives of chiefs, why do followers particpate?
As Redmond (Chapter 4) makes clear, individuis are
very reluctant to fight; in fact, an ambitous leader may
earn the indebtedness of individuis who must seek
revenge by organizing a raiding party on their behalf (see
also Spencer, Chapter 3). Presumably, leaders organize
these rads on the same basis as other activities that they
carry out, by calling to action those ndebted to them for
previous favors. Thus, warfare becomes possible once
individual jealousies and the desire for revenge become
linked to the political goals of self-aggrandizing leaders
who have established followings that they can cali upon
to implement their plans (see also Sillitoe 1978).
In more stratified societies, the participation of subordnate groups in warfare is less problematic. Rulers
compel participation through coercin, and they reward
participation by conferring promotions of status upon
those whose performance is outstanding. The link
between warfare and population pressure in stratified
societies is weak. This is clearly indicated by Pohl and
Pohl's (Chapter 13) observation that the population of

Elizabeth M. Brumfiel

Caracol grew by 325 percent in the 130 years after it

defeated Tikal. If Caracol was making war with a population of less than one-third of its capacity, it is difficult
to believe that the war was a consequence of population
pressure.

On the other hand, several anthropologists, working

from a variety of positions sympathetic to Marxism,
have suggested that factional conflict within the lite
stratum explains the intensification, modification and
decline of lite power in chiefdoms and agrarian states
(Webster 1975; Earle 1978, 1987; Cowgill 1979; Kristiansen 1981; Brumfiel 1983; Gailey and Patterson 1987;
Patterson 1991). This volume extends their arguments to
suggest that conflicts within (both commoner and lite)
strata interact with conflicts between strata to determine
the course of political development.
First of all, competition between non-elites provides
frequent opportunities for leaders to expand their influence and power. As Spencer (Chapter 3) and Redmond
(Chapter 4) indcate, leaders in lowland South America
increased their influence by assisting individuis
embroiled in personal disputes or blood feuds. Brumfiel
(Chapter 8) suggests that mediating competition
between calpulli and teccalli groups for houses, land,
titles, and other resources was an important service performed by city-state rulers in central Mxico prior to
Aztec rule. Pohl and Pohl (Chapter 13) indcate that the
Postclassic Maya lites, as patrons to their subjects,
resolved property disputes and defended community
resources against outside attack. And the military commanders of the pre-Inka Wanka prosecuted wars that at
least ostensbly defended community resources from
rads by neighboring groups (D'Altroy, Chapter 15). In
each of these cases, competition among non-elites motivated commoners to subordnate themselves to a political leader. The resulting algnments were factions.

Factional competition and class conflict

If, because of earlier research agendas, factions have
frequently been contrasted with corporate groups, our
own interest in social transformation leads us to contrast
factions with classes. Under conditions of class struggle,
society is divided by horizontal cleavages that seprate
internally solidary and externally competing strata. This
contrasts with a situation of factional competition in
which society is divided by vertical cleavages that unite
members of different strata and foster conflict between
members of the same strata.
Intra-class competition is a common theme in Marx's
writings on capitalist society. Marx (1977:266-7)
observed that capitalist society contained two marketdriven, intra-class struggles: on the one hand, "the
industrial war of capitalists among themselves" to maintain profits, on the other hand, the competition among
workers for employment. Marx maintained that intraclass competition is only overeme by class consciousness brought about by class struggle: "The seprate
individuis form a class only in so far as they have to carry
on a common battie against another class; otherwise they
are on hostile terms with each other as competitors"
(Marx and Engels 1947:48-9). Thus, factional competition and class conflict are presented as inversely related,
the former fading as the latter intensies. And class
At the same time, competition among political ehtes
struggle is given the greater explanatory weight: "The frequently moderates the intensity with which comhistory of all hitherto existing society is the history of moners are exploited. In the tribal societies described
class struggles" (Marx 1977:222).
by Clark and Blake (Chapter 2), Spencer (Chapter 3),
As Bloch (1983:18) points out, Marx's emphasis upon and Redmond (Chapter 4) leaders compete with each
class struggle has not been very helpful for anthropolo- other to supply benefits to followers, although the sucgists investigating the causes of change in classless soci- cessful leader will also have cultvated his ability to cali
eties. In the absence of class struggle, Marxist anthropo- in his debts at critical junctures. In chiefdoms and
logists have variously attributed social change in cty-states, rulers seek to finance their competition
classless societies to technological development and against rival lites by enlarging the size of their tributeenvironmental change (see Levine and Wright 1980); to paying populations. Although this is sometimes accomthe structural incompatabilities ("contradictions") plished through conquest, it can also be achieved by
between the forces of production, the relations of pro- offerng commoners prime agricultural land (Earle 1978)
duction, and the social and ideological superstructure or a low per capta tribute burden (Pohl and Pohl,
(Godelier 1977; Friedman 1975; Friedman and Rowlands Chapter 13) as inducements to settle. In pre-Aztec
1978); or to the conflict of interests between individuis Mxico, intense competition within the rulng class
who occupy subordnate statuses within society (women, enabled commoners to move from one city-state to
lineage juniors, etc.) and those who domnate them another, thus avoidng conditions of intense exploitation
(Bloch 1983:160). The dynamics of conflict among those (Hicks 1982; Brumfiel, Chapter 8).
The suppression of commoner residential mobility
in similar social positions has received little attention.

Introduction
may be a necessary condition for the existence of class
stratification in agrarian states. This could be accomplished through "social circumscription" (Carneiro
1970), as Pohl and Pohl (Chapter 13) suggest for the
Peten Maya. Or, it might be accomplished by the political unification of a regin, resulting in uniform conditions of exploitation for the commoner class. Seeing
the advantages of unification, ruling lites might voluntarily surrender their sovereignty to an expanding
regional state (Smith 1986; Hicks, Chapter 10).
Commoners can sometimes exploit intra-elite competition to their own advantage even without leaving
home. When plagued by oppressive rule, commoners can
support the efforts of some ambitious prince to overthrow the incumbent ruler. Class warfare can assume the
guise of civil war between noble factions (Fallers
1956:247; Sahlins 1968:92-3). Thus, the suppression of
intra-elite competition is a second condition necessary
for the emergence of class stratification. Mechanisms for
unifying the ruling class include fostering a homogeneous lite culture (Pollard, Chapter 7; Brumfiel,
Chapter 8), lite intermarriage (van Zantwijk, Chapter
9), the rotation of status-conferring ritual and political
activities (van Zantwijk, Chapter 9; Pohl and Pohl,
Chapter 13; Fox, Chapter 14), and the forging of
patron-client relationships between the state and individual members of the regional nobility (Pollard,
Chapter 7; Hicks, Chapter 10; Fox, Chapter 14;
D'Altroy, Chapter 15). Interestingly enough, a homogeneous lite culture, lite intermarriage, and possibly
the rotation of ritual responsibilities were present among
the Classic Maya, who never achieved regional unification (Pohl and Pohl, Chapter 13). Thus, intra-elite
patron-client relationships, backed by coercive forc,
emerge as the most important mechanism for securing a
unified ruling class among the relatively non-bureaucratic Aztecs and Inkas.
While factional competition affects the dimensions of
social inequality, class structure shapes competition and
alliance building (Lloyd 1965). In the tribal systems
described by Clark and Blake (Chapter 2), Spencer
(Chapter 3), and Redmond (Chapter 4), classes are
absent, and followers align themselves with the leader
who supplies them with the greatest immediate benefits.
These vertical alliances are shallow, extending only from
followers to the local leader. On the regional level, linkages are supplied by alliances between village leaders,
and these linkages are the most critical advantage that
incumbent leaders enjoy over aspiring rivals. Alliances
between leaders give incumbents greater access to exotic
goods and military assistance than is available to their

rivals. Spencer (Chapter 3) suggests that efforts by

incumbent leaders to regularize relationships within
their alliance network might lead them to deal preferentially with the heir of a deceased leader, initiating a
form of ascriptive leadership that could develop into a
permanent chieftainship.
Institutionalized tribute extraction in chiefdoms and
city-states makes available greater quantities of wealth
for lite competition and alliance building. Anderson
(Chapter 6) suggests that competition is most intense
when the material rewards associated with leadership are
greatest, and this is borne out by the high level of
factional competition within the lite stratum of the
chiefdoms and city-states described in this volume.
Internally, cise kinsmen struggle to control leadership
offices; externally, leaders struggle to gain higher positions in the regional political hierarchy. Their allies in
this quest are an unstable coalition of consanguineal and
affinal kin and commoners raised to noble rank as a
reward for valorous military service. These coalitions are
held together by the redistribution of tribute wealth to
noble followers and the allotment of segments of the
tribute-paying population to the leader's strongest
rivals.
But the intra-elite competition, as discussed above,
limits exploitation, creating a chronically underfunded
ruling class. To augment their incomes, leaders make
war upon their neighbors, and, as Redmond (Chapter 4)
documents so convincingly, warfare in chiefdoms
reaches an intensity that is clearly greater than that
found among tribal peoples. To survive both internal
and external competition, leaders place themselves
under the patronage of strong regional leaders. These
vertical alliances have greater depth (three or four levis)
and territorial range than the vertical alliances found in
tribal "big-man" systems. Anderson (Chapter 6) suggests that such systems are marked by a secular trend
away from intra-elite and inter-class relations based
upon display and redistribution toward the greater use
of forc.
The unification of lites in a regional state permits
very high levis of tribute extraction. This surplus flows
to the paramount ruler who establishes himself as the
primary supplier of sumptuary goods and military assistance to subordnate leaders. Thus, vertical alliance networks all converge on the state rulers while horizontal
alliances between local rulers wither away. State patronage permits subordnate rulers to enjoy a definite
in-group advantage over their local rivals. As in modern
systems of centralized patronage, the level of overt factional competition tends to be quite low.

Elizabeth M. Brumfiel

The position of local lites is further weakened when

the state creates new territorial units and administrative
offices filled by members of the state's ruling group
instead of by local rulers. Such policies result in the
severing of ties between rulers and ruled and strengthen
class stratification (D'Altroy, Chapter 15). Subsequently, these policies may give rise to disputes
between those who have traditional claims on resources
and those who derive claims from the new system, both
of whom must turn to the state for recognition of their
claims. The absence of local solidarity and the competition between local factions for favorable treatment
by the state weakens the local capacity for resistance
(D'Altroy, Chapter 15; see also Dennis 1987). The
greatest threat to these states is factional competition at
the very highest level, within the royal family (van
Zantwijk, Chapter 9; D'Altroy, Chapter 15).
As states disintegrate, considerable wealth and power
may become lodged outside the realm of political control.
Blanton (1983) points out that associations organizing
craft production or exchange tend to arise during periods
of weakened state control. Religious power may also be
lodged in more or less autonomous institutions such as
the priesthood during Postclassic times in Oaxaca and
possibly also the Maya lowlands (Pohl and Pohl, Chapter
13; see also Patterson 1985). To deal with these groups,
political lites are sometimes forced to admit their
members to political office so that the interests of outside
groups coincide with the interest of the political lite
(Lloyd 1965:98). At other times, such groups promote
factional competition among political lites to maintain
their own autonomy. Aligning themselves with different
lite factions, outside groups may succeed in transforming a tributary state into a broker state, in which factional
competition is overshadowed by competition between
functionally differentiated interest groups.

Factional competition and the archaeological record

Factional competition invoives two complementary processes: the construction of coalitions of support and
participation in political contests. In building coalitions
of support, leaders forge ties between themselves and
their clients, allies, and patrons. In political contests,
leaders exchange Information on the strength of their
coalitions and determine who will control contested
resources (see Bailey 1969). Coalition building and contests may occur simultaneously, as when leaders compete
to offer potential supporters the most attractive gifts.
Both alliance building and political contests leave distinctive imprints on the archaeological record.

Alliance building is frequently achieved through

exchange. Locally, the liberal distribution of gifts and
preferred foods is used to attract followers who are then
tied to the leader by their indebtedness for unreciprocated favors (Sahlins 1968:88-90). Regionally, balanced
gift exchange establishes a pattern of mutual aid between
allied leaders while asymmetrical exchange (involving
the movement of staple crops upward and the flow of
valuables downward) characterizes relations between
local lites and regional paramounts. The valuables used
in these exchanges are scarce and highly valued, usually
owing to their foreign origin or the quantity of labor
involved in their production (Drennan 1976:357). The
valuables are endowed with symbolic meanings that valdate the alliances under construction; furthermore, the
valuables are distributed in ritual contexts that further
vahdate the relations of alliance.
The intensity and organization of alliance building are
visible in the frequency and distribution of exotic or
highly crafted wealth items, preferred foods, and
feasting paraphernalia in prehistoric sites. Clark and
Blake (Chapter 2) cite the presence of finely finished,
elaborately decorated ceramies and maize to arge for
competitive coalition building on coastal Chiapas by
1600 BC. D'Altroy (Chapter 15) suggests that the high
concentration of butchered camelid bones and certain
jar and basin types in lite households are evidence of
lite sponsorship of feasts at Tunanmarca, Per. Since
there appear to have been few dietary differences
between lites and commoners, commoners were probably the guests at elite-sponsored feasts (see also Costin
and Earie 1989).
A more exclusive sphere of lite alliance building at
Tunanmarca is suggested by the restriction to lite
houses of metal working and metal artifacts. Metal was
probably used in gift exchanges that created coalitions of
support among political lites. In Barinas, Venezuela, a
similar restriction of polished stone jewelry (much of it
from extra-local serpentine) to lite contexts also suggests the existence of a seprate sphere of lite alliance
building (Spencer, Chapter 3).
The presence of non-local goods in non-elite contexts
might reveal another dimensin of alliance building. For
example, in Early to Middle Formative Oaxaca, the
uniformity among households of the sources of obsidian
used suggests that obsidian was distributed to all households from a single point (Winter and Pires-Ferreira
1976). This could be interpreted as evidence that obsidian procurement and distribution was used by a faction
leader to build a popular following (see Clark 1987 for a
discussion of how the procurement and processing of

Introduction
obsidian provided opportunities for political entrepreneurship in Middle Formative Mesoamerica).
Shifts in the frequency and distribution of prestige
goods, preferred foods, and feasting paraphernalia
provide Information on changes in the structure and
intensity of alliance building over time. Anderson
(Chapter 6) arges that the declining frequency of prestige goods in Mississippian chiefdoms marks the transition from leadership based on persuasin (which
required chiefs to build a mass following among commoners) to leadership based on coercin (which permitted chiefs to limit their attentions to a smaller group of
strong-arm men). In the Valley of Mxico and the central
Andes, a decline in the frequency of vessels used in
feasting in the capitals of previously autonomous states
reflects the suppression of political competition among
local polities by a powerful regional state (Brumfiel
1987a, Costin and Earle 1989).
Marriages also play an important role in alliance
building. Unfortunately, the archaeological record only
rarely preserves evidence of marriage alliance. Stone
inscriptions from the Classic Maya constitute one of the
very few instances where such records are preserved
(Marcus 1976). Pohl and Pohl (Chapter 13) suggest that
the Late to Terminal Classic shift in stelae subject matter
from marriage to warfare reects a secular trend in Maya
political factionalism from an earlier dependence upon
alliances among regional lites to a later pattern of
violent competition.
The construction of coalitions requires the production
of surplus wealth which can underwrite gift exchange
and feasting (Sahlins 1968, D'Altroy and Earie 1985).
Because the vast majority of production in agrarian
societies is household based, changes in the intensity of
factional competition should be marked by changes in
household size and composition. The initial stages of
coalition building ought to be marked by high birth
rates, polygamy, and/or the inception of dependent
labor within the households of faction leaders (Sahlins
1968:89, Coontz and Henderson 1986). Increases in
leaders' influence and power should be reflected in the
size and structure of a growing number of households as
the leader extracts increasing amounts of goods and
labor from a wider circle of clients and subjects.
Successful coalition building might leave its imprint
on settlement patterns. The size of the leader's settlement
may suddenly increase as it did during the emergence of
chiefdoms on the south Chiapas coast (Clark and Blake,
Chapter 2). The early expansin of San Jos Mogote in
the Valley of Oaxaca might also be an example of an
unusually large settlement created by successful coali-

11

tion building (Flannery and Marcus 1983b). The successful control of local rivals should be reflected in the
distribution of lite residences within a regin, as in de
MontmolHn's (1989:191-6) "Ehte Forced Settlement"
measure - the proportion of lite residences found at
political centers compared to the total number of ehte
residences in the political catchment controUed from
these centers. Alternatively, paramounts might favor a
policy of dispersing their rivals to a mximum extent
(Anderson, Chapter 6). A lack of control over rivals
might be indicated by evenly dispersed, tight clusters of
lite and commoner housing indicative of leaderfollower groupings well suited to factional competition.
Successful coalition building may also be evident in
constructions that by their size or complexity suggest
communal labor: raised fields and causeways in Venezuela (Spencer, Chapter 3), agricultural terraces and
defensive works in Per (D'Altroy, Chapter 15), and
fortifications and monumental architecture among the
Mayas (Pohl and Pohl, Chapter 13; Fox, Chapter 14).
The particular type of labor investment reflects different
strategies for competitive success. Agricultural intensification may improve the leader's ability to attract followers by sponsoring larger feasts or by supplying them with
improved lands (Earle 1978). Fortifications suggest that
warfare provided wealth that a leader could redistribute
to followers (Webster 1975). Monumental architecture
might involve manipulation of the symbols of group
unity: the ancestral or patrn deity. Such symbols would
be most prominent under conditions of competition at
the regional level as part of an effort to crate bonds
between leaders and followers that could not be easily
transferred to competing leaders. Ethnic symbols may be
manipulated with similar goals (Brumfiel, Chapter 8;
Pohl and Pohl, Chapter 13). Public architecture also
suggests efforts to impress a regional audience of
potential allies and rivals who use the size of the building
projects to judge the size and commitment of one's
following.
Competition on the regional level often takes the form
of warfare. Warfare is archaeologically visible in a
number of ways: physical evidence of violent death and
the taking of war trophies, defensive works, large settlement size, ephemeral site occupation, defacejnent of
public buildings and prestige goods, and depictions of
warfare in art and inscriptions. Drawing upon data from
Oaxaca, Kowaiewski (Chapter 12) suggests that territorially based competing polities on the regional level
might also be indicated by the occurrence of shrines,
boundary markers, and buffer zones between localities,
by roughly equivalent amounts of civic-ceremonial

12

Elizabeth M. Brumfiel

architecture across subregions, by the absence of organic

sohdarity between subregions, by the existence of subregional differences in settlement patterns and ceramic
styles, and by a settlement plan in regional capitals that
provides representation of geographically affiliated subregions. Many of the archaeological manifestations of
sociopolitical and spatial segmentation in the Valley of
Oaxaca are also to be found in the Classic Maya polity
analyzed by de Montmollin (1989).
Conclusin: implications for theory
This overview of factional competition and political
development has eonsidered a wide variety of ecological
and social variables. Resource productivity, geographical boundedness, long-distance
exchange,
warfare, kinship and marriage systems, rules of succession, class, corporate group structure, and ethnicity all
impinge upon the patterning and intensity of factional
competition and are, in turn, affected by it. The number
of variables discussed reflects the fact that competitive
advantage is pursued through the strategic manipulation
of many material and social variables; winning strategies
must be fabricated according to the resources available.
The relative advantages of extensive or intensive agriculture, attached craft production or long-distance
exchange, marriage alliance or warfare, ethnic assimilation or ethnic persistence vary according to local circumstances. In addition, the efficacy of different strategies
vares over time, depending upon the development of
resources, alliance systems, and competitive strategies as
consequences of earlier rounds of competition.
Given the great variation of competitive strategies
over time and space, no two cases are identical. While
the logic of factional competition provides an essential
principie for understanding the course of local history,
the study of factional competition is unlikely to reveal
any universal laws of cultural development. The variables involved are too numerous; the strategies for mobilizing resources are too diverse. Although factional competition provides a common mpetus to political
development, any particular sequence of development is
uniquely complex and contingent.
In the face of such complexity, it is necessary to altrnate between a subject-centered and a system-centered
analysis. A subject-centered analysis organizes ecological
and social variables by weighing them according to their
importance in specic competitive strategies. A systemcentered analysis reveis how the implementation of
strategies alters the quality and distribution of ecological
and social resources for the next round of competition.

A subject-centered analysis is also required in order to

account for the dynamics of political development. Political development should be attributed to the efforts of
individuis to advance their goals of material advantage
and social esteem by joining factions or class alliances
that prosecute their interests in competition with opposing factions or classes. Competition occurs within a
structure or matrix of ecological resources, social relations, and cultural rules and vales, which constrain
behavior but which also provide opportunities for
innovation. The growth of political complexity, marked
by the emergence of larger, more centralized polities
with greater degrees of vertical and horizontal differentiation, is an epiphenomenal consequence of the strategies
and counter-strategies employed by leaders and followers engaged in factional competition and lites and commoners locked in class struggle (see Clark and Blake,
Chapter 2).
The methodological individualism at the basis of this
formulation might be criticized as a projection of the
competitive, self-seeking, pragmatic ideology of Western
capitalist society. This is partially true, although the
problem is ameliorated by situating self-interested competition in specific ecological, social, cultural, and historical contexts. Even so, such a formulation might be
inappropriate for analyzing the more expressive and
solidary aspects of social life. However, an assumption
of competitive, self-seeking, pragmatic social actors is
eminently suited to the analysis of factional competition,
for, as Earle (1987:294) observes, factional competition,
which is inherently a competitive, pragmatic process,
requires a maximizing strategy (see also Bailey
1969:36-7).
The subject-centered/system-centered approach to
factional competition and political development
adopted in this volume contrasts sharply with the
exclusively system-centered focus of both cultural
ecology and structural Marxism, currently the two most
popular theories for explaining political change in
ancient societies. Both cultural ecology and structural
Marxism focus upon strictly bounded, culturally distinct social groups. In cultural ecology, the social unit is
defined by external boundaries, created by geographic
isolation and by the competitive relationships that
develop between populations under conditions of
resource scarcity (Sanders and Price 1968; Carneiro
1970; Ferguson 1984). In structural Marxism, the social
unit is defined by its internal organization of social
production and reproduction under a guiding set of
social rules and practices ("structures") (Friedman 1975;
Godelier 1977:63).

Introduction
Both cultural ecology and structural Marxism postlate the narrow constraint of human behavior and decisin making within these systems: by stringent considerations of energetic efliciency in cultural ecology (Price
1982:719) and by the Hmits of structurally determined
consciousness in structural Marxism (Godelier
1978:768). These same constraints oprate equally for all
members of society, implying a condition of cultural
homogeneity for human groups.
Based upon such assumptions, cultural ecology and
structural Marxism supply accounts of the growth of political complexity that are quite different from the one proposed in this volume. Cultural ecologists believe that
incipient political complexity occurs as a normal part of
the array of random, low-level cultural variation present
in all populations (Price 1982:724). However, political
complexity begins to grow only after population pressure
has created the need to intensify production within
groups and to compete for resources belonging to others
(Sanders and Price 1968; Carneiro 1970; Webster 1975;
Ferguson 1984; Johnson and Earle 1987:16-18). Through
either expansin or emulation, the beliefs and practices
that result in larger groups with greater ability to organize
production and warfare preval within a regin. The concentration of wealth, coercive power, and prestige in the
hands of political leaders contributes to managerial
capacity by giving leaders the ability to coordnate larger,
more complex populations (Webster 1975, 1976).

13

duction, and social and ideological superstructure)

resulting in structural reorganization (Friedman 1975;
Godelier 1977:63, Bloch 1983:154). Power is not constructed by individuis, but rather falls to certain people
as a consequence of the prevailing cultural rules,
especially those allocating resources, labor, and product
(Godelier 1978; Bender 1985). In the initial phases of
political complexity and social inequality, the conferring
of superordinate status occurs with the consent of subordnate groups, with superordinate status falling to
those who medate between humans and deities, i.e.
those who control the imaginary means of social reproduction through religious ritual (Friedman and Rowlands 1978; Godelier 1986:156-64; Bender 1985).

Like cultural ecology, this account of political development is deficient. Its principal difiiculty is that it
postlales consensus within the body politic prior to the
emergence of class and class struggle. While the consent
of followers is certainly needed by leaders in the very first
stages of political complexity (Clark and Blake, Chapter
2; Spencer, Chapter 3; Redmond, Chapter 4), evidence in
this volume and elsewhere suggests that coercin and not
consensus is a dominant motif in complex chiefdoms and
is absolutely pervasive in agrarian states.
Cultural ecology and structural Marxism difier from
the approach taken in this volume in yet another way.
They both postlate strict behavioral determinism while
this volume takes seriously the impact of human purpose,
This formulation can be criticized on two grounds. creativity, and choice. This difference derives from the
First, it proposes that population pressure and resource number of variables employed by each of the analyses.
shortages are necessary to initiate the process of political Both cultural ecology and structural Marxism employ
development and to maintain its progress. There is, models constructed of relatively few variables, which
however, empirical evidence to the contrary, presented necessarily limits human choice and agency to a few highly
in this volume and elsewhere. Second, in analyzing the redundant options. As a consequence, human behavior
concept of power, cultural ecologists focus attention on emerges as highly determined. In contrast, the studies in
the energy used to maintain competing populations and this volume consider the rich complexity of competitive
the Information flows used to organize them (e.g., White strategizing, which invoives the manipulation of every
1959; Price 1982; Wright 1969; Flannery 1972; Johnson conceivable ecological and social variable in complex and
1978). But they tend to ignore the relations of alliance contingent ways. In doing so, they broaden the dimenand dominance that brings organizational structures sions of human choice and leave room for the exercise of
into existence and maintain them. The manipulation of human agency. Thus, the willingness to consider the full
social relations is as important as the manipulation of complexity of specific historical situations restores the
natural resources in the process of political development, concept of agency to the study of social change.
a point which cultural ecologists tend to ignore (but not
always, see Webster 1975, 1976; Spencer 1982).
A cknowledgments
Structural Marxists present an epigenetic model of
political development (Friedman and Rowlands 1978).
Political change is not due to the goal-directed strategies
of political leaders but, rather, is a consequence of structural contradiction (that is, the incompatibility of the
interlinked forces of production, relations of pro-

Earlier drafts of this chapter were patiently read by John

Clark, Timothy Earle, John Fox, Roberto Korzeniewicz, Hattula Moholy-Nagy, Mary Pohl, and Glenn
Perusek. Their criticisms, comments, and line editing
shaped the chapter in many important ways. I am grateful for their help.

7
Ethnicity and political
control in a complex society:
the Tarascan state of
prehispanic Mxico
HELEN PERLSTEIN

POLLARD

The interrelationship between ethnic units and a central

political authority is known historically to have been
crucial to the operation of complex societies. In the
process of this interaction central authorities, particularly ruling lites, have created new ethnic groups,
have altered the attributes which define ethnic identity,
and have restructured relationships between ethnic
groups (Enloe 1980:17ff.). Political authorities have two
fundamental goals for the survival of their centralized
power: (1) the economic exploitation of populations and
resources, and (2) the protection of the integrity of the
state frontiers. In achieving both these goals ethnic
diversity can either faciltate or hinder lite action.
Centralized authorities can assure maximal access to
populations and resources when decisin making flows
from the top downward, according to principies established by dominant lites. Ethnic diversity often disrupts
this flow, by interposing local or regional leaders, who
acquire power not through their allegiance to central
hierarchies, but through positions of ethnic status. Decisin making may be undertaken to reflect the needs of
local populations or local lites at the expense of the
state. On the other hand, under conditions of rapid
territorial expansin, when large populations and/or
resources are being incorporated into a single political
unit, the existing lines of authority, legitimacy, and
social cohesin present n ethnically distnct populations
may provide central authorities with an infrastructure
of political and economic networks that can be tapped
to the benefit of the state. In a similar manner, the maintenance of the state's territorial integrity demands
populations wUing to defend that territory, and not
themselves act to foster rebellion against central auth-

ority. That can be achieved in at least two ways: the

acceptance of state ideology and legitimacy, usually
through the identification of individuis and groups as a
single social group, i.e., coirmion ethnicity; or the acceptance of common self-interest among ethnically diverse
populations who see their itmninent survival as dependent on subordination to a central authority. The second
is often found on active military frontiers or among
refugee populations fleeing conquest. To the state such
peoples can provide valuable service as warriors, messengers, spies, and long-distance traders, services that
may outweigh the hazard of desertion.
To a great extent our understanding of the evolution
of complex societies is based on societies known primarily or exclusively through archaeology. One
approach to studying the archaeology of ethnicity, that
of ethnoarchaeology, has primarily concentrated on
acephalous societies, attempting to test the regularities
in relationships between isolable "cultures" and artifact
distributions (e.g. Hodder 1979). A second major
approach, modeling the spatial and functional distribution of prehistoric ethnic groups on the basis of
ethnohistoric evidence, has the advantage of dealing
directly with complex, ethnically plural societies. The
primary disadvantage is its limitation to those societies
knowable through both ethnohistoric and archaeological Information. These societies, while representing
only a small sample of complex societies which have
ever evolved, nevertheless provide a basic resource for
the identification of those ethnic processes of significance in societal evolution and their detection
archaeologically. In the New Worid such an approach
has been used in the Andes to evalate the patterns of
Inka expansin (Morris 1982; Rowe 1982; Wachtel
1982, among others) and to genrate models of ethnic/
state interaction which can be applied to earlier periods
(Schaedel 1978).
The central defining feature of an ethnic group, selfidentification, and categorization by the governing lite,
is cleariy beyond the resources of archaeologically
derived research. Nevertheless, because ethnicity is generally associated with high rates of endogamy, the
sharing of clusters of beliefs and vales, marked by the
use of a common language, and is often territorially
isolable, the discemment of ethnic variation is not
impossible. As with all aspects of archaeological analysis, the introduction of ethnicity into the variables
studied must be done with care. Ethnic boundaries are
fluid, contracting or expanding with major political and
economic shifts in the society at large. Ethnic identifications coexist with other social identifcations, meaning
79

80

Helen Perislein Pollard

that occupational and class boundaries, for example,

may cross-cut ethnic groups. Different artifact classes, or
properties of artifacts may signal these different social
boundaries. Moreover, within a single multi-ethnic,
complex society the intensity of ethnic affiliation may
differ, and along with this the number and kinds of
markers used by a group to signal ethnicity may vary.
Finally, the archaeologist deals with units of time that
often compress ethnic "maps" and obscure the processes
of ethnic emergence and change.
Just as complex societies vary in their basic structures
and developmental trajectories, one must expect similar
variation in the structure and significance of ethnicity.
As part of the larger process of developing theories of
the nature of complex societies, one goal should therefore be the development of models of ethnicity, based on
various ethnohistorically knowable societies, and their
judicious application to earlier polities. With this goal in
mind, a model of the structure of ethnicity in the protohistoric Tarascan state of central Mxico (AD 14501520) is herein presented (Fig. 7.1).
The Tarascan territorial state in its sixteenth-century
form is renowned for its high degree of political
centralization and relatively unchallenged control of its

n
p AZTEC
H TARASCAN
y QUICHE
KM

Fig. 7.1 Protohistoric Mesoamerica.

territory (Gorenstein and Pollard 1983, Pollard 1993).

When viewed from the perspective of the geopolitical
core, the Lake Ptzcuaro Basin, these characteristics can
be related to the emergence, by the protohistoric period,
of a social system with a fully Tarascan identity, produced by the conscious subordination and replacement
of local ethnic/linguistic status as the basis for social or
political power. Despite clear indications of earlier
ethnic heterogeneity in central Michoacn (Relacin de
Michoacn [1541] 1956), by the sixteenth century the
population was self-identifying and being identified by
others as solely Tarascan (Relacin de Michoacn [1541]
1956; Suma de Visitas de Pueblos [1547-50] 1905; Relaciones Geogrficas [1579-81] 1987; Warren 1968, 1985,
among others).
Such subordination and replacement of ethnic variation distinguishes this system from the mainstream of
protohistoric Mesoamerica, which was characterized by
ethnic economic and political specialization and multiethnic social classes (Carrasco 1971; Zantwijk 1973).
Along the Tarascan military frontier, however, the
Tarascan polity was multi-ethnic, plural, and demographically largely non-Tarascan (Mendieta y Nez
1940; Brand 1943; Gorenstein 1974, 1985; Herrejn
Peredo 1978; Gonzlez Brespo 1979; Contreras Ramrez
1987). Taken together, these policies, the one emphasizing social homogeneity of the lite and a new common
Tarascan identity, and the other emphasizing plurality
and heterogeneity, appear to be in conflict.
However, these two distinct policies of ethnic assimilation and ethnic segregation, dominated community
interaction in geographically seprate zones of the
Tarascan polity. By 1520 they had resulted in the ethnic
boundaries recorded in the early documents and mapped
by Brand (1943; Fig. 7.2). Combining our knowledge of
the ecological, economic, political, and artifactual variation within this territory we can propose the following
model of the Tarascan state.
Zone of assimilation
This is the territory within which Tarascan was the
dominant language and cultural identity was ethnic Tarascan. It included at least two distinct regions.
Ethnic heartland
This is the zone within which the Tarascan political core
existed, within which Tarascan vales and norms were
held by the bulk of the population, and within which
there was a similar economic and settlement adaptation
to the regin (Stanislawski 1947). Thus, this zone defines
the regional marketing network of the core (Fig. 7.3) and

Ethnicity in the Tarascan state

81

Fig. 7.2 Sixteenth-century ethnicllinguistic boundaries. Based on Brand 1943.

the primary extent of Tarascan religin and ideology.
The population within this zone, especially the lite,
participated in a unitary social system dating to the
political incorporation of central Michoacn (13501440) and the emergence of Tarascan identity (Gorenstein and Pollary 1983). The geographical extent of this
zone is largely co-terminous with the regin within
which Tarascan remained the dominant language as late
as 1750 (West 1948; Fig. 7.4).
Within this heartland zone there are known to have
been several cultural/linguistic groups in the period
before the emergence of the Tarascan state {Relacin de
Michoacn [1541] 1956). These groups were politically
autonomous, and most were socially differentiated.
Within the Lake Ptzcuaro Basin alone there were at
least four such groups, including proto-Tarascan
speakers, local Nahua speakers, and two seprate
groups of "Chichimecs" who had, at different times.

migrated into the regin during the Postclassic (AD 1000

and later) (Relacin de Michoacn [ 1541 ] 1956; Relaciones
geogrficas [1579-81] 1987; Brand 1943; Beltrn 1982;
Gorenstein and Pollard 1983). Each group was identified
by the worship of specific deities, the performance of
specific rituals, and the wearing of specific status markers
(Relacin de Michoacn [1541] 1956). The political unification of this zone was accompanied by the emergence of
Tarascan ethnicity, marked by the universal use of the
Tarascan language and the incorporation of all previously autonomous, culturally distinct groups. Political
incorporation, and the ideology of power that accompanied it, are seen as primary in creating the homogeneous Tarascan zone that is documented in the sixteenth century. This rapid assimilation in the Ptzcuaro
Basin from A D 1250 to 1400, and in the surrounding highlands from A D 1350 to 1440, was probably made possible
by at least two factors: (1) the bulk of the heartland

82

Helen Perlstein Pollard

Fig. 7.3 Protohistoric Tarascan market and tribute regions.

was probably of "proto-Tarascan" culture, and (2) the >Warren 1968; Relaciones Geogrficas [1579-81] 1987;
groups in question were not politically unified with a his- IPonce [1586] 1968; Brand 1943:51-2). The regin repretorical pattern of multi-ethnic complex society (Fig. 7.5) ssented an entrance of Tarascan identity into fundamentally
different resource zones, dominated by temr
Zone of active assimilation
fprate and tropical ecosystems. The degree of
This zone, absorbed into the expanding state only after aassimilation by 1520 clearly varied, and, as Brand found
1440 (Pollard 1993), was increasingly basic to the main- i:in the northern Tarascan zone, Tarascan ethnicity could
tenance of Tarascan lite society. While populations vvanish within decades of European control.
were relatively low in density and widely dispersed
(Warren 1968; Gonzlez Crespo 1979), many resources
Combining the data from the Guzman relaciones and
basic to lite identification came from this zone, includother sixteenth century sources, it would appear that
ing tropical fruit, cacao, cotton, copal, jaguar skins,
the Tarascans had managed through conquest, colotropical feathers, gold, silver, copper, and tin (Pollard
nization, and acculturation to impose their language
1982; Gorenstein and Pollard 1983; Pollard 1987). In this
over a presumptive Teco rea from Jacona-Zamora
regin local populations began assimilating to Tarascan
down the valley of the Duero and into the Teco state
identity only after their conquest by heartland Tarasof Coinan and the Coca state of Cuitzeo. During the
cans. Tarascan had become the dominant language at
first half of the sixteenth century Cuitzeo was prethe time of Spanish contact {Minas de Cobre [1553] in
dominantly Tarascan in speech, and Tarascan was

Ethnicity in the Tarascan state

liifiiii I '

II

>

11 >,

83

Fig. 7.4 Eighteenth-century Tarascan language distribution. Based upon West 1948.
spoken to a considerable extent in Coinan. However,
in the next one hundred years Coca and Teco reasserted themselves only to be supplanted by Spanish and
Mexicano. (Brans 1943:57)
Zone of ethnic segregation
The zone includes both ethnic enclaves located within
the zone of assimilation and those large territories along
the military frontiers within which were a variety of
ethnic groups. It is possible that there may have been
ethnically distinct populations within the heartland
which held specialized occupational roles. This kind of
segregation was known in other parts of Mesoamerica,
although the documentary sources do not specify this for
the Tarascan domain.

Rather, ethnic enclaves within the zone of assimilation

occupied distinct communities, primarily as political
refugees. One group of such communities was occupied
by Matlatzincas escaping Aztec domination of the
Toluca Basin. They were settled in the Charo-Undameo
zone and were referred to as pirindas {Relacin de
Michoacn{\5AV\; Relaciones Geogrficas [1579-81]
1987). This is a Tarascan term meaning "those in the
middle," and referring to the physical placing of these
settlements within the Tarascan heartland. These communities may have been established on patrimonial land
belonging to the king or on land that was previously
unsettled (as per A G N Hospital de Jess, 1635, leg. 29,
vol. 51, exp. 34 f. 55). According to Matlatzinca sources,
the fleeing groups included members of the Matlatzinca

84

Helen Perlstein Pollard

Fig. 7.5 Tarascan ethniclpolitical structure.

nobility (settle at Charo), others of the lower nobility
(settled at Undameo) and commoners (Quezada
Ramrez 1972:51). Matlatzinca communities were also
established in the tropical zone at Huetamo {Relaciones
Geogrficas Cuseo [1579-81] 1987) and other Otom
groups fleeing Aztec rule were settled in the Balsas
zone, perhaps near Huetamo {Relaciones Geogrficas
Necotlan and Taimeo [1579-81] 1987).
Documentary sources suggest that in the case of
Matlatzinca in the Charo-Undameo zone, the communities were administered as a group, and headed by Charo
(Quezada Ramez 1972:43; Warren 1985:283-5). As the
location of the highest ranked Matlatzinca nobility, the
choice of Charo as administrative center imples little
Tarascan "medding" within the Matlatzinca society, and
a retention of Matlatzinca defined status and authority.

Along the military frontiers ethnic enclaves often provided their tribute in the form of specialized military
service {Relacin de Michoacn [1541] 1956:14, 248;
Relaciones Geogrficas [1579-81] 1987; Brand 1943:54;
Carrasco 1969:219). Many of the frontier centers and
fortified towns were multi-ethnic, and non-Tarascan in
origin, including as many as four different ethnic groups
in addition to small Tarascan communities sent to
colonize the center. Taximaroa is first referred to as an
Otom village {Relacin de Michoacn [1541] 1956:154),
and Tuzantla is known to have been Ocumo n Tarascan
(ucum), which means Otom (Warren, pers. comm.
1989, based upon the sixteenth-century Tulane dictionary, pt. I I , f 126v). These communities were administered separately by ethnic group. Thus the Tarascan
governor sent out to Acambaro was n charge of the

Ethnicity in the Tarascan state

Tarascan community only (Relaciones Geogrficas
Celaya [1579-81] 1987). Local lords of each ethnic group
were selected, with the approval of the Tarascan king, to
administer their own communities. When they fought in
Tarascan military campaigns they remained within their
own military units, although serving under Tarascan
leaders. These non-Tarascan ethnic groups spanned the
political borders of the state. They were able to serve as
buffers along active military lines and as cultural brokers
between Tarascan and non-Tarascan communities.
Political implications
Within the heartland local leaders articulated with the
central administration directly (Caravajal Visitation
[1523-4] in Warren 1985:Appendix A; Relacin de
Michoacn [1541] 1956; Gilberti [1559] 1975; Lagunas
[1574] 1983). This regin appears to have been under the
direct control of the political capital. All local leaders
were approved by the Tzintzuntzan lite, and they could
be replaced, and their decisions over-ruled by the king.
Their loyalty was assumed, and intervention in local
affairs was eonsidered unusual (Relacin de Michoacn
[1541] 1956:201-2). Such loyalty to the Tarascan royal
dynasty was repeatedly demonstrated during the early
colonial period (Warren 1985).
Within the second zone, that of active assimilation, the
loyalty of these peoples to the core lite would have been
best assured by a policy of gradual assimilation into the
Tarascan social system and Tarascan ethnic identity.
One mechanism of assimilation may have been the
administrative centers of the frontier provinces. There
were four such centers, each administered by a highranking member of the Tarascan lite. They served as
political centers for those settlements outside the direct
administrative control of Tzintzuntzan. The exact
location of only one is known, Jacona (also spelled as
Xacona) (Relacin de Michoacn [1541] 1956:195),
although I have placed on the map the settlements most
often believed to have been the other three administrative centers. The precise location of the centers is less
significant than the indication that all were located at the
limits of the ethnic heartland, on the earlier political
frontier, and helped administer the zone of active Tarascan assimilation. The local presence of central Tarascan
administration must have fostered cultural interchange,
and perhaps also, cultural assimilation. If these administrative centers were also locations of markets, such interaction would have been still greater.
A second mechanism of cultural assimilation in this
zone was the extensin of the basic pattern of centralized

85

political authority. Despite the existence of regional

administrative centers, the Relacin de Michoacn indicates that local leaders in the zone of assimilation were
directly appointed by the king, as in the heartland. Direct
control of local decisin making is corroborated in
several colonial testimonies, as in the sending of a judge
by the king to settle disputes within the community of
Tetlaman, near Tepalcatepec (Carrasco 1969:219: see
also Warren 1968, Relaciones Geogrficas [1579-81]
1987). This direct penetration of local authority in
regions far from the Ptzcuaro Basin is understandable
as both (1) the extensin of a pattern historically successful in the political core and adopted during the formative
phase of Tarascan political unification, and (2) the
primary mechanism of cultural assimilation. Local lites
gained prestige and power by identifying with the Tarascan social structure. Local village leaders, angamecha in
Tarascan (caciques or seores in colonial documents),
were people who wore lip plugs given them by the king
(Lagunas [1574] 1983:22, 221), and who were thus elevated to noble status. By controlling access to local
political offices the core lite also controlled the local
elite's access to power and prestige, now definable on
Tarascan terms.
This flow of authority from the center to the village
was suppprted by the fundamental system of land and
resource ownership. All land titles within the Tarascan
domain were justified by having come from the king.
This even included agricultural lands, fishing rights,
mineral resources and hunting territories (Beltrn 1982;
Carrasco 1986; Pollard 1987). By extending the Tarascan
political ideology throughout this zone, non-Tarascanbased access to resources and social status became
illegitimate and over time would have become irrelevant.
A third policy fostering assimilation was the resettlement of Tarascans from the heartland into this zone.
Such resettiements are known to have occurred from
Jacona to Jiquilpan, to Turecuato, and to Periban (Relaciones Geogrficas [1579-81] 1987, Stanislawski 1947).
Smaller resettiements are known to have been made
within the heartland and along the military borders,
apparently to enlarge local populations in the face of
active military threat (Stanislawski 1947).
Given the economic and political bases for the
pressures of assimilation within this zone, there must
have been great variability in the rate of cultural/linguistic assimilation. Zones of particular economic or political interest to the core would have had the greatest
interaction with Tarascans, and may have been targeted
for absorption. Thus, portions of the zone of assimilation near Jacona in the west, and Zinapecuaro in the

86

Helen Perlstein Pollard

east, two strategic itiilitary borders, had become heartland regions by 1520. Within the zone of active assimilation, the retention of Tarascan as the primary native
language in the southeastern portion of Michoacn in
1750 attests to the relatively high degree of assimilation
reached in this important metal-producing regin
(Pollard 1987).
Within the zone of ethnic segregation, along the military borders, loyalty to the Tarascan lite was assured in
exchange for the security provided by the state military
structure. Administrators were sent from the political
core to articlate with the local population and insure
this loyalty. Nevertheless the populations were regarded
as subject allies, rather than subjects, and tribute
included war captives and slaves (Relaciones Geogrficas
[1579-81] 1987). The inclusin of zones of segregation
within the state was highly desirable; the risk to the state
of losing their ultmate loyalty was balanced by the
benefits to the state of using these populations for military support, sacrificial victims, and economic brokers
with neighboring societies. Carlos Herrejn Peredo
(1978) has detailed the valuable role played by Matlatzinca and Otom communities, both along the eastern
border and at enclaves at Charo-Undameo, in the Tarascan repulsin of the 1476-7 Aztec military campaign.
Such support brought harsh Aztec reprisals in the
Toluca valley itself, and resulted in additional refugees
fleeing to Tarascan territory.
With the demise of the autonomous Tarascan
kingdom in the early sixteenth century, these pooriy
articulated zones separated. The rather "thin veneer" of
Tarascan identity in the zone of active assimilation was
easily eroded as historical and ecological differences
took precedence. The cohesive unit remained the fundamental cultural and economic unit, the ethnic heartland.
Archaeological implications
Tarascan archaeology has been dominated by the often
fruitless search for a cultural identity synonymous with
the protohistoric state, as revealed by artifacts and structures. The model of ethnicity proposed here predicts that
there will be no commonality of artifacts and architecture over the territory of the Tarascan state, both
because material culture often varies in ways unrelated
to political control, and because different communities
and regions articulated with the state in greatly differing
ways.
The Tarascan heartland should be characterized by a
common lite culture, represented by similar artifacts
and architecture defining lite status. Common religious

beliefs should be visible in design motifs, ceremonial

architecture and individual artifacts. On the basis of the
archaeological manifestations of lite culture at the Tarascan capital of Tzintzuntzan, and those mentioned or
illustrated in the Relacin de Michoacn, it is possible to
specify the following items: (1) ear and lip plugs
(especially of green obsidian, often inlayed with turquoise or other imported material); (2) cotton garments
(especially those embroidered with feathers, and with
copper bells); (3) Tarascan polychrome pottery (often
with negative painting) in a variety of unusual forms
(spouted vessels, loop- and basket-handied vessels,
miniature bowls, highly polished plates) with painted
motifs of double spirals, hatching, checkerboards, lines
of dots and bands on vessels slipped in red, cream, white,
grey, and pink; (4) Tarascan ceramic pipe forms; and (5)
state architecture associated with the Tarascan patrn
deity Curicaueri, better known as the key-hole-shaped
pyramids or ycaias. These characteristics are documented together in the forty-one burials which have been
excavated adjacent to the ycatas at Tzintzuntzan and in
surveys of Tzintzuntzan, Ihuatzio, Zacapu, and other
ethnohistorically known Tarascan centers (Rubn de la
Borbolla 1939, 1941; Moedano 1941; Gali 1946; Pollard
1977; Castro-Leal 1986; Cabrera Castro 1987; Macias
Goyta and Cuevas Garca 1988). Other archaeologically
detectable characteristics may well prove valuable in the
future. For example, the Relacin de Michoacn depicts
clearly different house forms for the lite, larger, with
dressed stone foundations and painted wooden supports. The single resdence excavated at Tzintzuntzan
(and not adjacent to the pyramids) may represent such a
house (Cabrera Castro 1987). On a wider level the
common ecological adaptation of these communities
and the shared economic networks should be reflected n
common sources for localized resources, such as obsidian, lime, and salt. On the other hand, the wide distribution of ceramic clays and the lack of (known) mass
production of planwares and non-elite ceramies indcales that while manufacturing technques were widely
shared, specific ceramic types were not.
Within the zone of active Tarascan assimilation, the
overriding factor is variation. A l Huandacareo, a
regional administrative center on the northern shore of
Lake Cuitzeo, founded after the Tarascan conquest of
this zone, burial reas suggest the nature of interaction
between Tarascan lite, local lite, and local commoners
(Macas Goyta 1989). Within the public center there are
lite burials in formal tombs with grave goods, especially
ceramies, that are identical to lite goods from Tzintzuntzan. There is a second zone of mass burials of

Ethnicity in the Tarascan state

sacrificial victims with Tarascan ritual goods. A last zone
of shallow pit graves with predominantly local undecorated ceramies is believed by the excavator, Macas
Goyta, to represent local non-Tarascans buried separately, but within the state-admnistered ceremonial
center. While it is impossible to tell if the lite burials
represent core Tarascans staffing the center or local lite
assimilating to Tarascan ethnicity, the absence of a distinct burial zone for non-Tarascan lite suggests the
latter.
The existence of Tarascan lite polychrome ceramies
has been documented at numerous settlements
throughout this zone, sometimes associated with burials
of individuis with ear and lip plugs adjacent to possible
ycatas (Noguera 1944; Moedano 1946; Gonzlez
Crespo 1979; Maldonado 1980). Smlarly there may
have been a sharing of religious symbols, although these
may simply be on the level of pan-Mesoamerican philosophy and beliefs. The ecological adaptations within
this zone, being distinct from those of the heartland,
suggest that basic differences in artifact inventory, settlement patterns, and economic patterns may have perssted despite ethnic/linguistic assimilation. Only the
increasing integration within Tarascan marketing networks may reveal the assimilation process, unless the
regin was targeted for Tarascan exploitation. Thus,
within the zone of copper production, groups of ethnic
Tarascans were moved south to work in the mines of the
Balsas, and smelters were sent to the major smelting
centers (Pollard 1987). Here should be found clear evidence of ethnic Tarascan artifacts, architecture, and
technology (especially that directly related to mining and
smelting). The critical factor in this zone is the lack of
significant time depth to the Tarascan association.
However, the absence of outside administrators or a
local lite within a community may mean the absence of
any indications of lite culture and architecture, most
easily recognizable indications of the protohistoric Tarascans. Thus even the political changes associated with
Tarascan expansin will probably be difficult to discern
in material culture.
Within the zones of ethnic segregation, the archaeologist fares both better and worse. Here local or
regional traditions would have persisted. This persistence has been archaeologically documented along the
northeastem border in the form of continuity of ceramic
types and house forms (particularly storage bins), and
ceremonial architecture (Gorenstein 1985; Contreras
Ramrez 1987). It is often difficult to discern when the
zone of segregation has ended and the political border
has been breached. On the positive side, there may be

87

non-local, heartland Tarascan artifacts and architecture

associated with small numbers of lite stationed along
the frontier. The presence of military fortifications in this
zone is the best artifactual evidence of this ethnically
di verse zone (Gorenstein 1985), as it was a policy to
associate military frontiers with non-Tarascans. The
lack of articulaton of this zone with Tarascan local
marketing networks, and the dominance of pre-existing
networks may well mean that goods circulatng within
this zone in fact derive from sources not utilized n the
rest of the territory.
One of the ethnic enclaves located within the Tarascan
core has been solated solely through archaeological
analysis. At the Tarascan capital, Tzintzuntzan, one
locality of the city contained relatively large quantities of
a clearly non-local ceramic ware. Research along the
northeast border indicated this ware was the dominant
form there during the protohistoric, probably of an
Otom-speaking group (Pollard 1977; Gorenstein 1985).
The presence of this border population in the Tarascan
capital suggests that future archaeological research will
reveal a more complex pattern of ethnic relations than
that suggested by this model.
Ethnicity and the expanding state
The model of ethnic structure which has been proposed
was the outcome of the twin Tarascan policies of ethnic
assimilation and segregation. Within the central regin,
the protohistoric period population was characterized
by a social system with a fully Tarascan identity. Along
the frontiers, the state encompassed a multtude of distinct ethnic groups, with varying loyalty to the ruling
dynasty. Between these two zones was a large territory in
the process of becoming Tarascan. Its existence reflects
the dynamic nature of the protohistoric period in
western Mxico, and the rapid rate of cultural change.
As a result, the borders of these zones can only be
approximated, as they must have changed repeatedly.
This ethnic strategy differed from that of the Tarascans' contemporary neighbors, the Aztecs, among
whom ethnic diversity was woven into the basic
economic and political fabric of society (Carrasco
1971b; Zantwijk 1973; Hassig 1984). Robert McC.
Adams (1979:64) has suggested that an important reason
for the absence of policies of cultural assimilation may
have been the overwhelming demographic dominance of
the core population over its peripheries. There is no
Mexica ethnic heartland, but a zone of culturally similar
populations sharing the Basin of Mxico, often sharing
individual communities, for their mutual (if not equal)

88

Helen Perlstein Pollard

benefit. Traditional ethnic markers included language,

dress and adornment, marriage patterns, deities and
ritual, and even occupational specializations. Here
ethnic boundaries were maintained, even cultvated, by
the Mexica lite, although the markers and significance
of the boundaries surely changed as Mexica power
increased. Beyond the political core of the state, Mexica
policy was largely one of indifference to ethnic dstnction, with the result that conquered communities,
regions, and states retained previous ethnic boundaries
unaffected by Mexica administrative units or economic
networks. In such circumstances the possibility of rebellion was countered by the ability of the core lite to field
large armies of internal re-conquest. In the Tarascan
case such demographic disparities were clearly influential in the early conquest of the Balsas and Tepalcatepec
Basins. But the Cuitzeo Basin population, for example,
could have easily raised armies to challenge Tzintzuntzan-based control effectively, had the regin ever
been politically united. Moreover the Aztec threat on the
eastern Tarascan border, along with a militarily active
western and northern border, spread heartland Tarascan
warriors across a wide landscape, and limited their
ability to field large armies wiihin their territory at the
same time.
While Tarascan ethnic strategy may have been distinctive in sixteenth-century Mesoamerica, it exhibits
many characteristics common in the structure of ethnicity of complex societies. The policy of assimilation,
and its implementation by control over positions of
power and prestige, is a fundamental component of the
military strategies of many expanding states, including
the Inka (Morris 1982; Rowe 1982; Wachtel 1982), the
Romn (Salmn 1982), and the Assyrian (Larsen 1979),
among others. One of the best-documented cases is that
of the Romanization of central Italy (Salmn 1982). The
spread of Romn ethnicity took more than 300 years to
be achieved in central Italy, and was a product of specific
policies of the expanding Romn state. A consistent
feature of this process was the direct administration of
peoples by Romn law and the severing of communities
from their ethnic homelands.
The high degree of political centralization found in
such polities may be reinforced by an lite perception of
the role of ethnicity as an integrating mechanism. In
those cases, like the Tarascan, where the early expansin
of the state involved the incorporation of societies less
politically complex than itself, direct rule from the
center was established. It is quite possible that cultural
assimilation was at first an unintended consequence of
centralized administration, only later recognized for the

benefits it provided the state. By promoting assimilation

in the larger Tarascan territory, the lite of the geopolitical core could better insure the economic exploitation of this zone, uninterrupted by rebellions and retaliatory campaigns. This exploitation was increasingly
necessary to feed the growing population of the Ptzcuaro Basin and maintain the conspicuous consumption
of the state (Pollard 1982, 1987). From this perspective,
the ethnic strategy, whether or not recognized by the
central dynasty as such, was an adaptive mechanism,
increasing the stability of the Lake Ptzcuaro Basin
ecosystem.
Ethnic segregation along military borders is also
common, particularly the use of such populations in a
military capacity. Enloe (1980:25-9) has termed this the
"Ghurka Syndrome," characterized by the maintenance of ethnic boundaries in regions which are usually
mountainous, along historie invasin routes, and geographically distinct. In the Tarascan case such groups
served not only military functions, but also, I believe,
economic ones as well. Tarascan long-distance merchants traveled to the borders of the state, including
Taximaroa, to acquire elite-related scarce materials
produced in other regions of Mesoamerica (Pollard
1982). Some of these exchanges may have been directly
with Aztec merchants, but many were surely with
Otom or Matlatzinca traders operating within their
traditional territory, now spanning a major political
frontier. The existence of such trade s suggested by the
appearance of Michoacn maize and chili n the Tlatelolco market, as recorded by Sahagn [1569] 1950-69,
(Bk. 10, pp. 66-7).
One pattern of expanding complex societies invoives
rapid territorial increase followed by periods of consoldation and absorption of the conquered territory. The
political unit soon after conquest is multi-ethnic and
heterogeneous and is often termed an empire (Larsen
1979). The assimilation that follows expands the center,
absorbing increasing proportions of the periphery, until
the resultant polity is best viewed as a territorial state.
This appears to be the pattern followed in the evolution
of the Tarascan state. In AD 1520 the Tarascan state was
actively in the process of this internal assimilation,
reflecting the strong geopolitical and economic motivations of expansin and the high degree of
centralization of administrative functions which made
such an ethnic strategy both desirable and practicable.
Modern Tarascans who cling to their ethnic heritage,
albeit in greatly modified form, are only the latest of a
people who have long recognized the political consequences of ethnicity within complex societies.

other than as Malthusian competition for material

resources - i.e., biological reductionism. Darwinian-like
competition removed the dysfunctional social parts
while the evolutionary clock ticked uniformly upward in
the spiral of cultural evolution. According to such linearity, environmental resources as exploited by technology determined amounts of food, which in turn determined community size, whose economy shaped political
organization and ideology (de Montmollin 1989:8-9).
In contrast, the present volume examines a political
dialectic, arguing that ideology and social organization
often change as a result of political action. Rather than
being prime movers, technology, land, and labor are
JOHNW. FOX
simply factors of production manipulated to promote
factional interests. While native American technology
(e.g., obsidian tools, raised-field agriculture) was essentially unchanged from the Formative, its use by social
labor differed within different political contexts. As
In aboriginal American societies, factions competed for Clark and Blake (Chapter 2) arge, technoeconomic
power, prestige, authority, and material benefits. A development and demographic change are consequences
counterpositioning of structurally similar corporate of factional competition. People are a major production
groups, often in pairs, generated factional competition. factor, and political strategies involve manipulating
This coUection provides case studies on factional com- social relations to increase benefits.
The Quiche example (Chapter 14) shows that political
petition in a variety of environmental settings and on the
methods for disceming factionalism across a gamut of strife can explain transformed social relations where
environmental variables fail. Urbanism resulted from
social fields.
intrusive lineages who nucleated for protection, rather
than to exploit ecological symbiosis or arid soils with
Theoretical approaches to conflict and change
irrigation, or exchange valuable natural resources.
A steady theoretical stream has attempted to explain External pressure and intemal factional manipulation
exogenous cause and culture effect (adaptation) in the gave form and direction to the Quiche "coalition."
evolution of simple to more complex societies. In the
Because factionaHsm pervades all groups, from housepositivist tradition, with Newton drawing analogy to the holds to social classes, we focus upon incompatibilities,
clock, a principal machine of his day, eighteenth- and tensions, and contradictions within and between social
nineteenth-century social theoreticians sought the forces clusters. Competition and conflict occur at social boundthat set the three-fold (savagery, barbarism, civilization) aries, for instance between lineages within clans or moie"evolutionary clock" in motion (e.g. climate for Montes- ties. In this view, states comprise pyramidally nested
quieu). During the 1920s to 1950s, a "dynamist social segments cross-cut by innumerable factions. The
approach" focused on the basic "tensions inherent in double-edged phenomena of intemal personal goals and
any society" (Balandier 1970:17-18), but conflict was external exigencies shape factions and their behavior.
said to improve social functioning. Newton's smoothly
Overall, two directions of dialectical transformation
running clock was transposed synchronically; groups are evident. First, competitive strife among social equivstrove teleologically to maintain the well-oiled social alents in a kinship mode of production (Wolf 1982)
machine for the greater social good.
results in lateral social movement and simple quantitaFor cultural ecologists, environmental stress was the tive change. For example, segmental kinship groups
catalyst to evolution. In essence, Newton's smoothly bonded or broke apart as seprate building blocks with
functioning system was recast as trophic exchanges and options to relocate. Second, struggle among peoples of
competition. But "systems models cannot explain contrasting social status (classes) resulted in qualitative
chronic problems generated by the very operation of the changes of organic solidarity (e.g., Zhao 1986, 1989).
system as constituted, such as civil wars, succession For example, underclasses assumed new tasks in prodisputes or tax evasin" (Gailey and Patterson 1987:5) duction and eventually formed new ethnic identities in a

17
Conclusions: moietal
opposition, segmentation,
and factionalism in New
World political arenas

199

200

John W. Fox

tributary mode of production. Factions occurred within

the same social strata. For example, families of lite
Aztec or Inka vied for positions of authority within
the capital (Zantwijk, chapter 9; D'Altroy, chapter 15).
Factions also recruited across social strata. Patronclientage groupings occurred in provincial communities
allying Aztec or Quiche patrons with local constituents
(Brumfiel, chapter 8; Hicks, chapter 10; Fox, chapter
14).
Endogenous processes within social arenas resulted in
discontinuous and often non-progressive change. For
example, rebellion was a recurrent process for segmentary states in the New World and Africa (Gluckman
1956). Implicit uniform change gives way to the irregularities of punctated evolution and even devolution (Fox
1987:259). For example, the Postclassic Mayan states
were far smaller than their Classic counterparts. But like
quantum mechanics, with a multitude of seemingly
random interactions of particles and waves in specified
fields with uncertain outcomes, factional competition
results in a myriad of possible evolutionary outcomes,
each predictable to only a minor degree.
The propelling forc for change is competing groups,
rationally maximizing "utility," whether as direct profits
or as increased prestige which can then be parlayed into
economic advantage (Schneider 1974, Cohn 1976:22).
Symbolic systems of thought both channel political
actions and are transformed by political action. Even
though rules may stipulate access to authority and
strategic resources, brokers incessantly circumvent the
established rules through actions of personal power and
by so doing crate new social formations and new rules.
Specific political actions transform corporate identities
and mythic charters. As a consequence, there is constant
discord between the somewhat necessarily ambiguous
systems of ideas (the ideal model) and the perpetual real
political contestations within and between corporate
bodies and factions.

Social charters and symbols.

People uniting to advance special interests require insignia and distinctive practices to distinguish themselves
from their rivals and to suggest a shared heritage. Myth
furnishes a charter and explains inequalities in power.
For example, Mesoamerican myth paired brothers who
allied against common foes (e.g., Junajpu and Xbalanque of the Maya) or who contested one another for the
spoils of conquest (e.g., Quetzalcoatl and Tezcatlipoca).
Such bipolarity bespeaks day-to-day factionalism, ideologically linked to the calendar. Acquiring the identity of

the sun and the concomitant right to maniplate relations within the body politic pervaded Mesoamerican
mythology.
The Classic-Postclassic transition saw peoples from
collapsed polities migrate and coalesce with indigenes to
form new ethnic bodies. The Codex Zouche-Nuttall
defines relations between new factions that emerged
from the collapse of Classic polities in the Mixteca Alta
(Byland and Pohl, chapter 10). In central Mxico, the
feud between the peaceful Quetzalcoatl and the warlike
Tezcatlipoca reflects deeply seated schism, followed by
political collapse and refigured polities in new locahties.
Of dual authority, one figure (Quetzalcoatl) furnished
authority based on genealogy; the other (Tezcatlipoca)
embodied the authority of warriors (of indigenous
descent?). Zantwijk (chapter 9) arges that the Aztecs
maintained bipartite factionalism with the warrior Huitzilopochtli pitted against those under Tlaloc who persisted in the "Toltec" lifeways. Clearly, claims to a
Classic anden rgime were a tactic for strengthening
positions to maniplate subjects and their production.
The Toltec connection, whether mythic or legendary,
became "grounds for material self-interest" (see Worsley
1984:240).
For the Postclassic Maya, ethnohistory begins with
the appearance of Quetzalcoatl (Kukulcan), which provided a justification for inequality between invading
Putun and local Maya. In mythology and polities, this
Feathered Serpent transformed from Venus to the sun
(Nacxit) both in the Yucatn and in the Quichan highlands (Jakawitz/Gakawitz [Venus] to K'ucumatz [Sun]).
K'ucumatz was clearly a Quetzalcoatl reincarnate; both
relied upon airborne flight to vanquish adversarles, and
both revenged a slain father. Asserting supernatural
power in divine kingship allowed one lineage to further
its ambitions over rivals.
Political groups rechartered ethnic and mythological
identities as power was acquired or lost. While polities
discussed in this volume endured four to thirty generations, ethnic groups continually merged or were absorbed and redefined. Pollard (chapter 7) for the Tarascans,
Brumfiel (chapter 8) for the Otom, and Byland and Pohl
(chapter 10) for the Mixtee and Zapotee demnstrate the
fluidty of ethnicity. After subjugation by the Aztecs, the
Otom wore distinctive labrets less and opted for Aztec
adornment. When marrying. Mixtee and Zapotee
nuptials would wear half the face paint of the other
group. Zantwijk (chapter 9) explains that among the
Aztecs corporate membership and ethnic identity could
be clamed through either paternal or maternal lnks,
depending upon the potential advantages of each. Cer-

Conclusions

201

tainly ethnic malleabiHty allowed access to strategic (chapter 3), Anderson (chapter 6), and Kowaiewski
(chapter 11) discuss how factional competition incorpopositions within unitary states.
Within the segmentary Quiche state, conquered rales specific lopography to gain access to resources.
peoples were recruited to become ethnic Quiche (e.g., the Within siles and regions, the cenlrality and size of buildNijaib). Outsiders (Cotuja) were also incorporated ings is often proportionate lo the power exercised. For
according to the symbolism of the calendar. Alignment example, as a Quiche lineage increased its power, it
to celestial bodies justified degrees of privilege from increased its cenlrality of location.
From settlement pattern alone, a basic bipolarity
inter-lineage maneuvers.
The ethnohistories codified social status in "tribal" (moiety structure?) is discerned at many Mesoamerican
histories that commenced with migration from a pri- sites, across three millennia. Clearly, dichotomous facmordial homeland guided by a patrn deity. Principal tionalism underpinned transformative movement in
enemies of allied subgroups were listed and differential Mesoamerican civilization. Thus, Preclassic Olmec La
access to power was sanctioned by those who became Venta reveis two plazas north of a central circular
pyramid (Heizer, Graham, and Napton 1968), perhaps
guardians of tradition (Balandier 1970:34).
representing kinship groups that once vied for control.
Similarly, at Cuicuilco, a wide circular temple bifurcates
Factionalism in archaeology and ethnohistory
two circular residential clusters, possible descent groups
The ethnography of factionalism discussed by Brumfiel (Marquina 1964).
(chapter 1) stresses the informal linkages between power
During the Classic period, the ritual centers of Uxmal
brokers and their clients to gain advantages over equi- and Copan were bipartitioned into a lower northern
poised "clientage" groups, that is, competition in plaza and a southern elevated lite complex through a
zero-sum settings where leader/followers ally to enhance low-lying open-ended ballcourt. The ball game
mutual self-interests. However, the basis for identifying expressed competition and alliance between antagonists/
the ever-fluctuating factional bond in archaeology is protagonists who were grouped as moieties of the day
barely developed.
and of the night. The ethnohistoric record suggests that
Fortunately, ethnohistory adds events of "official the elevation differences refer to political "niches"
strife" back to the AD900s generally, and in Maya achieved by each "moiety"; the higher groups possessed
hieroglyphics to A D 200. Accordingly, groups and ritual power and the lower groups military might (see
leaders are named, and social exchange (marriages, wars, Balandier 1970:81).
alliances, patron-client bonding, commercial transSimilarly, pairs of temples at Teotihuacan, Tula, and
actions, and tribute payments) gains illumination. Tenochtitlan reflected binary groups in highland Mxico
Therefore, ethnohistory may "bridge" conflict theory (see Zantwijk, chapter 9), seemingly as dual authority
and inert archaeological/environmental residue. And between conquerors and indigenes (see Fox 1981:331-4).
ethnographic analogy provides a lens as to how factions Through the millennia, then, as centers experienced
genrate tensions.
administrative breakdowns, and the meaning of
Writing and power correlate; nonstates lack formal authority was reformulated, the size and cenlrality of the
history. In segmentary states, like those of the Maya, the temples and ballcourts steadily transformed, reflecting
almost exclusive proprietors of historical knowledge new allempls at channeling factional competition.
were the one or two highest ranked lineages. Therefore,
The remainder of this chapter focuses upon the social
ethnohistory was the prerogative of power brokers transformalions wrought by factional competition,
whose written discourse concemed dealings with other framed spatially. It follows the volume formal, examinehtes in coups, conquests, and rituals of alHance. Rarely ing factional struggle with various outcomes, contingent
are clients in subordnate positions named. Also omitted upon the size, complexity, and specific interactions: (1)
are defeats, and endemic intrigue, and the usurping of allegiance to war leaders, then linkage within mltiple
power and authority. The customary behind-the-scenes village networks; (2) at times cryslallizing into hereditary
intrigue and manipulation by power brokers and their ranking; (3) accompanied by pervasive inter- and intraconstituents usually escaped transcription.
chiefdom conflict; (4) with segments that fissioned to
But factionalism is inherently spatial. The geometry of new territories (e.g., Mississippian peoples); and (5)
power relations is borne out in arrangements of organizational shifts to ranking in segmentary (Maya)
monumental architecture and in proximity to key states and stratification in unitary states (e.g., the Aztecs
natural resources. Clark and Blake (chapter 2), Spencer and Inka). While coverage spans Mississippian peoples

202

John W. Fox

from the north and Andean peoples to the south,

Mesoamerica emerges as a geographical and methodological focus.'
Transactional analysis - networl(ing, warfare, and
hereditary leadership
The first section of this volume (chapters 2-6) traces
transformalions of egalilarian and incipiently ranked
groups. Competition among equipoised groups with
undifferentiated kinship-based production units saw
war leaders formulating patron-client linkages. Threals
from opposing villages necessitated group mobilization,
and strategic action by individual self-aggrandizers or
big men established personal constituencies. Village
war leaders attempted to reproduce leadership among
their offspring - to pass on privileges they had achieved
- usually without success beyond one or two generations.
For the Circum-Caribbean, where raiding was
endemic, Clark and Blake (chapter 2) and Redmond
(chapter 4) propose how local war leaders became
regional chieftains. They apply transactional analysis to
the archaeological record, construing personal interaction as contests for enhancing power through struggle,
scheming, and temporary alliance. Arenas comprised
face-to-face interactions within villages and regional networks. Leadership was based upon the prestige of personal achievement in fighting and leading raids and the
distribution of booty from raids, compelling clients'
loyalty. Thus, bellicose external pressures preceded
social/material exchange that cemented clients to
patrons in mutual self-interest. Attempts were made to
perpetate leadership to offspring by transferring booty
from raids, such as trophy heads, women, weapons, and
ornaments to coming-of-age sons to enable them to
recruit a following. However, prestige was usually dissipated with loss-of-face in unsuccessful raids.
As contact with outsiders increased, the big-man
figured as a broker between his own village and other
villages. On the regional level, a village big-man offered
military assistance, prestige goods, and marriage partners to his counterparts within region-wide networks.
The big-man created relations of reciprocity in an everfluctuating web of cross-linkages. A big-man operated
on a regional level beyond the leveling sanctions of his
own village, where more ambiguous norms were easier
to maniplate. Applying this model, Clark and Blake
(chapter 2) chart the spread of maize cultivation and
ceramies along region-to-region networks during the
rapid transition (c. 2000-1500 BC) from late Archaic

egalitarian villages to Formative chiefdoms in coastal

Mesoamerica.
In more complex fields of interaction, the chiefdom
ossified a mltiple village network into a single polity. In
this step-by-step transformation, Redmond (chapter 4)
theorizes that military aggression intensified into a yearround activity for acquiring land, prestige goods, and
war captives. External threats reinforced status, and
authority crystallized with special privileges to estates,
the labor of kin and non-kin, and luxury goods from
afar. The gift-giving of the village big-man became redistribution where different kin units transferred produce
to the chief and his retainers.
That the precocious Olmec chiefdom took form
through raiding is supported by purposely decapitated
statues at the earliest Olmec center, San Lorenzo. There
(1) raiding insurgents may have severed the stone heads
of ancestors or deities apparently as trophies; or (2) the
local population released supernatural power through
ritualistic severing of heads (Grove 1981), recalling the
political decapitation of Junajpu when the Ajaw Quiche
lost power to the Cawek (chapter 14). In any event, the
enormous carved heads, presumably of chiefs, personied institutionalized, dynastic power.
Helms (chapter 8) reconstructs from ethnohistory how
links between chiefs and clients were forged in the Panamanian isthmus at European contact. There, competing
paramount chiefs (quevis) were supported by their noble
relatives (sacos), chiefs subordinated through military
forc, and honored warriors of commoner status
(cabras). Chiefs controlled trade routes to acquire for
distribution rare, mystical, or awe-inspiring objects from
great distances, and further obliged loyalty by allowing
supporters access to favored hunting and fishing territories. In summary, competition for high position verticalized relations, with the chief extracting the labor of
kinship groups and distributing precious items and perquisites to his supporters.
In a parallel case, Anderson (chapter 6) shows that
early Mississippian chiefs maintained their positions
with scarce goods acquired from networks crisscrossing
the entire Eastern Woodlands of North America. The
French in the eighteenth century chronicled a social rank
rotation between nobles and commoners. Newly elevated nobles supported their positions by redistributing
luxury items. However, as competition intensified, and
clients were permanently subordinated, rival factions
fissioned to new localities. In short, Helms and Anderson arge that internecine competition necessitated
power symbols and lucrative perquisites to improve
position and not the reverse, as implied in the "trade"

literature. Spurred by intra-community competition,

social interchange preceded economic exchange, as
emphasized by Clark and Blake, and Redmond.
Tensions and transformations within segmentary and
unitary states
Cise kin of formal leaders embued with ascribed status
held together power cores in both segmentary and
unitary states. However, the core often bifurcated; two
factions asserted privilege premised on genealogy to distinguish themselves from out-groups (see Salisbury and
Silverman 1977:8-9). The Mayan, Tarascan, Aztec, and
Inka states graduated successively outwards from the
royal cores through the less powerful provinces. Specific
tactics for gamering power, known from ethnohistory,
established inequality based on seniority and descent
from the founding ancestor, on controlling productive
lands, and on exploiting the labor of subordinated
peoples channeled into tributary agricultural and craft
production and military service. However, qualitative
differences in size and continua of stratification warrant
a distinction between less complex segmentary states and
more complex unity states.
In segmentary states, rivals were more autonomous
economically and politically. Coalition building, gift
giving, and rituals of alliance, like feasts, figured as
prominent strategies for enhancing the standing of one's
faction. Although slightly smaller in size, provincial
seats often mirrored socially and architecturally the
single capital at the political, geographical, and cosmic
center of the state. At Utatlan, for example, the Cawek
were able to use cosmic symbolism to strengthen alliances with the thirteen calendric peoples of the colones.
The conquering lineages fissioned and settled among
outlying folk who supplied tribute, fidelity, and military
service. Kinship, whether real or fictive, was the basis for
rank and production. The intrusive lite furnished
correct pedigree and the autochthons, having intermarried with the intruders, eventually claimed the "noble"
genealogy of the founding Feathered Serpent. Paradoxically, egalitarian ideologies prevailed. Competition
among rival kin groups resulted in centrifugal segmentation and fission, rebellion and fusin (intermarriage).
Since the hierarchy produced little more than "cult"
Information for alliance building, even the forced linkages were prone to rupture.
The Maya, as analyzed by Pohl and Pohl (chapter 13),
afford a particularly long developmental history of segmentary political formations during two millennia. Certainly the Venus and sun masks on Late Preclassic

temples attest to power and alliance intertwined with

the calendars. During the ensuing Early Classic,
warfare intensified from raiding to pitched battles
between neighboring city-states only 40 km apart (e.g.,
Tikal/Uaxactun, Calakmul/El Mirador). Accordingly,
Calakmul was vacated during various episodes of the
Early Classic period, while the population of a temporarily defeated Tikal may have been forced out. By the
Late Classic, the tributary mode of production had
subverted the redistributive economy. Continuous competitive spirals resulted in the growth of the lite among
rival polities. Evidently, in each state "vertical" factions linked urban lites with rural supporters, who
shored up the hierarchy under stress. The peripheral
groups became empowered and "married up," thereby
also claiming within a generation or two prerogatives
based on genealogy. The Tulns confederated citystates and rotated authority, so that none rose to ultmate power.
As less enfranchised groups grappled for power and
privilege, thematic content on the stelae shifted from the
genealogies of god-kings to gladiatorial combat and
capture. Prestige focused on achieved warrior qualities
rather than on ascribed genealogy. And when neighbors
were seized, ancestral lineage symbols were desecrated,
recalling decapitation of ancestral Olmec statuary.
Newly constructed twin radical-pyramids at the edge of
Tikal bespeak realignment of newly emerged political
groups based on the calendar (Becker 1983). Agriculture
intensified, and emphasis on starchy staples increased
nutritional deficiencies; the economy degraded the
environment.
By the Terminal Classic, provincials seized power;
some 40 percent of the stelae were erected in small
outlying settlements. However, all succumbed to sudden
organizational failure and concomitant demographic
catastrophe. Within this power vacuum of the collapsed
peer polities system (Tainter 1988:202), only scattered
lineages migrated to the Yucatn and to the Guatemalan
highlands. The Maya demnstrate that unitary states do
not evolve from unremitting population growth and
heightened competition (cf. Sanders and Price 1968).
The Quiche were refugees of the Tulns who forged
triadic alliances, cemented by wife exchange (chapter
14). At first, Utatlan had few rivals, so the political field
was notably diminished and investment in hierarchy was
less necessary (Tainter 1988:117). Later, pressed into
tighter confederation, the Cawek lineage formalized its
ranking on a calendrical basis so that Utatlan became a
versin of a Tuln. However, Quiche lineages fissioned
to the provinces, further duplicating offices of the

204

John fV. Fox

capital. The colonial lineages formed factions with the

indigenous lite, recalling the peripheralization of power
during the Late Classic. Power diffused among so many
centers eventually fragmented the state. The outer rings
of the Quiche state came to constitute a seprate political
field, and much of the outer ring eventually rebelled.
Three fairly counterbalanced states became locked into
wars of attrition, where just two centuries earlier only
the Quiche held sway.
Both the Quiche and Tarascan states formed three
concentric circles of successively less powerful blocs ritually tied to a capital. Colonization resulted in "ecological symbiosis," and not the reverse (i.e., groups,
symbiotically bound together for economic exchange,
creating a unitary state). For the Tarascans (chapter 7),
centers like Jacana and Xacana in even the most remote
concentric ring served as secondary capitals for Tzintzuntzan. The local ethnic lite became "Tarascanized,"
through intermarriage with the colonists in these centers.
In contrast to segmentary states, power within unitary
states was centralized within single capitals, supported by
powerful judiciary and military institutions. For example,
the grandiose pyramids at Teotihuacan seem to reflect
coercin of labor for construction, rather than "negotiated" labor for semi-autonomous Mayan lineages. Teotihuacan was also an enormous metrpolis ten times
larger in population than Utatlan, where perhaps a third
of the regional population "forcibly" resided. Teotihuacan's single orientation 15'A degrees E of N reflects the
unin of a more "secular" military faction, associated
with the Temple of Quetzalcoatl, the palaces, and the Ciudadela, with a more "ecclesiastical" faction patronized by
Tlaloc at the gigantean Sun Pyramid (Milln 1988:10912). However, bespeaking factional bifurcation, Teotihuacan "celebrated the creation of the universe . . .
through a series of dual oppositions..." as evident in the
two massive pyramids (Milln 1988:112-13). Milln suggests that power passed from an older line of rulers under
Tlaloc to a newer group under Quetzalcoatl.
In short, mechanical solidarity of Mayan segmentary
organization qualitatively differed from the more
organic solidarity of Teotihuacan and its economic
interdependences. In the latter, lineage and ethnic identities were remolded within social classes, and economic
flow was largely centripetal to an enormous capital. Both
Teotihuacan and Tenochtitlan were easily several sizes
greater than their provincial centers, and had no serious
political rivals.
Nevertheless,
inherently antagonistic
interests
between the capital and the provinces resulted in transformation. For example, Sanders (1965) suggests that

the provincials at Tula, and perhaps the approximately

equidistant Cholula, Cacaxtia, and Xochicalco, may
have conspired with factions within the capital itself to
sack Teotihuacan, as inequality increased within urban
social classes (Milln 1988:149-58). The fiery cataclysm
was comparable to the destruction of the symbols of
genealogical authority at Olmec San Lorenzo and
Mayan Copan; only the ritual core of Teotihuacan was
dismantled. Similarly, the Totonacs, Tlaxcalans, and
Otom, as provincial "factions" within or bordering the
Aztec empire, readly allied with the Spanish under
Corts to overthrow Tenochtitlan.
The Aztecs llustrate the rapid transformation of segmentary lineage alliance by absorbing unitary state institutions long entrenched among their neighbors. The
early Aztecs resemble the Quiche in that they (1)
migrated from a thinly populated frontier (Aztlan) east,
with a stopover at Tuln; (2) were led by four "big-men"
(teomamas), who carried the icons of the patrn deities
for the lineage groups (calpulltm); (3) married into the
Toltec lineage at Culhuacan as mercenaries, who provided their first ruler (tlaloani), as the first Quiche ruler,
Cotuja, transferred from the direct Toltec line at Tuja;
(4) formed a triadic state (Triple Alliance).
Even after the Aztecs became a unitary state, segmentary ordering to the solar calendar persisted; with parallels to Utatlan (1) Tenochtitlan was quadripartitioned
and its three causeways led to its closest allies, as the
three stairways of the Cawek's temple aligned to the
three major lineages; (2) the twin temples housed the
tribal solar war god (HuitzilopochtU) and the indigenous
rain god (Tlaloc), and the circular Temple of Quetzalcoatl aligned them to the sunrise on the equinoxes
(Aveni 1980:246-9), comparable to the temple for the
Quiche tribal solar god (Tojil) opposite the indigenous
Nijaib god (Awilix), which from the circular temple of
the feathered K'ucumatz, aligned respectively to the
sunrise and sunset on the equinoxes.
The artifact distributions and kinship reckoning
delineated by Brumfiel (chapter 8) and Zantwijk
(chapter 9) reveal ethnic mutability in provincial towns
under Aztec rule (e.g., Otomi Xaltocan). Stratification
developed with Aztec scions forging clientage relationships with subordnate Otomi. Hicks (chapter 10) provides a step-by-step reconstruction of how the Aztec
established indirect rule by playing one ethnic faction off
against another, so that one group of provincials became
elevated through direct ties to Tenochtitlan. These chapters demnstrate how ethnic customs were reordered
with power relations, rather than persisting as enduring
traditions.

Conclusions

205

The ethnohistories were also reformulated. While the agricultural base for American civilization transferred
Aztecs, Otom, and Tarascans all claimed Chchmec from one regin to the next along factional linkages.
ancestry, Brotherston (1974) questions whether the Specific agricultural practices, ecological symbiosis, and
Aztecs fabricated aspects of their migration epic. Cer- population density are all social "effects" of factional
tainly, much of the early Aztec history was rewritten competition.
under Itzcoatl (1426-40) to legitimize factional control
This volume examines peoples divided along the lines
of state institutions.
of kinship, ethnicity, and class. The ranked societies of
D'Altroy (chapter 15) describes the intense com- Panam, Venezuela, Mississippian North American, and
petition among chiefdoms that prompted the meteoric Mayan Mesoamerica were transitional between kinrise of the Inkas to unitary statehood. The Inka came to ordered (segmental) and civil (stratified) society, with
rule twelve million subjects in a mere century, a clear both kinship and tributary modes of production. In all
example of "punctuated evolution." D'Altroy traces the of these fields, self-aggrandizers recruited followers to
emergence of ruling chinchecona who, like the tribal increase power, prestige, and material advantages.
leaders described by Clark and Blake, Spencer,
In overview, leadership took form from raiding,
Redmond, and Helms (chapters 2 5), secured their poli- feasting, and other services abetted by intra- and intertical prerogatives with the spoils of war. By Wanka I I village gift-giving among counterpoised villages. Leadertimes, endemic warfare forced bifurcated populations to ship crystallized in chiefdoms where the products of
nucleate into elevated defensible locations. There, kinship production were redistributed through the chief
households of power brokers with attached metalwork- to outlying support and production groups. Lastly, a
ing functionaries distributed sumptuary goods to sup- ruling class with the hegemonic mechanisms of the state
porters. Urbanism in turn intensified cultivation to appropriated social labor in a tributary mode. People
support more warriors and craftsmen.
emerge in these case studies as the principal factor of
In contrast to the Mesoamerican states, the Inka production. More citizens meant more food and craft
thwarted alliances and rebellion among foes by resettling producers, and more warriors. Urbanism was a political
them to the far corners of the realm and maintaining phenomenon; wealth, power, and people were concenethnic speech and dress to accentuate provincial status. trated for intensified manipulation. The Maya, Inkas,
Core administrators thereby averted the formation of and Aztecs were in the business of moving and realigning
vertical factions linking urban lites and rural producers, people to suit the needs of a core group. Nevertheless,
as in the Mayan and Aztec states. Unlike the Quiche, even in new territories the processes of factionoutlying garrisons functioned as loyal "puppets" of the formation soon aligned functionaries of the core group
state. Nonetheless, bipartite factionalism developed. The with their subjects in vertical patron-clientage grouplast aboriginal stand in 1527 pitted two enormous fac- ings. Upward mobility was achieved through the conduit
tions, led by Atawalpa and Wascar, in a bitter civil war of factionalism.
to advance one as the god king. This was factionalism
In general, since the core group must justify the
writ large, which doomed the Inka polity as it had the extraction of surplus labor or face open resistance, ideolbipartite Mayan and Mexican states.
ogy and mythology are subject to manipulation by
opposing factions. For example, the Maya assigned conquered peoples calendrical identities; therefore, asymConclusions
metries in relations of power, authority, and labor were
The authors in this coUection see factional competition said to be ordained by cosmic design.
In this view, the state amalgamated peoples with conas an internal dynamic spurring sociocultural transformation. Unlike the progressive models of the cultural flicting interests who vied in subterfuge. Ethnicity,
evolutionists or the ecological systemists, we do not offer kinship, and class defined the mosaic of overt contending
external agency to explain change, whether linear groups; factions cut across them. Consequently, ethnic
advance, demise, or reconfiguration. Melding ethnohis- distinctions began to blur (e.g., the Otom) as kinship
tory, ethnography, and archaeology, these essays docu- relations transfigured under the tribute demands of a
ment cases of transformation from endogenous com- dominant class (e.g., the Aztecs). Factional connections
petition. The often cited factors of population growth/ were used to move up socially. Paradoxically, the Inkas
urbanism, hereditary leadership/stratification and eco- rigidified ethnicity for similar reasons.
logical symbiosis/trade emerge as consequences of facAcross the gamut of social formations, two groups
tionalism, not as its preconditions. Indeed, even the generated a social forc. In settlement patterning.

206

John W. Fox

duality is evident among the incipient Olmecs through

the Inkas and Aztecs, and was mythologized in the
Quetzalcoatl-Tezcatlipoca saga of endemic competition.
Ethnohistory commences in the A D 900s, with the flight
of peoples throughout Mesoamerica who needed charters
for new political formations and to legitimize inequalities.
What then was the basis for political action? Allied
peoples nested within wider horizontal or vertical alliance groups seeking to secure mutual advantages. But
spoils were divided, and this provided potential for
further conflict. New factional relations continuously
transformed political structures.

Notes
1 Mesoamerican ethnohistory describes conflict in
greater detall than in other American contexts, and
Mesoamerica connected the Mississippian north with
at least Panam south, defining a macroregional
sphere of interaction. Thus, comparisons are made for
regionally distinctive peoples who were linked in
social exchange.