Gender ideology in late formative western Mexico

Material Information

Gender ideology in late formative western Mexico a study of ceramic representations
Logan, Melissa Kristine
Publication Date:
Physical Description:
vii, 363 leaves : ; 28 cm


Subjects / Keywords:
Gender identity -- History -- Mexico -- Jalisco ( lcsh )
Indians of Mexico -- Social conditions -- Mexico -- Jalisco ( lcsh )
Indian pottery -- Mexico -- Jalisco ( lcsh )
Pottery, Prehistoric -- Mexico -- Jalisco ( lcsh )
Gender identity ( fast )
Indian pottery ( fast )
Indians of Mexico -- Social conditions ( fast )
Pottery, Prehistoric ( fast )
History -- Jalisco (Mexico) ( lcsh )
Mexico -- Jalisco ( fast )
History. ( fast )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )
History ( fast )


Includes bibliographical references (leaves 351-363).
General Note:
Department of Anthropology
Statement of Responsibility:
by Melissa Kristine Logan.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
228298180 ( OCLC )
LD1193.L43 2007m L63 ( lcc )

Full Text
Melissa Kristine Logan
B.A., University of Northern Colorado, 1998
A thesis submitted to the
University of Colorado at Denver and Health Sciences Center
in partial fulfillment
of the requirements for the degree of
Master of Arts

This thesis for the Master of Arts
degree by
Melissa Kristine Logan
aran Mineage

Logan, Melissa Kristine (M.A., Anthropology)
Gender Ideology in Late Formative Western Mexico: A Study of Ceramic Representations
Thesis directed by Assistant Professor Christopher S. Beekman
It is believed that pre-Columbian populations in West Mexico shared in the greater
Mesoamerican tradition of cultural complexity, and were in the process of becoming politically
centralized. However, gender is not often considered in these studies of sociopolitical
organization. This results, first, from the use of systemic approaches which do not account for
agency, and second, from the persistence of an androcentric perspective in which men are
assumed to be the only visible social actors. Complex interconnections between all
dimensions of the social structure necessarily implies gender is intimately involved in
processes of cultural development and change. This project demonstrates how human
imagery in indigenous representations links the dynamics of gender at the local level to larger
issues of sociopolitical organization at a regional scale. The presence of gender infused
human imagery coupled with evidence of growing social complexity in Late Formative (100
BC AD 200) central Jalisco, Mexico provides an ideal setting to address these issues. High
concentrations of human activity within the Tequila valleys have defined it as the core of
sociopolitical change during this time. An evaluation of biological sex and gender within this
area and throughout the remainder of West Mexico provides an indication of the integrity of
gender ideology across the region in a political setting. The manifestation of characteristics of
gender and other social statuses in human imagery as poses and gestures, adornment, and
objects possessed establish a basis for biological sex-attribute associations and gender
definitions. Results reveal gender is not perceived uniformly throughout West Mexico as has
been previously assumed. Instead the region is divided into two distinct areas, one in which
gender relations are viewed as complementary between men and women as individuals, and
the other in which gender is organized into a dynamic system of complementary and
hierarchical relations within and between individuals. Without an analysis including gender
such important organizing principles of the social structure will not be recognized.
This abstract accurately represents the content of the candidates thesis. I recommend its
CKristopher S. Beekman

I dedicate this thesis to my parents, Donna and Uwe, who have inspired and encouraged me
to follow my dreams no matter how difficult they are, or how long they take to achieve. I also
dedicate this to my husband, William, for his support, patience, and understanding, but most
of all, for his many selfless sacrifices which have helped me to complete this thesis.

There are many generous people who have assisted me in the completion of this thesis that I
wish to thank. First, many thanks to my advisory committee, Chris Beekman, Tammy Stone,
and Sarah Nelson, for their guidance and support throughout my research. Their
contributions and insight have been invaluable. I would also like to express my appreciation
to Steve Holen of the Denver Museum of Nature and Science, and Margaret Young-Sanchez
and Patricia Tomlinson of the Denver Art Museum, for granting me access to the West
Mexican figures in their museum collections. A special thanks as well to Paul Pennock,
Barbara Ferber, and the staff at Zeiler-Pennock, Inc. for not only the opportunity to pursue my
degree, but their continual support through its completion. And finally, a most heartfelt thanks
goes to my family and friends for standing behind me and encouraging me throughout this

Figures ..................................................vii
Tables ................................................. viii
1. INTRODUCTION ............................................1
3. REGION .................................................24
4. METHODS ................................................49
5. ANALYSIS ...............................................81
6. CONCLUSIONS............................................118
A. MALE-FEMALE PAIRS......................................128
B. CATEGORY RESPONSES ....................................306
C. DATASET................................................349
D. STATISTICAL TEST RESULTS...............................350

3.1 Shaft and Chamber Tomb Distribution ............................................27
3.2 Study Region and Relative Substyle Locations....................................38
3.3 Substyle Dates from Figure Sources and Date Range for
Sociopolitical Change Based on C14 of Excavations..............................41
3.4 Relative Substyle Locations with Respect to Core and
Periphery Regions..............................................................42
4.1 Female Collapsed and Seated Kneel ..............................................63
4.2 Typical Female and Male Mesoamerican Garments...................................67
4.3 Clothing Motifs.................................................................69
4.4 Spatial Arrangement Schemata ...................................................71
5.1 West Mexican Garment and Examples of the Quechquemitl ..........................98
5.2 West Mexican Garment and Examples of the Cape (Tilmatli)........................99
5.3 Garment Positions on West Mexican Figures......................................100

4.1 Number of samples by style ..........................................................53
4.2 Number of samples by substyle .................................................54
4.3 Gender type definitions..............................................................80
5.1 Original data collection categories with associated recoded
categories for SPSS ................................................................83
5.2 Pan-regional biological sex-attribute associations ..................................87
5.3 Biological sex-attribute associations for individual substyles
in the periphery region.............................................................94
5.4 Biological sex-attribute associations for the combined substyles
of the core region ................................................................104
5.5 Pan-regional attribute associations for figures without visible
sex characteristics................................................................107
5.6 Attribute associations for figures without visible sex characteristics
in the Ixtlan del Rio substyle.....................................................108
5.7 Alternate gender typology based only on biological sex associated traits .........112
5.8 Gender counts.......................................................................113
5.9 Gender pairing......................................................................114

Gender is not usually addressed in studies of sociopolitical organization, especially those
that examine processes of political centralization. The exclusion of gender from these
analyses has been two-fold. First, archaeologists tend to use systemic models for evaluating
sociopolitical organization which do not address the role of individuals or agency. Secondly,
the androcentric bias of the discipline has ignored the issue of gender by assuming males are
the only visible actors in the archaeological record.
Recent paradigmatic changes in the discipline emphasizing the importance of issues of
agency and individual action have demonstrated the inability of such models and approaches
to accurately describe sociopolitical organization. Cultures are no longer being viewed as a
collective system, but rather as a dynamic entity composed of active individuals and groups
with their own ideas, choices, and goals.
With the focus shifted from the group to the individual, critical issues of individual agency,
such as identity and ideology, have been recognized. As a result, archaeologists have been
challenged to consider the place of the individual within the sociopolitical development of
complex societies (Dobres and Robb 2000). Studies have begun to evaluate strategies of
individuals at the local level (Dobres and Robb 2000), however, they tend to focus on the use
and manipulation of religious or sacred ideologies in political and economic contexts.
Of the different forms of identity examined, gender has been given the least attention in
archaeological analyses of political centralization. Ethnic, religious, and political identities
have been, and still are, of major importance in studies of complex societies. In drawing
conclusions from a limited number of variables, these approaches have not fully taken into
account the complex interconnection of other social dimensions such as gender, ethnicity, and
class that result in highly diverse and fluid gender ideologies, identities, roles, relationships,
and power dynamics (Spencer-Wood 2007:30).
Studying gender in complex social organization is more than simply finding women in the
past (Nelson 2007:vii). It is about examining the dynamic relationships between the different
genders, and how they define and shape the identity of each (Brumfiel 2007:4). Collectively
examining men, women, and other genders additionally informs on the importance of gender

as an organizing factor of the social structure (Brumfiel 2007:4). The intimate interaction
between gender, other social elements, and the larger social structure necessarily implies
gender is as integral to sociopolitical organization as religion, politics, and economy.
Incorporating gender into studies of cultural change, such as political centralization, not
only allows the relative fates of men, women, and other genders to be assessed (Brumfiel
2007:4), but also reveals how different dimensions of gender affect, and are affected by,
changes in political organization. Although at first glance gender may appear trivial in
sociopolitical transformations, it is in fact quite enmeshed in both the development and
functioning of complex societies.
Chapter 2 lays out the theoretical basis for this thesis, in which I discuss the current state
of gender in studies of sociopolitical complexity, and suggest how a gendered approach could
further these types of studies. Early attempts at addressing gender and complex social
organization either focused on contradicting the universal assumption that female
subordination occurred with the rise of the state, or explained why and how women became
devalued during this process. However, gender relations have been shown to be more
complex than this. Not only can they change in some aspects of society, while remaining the
same in others, they are not always viewed as hierarchical. Additionally, gender is not a
bounded social identity, but is, rather, constructed by its complex interaction with other social
dimensions such as status, class, age, and ethnicity (Gero and Scattolin 2002:160; Spencer-
Wood 2007:30). Therefore, addressing gender in complex social organization should be
focused on understanding the dynamics of the status of different genders with respect to one
another, and if gender inequality is present, what is its relationship to the other social and
political structures of society (Nelson 1993:299, 2003:1, 2006).
In light of these changes, new models and ways of addressing agency are being
developed (Dobres and Robb 2000). However, archaeologists continue to be hindered in
studies of gender by equating biological sex with gender, and viewing gender ideology as
universal and hierarchical. If archaeologists hope to gain a more representative view of past
societies in general, and processes of centralization in particular, then it is critical they
recognize that persons of other genders, such as women, as well as other minorities are
active and visible in the archaeological record (Spencer-Wood 2007:32). Additionally, they
must acknowledge that gender is not as simple as the binary based classifications of male
and female, nor are all aspects and relations in a gendered ideology organized
complementarily or hierarchically (Gero and Scattolin 2002:158).

Gender, because it is part of the interconnected social structure in which all members of a
society participate, can not only inform on its role in society, but also on those of other social
aspects, such as politics, religion, and economy. In the same respect, these social elements
can, in turn, inform about gender. Through the examination of gender ideology in indigenous
representations of human images, we can begin to understand the place of gender in
sociopolitical organization, its relationship to other social identities, and how those articulate in
processes of political centralization at both local and regional levels.
At the micro scale focus is directed towards determining how steadfast the boundaries are
between different genders, and under what conditions are these lines blurred or crossed.
Macro analyses specifically address how these gender definitions and relations are shared or
unshared throughout the region.
Chapter 3 introduces pre-Columbian Mesoamerica as an ideal location to study gender
because of the abundance of human imagery and evidence for the development and
presence of dynamic and complex sociopolitical systems. West Mexico, in particular, is well
suited for this because it is a region that is believed to have undergone significant
sociopolitical changes between 100 BC and AD 200 (i.e. Late Formative period) (Beekman
2007a; Beekman and Weigand in press:7, 8), that were accompanied by the public use of
human figural sculpture in mortuary ritual.
Current models suggest the presence of a lineage-based form of social organization
during this time, that is either reminiscent of multiple competing chiefdoms (Butterwick 1998,
2004), or a corporate power sharing strategy in which power is distributed among different
lineages or groups of lineages at various levels (Beekman 2007a, 2007b). These models are
set within a regional context of core-periphery interactions that have been described as
representing increasing political centralization (Beekman 2000), as well as a state-like form of
social organization (Weigand 1985, 2000).
The shaft and chamber tombs of ancient West Mexico have yielded an abundance of
ceramic figural sculpture in the form of paired figures, referred to as male-female couples.
Ceramic pairs not only display emic perceptions of gender, but also contextualize them within
the accepted social relations. The production of these pairs in eleven different locations
throughout the region each displaying local variations provided a means by which to compare
and contrast intra-regional gender ideologies at a regional level. This connection between
micro and macro scales shows how gender plays into not only local sociopolitical
organization, but also the larger social system within the region.

The demand for these paired figures by art and antiquity collectors has led to rampant
looting in the region, and as a result, many couples exist today in private or museum
collections. Unfortunately, there is much opposition to the use of such artifacts which are
unquestionably looted and lacking provenance. Concerns focus on issues of artifact context
and authenticity. I would like to propose that these pieces can not only be contextualized
within the social reconstructions offered by archaeology, but also that their authenticity can be
evaluated using scientific methodology.
Chapter 4 describes the methodology of this analysis, beginning with an introduction of
the artifacts comprising the dataset. Of specific concern, is where the specimens came from,
and the types of individuals represented in the human imagery. This is followed by a detailed
discussion of the attributes that indicate gender and other social information on the figures.
These include body poses and gestures; various forms of body adornment, particularly
clothing, body painting, tattooing, scarification, jewelry, and headdress and hairstyles; and
different objects held by or associated with each individual.
Also of importance is the identification of the biological sex of the figures. This is
determined by the presence or absence of primary and secondary male and female sex
characteristics. Although biological sex is identifiable on many of the figures, it can not be
determined for all of the individuals. The presence of a swollen belly or a garment used for
male genital protection are a more indirect means of assessing the biological sex of the figure.
These two additional attributes could not only confirm the sex of an individual with visible sex
characteristics, but also potentially suggest the sex of those without them.
Chapter 5 discusses the use of chi-square tests for determining which attributes
consistently associated with biological sex. Attributes that repeatedly associated with males
and females are considered to be characteristic of the man and women genders respectively.
Since other gender forms are visible either by individuals possessing traits of both genders, or
by an individuals biological sex being contradicted by the presence of traits characteristic of
the opposite sex (Munson 2000:133), gender is assigned to the figures based on the
combination of biological sex characteristics and male and female traits.
Chapter 6 focuses on interpreting the results from statistical testing, which reveal that
gender was not perceived uniformly throughout West Mexico. In fact, within the more
centralized part of the region where the highest population concentration was, gender was
viewed as a dynamic system of complementary and hierarchal relations both between and
within individuals, while the areas located further away from this central region saw men and

women as complementary genders.
Although gender complementarity characterized the periphery region, each cultural unit
perceived this relationship in its own unique way. In one area, the pairs were utilized to make
the distinction between the roles of men and women, while at the same time emphasizing the
importance of their complementary union. In another, the pairs not only strongly differentiated
between men and women despite their complementarity, but also displayed other aspects of
their social identity.
The perceptions of gender and other social relations appear to be most complex within
the central region. Men are not defined by specific gender roles, but they do display symbols
of rank and status. Women were afforded much prestige for their reproductive and productive
abilities. Regarded as vital elements for the reproduction of society, women not only held a
high status, but were also tightly integrated within the social structure. Males needed to work
to achieve the reproductive status of women, which was done through attaining a dual-
gendered identity. Only through possession of masculine and feminine aspects could a male
be joined with a women to either produce children, or perpetuate a lineage. Complementarity
between men and women is achieved not only through the dual-gendered individuals, but also
through the pairing of two dual-gendered figures together, in which a complete union between
a man and woman occurs through each individual bringing elements of both.
The dynamic gender relations within the centralized region suggest the joining of the two
figures in the ceramic pairs is much more complex than a male-female marriage couple. Both
biological and social elements play an important role not only in the relationship between the
two figures, but also how each figure's gender and identity is defined.
The establishment of official ideologies, particularly related to gender and social rank and
status indicate the core may be more centralized than the periphery. Disunified ideology
within the core presents several potential scenarios, including the misidentification of the core
itself, as well as interacting, but independent social structures. A uniform gender ideology
throughout the remainder of the region points to the possibility that it represented the
dominant ideology prior to the changes within the centralized region.
In the pages that follow, I demonstrate how to make this link between the dynamics of
gender at the local level and larger issues of sociopolitical organization at a regional scale. I
show how human imagery in indigenous representations can be evaluated to inform on
gender ideologies. Through the analysis of symbols of identity and status, and the social
context within which the imagery exists, we can view emic perceptions of the interplay

between gender, class, ethnicity, and age in the formation of social identities, and the
relations, roles, and responsibilities associated with them. Viewing gender ideologies as
universal, hierarchical relationships between only men and women, and equating gender with
biological sex can no longer provide a realistic representation of gender, nor can it draw the
appropriate link between micro and macro scales. Only through conceptualizing gender in
ways that honor how gender performances and ideologies operate in the real world (Gero and
Scattolin 2002:161), can issues of political centralization and social complexity be completely

The development of political centralization is a heavily debated and extensively covered
area of anthropology, however, the topic of gender is repeatedly excluded from such
analyses. This is due, in part, to the continued use of systems-based approaches for the
analysis of sociopolitical change by archaeologists. In emphasizing group and not individual
change, these models focus on macro scale patterning of regional relationships and
interactions, and not micro scale strategies of individuals. The persistence of these
approaches derived from a male dominated perspective in which elite white mens roles,
activities, and viewpoints [are] represented as the genderless norm (Spencer-Wood
2007:32), distort reconstructions of social organization. Such models assume only the
activities and behaviors of males are visible because they are the important members of
society. Contributions by women, as well as other groups, such as children, are completely
ignored by these models (Spencer-Wood 2007:32). Gender, then, has in two ways been
overlooked in these past approaches to sociopolitical change. Through the continued use of
systemic approaches to sociopoltical change, archaeologists continue to ignore the role of the
individual, which includes such issues as gender. Additionally, these systemic models which
are the product of a male-biased discipline are not only flawed in their afformentioned focus
on the group, but also in their assumption that males are the only actors in these groups.
The persistence of implicit assumptions associated with these approaches that tie gender
to biological sex, and see only male involvement in the public political sphere have
perpetuated the development and use of models denying the role of gender in complex
sociopolitical processes (Key and MacKinnon 2000:118). Androcentric models of state
development, such as the one proposed by Frederick Engels (1972 [1884]), attributed female
subordination to the rise of the state. This decline in status for women, which he proposed,
became the end-all be-all result of state development (Nelson 2003:2). Even early feminist
models accepted this inevitable consequence and set out to understand and explain why it
was occurring (Reiter 1977:5).
Recent paradigmatic changes emphasizing the important role of the individual in cultural
change have provoked questions regarding the applicability of these older models of

sociopolitical change. With the focus shifted from the group to the individual, critical issues of
individual agency such as identity and ideology have been recognized. As a result,
archaeologists have been challenged to consider the place of the individual within the
sociopolitical development of complex societies (Dobres and Robb 2000). New studies have
begun to evaluate strategies of individuals at the local level (Dobres and Robb 2000),
however, they tend to focus on the use and manipulation of religious or sacred ideologies in
political and economic contexts.
In drawing conclusions from a limited number of variables, these approaches have not
fully taken into account the complex interconnection of other social dimensions such as
gender, ethnicity, race, and class that result in highly diverse and fluid gender ideologies,
identities, roles, relationships, and power dynamics (Spencer-Wood 2007:30). This
multifaceted relationship necessarily implies that any dataset recovered from the
archaeological record will yield information on a variety of social aspects including gender. It
is, therefore, not only possible to examine gender in processes of centralization, but it is also
essential to conceptualize [the] subject matter in ways that honor how gender performances
and gender ideologies actually operate in the real world (Gero and Scattolin 2002:161).
Gender ideology is most frequently depicted in material culture as artistic representations
of human figures. By its very definition, art is a material expression of mental constructs that
are formed by the social structure in which one operates. As such, the imagery they display
can provide an intimate view into how gender was perceived by past populations. With this
type of access to the social structure, it is then possible to attempt to reconstruct gender
relations within different political situations.
Gender and Politics
Surprisingly, Engels' (1972 [1884]) model not only discouraged gender studies in political
contexts, but also promoted them. In failing to recognize the dynamics of gender, he both
skewed perceptions of gender relations and devalued the status of women. However, Engels
was one of the first to really address the issue of gender in political organization. His model
reflects an early attempt at placing gender within a political context, and it has provided a
base point from which feminist models could grow.
In The Origin of the Family, Private Property, and the State, Engels (1972 [1884]) details a
model of evolutionary social development in which the matrilineal egalitarian organization of

primitive societies is replaced, through the process of state formation, by a patriarchal gender
hierarchy of female subordination. This is brought about by the development of private
property and institutionalized by means of the monogamous family (Silverblatt 1988:430).
Influenced by his western perspective, Engels (1972 [1884]) linked gender roles with
reproductive tasks and defined a sexual division of labor based on biology which delegated
women as dominant in the home (i.e. private sphere), and men over the forest (i.e. public
sphere). This created a universal and uniform gender ideology for men and women.
Engels (1972 [1884]: 100) believed that in egalitarian societies male-female social/sexual
relations existed in the form of a group marriage in which consanguine kin were common
spouses to one another, creating a competitive free situation in which all females were widely
available to all males. However, such an androcentric perspective fails to recognize that
males were equally available to all females. At first, these relationships were more
multidirectional across age and generation lines, then later, they became more vertically
restricted within generations and excluded close kin (Engels 1972 [1884]: 100-103). Under
this form of organization, matrilineality and matriarchal descendence dominated because a
lineage could be easily traced through females, as one is always certain who their mother is,
but much less so of the identity of their father (Engels 1972 [1884]: 106).
Over time this form of organization became more difficult to adhere to as new social
mores prohibited close relative relations and natural selection revealed the advantages of
producing children with very distant kin or non-blood relatives (Engels 1972 [1884]:111, 112).
Offspring produced by non-related mates were more likely to survive than those born of
unions between closely related persons in which stillbirths and genetic birth defects were
more frequent. These new restrictions on sexual partners led to the scarcity of females by
reducing the pool from which to draw from, thereby creating competition between males for
females (Engels 1972 [1884]:111, 112). While Engels emphasized the decrease in number of
females as potential mates for males, the number of males available to females would also
have declined because relatives of the opposite sex were ruled out as possible mates.
Intentional pairings, then, ensured mates would be provided for both males and females alike.
Initially it seemed that a paired marriage offered the greatest benefits for women, and Engels
(1972 [1884]:116, 117) in fact, argues that it was female driven. The more traditional sexual
relations lost the naive primitive character of forest life, owing to the development of economic
conditions with consequent undermining of the old communism and growing density of

population, the more oppressive and humiliating must the women have felt them to be, and
the greater their longing for the right of chastity, of temporary or permanent marriage with one
man only, as a way of release (Engels 1972 [1884]:117).
Later, however, it became apparent that males gained the most from this new
relationship. The relatively monogamous relationship of the paired marriage for the first time
allowed paternity to be quite easily identified, and with the development of private property
this allowed males not only the ability to possess wealth, but also the means by which to pass
it on through generations, albeit through their mother's blood relations (Engels 1972
[1884]:118, 119). Females, however, still retained their right to the children, and therefore,
descent remained matrilineal. It was only after males overthrew the mother right-descent
in the female line and the matriarchal law of inheritance, that patrilineality was able to be
established (Engels 1972 [1884]:120). Patrilineal descent could only persist within the context
of a strictly monogamous marriage that ensured a man's wifes offspring were without a doubt
his own. This was achieved through the institutionalization of the monogamous family in
which the exclusive supremacy of the man [was] established, as the man took command in
the home [and the] woman was degraded and reduced to servitude; she became the slave of
his lust and a mere instrument for the production of children (Engels 1972 [1884]:120-121).
While acknowledging the inability of knowing how and when patrilineality took hold, other than
it may have resulted from a simple decree...that...future offspring of the male members
should remain with...their father" (Engels 1972 [1884]: 120), Engels (1972 [1884]:96, 121, 125,
131) did believe that it occurred concurrently with the evolutionary development of civilized
society which he directly equated with state formation using Greek and Roman examples.
For many years this model was accepted not only as a universal of state development,
but also the ultimate and inevitable reality associated with increasing political complexity.
Early feminist studies focused on this question of origins and [the] perpetuation of gender-
linked power hierarchies (Reiter 1977:5), specifically trying to understanding how and why
female subordination occurred with state formation. Feminists were faced with the task of
determining the conditions under which parallel forms of organization between women and
men [were] subsumed into one [that was male], specifically how leadership roles and
activities [that were] formerly associated with women [became] devalued or obliterated
(Reiter 1977:14).
Many of the first frameworks retained Engels evolutionary perspective. While feminists

critiqued his lack of attention to the important role a societys unique cultural history played in
shaping the processes leading to this change, they continued to ignore it as well. The models
presented alternative processes by which women became inferior, such as the undermining of
kin group authority (Sacks 1979) and giving political character to existing kin-based
organization (Gailey 1987). However, these processes also were assumed to be universal
and applicable to all societies.
Christine Ward Gailey (1987) argued that women became subordinate with state
development through the processes involved in class formation. In removing themselves from
the production of subsistence goods through the control of labor, emerging elites alter the kin-
based division of labor by separating the facets of gender and age from kin defined identities
(Gailey 1987:16).
In kin-based societies the social value of a person is determined by the collection of their
different identities, where the actual maleness or femaleness of an individual is only one of
many identities used to define an individual's social role. As such, gender relations are an
integral part of an individuals social identity (Gailey 1987:9). Being a male or a female may
suggest a variety of potential roles or relationships of an individual, but it does not specify their
particular social status (Gailey 1987:9). For women, in particular, this means their role in
sexual reproduction is merely an aspect of being a woman.
Through state formation, the roles of individuals are divorced from kinship obligations and
become defined solely by gender due to the disruption of social continuity in kin groups
brought about by an emerging ruling class (Gailey 1987:14). In class and state formation,
peoples functions in the division of labor come to be discernible with reference to categories
of gender, age, and skill, abstracted from their particular kinship connections and meanings
(Gailey 1987:16). As people become identified independently of kinship-for instance, as
constituents of a class-biological differences or functions (as defined in the culture), rather
than social identities, become increasingly important (Gailey 1987:16-17).
As both producers of goods and people, women are most heavily affected by this. Their
reproductive ability which was previously a single dimension of their person, has become the
primary and sole defining aspect of their identity and status (Gailey 1987:22). With state
development, controlling womens reproduction is essential to the reproduction of society,
specifically the elite classes. To gain control of womens reproduction, women are afforded a
lower status as they become defined solely by their reproductive ability, while men remain

defined by their collective identities (Gailey 1987:22-23).
These early models also retained Engels perception of a universal gender ideology, some
further supporting female subordination with state formation, while others arguing womens
status either improved (Ortner 1981) or remained the same after this transition. Sherri Ortner
(1981) believed womens status tend[ed] to be higher in [stratified] societies than in simpler
societies" because virginity provided a way for women to achieve a higher status (Ortner
1981:401). By downplaying the uniquely feminine capacity to be penetrated and give birth to
children, a virgin could hold high status, and in essence escape the low status associated
with womens role as a producer (Ortner 1981:401). In this way, some women, those who
were virgins, came to be held in higher esteem with the rise of the state, rather than becoming
devalued by it.
Regardless of whether or not womens status improved with state formation, gender was
still being viewed as a single all-encompassing ideology. Both Karen Sacks (1979) and
Christine Ward Gailey (1987) observed much ambiguity in the gender ideology of the
populations they studied. Not only was it not consistent and uniform (Gailey 1987), but it was
often contradictory (Sacks 1979). What this suggested was that the assumption of a universal
ideology was incorrect. The dynamic nature of gender as a cultural construct made it
impossible for it to exist.
New research was revealing that gender relations were not as simple as society-wide
intensification of existing differences or the development of differential status with state
development, rather changes could occur in some aspects of society while remaining the
same in others. Furthermore, gender relations did not always exist in a hierarchical fashion in
each part of society either. They were often viewed as complementary.
Gender complementarity was originally seen as an alternative to gender hierarchy (Joyce
1996:180), where males and females did not stand in opposition to one another, but rather
functioned together in an interrelated relationship (Gero and Scattolin 2002:156-157; Joyce
1996:180). Each gender brought forth an equally important, but fundamentally different social
aspect and/or responsibility necessary for the proper functioning of a household. Without the
contribution by either, society would be less than complete.
Complementarity!, in this way, ] constructed by and expressed in distinctive
productive and ritual roles. These gender codes rely not on biological features,
but on the actions of men and women as culturally constituted within...society
(Joyce 1996:181).

This represents a form of gender balance in which stereotyped male and female actions
complement...each other (Joyce 2007:204). Complementary gender relationships have been
argued to exist throughout the ancient Americas. Through her work with figurine imagery in
pre-Hispanic Central America, Rosemary Joyce (1993:260, 266) has been able to determine
that within these societies males and females held different but complementary roles that
were negotiated through the use of public representations of human images.
Using gender imagery in Classic period Maya monuments, Joyce (1996) has also
expanded on this concept by arguing that paired depictions of males and females not only
emphasize the complementarity and interrelationship between different genders, but also the
duality expressed within a single individual. Complementarity between males and females is
reflected in their relative placement with respect to one another (Joyce 1996:172-176), and is
evident on individual representations as the presence of costumes associated with both males
and females (Joyce 1996:182). This is often seen in Mesoamerican imagery in which deities,
legendary figures, and even contemporary actors were understood to act as necessary halves
of gendered wholes (Joyce 2007:203). Expressed like this, focus is not on individual and
bounded gender identities but on the complementarity... embodied [by the person] and the
power that [is] derived from possessing aspects of both genders (Gero and Scattolin
2002:157). Regardless of whether or not two separate people or a single person in two
gendered roles is being depicted, it is the duality of the gender identity of the participants in a
single action that is being stressed when gender is perceived as complementary.
Issues of gender were further complicated by the dynamics of gender as a cultural
construct, which enables it to be manifested differently throughout an individuals life, as well
as in different social situations. Feminists were faced with the increasing realization that the
status of women could not be defined because [g]ender is not a bounded, static
phenomenon but is rather a constructed set of relationships embedded in other cultural and
historical institutions and ideologies, such as status, class, age, ethnicity, and race (Gero and
Scattolin 2002:160). Theoretical critiques in anthropology at the time regarding critical issues
of the applicability of universal models and the associated effects of denying the importance of
unique cultural history only confirmed doubts feminists were having on the orientation of their
approach (Silverblatt 1988:441-444, 452, 454).
These factors led to a paradigmatic shift, in which the former focus on womens loss of
status was redirected towards understanding the dynamics between gender and political

organization (power) through examination of the status of different groups of women with
respect to one another, as well as compared to men (Nelson 2003:1, 2006). It should not be
assumed that gender hierarchies and female subordination arise with state development.
What should instead be determined is whether states necessarily entail gender inequality
(Nelson 1993:310), and if so, what is its relationship to the other social and political structures
of society (Nelson 1993:299). It may be found that a gender hierarchy does not exist, or that
other statuses take precedence over that of gender (Nelson 1993:299). Gaining an
understanding of gender relations can aid in understanding the specific processes of political
centralization, even if gender alone cannot be used to indicate the presence of a particular
form of social organization. Today, the older models suggesting strong cross-cultural trends
in gender-politics relations have been largely eliminated and current approaches focus on
these new perspectives.
Joan Gero and Cristina Scattolin (2002), in Beyond Complementary: New Definitions for
Archaeological Gender Relations, critique current archaeological approaches to gender and
bring to light important issues that archaeologists must consider and attempt to account for
when doing gender studies. As the title of their article suggests, they want archaeologists to
move beyond dichotomous assumptions of gender complementary and hierarchy that lump
together too much variability (Gero and Scattolin 2002:158). Gero and Scattolin (2002:170)
that binary classifications of gender systems into complementary and
hierarchical classes and the recognition of male dominance or greater male
status within hierarchical systems neglect critical areas of gender arrangements
where authority and autonomy, control and knowledge, are distributed in a more
fine-grained manner among gendered participants. Crosscutting, sometimes
contradicting, the ideologies of prestige, power, and status are practical
arrangements taken on by gendered personnel that often give significant control
over resources and independence of action to the dominated gender. Not only
do we fail to see how action and responsibilities are distributed across genders in
such analyses, but characterizations of sociocultural systems in such gross,
generalizing manners fail to illuminate how gender arrangements are integral to,
underpin, and substantiate basic social structures.
Such a perspective lends itself well to approaching gender relations and gender diversity
in terms of heterarchy (Levy 2007:192-195). Heterarchy refers to the idea that the different
elements of the social structure can be ranked in a variety of different ways, including
hierarchically, depending on different spatial and temporal conditions (Crumley 1987:158,
163). Janet Levy (2007:192-193) argues that rather than accepting dichotomous models of

male and female organized in opposition to one another, what should instead be done is
embrace the diversity that exists in gender and the multitude of ways in which it can be
organized. A heterarchical approach emphasizes lateral social differentiation which in turn
requires the recognition and examination of both vertical and horizontal social relations at
various scales (Levy 2007:200, 201). By adopting the concept of heterarchy, an
archaeologist is encouraged to look not at classifications (e.g. chiefdom/state or
corporate/network, although these can be helpful concepts at times) but rather at cross-cutting
variables of status, influence, and power and the fluctuations among and between them
(Levy 2007:200). Such variables include gender, age, class, ethnicity, race, lineage, as well
as, ritual influence, economic control, and political power (Levy 2007:200).
Despite the advances made by new perspectives in gender studies, analyses tend to
focus on either pre- or post-state conditions while ignoring the transition in between. It has
been argued that unstable times created by these radical changes in social organization
present far greater flexibility in social roles associated with gender identities than during
periods of political stability (Key and MacKinnon 2000:109, 118). Studies of ancient Maya
burials in Mesoamerica suggest that among societies increasing in political centralization
there are variations in grave goods, treatment of the deceased, cause of death, as well as
grave construction and location, that reflect the differing degrees of gender inequality and its
negotiation associated with sociopolitical changes (Key and MacKinnon 2000:112, 113).
Archaeologists must begin to address this transition because it is within it that gender-power
dynamics become heightened and most apparent as individuals vie for power by reifying,
altering, and redefining identities to improve their social position. Only through examination of
this shift are we be able to see how aspects of gender affect and are affected by changes in
political organization.
My goal here is to embrace the possibility of the coexistence of gender hierarchy and
gender complementarity, and to recognize that gender types can be defined by many different
factors without giving precedence to biological sex or sexual preference even though they are
a part of gender. Identification of gender then is not about defining specific gender roles, but
is rather focused on the nature of gender boundaries, particularly determining whether they
are clearly delineated or more blurred between males and females. What I attempt to do
through this analysis is define and explain which societal aspects are organized in which ways
(Gero and Scattolin 2002:158). More specifically, I consider how, and in what realms, gender

systems are organized hierarchically or complementarily (Gero and Scattolin 2002:159) within
societies undergoing processes of centralization.
Gender and the Archaeological Record
Gender is often used as a contrasting term to sex, to depict that which is socially
constructed to that which is biologically given (Nicholson 1994:79). However, new thought
has suggested that these concepts are more interrelated and less distinct than had been
previously believed (Nicholson 1994:79). There differences can be resolved by recognizing
that biological and social aspects interact to form each other (Errington 1990:14).
[Although bodies have undeniable physical characteristics, we as human beings
make sense of bodies through cultural means. We interpret physical
characteristics as signs-of age, gender, race, status, ethnicity-that indicate the
persons inner self (intelligence, abilities, morals, personality, etc.). These signs
lead us to expect certain kinds of behaviors (Errington 1990) (Munson 2000:128).
Gender is constructed when physical traits of sex are culturally elaborated and given
meaning within a culture (Munson 2000:128). Every culture defines gender by a unique set
of socially determined characteristics, such as physical attributes, roles, behaviors, and rules
for participation within the social structure (Eicher and Roach-Higgins 1992:17; VanPool and
VanPool 2006:54). These vary from culture to culture, and include biological sex to some
degree or another, but do not require it to be the sole determiner of gender. Commonly, third
and fourth genders reflect identities of those individuals whose biological sex does not
correspond with gender (Munson 2000:128; VanPool and VanPool 2006:54). The berdache is
probably the most well known and frequently cited example of this. More than one-hundred
North American Indian cultures recognize a third, and often fourth gender, referred to as the
berdache in which an individual, usually a male, assumes the traits and roles of the opposite
sex (Callender and Kochems 1983:443, 445). Among some of these groups berdache also
includes females who take on characteristics of males (Callender and Kochems 1983:443).
As a cultural construct, gender can not only be defined in many ways, but it can also
change its expression and manifestation over the course of an individuals life, or even in
different social contexts. Significant gender transformations have been recognized to coincide
with different periods of biological development, often tied to the reproductive life cycle, and
are represented as the transition from childhood, to adulthood, and then to old age (Joyce
2000b; Lesick 1997). Gender, and the associated roles, responsibilities, and relations, are

conceptualized differently within each age bracket (Joyce 2000b). Among the Aztec of late
Pre-Columbian central Mexico, boys and girls, are not considered gendered until they have
made the transition to adult status (Joyce 2002:82-83; for a detailed discussion see Joyce
2000b). For males, this occurs upon taking captives in war, and for females, it is marked by
the delivery of a child (Joyce 2002:82-83). Figurine studies among early populations in
southern Mexico have revealed that specific hairstyles were highly indicative of women's
transition between different gendered statuses. The most elaborate hairstyles were worn by
young women of a marriageable age, while the simplest were depicted on married women
(Flannery and Marcus 1998:38).
Although gender has been widely examined in feminist studies in cultural anthropology,
archaeological studies have been much slower in doing the same (Gero and Scattolin 2002;
Nelson 1993:297-298). In general, archaeologists have only minimally addressed issues of
gender, and have discussed even less so its place in the development of complex societies.
This is due, in part, to the use of systemic approaches for cultural analysis, in which the
individual is subsumed within the collective, unable to be recognized separately from the
group and incapable of affecting the overall condition of society. It is only with the
development of current theoretical insights that the importance of the individual, and related
issues such as identity, have began to emerge for archaeologists. Of the different forms of
identity that have been examined, gender has been given the least attention in archaeological
analyses, especially in complex sociopolitical development (e.g. the development of
centralization). Ethnic, religious, and political identities have been, and still are, of major
importance in studies of complex societies.
Another reason archaeologists have shied away from gender studies may be due to the
type of data available in the archaeological record which can make addressing some cultural
aspects more difficult than others. Feminists who deem archaeological studies of gender as
too generalized (Gero and Scattolin 2002:158) often forget the limitations that are posed by
the different sources from which they draw their data. Anthropologists, including many
feminists, have been able to examine more complex societal issues through the use of
ethnographic and ethnohistoric data which provides them with a richer source of information,
in a sense, because they can converse directly with their informants and/or consult written
text. Even with occasional access to recorded history, archaeologists will never have the
luxury of living informants with whom they can directly discuss such issues.

Despite the fact that these individuals deal with different sources of data, it does not mean
that one field of study is more suited to look at issues of gender than the other. It merely
suggests that they have to address different aspects of gender and ask different questions of
their datasets. Each has advantages and disadvantages that the other does not for obtaining
certain information based on the type of data available to them. This should be viewed
positively because each approach brings forth information unattainable by the other, and in so
doing, creates a more complete picture of what gender is and how it operates.
Archaeologys contribution, which is important for this study, is two-fold. First, it allows
changing social roles to be evaluated in contexts that no longer exist today, and second, it
allows us to look at the distribution of the thoughts and ideas of people across the landscape.
The original sociopolitical development of societies is one context in which we cannot gain
information by studying modern populations because in the face of globalization no group
exists today in isolation from outside social or political influences. Therefore, we cannot
observe a culture under the conditions in which these early populations were developing.
Additionally, such sociopolitical changes as the initial development of centralization, are no
longer occurring in the world.
To understand the dynamics of such developments, we as archaeologists are able to look
at different aspects of social organization through the use of material culture. Because
material culture is a physical manifestation of intangible concepts, like ideologies, you can
look at the way the objects are placed temporally and spatially (in the archaeological record)
and get an understanding of the operation of the social structures and the relationships that
existed not only within them, but between them as well. Artistic depictions are one form of
material culture in which gender ideology is often conveyed.
Since many of the characteristics of gender are visually manifested upon the human
person as clothing, ornamentation, hairstyles, and objects in their possession (Munson
2000:128; VanPool and VanPool 2006:54), it would stand to reason that it is possible to
identify gender on artistic representations of humans who possess the same characteristics.
Even if the figures do not accurately depict everyday realities, they do reflect the artists
perceptions of biological sex and gender (as well as other social categories) that have been
shaped by their social structure (Munson 2000:128).

Identifying Hierarchical and Complementary Relationships
The evaluation of gender and political centralization in the archaeological record is two
part. At the micro level, identity marking (specifically of gender types), can be used to
evaluate the differential status and relationships that exist between various forms of identities
(i.e. gender ideals/ideology) within a group or community. At the macro level, the extent to
which these gendered statuses and relations (i.e. gender ideals/ideology) are shared
throughout a society can point to the types of relationships that exist between the different
groups or communities.
Marked Identities
One way differential status is recognized in the archaeological record is through the
presence of marked identity forms. The idea of marking forms to indicate difference
originated in linguist studies during the 20th century with Ferdinand de Saussure. He
suggested basic word forms, such as man, are considered natural or normal because of
their more general and non-gender specific meanings (Saussure 1974; Trask 1999). Marked
word forms then, such as woman, which are essentially modified versions of the basic form
with a more specific meaning (i.e. usually an attachment to a gender type), are considered
unnatural or abnormal. It is the very nature of the marked form (i.e. its modification and
specification) that makes it less than" the basic form of the word because it can only be used
in very specific instances, whereas the basic form can be more universally applied (e.g.
Americans versus African-Americans, Native Americans, etc.).
When this idea was translated into archaeological study, forms of identity became the
subject of markedness. However, the same basic principles still applied. One of the most
common applications of identity marking in archaeology is in studies examining ethnic
interaction, especially in situations of initial contact. When outside migrant populations come
into indigenous communities for the first time, because they are different, they can become
marked as the other" (Stone 2003:35). The indigenous ethnic identity which is familiar and
has existed as the only identity within the community is deemed the normal or natural way to
be. Therefore, there is no need for this identity to become marked. The migrant ethnic
identity, however, is easily marked because it is usually visually different from the indigenous
one. Physical differences between the indigenous and migrant populations become the
source of ethnic marking and the us/them distinction (Stone 2003:35). In this particular

analysis it is gender, rather than ethnic identity, that is of most concern. Although Western
linguistic examples often associate marked forms with the female gender, I would argue that
any form of gender may become marked depending upon its lack of normality.
Regardless of the form that is marked, marking will always create an unequal relationship
between the different forms based on their degree of normality. Within this vertical hierarchy
the normal form is placed at the top while all other forms are ranked below it. As in
linguistics, the marked form is modified and assigned specific traits, but the normal form is
not. However, the implications of this are much more significant when applied to identities
held by individuals. Placed in this opposition, those who possess a marked identity assume a
subordinate position of power and are afforded a lower status. Gender marking, therefore,
should be a fairly reliable indication that a gender hierarchy is present.
The absence of marking has important implications for gender as well. If forms are not
explicitly marked within a particular sociopolitical context, then neither gender is associated
with traits distinct from the other. The lack of marking may indicate that gender is not being
signaled through the artifacts examined, or it may be that gender is unimportant in that
particular sociopolitical context. Lack of importance of gender in an aspect of society is
usually concluded when there is no evidence supporting the assignment of roles to individuals
based on gender. Imagery and ideology would then suggest that roles are defined by other
forms of identity.
If complementarity exists within a specific social aspect, then both genders are associated
with a unique set of traits. This relationship tends to be seen repeated in imagery, ideology,
and/or mythology that depicts the importance of the unique roles of each, and how they are
bound together into an important function within society (Joyce 1993).
Marking becomes most apparent both in society and within the archaeological record
when the forms are repeatedly assigned the same attributes in prominent contexts. Artistic
depictions, by their very nature, are an intentional and highly visible form of communication
that associate different forms of identity with different attributes.
Questions to be asked in the micro analysis include 1) are tasks/roles segregated by
gender or are they integrated across gender lines?, 2) if tasks/roles are well defined by gender
boundaries are these sometimes violated by non-gender criteria such as kinship or community
of birth/residence?, and 3) how are the gendered figures segregated by specific roles and how
at the same time are these gendered figures integrated by activities that crosscut gender

categories (Gero and Scattolin 2002:162, 163, 168)? To address such a complex and
dynamic issue in an equally complex and dynamic context requires a study area that is not
only growing in political complexity, but is also rich in gender imagery.
Shared Ideologies
The degree to which an ideology is shared throughout a region can indicate interaction
between different areas, the degree of integration within a given area, or the extent of
influence or control that one area has over another. An increase in all of these aspects within
a region may suggest the presence of processes of centralization.
For the purpose of discussing patterns of ideology that accompany centralization, I use
the terms core and periphery as a means of distinguishing between the relative regions
involved. However, I am in no way implying the interregional interaction between the core and
periphery is a dominant relationship. Chase-Dunn and Hall (1991:19) have proposed the
concept of core-periphery differentiation to describe such a relationship in which the core and
periphery are different, but are dynamically interlocked. Here, core refers only to the more
centralized area of the region marked by higher concentrations of archaeological materials,
possibly indicating increased activity and population density. The periphery then is defined as
all the remaining area within the extent of the region (under discussion) that exists outside the
Although interaction may be occurring between the two regions, this does not necessarily
mean that practices and ideas are shared (Stein 1998). Each region can maintain their own
separate ideologies despite the information, goods, and people crossing the boundaries (Stein
1998:155). Conversely, the presence of core practices or ideologies in the periphery does not
necessarily imply core dominance over periphery elites (Stein 1998:166). Agency in the core
and periphery, reflected in choices and goals of individuals and groups, works to both
incorporate and inhibit social influences by structuring internal and external interactions (Stein
1998). Therefore, evaluation of the macro scale is equally as important as evaluating the
micro scale. Questions to be addressed in this analysis regarding gender ideology at the
regional level are 1) is gender ideology shared within the core region?, and 2) is gender
ideology shared between the core and periphery regions?
Political centralization is not simply an either-or condition, rather it is a process. As such,
centralization can be placed on a continuum and interpreted in varying degrees. In areas

where political centralization is beginning to occur, one would expect to see differences
among ideologies within the most centralized region or core area. This is because the polities
or groups vying for power each have their own ideologies. Early on, competing ideologies are
prevalent within emerging cores, because no one ideology has been established as dominant
yet. Although ideologies are manipulated and negotiated during centralization, it does not
necessarily imply that all ideologies are a part of this process.
As political centralization progresses, and the core becomes more consolidated, a more
widely accepted state ideology may exist. Resistance may be present in the core, however,
it is likely found among lower status elites or commoners, and would not be as visible overall
because elites are in control of the dominant ideology. As a result, one would expect to see
many similarities among official ideologies within the core. Similarities with ideologies beyond
the boundaries of the central area vary in possible patterns based on the cores ability to
project power over great distances.
The core, is first and foremost, limited by distance in its capacity to influence the periphery
(Stein 1998:162). Although the cores ideological power is much less subject to this because
of its high value and low cost of transport (Stein 1998:166), distance from the core to some
degree still limits the speed and movements of ideology to locations in the periphery.
Distance associated with natural geographic barriers can further hinder the cores ability to
establish its influence. Alone, geographic terrain can act as a limiting factor as well, causing
communities close in distance to the core, but divided by complex geographic features, to be
very dissimilar from the centralized region.
If the challenges of distance are overcome by the core, they are still faced with locally
based agency by polities or groups in the periphery (Stein 1998:155). Political dominance is
not merely obtained by the cores ability to exert power over the periphery, but is rather the
result of a choice by peripheral elites to participate in a relationship with the core and accept
those ideas. If the core is highly influential, then the peripheral regions may emulate its
ideology. In the face of political centralization, the peripheral elites may choose to do this as a
means of increasing their power. Rather than run[ning] the risk that a rival faction will do the
same to gain the advantage in the local political arena, peripheral elites will give up some
autonomy in exchange for the benefits of a relationship with the core (Beekman 2000:402).
If, however, the politically centralized cores degree of control over peripheral locations is
weak, then its influence should decrease greatly outside of the central area. In this scenario,

ideologies are either not shared with those in the periphery, or they decrease in similarity
between the core and periphery as distance from the central area increases. Discontinuity
between the ideology of the core and periphery also may indicate that ideology is used as a
means of maintaining autonomy or actively resisting elite sanctioned ideals (Brumfiel 1994;
Yoffee 1988).
When processes of political centralization are not present, one would expect to see few, if
any, similarities among the ideologies of different groups throughout a region. Should
ideology be shared by one or two of the regional variants, this might suggest it is used to align
different regions, or that the similarities simply reflect the interaction and social relationships
that existed among politically autonomous (kin or lineage) groups. It may be observed that an
ideology is unique to certain groups or areas within the region. This would indicate that there
are locally interpreted differences in ideology.
Mesoamerica provides an excellent opportunity to address processes of centralization
using a gendered approach. Pre-Columbian West Mexico specifically, offers not only an
abundance of well preserved artistic representations of gender imagery, but is also believed to
have undergone processes of political centralization concurrent with their manufacture and
use (Beekman 2007a; Beekman and Weigand in press). Through the consideration of how
gender is depicted in human imagery insight can be gained into the dynamic relations of
gender and politics at the local level, and the regional as well. Evaluation of gender within
regional variants of art focuses on individual community perceptions and strategies, while
comparisons among them illuminate pan-regional ideological patterns.

This study focuses on the western portion of Mesoamerica referred to as West Mexico.
Mesoamerica is one of the seven places in the world where civilizations were believed to have
independently developed. This major transition in human lifeways (and social organization)
included, but not necessarily in this order, the adoption of sedentism, the transition from
foragers and collectors to agriculturalist and then the intensification of agricultural practices,
the rise and institutionalization of stratification, and the creation of states and empires through
processes including urbanization, centralization, and specialization.
Although West Mexico possess many of its own unique characteristics, it does appear to
share in many of the Mesoamerican traditions including world views, symbolism and imagery,
and architectural types. Some have shown that Mesoamerican perceptions of the universe
have been repeated in not only architectural layouts in West Mexico, but also in ritual
practices and religious symbolism (Beekman 2003a, 2003b, 2007a, 2007b). Others have
identified the presence and use of ballcourts in West Mexico (Day 1998; Weigand 1985,
1991:76, 2000). Ballcourts in this region were constructed like others throughout
Mesoamerica, and it appears the game had similar rules and significance. Evidence of this
can be seen in the diorama models from West Mexico depicting figures actively participating
in and attending ballgames. Similarities in dress and costumes, as well as elite and religious
status symbols have also been recognized between West Mexico and the rest of
Mesoamerica (Anawalt 1981, 1998; Furst 1998; Graham 1998).
Study Area
Western Mexico borders the Pacific Ocean and is composed of the states of Nayarit,
Jalisco, and Colima. This region makes up a portion of the eastern edge of the Pacific Ring of
Fire, a highly active plate tectonic region in which earthquakes and volcanic eruptions are
frequent (Monroe and Wicander 1995:291). Situated directly upon the Middle America ocean
trench, a point where the oceanic plate is being slowly subducted below the continental plate,
western Mexico displays many characteristics of an extremely active geologic landscape
(Monroe and Wicander 1995:306-309). This is evidenced today by the presence of uplifted

volcanic mountain ranges, large obsidian deposits from ancient lava flows, and high ash
content in soils (Weigand 1985:55, 56). The coast is composed of alluvial material deposited
by river drainages from the mountainous regions further inland. Tablelands transition the
coastal plains into the mountain regions (Townsend 1998b: 17). Mesquite and acacia brush
cover these lower elevations, while oak and pine forests are characteristic of the higher
elevations. The mountains, which make up the majority of the region, are riddled with shallow
marshlands and dried lake basins. The combination of complex topography and a
considerable range of climates in West Mexico, has resulted in an abundance of plant, animal,
and mineral resources within the region (Townsend 1998b:16-17; Weigand 1985:55).
It was within a setting much like this, between 100 BC and AD 200, that the development
of a complex society was occurring in this region. This emerging culture has long been known
for their elaborate burial offerings which were interred in shaft and chamber tombs. These
tombs are entered through a long vertical shaft roughly three feet wide (Furst 1978:20), and
ranging anywhere from one and a half to twenty-one meters in depth (Christopher Beekman,
personal communication 2007; Butterwick 2004:13; von Winning 1974:16), with most
averaging between two and four meters (Butterwick 2004:13; Weigand and Beekman
1998:39). At its base, a narrow horizontal tunnel leads to one or two chambers opposite each
other, often oriented along a north-south axis (Beekman and Galvan 2006:260; Furst 1978:20;
von Winning 1974:16). Many of the shaft and chamber tombs conform to this typical boot-
shaped design (Beekman and Galvan 2006:260). Bodies within these tombs have been
found placed horizontally in the center of the chambers, or occasionally piled in corners, which
suggests the tombs were used either for multiple interments, or the burial of several dead
together at one time (Butterwick 2004:13; von Winning 1974:16). To maneuver the bodies
down the shaft, they were bound in mats, tied to a board, and lowered by a rope (Butterwick
2004:13), evidence of which can be seen in the human figural sculpture associated with the
tombs. After the bodies and offerings were in place, the shaft was filled with dirt, and a slab of
stone was placed over the entrance (Easby and Scott 1970:118). Although these entrances
were rendered invisible through this process, they were sometimes marked with a small
mound of dirt ora stone ring (Butterwick 2004:13; Easby and Scott 1970:118).
A wide variety of these tombs have been found in the lake and river basins of the
mountainous volcanic region forming part of the Sierra Madre Occidental, which describes a
wide arc from the Rio Grande de Santiago in the State of Nayarit, across central and southern
Jalisco, to the Rio Coahuayana in the southern boarder area of Colima and Michoacan

(Holsbeke and Arnaut 1998:61) (see Figure 3.1). Recently, shaft and chamber tombs have
been identified just north of Puerto Vallarta (Mountjoy and Sandford 2006), as well as in the
Bolanos River Valley, which extends from southern Zacatecas into northern Jalisco (Cabrero
1989) (see Figure 3.1). However, the most monumental shaft and chamber tombs, those
more than four meters deep and with multiple chambers, are located within the Tequila valleys
of central Jalisco (Weigand and Beekman 1998:39, 40).
Jalisco, which is centered between Nayarit to the northwest, and Colima to the southwest
contains the highest concentrations of human activity of prehispanic populations in western
Mexico (Weigand 1985:70, 2000:44). The greatest density exists within the Tequila valleys,
and diminishes as one goes beyond this rough geographic boundary. The Tequila valleys
extend from their most northwestern point, close to the Nayarit-Jalisco border and east of the
modern town of Ixtlan del Rio, to their most southeastern point, just beyond the modern city of
Tala (Weigand and Beekman 1998:37). These valleys are composed of many natural shallow
lakes and reservoirs surrounded by high volcanic mountains (Weigand 1985:54-55). The
combination of lacustral deposits and volcanic ash have provided the region with fertile soils
that have been exploited by humans today and in the past (Weigand 1985:55).
It is these high population concentrations in the valleys around the Tequila Volcano that
have been used to define this region as the core of sociopolitical changes occurring during the
Late Formative and Early Classic periods (Beekman 2000; Weigand 1985). The periphery,
then, encompasses all of the areas using shaft tombs beyond the mountains ringing the
Tequila Valleys (Weigand 2000:44). This includes the rest of Jalisco, the bordering states of
Nayarit and Colima, as well as, the western edge of Michoacan and the southern portion of
Zacatecas (Bell 1974; Cabrero 1989; Holsbeke and Arnaut 1998:61; Mountjoy and Sandford

Created from Cabrero 1989; Holsbeke and Arnaut 1998:209; Kan et. al.
1989:11; Mountjoy and Sandford 2006; Weigand 1985:65.

Early Views on West Mexican Sociopolitical Complexity
Studies in the past focused heavily on the ceramic art within the shaft tombs (Weigand
1985:51-52), but overlooked the associated monumental architecture and settlement systems
that covered the region (Weigand 1985:51-52). It was only with the work of those like Phil
Weigand, Arturo Oliveros, and Javier Galvan, in the late 1960s and 1970s, that the focus was
shifted and the complexity of these populations recognized. These individuals took the
emphasis off the use of decontextualized artifacts for data collection, and turned rather, to the
use of the contextualized archaeological record, which provided a more concrete and reliable
source of data. Since then, others have proposed models of a more complex social
organization existing prior to the Tarascan Empire (Beekman 2000; Butterwick 2004).
The work of these individuals has not only changed the traditional views of West Mexicos
prehispanic populations, but also their importance in the greater Mesoamerican tradition.
While current research is continuing to explore the intricacy of these populations, studies have
only just scratched the surface. Archaeological work in West Mexico over the last twenty
years has included settlement pattern studies (Weigand 1985), analyses of social organization
and degree of political centralization (Beekman 2000, 2003a; Butterwick 2004; Weigand
2000), evaluation of mortuary practices through burial excavation (Lopez and Ramos 1998;
Pickering and Cabrera 1998), analyses and interpretation of ceramic art (Butterwick 1998;
Day 1998; Furst 1998; Graham 1998; Townsend 1998c), an assessment of regional trade and
interaction with other Mesoamerican cultures (Anawalt 1998; Weigand 1992), and
examination of the relationship between the natural resources and human populations of the
region (Schondube 1998; Valdez 1998). While these studies have expanded the view of
prehispanic populations and contributed much knowledge, there is still more information that
can be obtained from the available data.
Current Perspectives on Sociopolitical Complexity
Part of the work done since the 1970s has been geared toward developing a more reliable
archaeological sequence for the region. Prior chronologies have been based on tenuous
connections between museum collections or looted materials in private collections, upon
isolated artifact types, upon similarities to material in other regions, and upon relative ordering
of materials with often little supporting stratigraphic and radiometric data (Beekman and
Weigand in press:1). Recently completed, this new sequence was created using only

excavated data (Beekman and Weigand in press: 1). This has untangled issues of
contemporaneity, and made it possible to make more confident statements about the
developments and changes that were occurring throughout the region (Beekman 2007a).
One instance in which this has been invaluable is in addressing the changes seen in the
archaeological record between the Late Formative and Early Classic periods (100 BC- AD
200) (Beekman 2007a:2; Beekman and Weigand in press:6).
The Late Formative is marked by intensive use of shaft and chamber tombs, and the
abundance of both local and imported mortuary offerings associated with them (Beekman
2007a:2). There is much variation among the shaft tombs (Beekman and Galvan 2006:260),
some of which may result from distinctions being made between high and low status burials.
The larger, deeper, and more complex tombs exist within the central communities (Beekman
2007b: 12), and are usually associated with ballcourts and a circular form of public architecture
known as guachimontones (Weigand 1996), in which a central plaza is surrounded by an
even number of rectangular structures (most commonly eight) (Beekman 2007b: 13). This
circular form of architecture is distributed in settlement hierarchies across the landscape
(Weigand 1985). The larger elite tombs are usually multi-chambered, contain multiple
interments, and include imported high status goods (Beekman 2007b:3-4). Simple shaft and
chamber tombs and pit burials are found in smaller communities surrounding the larger
settlements (Beekman 2007a:3-4). Burial goods in these simple tombs are fewer in number
and less elaborate than those in the centers (Beekman 2007a:4). Social hierarchy [is]
evident in tomb wealth, the varying sizes of surface architecture, and the dramatically different
scales across ceremonial centers (Beekman 2007a:4).
By the Early Classic shaft and chamber tombs are almost nonexistant. Those that are
found are smaller in size, less elaborate, and rarely associated with architecture (Beekman
2007a:5-6). Even the mortuary offerings have declined in numbers (Beekman 2000:399).
Overall, there is much less diversity between the tombs used throughout the region during this
period (Beekman 2007a:5). The rapid construction of the large architectural complexes that
was occurring in the Late Formative levels off during the Early Classic period, and
construction in the guachimontones is directed more toward maintenance, rather than growth
(Beekman 2007a:6). These changes in the archaeological record have been used to support
interpretations of sociopolitical organization at both macro and micro scales. Micro scale
models have focused on the local strategies of elites for power (Beekman 2000; Butterwick
2004), while macro scale models have evaluated the regional patterning in settlement

variations (Beekman 2000; Beekman and Weigand 1998; Weigand 1985).
Models developed prior to the new chronology and the discovery of the contemporaneity
of shaft tombs and circular complexes, have used settlement pattern models (i.e. core-
periphery and world systems models) to describe the more microscale workings and
sociopolitical organization of these populations. Models based on this top-down approach
have suggested the existence of a fairly centralized sociopolitical form of organization, which
is somewhat state-like in many aspects (Weigand 1985). This is most importantly indicated
by the presence of monumental architecture and hierarchical settlement patterns, but also by
the presence of specialization (i.e. social and artisanal specialization), intensification in
sociopolitical organization, agriculture, and population, and urbanization (i.e. 1)
institutionalized stratification, 2) presence of managerial elite, 3) increase of population into a
nucleated area large enough to sustain economic and political regime, and 4) control over
rare resources) (Weigand 1985:88, 89). Architecture is viewed as key to understanding
sociopolitical organization in this model because it embodies the culture, politics, ideology and
technology of a society (Weigand 1985:90).
Within this model the region is defined by a more classical core and periphery relationship
(as defined by Wallerstein 1974) that is linked economically through trade and exchange of
resources (Weigand 2000). The core region is defined by the presence of a unique
architectural style and higher population concentrations, and is held together by politically
controlled cermonialism, evident by the shared use of monumental architecture (Weigand
1985, 2000).
The ecological diversity of the Tequila valleys stimulated the development of the socially
complex core region during the Late Formative and Early Classic periods (Weigand 1985,
2000:43). During the Late Formative variation in the shaft tombs and associated surface
architecture suggested the emergence of complex social activities and social hierarchy
(Weigand 2000:47, 48). At this time, monumental shaft tombs symbolized power and prestige
throughout the region (Weigand 2000). The presence of formalized architecture suggested
the development of distinctive ideologies and elites to control them (Weigand 2000:49), while
the abundance of ceramic sculpture and vessels, as well as obsidian jewelry suggested the
presence of specialized craftspersons within the core region (Weigand and Beekman
The transition to the Classic period reflects processes of intensification, despite the fact
that the societies that existed between the two periods were quite different (Weigand 2000:

49-50). They are considered to be of a continuous cultural tradition though, because they
retained a continuity in expression within the same cultural heritage (Weigand 2000: 49-50).
During this period, monumental architecture replaced monumental shaft tombs as symbols of
power (Weigand 2000). In addition to the intensification of development there is also a
demographic implosion in the Tequila valleys (Weigand 2000:50). Communities in the
surrounding areas diminished, as their populations migrated to the core (Weigand 2000:50).
Settlements that remained at the outer reaches of the Tequila valleys were limited to locations
for valuable resource procurement (specifically salt and obsidian) or were fortifications at
passes to guard access to the core (Weigand 2000:49, 51, 53). This population influx into the
core region is believed to reflect the process of nucleation, in which the nucleus of a regional
social system is developed (Weigand and Beekman 1998:41).
The shift from monumental shaft tombs to monumental architecture, in conjunction with
demographic implosion, suggested the rise of social stratification within the core, organizing it
as either a single state or collection of states (Weigand 2000:50). The disappearance of shaft
and chamber tombs also likely suggested the discontinued use of ancestor veneration, as well
as the development of state defined positions of authority (Weigand and Beekman 1998:42).
The formalization of human imagery, in which the unique personalized characteristics of the
Late Formative figures was lost, further suggested the importance of the offices, or positions
people held, rather than the individuals themselves (Weigand 2000:50; Weigand and
Beekman 1998:42). Strain on the resources caused by the population implosion in the core,
along with environmental changes, required these populations to respond by intensifying
agricultural practices, through the development of terraced fields and chinampas (Weigand
1974, 2000). The core was defined as a political landscape that contained a hierarchal
pattern of settlements and habitation areas (Weigand 2000:54), which suggested it was
becoming urbanized (Weigand 2000:55). In this model, the more consolidated core shared a
state ideology, evident in the widespread use of first, monumental shaft and chamber tombs
and then, complex monumental architecture (Weigand 2000).
The more recent micro scale models focus on taking a bottom up approach to explaining
sociopolitical organization within the region. They examine local strategies and explain how
the activities of individuals produce the patterning seen at the regional scale. These models
suggest social organization was lineage-based. Two lines of archaeological evidence have
been used to support this argument. The first, is based off of studies examining the
construction of individual structures in the guachimontones (Beekman 2007b:26). It has been

found that the materials and techniques used in each vary between them despite the fact they
exist within a single architectural complex (Beekman 2007b: 18, 21,25). This is believed to
reflect the building of the individual structures by different lineages (Beekman 2007b: 18, 21).
The second line of evidence comes from osteological analyses of the human remains in the
Huitzilapa tomb (Pickering and Cabero 1998). The results of which show that the individuals
interred in the tomb were closely related, and may have belonged to the same lineage
(Beekman 2007b:9).
At the local level, lineages competed with one another for control over valuable resources.
Through self-aggrandizement during rituals, elites are argued to have either been seeking to
increase the amount of land they had ties to (Butterwick 1998, 2004), or the amount of human
labor they had access to (Beekman 2000).
In one scenario, the region was believed to have been composed of a hierarchical
network of competitive lineage-based chiefdoms tied together by mortuary feasting. Lineage
elites competed for power, by reinforcing their ancestral ties to the land. Members of founding
families held the highest status because they held the initial claims to the land and its valuable
resources (Butterwick 2004:11). Those who belonged to families who came later to the region
were afforded lower statuses as they did not hold priority claims to the land. In chiefdoms,
where power is inherited, showing ties to ones ancestors is essential in maintaining that
power. Group membership was determined by tracing successive generations back
to...founding ancestors (Butterwick 2004:11). Repeated use of certain places on the
landscape and of the shaft and chamber tombs reinforced ancestral claims to the land and
secured its inheritance for future generations.
Control of land was essential, because it possessed the very resources necessary for
feasting. Specifically, maguey agave, which was used to make octli, an alcoholic beverage
consumed in many ritual ceremonies in Mesoamerica (Butterwick 1998:103, 105). Feasting
was the elites source of power, as it not only allowed them to expand their social network and
establish political alliances (by means of reciprocal gift exchange), but most importantly, it
provided an opportunity for public ancestor veneration, which legitimized claims to the land.
Ancestor worship existed as a shared belief system and common ideology among the
different lineages (Butterwick 2004:11). The quantity and quality...[of the burial offerings]
placed in the tomb [during mortuary feasting] were a measure of the wealth and status of the
deceased and of their kin (Butterwick 2004:13). Aside from its ritual function, feasting...[was
a] mechanism for the redistribution, reciprocation,...[and] circulation of wealth and food

surplus (Butterwick 1998:90). As such, feasting was a setting for social, political, and
economic negotiation. The size of each chiefdom then, was determined by the local elites
success at feasting. As a result, some lineages were larger and had control over more land
than others. Construction of guachimontones over existing tombs containing ancestors
further emphasized these differential claims to land and surrounding resources (Butterwick
2004:23). Variations in size, not only of the architectural complexes, but also their individual
structures, reflected the relative rank and social status due to antiquity or allocated
resources between the lineages (Butterwick 2004:24, 25). It is this inequality between the
chiefdoms that is reflected on the landscape as a ranked and hierarchical settlement pattern.
The alternative scenario is that the sociopolitical organization of the region was based on
a corporate power sharing strategy (Beekman 2007b). Where political power lies in the size
of the social network (Beekman 2000:388). By increasing the size of ones social network, the
more people one will have to call on when they need them (Bender 1981). Power sharing
occurred among lineages at various social levels (Beekman 2007a:7). Essentially, power was
distributed among a number of different lineages and groups of lineages who competed with
one another for that power (Beekman 2007b). While competition for power existed, each
group was kept in check by a set of guidelines, an ideology, about how power could be
attained, utilized, and held, that was built into the social structure itself (Beekman 2007b:3).
This kept any one group from gaining exclusive power for a long time over the others. Based
on their success at self-aggrandizement different lineage elites or lineages would rise to
power for a short period of time, and then lose power to others who were better aggrandizers.
This can be thought of as a cyclical pattern in which different lineages and lineage elites
would rise and fall in power with respect to one another based on how successful they were at
During the Late Formative competition for power was primarily occurring between different
lineages within the lineage groups during mortuary ritual (Beekman 2000). Lineage elites
would compete for followers whom they could use to build elaborate tombs and manufacture
prestige (burial) goods, which in turn, would be used for future competition (Beekman 2000).
While competition may have been occurring in shaft tomb rituals, the community was being
reinforced in public agricultural rituals (Beekman 2007b:28). It is likely then, that competitive
mortuary performances were limited to a relatively private setting, as emphasis on community
cooperation would have discouraged it (Beekman 2000, 2007b).
Cooperation by the group was essential for ensuring agricultural success and community

survival (Beekman 2007b:27). By structuring the community around the guachimontones
that reflected their conception of the universe, each lineage knew their place within it, and the
importance of their participation in community agricultural ritual (Beekman 2007b:27). While
different lineages gained temporary power within the lineage groups, the social
complementarity that was established through these agricultural rituals, is what kept a single
lineage from becoming the dominant holder of power (Beekman 2007b).
With the transition into the Early Classic, competition between different lineages through
mortuary ritual appears to come to an end, and focus becomes placed entirely on competition
between groups of lineages at great distances through highly visible public agricultural ritual
(Beekman 2007a:6). This is likely the result of a change in ideologies and strategies for
supporters. Although...mortuary ritual [of the prior period] may well have provided prestige to
elites, it was probably not too effective at integrating larger groups at the community level or
beyond (Beekman 2000:394). To bring together the communities (i.e. initiating
centralization), lineage elites shifted the use of sacred space in the guachimontones from
local elite competition in mortuary ritual to that of regional participation in highly visible
agricultural ritual.
During the Late Formative the emerging core would have been characterized by the
competing ideologies associated with the mortuary rituals of different lineage elites (Beekman
2000). As the transition was made into the Early Classic period the more consolidated core
may have attained a more established ideology associated with public agricultural ritual that
was shared by the lineage groups (Beekman 2000). It is the combined strategies of lineage
elites for power at both the local and regional scales (i.e. competition within and between
groups) during this transition that gave rise to the core-periphery pattern and settlement
hierarchy that are recognized across the landscape (Beekman 2000).
It should be noted here that at the time this model was presented, the author believed
these ritual types (i.e. mortuary and agricultural) to be sequential, but today they are being
recognized as contemporaneous (Christopher Beekman, personal communication 2007).
These differences are now the result of changes in context, rather than chronological changes
(Christopher Beekman, personal communication 2007). Additionally, more recent articles by
Beekman (2007a, 2007b) show that his focus is shifting more towards corporate issues and
limits on power, and less on competition between elites.
If chiefdoms predominated during the Late Formative and Early Classic, then it is likely
that gender ideology would be quite similar throughout the region, as the chiefdoms were

unified by a shared world view and belief system (Butterwick 2004:11). Art integrated these
lineages by communicating the communal ideology associated with ancestor worship
(Butterwick 2004:12). The depiction of distinctly male and female figures in ceramic sculpture
pairs, believed to be lineage founders (Butterwick 2004:11), suggests that concepts of gender
were also shared among these groups.
If a more centralized form of sociopolitical organization existed instead, during this time,
then it is possible that gender may be shared throughout the core, and potentially with the
periphery region. A fairly consolidated core region would likely have established a dominant
state ideology. While ideologies do become more uniform within highly centralized regions,
it is not usually necessary to control or manipulate all forms of ideology. Therefore, if gender
is an ideology important to political centralization, then it should follow the general trend of
becoming a dominant ideology. However, if gender ideology is not essential to this aspect of
social organization, then it will likely be represented differently throughout the region.
Only within the earlier developments of political centralization would we expect to see
variation in gender ideology throughout the core. Within an emergent core, competing
ideologies always exist because no one ideology has been established as dominant yet.
Each lineage or lineage group would differentially express their own distinct ideologies, as
they manipulated them in attempts for power.
Ceramic Figures
At the heart of an unprecedented scale of looting, the shaft and chamber tombs of ancient
West Mexico have yielded an abundance of ceramic figural sculpture, most of which exists
today primarily in museum and private collections. The unique diorama models, and the
small, solid and large, hollow figures of this region, have been described as representative of
both the naturalistic and supernatural worlds (Furst 1998:169). The diorama models are
unlike any other artistic expressions of human imagery found within Mesoamerica. The
scenes illustrated in these models range in size from small scale activities involving two or
three figures, to ones that show large scale participation of entire communities. These larger
constructions depict ceremonial gatherings, such as ball games (Day 1998), feasting
(Butterwick 1998, 2004), and burial rituals, amidst detailed architectural structures. Besides
the diorama models, individual figures, of both solid and hollow construction, have been found
in association with the tombs. Individuals make up a large number of the human effigies in
western Mexico, and are often the most common themes within many of the substyles. These

figures portray, what are assumed to be mainly adult persons, in a diverse range of roles
including mothers, warriors, food preparers, musicians, priests, shaman, and ball players.
Also among the individual figural sculpture, are the highly debated hunchbacks and horned
figures (Furst 1998).
The third, and final form, of figural sculpture from ancient West Mexico are the paired
male and female figures. Although male-female ceramic pairs only exist within a few
Mesoamerican cultures, the depiction of males and females together in a single artistic
context is a common theme throughout Mesoamerica. Male-female ceramic pairs within West
Mexico standout amongst other pre-Columbian ceramic sculpture, not only because of their
unusually large size, but also the attention to detail in the construction and decoration that
extends to incising individual strands of hair, teeth, and nails on the toes and fingers of the
figures. Males and females alike display various symbols of status, such as body paint and
tattooing, clothing, and adornments, including headdresses, necklaces, leg and arm bands,
earrings, and even noserings. Males and females may exist as separate figures in these
pairs, however, many figures are physically attached into a single sculpture. The two figures
can be joined either side-by-side or back-to-front with one or two arms wrapped around the
other. Additionally, they are often attached at other places on the body, such as the torso and
The modern states of Nayarit, Jalisco, and Colima have provided the traditional style
distinctions for ceramic figures on the basis of broad scale stylistic differences (von Winning
1986:7). Due to the large degree of variability in not only figure design and construction, but
also their geographic location, this older typology has limited utility in defining the numerous
figure types. Current classifications, codified in the Chicago Art Institute volume Art and
Archaeology of the Unknown Past (Townsend 1998a), have acknowledged and attempted to
address this by subdividing the styles into smaller substyle designations based on more
minute stylistic similarities between the pieces. At the present time, there are eleven
substyles recognized throughout West Mexico: Ixtlan del Rio, Lagunillas, San Sebastian,
Ameca-Etzatlan, El Arenal Brown, San Juanito, Tala-Tonala, Zacatecas, Tuxcacuesco,
Coahuayana, and Comala (see Figure 3.2).
Current chronological sequences for central Jalisco suggest political authority was being
established within the region between 100 BC and AD 200 (Beekman 2007; Beekman and
Weigand in press:7, 8). The ceramic figural sculpture associated with the shaft and chamber
tombs date either nicely within the range, subsume this range within in their own range, or

overlap the beginning or end of this range by at least 100 years (Figure 3.3). Within the
Tequila valleys, or the core region, the figures are more precisely dated to this period using
archaeological excavation data (Beekman and Weigand in press). As such, it is expected that
the ideology displayed by the figures reflects this turbulent and unstable period when new
rules are being formed.
Through the examination of gender ideology in these indigenous representations, I hope
to gain an understanding of the role of gender in the transition between the Late Formative
and Early Classic periods. I intend to evaluate the possible scenarios of sociopolitical
organization proposed for the region by determining the kinds of gender relations that existed
within these populations during this time.

Figure 3.2.
1, Ameca-Etzatlan; 2, El Arenal Brown; 3, Coahuayana; 4, Comala;
5, Ixtlan del Rio; 6, Lagunillas; 7, San Juanito; 8, San Sebastian;
9, Tala-Tonala; 10, Tuxcacuesco; 11, Zacatecas. The relative substyle
locations were determined by compiling information from the works of others
who note the presence of specific figure substyles at different archaeological
sites and/or near modern cities and towns (Bell 1974; Butterwick 2004:32-35;
Cabrera and Lopez 2002; Dwyer and Dwyer 1975:39; Easby and Scott
1970:114,118-141; Gallagher 1983:15,23,25,39,41 -42,85-86,105-107;
Holsbeke and Arnaut 1998:135-136,209; Kan 1989:51; Kan et. al. 1989:11;
Lopez and Ramos 1998; Meighan and Nicholson 1989:32-40; Mountjoy and
Sandford 2006; Townsend 1998a: 12-13; von Winning 1974:8-12; von Winning
and Stendahl 1969:60-68).

Male-female ceramic pairs are particularly valuable to this study because they provide a
unique context in which the manipulation of gender ideology can be observed. Not only do
they reflect a broadly held emic perception of different gender forms by the artisans or their
clients, but also the socially accepted relationships between those genders in context with one
another. Additionally, regionally distinct styles of these pairs can be roughly placed within the
defined core and periphery areas allowing ideology to be compared between the two (see
Figure 3.4). As such, this particular artifact type allows gender ideology to be evaluated at
both the local and regional level.
In this study, the core substyles include Ameca-Etzatlan, El Arenal Brown, San Juanito,
and Tala-Tonala. The Zacatecas, Comala, Coahuayana, Lagunillas, Tuxcacuesco, Ixtlan del
Rio, and San Sebastian substyles represent the periphery. While the San Sebastian and
Ixtlan del Rio substyles could have been incorporated into either the core or periphery, due to
their transitional location between the two regions, I argue they are periphery substyles. The
placement of the San Sebastian type site on the very edge of the Tequila valleys near a
corridor to another outside valley, and the distribution of other pieces further northward along
the Nayarit-Jalisco boarder (Gallagher 1983:86), suggests this substyle is more likely
associated with the periphery region. Although the type site for the Ixtlan del Rio substyle is
located in southeastern Nayarit, the distribution of the figures extends over and along the
Nayarit-Jalisco boarder, and largely overlaps the San Sebastian figure distribution. The Ixtlan
del Rio substyles similar, but slightly more distant location from the core suggests this
substyle too belongs in the periphery. Additionally, the San Sebastian and Ixtlan del Rio type
sites are located on the western side of what would have been a large lake separating them
from the population concentrations around the Tequila Volcano, which are found on the far
eastern side of the lake. It is likely that interactions across this lake would be more difficult,
and as a result, would probably be less frequent. Therefore, influence of the core region on
these locations would be more reflective of those in the periphery. Although distance is
ultimately what makes the periphery substyles different from those in the core, transportation
limitations across the lake, even at a close distance, could also result in the exclusion of the
San Sebastian and Ixtlan del Rio populations from the core's influence.
Even though focus is on male-female pairs, simple visual observation of individual figures
and diorama models can also provide important supportive evidence. Individual figures offer
additional information on idealized depictions of symbols of status and identity with different
genders, and can be used to confirm or question different gender-symbol patterning observed

in the pairs. Diorama models, although they often do not depict enough detail to determine
gender relationships, provide a single instance of interaction in which multiple figures are
adorned in different attire actively using different objects. This can aid in understanding the
meaning and function of specific attributes displayed by the pairs.

Ixtlon del Rio
ac x>
*00 MO xn !> o 100 200 300 m soo x> no # soo iota itoo 12001000 mooimo
Son Sebastian
Son Juonito
(Buffer*** 2004; Omve* Museum of Hrtum ond Science 2006; Tovmimd >990)
(I rtropcrton ktoeum of Art 20C5)
(Gardner Museum of Corvmc Art 2005; To Humontas Oesewee Center 2005)
(UWt Museum f Fine Arts 2005)
(Hoffbeke ft Amort 199. Us An^rtJ County Hmw* of Art 2005;
(0mm Art feseun 200)
fljmor IfcMum {ConAnor lignum of Ceramic Art 2005)
(I rtUroiek 2004)
(Xoa d d. 1969)
{tat* 19
(HwSoom Museum 200$)
(8uffemk 2004)
(W4 1998)
I W4)
(Kubfcr 1966; Toanser* 1996)
- (Teufct 1988)
(Br*d 1929)
(Fast 1971}
(Oeyw k Dwyer 1975)
(Dockrtodg 1975)
(Meiftteim ft 1972)
(fwikeim at d. 1972)
I 199)
(Docksteder 1964)
(Buffniek 2004)
k Amod 19%)
(Tint 1978)
- (Dwyer t Dwyer 1975)
(r - (Bat 1974)
(Cosby 1970)
(Daw Art Muuem 2006)
(Hdsbekc i Amowl 1996. los Arqcto County Museum of Art 2005}
{Kan at of 1989; Ton stand 1998)
{Dwyer 6 Dwyer 1975)
. (kfuvwopoC* hetftde of Ms 20Q5)
(Townsend 1996)
(iWitilii k Amdat 199; Townsend 1998. von Mwwig 1974)
(Cokfalen 1988)
l 1986)
{Dwyer 6 Owyer >975)
(WesOwn el of. 1972)
.... {non Mnnirq 1969)
({ Hby 1970)
{HoMoka A Amort 1996. Km rt d 1969. lot Vxjrts County Museum of Art 2005)
- (The frtt*h Mwtaum 2005)
(Towneend 1998)
{Dwyer ft Owyar 1975)
* (von dnrwq 1974)
Hrrmq 1966)
(Dwyer ft Dwyer 1975)
(von afrunnq 1974)
(Buffer** 2004)
(Dwyrr ft Dwyer 1975)
(HobMke ft Amowl '99S. von Wtwrtq 1974)
(Denver Mumm* of Noture ond Scene* 2006; Hctsbefca ft Araout 1996)
(Ko *. d. 1989)
. (wr nn*9 197)
(Buttennck 2004; Townsend 1998)
400 500 200 > 00 0 100 200 500 400 500 (00 700 800 900 1000 1100 12001J00 1400 >500

1, Ameca-Etzatlan; 2, El Arenal Brown; 3, Coahuayana; 4, Comala;
5, Ixtlan del Rio; 6, Lagunillas; 7, San Juanito; 8, San Sebastian;
9, Tala-Tonala; 10, Tuxcacuesco; 11, Zacatecas. Core region delineated using
Weigand and Beekman 1998:37, Figure 3.

Controversy over Decontextualized Artifacts
There is much opposition to the use of museum and private collections for archaeological
study, especially when dealing with materials such as those from ancient West Mexico, which
are unquestionably looted and lacking provenance. Some consider the use of these
resources to encourage further collecting, looting, and forging (Braun 1993:253, 268). While
this link can always be argued to exist either directly or indirectly, archaeologists should not
ignore the abundance of information that can be gained from existing collections.
The work of museums and archaeologists have always been a double-edged sword,
where on the one hand they provide worldwide information to people that under normal
circumstances they would be unable to have access to, but then on the other in seeing these
rare and unique antiquities people are stimulated to collect them (Braun 1993:253). Should
we then cease to do archaeology because it encourages negative, as well as positive public
interest? I would argue against this, and suspect that many others would as well.
Long before archaeology and museums existed people were collecting antiquities to
satisfy their personal and social interests (Boone 1993b:315-318). Private collectors have
been promoting looting in and of themselves through competition with one another to improve
their collections for years (Braun 1993:254). The demand for new and more extraordinary
artifacts encouraged the development of a market to supply the needs of collectors (Boone
1993b:335) and in response, individuals and corporations arose to facilitate not only the
acquisition, but also the sale of antiquities. All of this has taken place without the involvement
of museums or archaeologists. Future looting and collecting will continue to occur without
their influence as well, simply because private collectors want to refine their holdings.
Although the initial development of museums admittedly furthered existing looting and
collecting by providing a means of income for dealers and local populations, especially in
poorer third world countries, most institutions today no longer participate in this due to ethical
issues of artifact ownership and acquisition (Messenger 1993:291-292, 295). A refocusing of
the discipline during the 20th century from describing] culture by its material
getting] at the laws of human behavior, to explain how and why cultures change" also
directed archaeologists away from collecting artifacts and objects of art (Boone 1993b:330).
Many of the concerns about using museum collections for archaeological research stem
from the fact that their holdings often contain private collections (which have promoted
looting). This is especially true of art museums who obtain their collection from collecting

those of others, as opposed to anthropological museums which acquire their collections from
fieldwork (Boone 1993b:336-337). Although private collectors do promote looting, they will
continue to collect regardless of whether or not a museum ends up with their collection. It is
better that private collections get to a place, such as a museum, where they can be available
and studied, rather than having them be kept hidden away. Knowing that museums only
purchase collections that have been legally acquired (Boone 1993b:337), individuals who
intend to leave their collections to museums are more inclined to purchase only legally
collected items, thereby eliminating participation in the black market, and potentially
decreasing the demand for looted goods. Additionally, private collectors often allow museums
and archaeologists access to their collections either for display or study. This is another
means by which materials typically unavailable to the public, are made accessible to them.
There is no denying that the collections currently held by both museums and private
individuals have been assembled in part through looting, however, we cannot dwell on the
wrongs that have been committed. Rather, what we should do is look at the positives that can
be drawn from these collections, like using them for scholarly analysis to improve our
understanding of the past, and strive for a future in which looting is eliminated. We should,
however, not push aside what can be gained from these collections, just because they were a
product of a behavior that is no longer considered acceptable.
Strides towards resolving one half of this equation have already been made. Laws
throughout the world governing acquisition, collection, and sale of antiquities have been, and
are continuing to be enacted, but this only addresses the future. What about the past? We
must draw something positive from these negative actions; we cannot in good conscience
ignore the multitude of artifacts that can provide information on our past. It is simply unethical
to not use these collections.
Using museum and private collections for archaeological study is not only a means of
preserving the past, but is also a way of maximizing the amount of information that can be
obtained from resources already removed from the archaeological record. By drawing off of
existing data collections archaeological artifacts...[that are]... nonrenewable resources [can
be] preserved" (Boone 1993b:336).
Theoretically this approach would minimize the unrelenting problem of site destruction
(Messenger 1993:308), by eliminating the need to tap the archaeological record to obtain a
dataset. While it is acknowledged that the archaeological record will continue to be destroyed
for the sake of other studies (as well as looting beyond our control), using extant databases

that have been ignored or forgotten reduces this occurrence and at the same time still
provides new information to the discipline. Additionally, making a first pass at a topic using
museum collections, even those lacking provenance data, can lend us to adopt more refined
and advanced research designs when we go to the field instead of duplicating research that
could be done without ever disturbing a site. In this way museums can again be the
fountains of research they once were, and their collections...the sources of knowledge they
were originally intended to be (Boone 1993a:5).
refocusing of archaeology [in the 1960s] had a profound effect on how
museum [and even private] collections were viewed. Where earlier
archaeological collections were the foundation of the discipline, they became
by the 1970s almost an embarrassment, a reminder of previous excess.
Many of these collections were haphazardly gathered, or they were gathered
when less attention was paid to details of association-at a time when
numbers were more important than precise location. Lacking secure
contexts, the artifacts had been robbed of much of their meaning.
Archaeologists thus did not value their study as before (Boone 1993b:330).
This perception that artifacts are incapable bearers of information, especially when they have
no secure provenance (Boone 1993b:336), has made many hesitant to utilize the artifacts of
ancient West Mexico for archaeological interpretation, and those (Furst 1974, 1998) who have
attempted to draw information from these pieces have been criticized for doing so (Robert
Pickering, personal communication 2005). Recently, archaeologists of this region, in following
art historians, have come to recognize and embraced these objects of art [as] worthy to
study (Boone 1993b:331,336). They are returning to artwork now that it can be
contextualized within the social reconstruction offered by archaeology (Butterwick 1998,
2004). Some have already spent much effort in the field trying to place these objects...within
the context of a more developed anthropological discussion as to their cultural role in [these
ancient] societies (Weigand and Beekman 1998:38).
The approach taken in this research falls well within this trend as there is a focus on
rescuing contextual data by looking at paired ceramic figures that should have been found
together. Despite the fact that looting has removed the pairs from their original context, the
two figures are almost always kept together. Looters make a conscious effort to maintain
matched pairs as they hold a higher value in the black market (Christopher Beekman,
personal communication 2004). In seeking a greater profit, the looters greed has
inadvertently allowed many of the figures to remain associated with their mate in collections

today. It is believed that while these artifacts may lack provenance they retain an aspect of
their original context by remaining together as they would have been in the archaeological
The context provided by each figures association with the other in the pairs allows us to
obtain idealized emic perceptions of gender, which in turn, can be used to make statements
about the sociopolitical organization in the areas from which these figures come from. The
very nature of these figures has given us a unique opportunity to take an intimate look inside
the minds of these ancient populations, that cannot be diminished by the effects of looting.
Even in locations where looting and forging are not an issue, such information can rarely be
The use of decontextualized artifacts for archaeological analysis not only makes many
scholars nervous, but also stimulates questions regarding the authenticity of the pieces
involved. Most people readily accept artifacts found in provenance to be authentic, however,
they are less eager to do so with objects that have been removed from their original locations
by looters. Unfortunately, looting has become so prevalent in some regions that the only way
archaeologists can study past cultures is through the examination of these artifacts in private
and museum collections. To allay concerns of authenticity in these situations, a number of
methods can be employed to evaluate the genuineness of the artifacts.
Some people prefer scientific approaches that address authenticity at the microscopic
level, by evaluating the nature of the materials used either on or within the object, and the
chronological sequence of their deposition or relative age. Two scientific methods, in
particular, have been applied to the West Mexican figures. Both analyses involve examination
of the deposits frequently found on their surfaces. Many figures from West Mexico, because
they come from a burial context, are covered in various types of insect and mineral deposits
that can be used as means of determining the authenticity of pieces. Manganese dioxide
deposits show up as black spotting that ranges in size from small dots to large patches that
cover much of the figure (von Winning 1974:77). Often forgers apply black paint to their
reproductions to make them appear more authentic. Black spattering on figures can be
scientifically tested to determine if the deposits are genuine or if they are paint markings.
Insect pupa may also be present on figures, which is typical of objects that have been buried
for long periods of time (Pickering and Cuevas 2003). Microscopic analysis can determine if

such deposits are present (Pickering and Cuevas 2003). Forgeries, then, do not display
evidence of insect activity on their surfaces.
While scientific methods may provide more reliable indications of authenticity, they cannot
be applied in every case. Scientific methodologies require specific materials and conditions
for testing, and often involve much time and money to perform the analyses. Additionally, it
may be impossible to determine an artifacts authenticity using scientific methodology
because it can require partial or complete destruction of the piece. As a result, many
determinations of authenticity of artifacts are based on the visual observations (at the
macrolevel) and judgement calls of professionals, who are familiar with the pieces in question.
This is not always the most reliable way of determining authenticity, but it is by far the least
expensive and time consuming. It also does not require physical access to the pieces, nor
any sort of destructive testing processes.
Any evaluations of authenticity on my part in this analysis were performed in this manner,
due to the benefits mentioned. However, these occasions were rare as authenticity was
typically addressed by the various source locations from which I obtained the figures for the
dataset. Books and other publications usually discussed the issue of authenticity in the
foreward or prologue, noting whether the pieces were authentic and what form of
methodology was used to make this determination. Unless stated otherwise, it was assumed
that pieces in museum collections were authentic. These collections included both those
accessible online, and at the Denver Museum of Nature and Science and the Denver Art
Museum. The West Mexican collection at the Denver Museum of Nature and Science had
been previously analyzed for authenticity by mineral and insect pupa analysis (Pickering and
Cuevas 2003), so they had a fairly detailed recorded on the authenticity of their pieces.
Additionally, the Denver Museum of Nature and Science possessed detailed photographs and
authenticity evaluations for a number of pieces that existed in other museum collections that I
had access to online. The Denver Art Museum too had quite extensive information not only
on the authenticity of their pieces (the method of testing was not noted), but also on any
reconstruction or damage that may have occurred to them as well. The only time I was
required to make a determination on authenticity was when no notation was made in a
publication or for online antiquity dealer collections. While I made the final decisions, I did
consult with Christopher Beekman, who is familiar with the ceramic figures of the region, on
their authenticity.
Although I have taken the upmost care in attempting to include only authentic figures in

this analysis, there is always the possibility that some may be present in the dataset due to
incorrect determinations on either my part or that of others. However, even if some of the
pieces in this dataset are in fact forgeries, due to the nature of my analysis, they should not
badly skew the results because the reproductions are modeled off of real figure types and
designed with the intention of being mistaken for real ones (i.e. use same colors, patterns,
poses, objects, and adornments as authentic pieces). Good forgers are familiar with the
attributes that are typical for not only the different styles, but the various types of individuals
that are depicted. No good forger would deviate from the expected patterns if their intent is to
sell the figures as authentic ones. Similar materials, construction techniques, and surface
treatments (i.e slips) would be used in the manufacture of forgeries. The same amount of
attention would also be payed to figure detailing, including the poses, gestures, adornments,
and objects associated with each individual. In fact, it has been noted that there is such a
high level of expertise in faking...[West Mexican] figures, it is often only when there are
inconsistences in the iconographic patterns that the pieces can be identified as forgeries (von
Winning 1974:76-77). As such, I should be able to assume that the attributes I am concerned
with for this analysis are similar on both authentic and forged figures. Additionally, I am
pursuing a statistical analysis that focuses on samples and not on individuals. These tests
are designed to detect the patterns exhibited by the majority of the sample. Individual cases
that are an exception to the rule, while they are still factored into the analysis, are so small
statistically they are not visible in the results. Therefore, if a few of the pieces are not
authentic, they are likely to be undetectable.
The following chapter outlines the methods applied in this analysis. The first sections
introduce the dataset by focusing on where the figures came from, and the kinds of individuals
represented in the pairs. This is succeeded by a detailed presentation of the types of
information collected on the figures. Finally, the chapter closes with a discussion on how
gender was determined.

Data Collection
Three hundred and seventy-four (374) male and female figures, or one hundred and
eighty-seven (187) complete pairs, comprised the dataset for this analysis. Eighty-seven
percent of the sample (162 pairs) belonged to museum and private collections throughout the
world and were viewed in photographs in books and online. Collections at two local
museums, the Denver Museum of Nature and Science, and the Denver Art Museum provided
figures that made up five percent of the sample (10 pairs). The close proximity of these
museum collections not only offered a first hand look at the figures, but also reduced time and
money expended in data collection. Figures not found in museum collections or in scholarly
publications, were acquired from online inventories of art galleries and antiquity dealers.
Specimens from these sources made up eight percent of the total sample size (15 pairs).
Most of the male-female pairs represent fully gendered adult figures who are in the prime
of their reproductive lifecycle. To become fully gendered, physical differences observable at
birth had to be transformed...into socially interpretable forms (Joyce 2000b:480) Through
processes of socialization and materialization the adults depicted in the pairs have come to
reflect the socially accepted gender types (Joyce 2000b:481). This is evidenced by their
ability (i.e. depicted with children) or potential ability to reproduce (i.e. shown with mature sex
characteristics), and the displaying of hair adornments, costume, and ornamentation distinct
to adult social identities, specifically reproductive and labor roles (Joyce 2000b:474, 476).
It is not uncommon to find the presence of young children or infants within this paired
context. Youth are easily distinguished from adults as they typically display minimal markers
of identity (Joyce 2002.82-83), such that they may wear only a simple headpiece, a single
piece of clothing, or have some body tattooing, but have no discernible sex characteristics. In
other instances, children do not display any signs of status (i.e. body adornment or clothing),
but instead are depicted with underdeveloped sex characteristics, indicating that they have
not yet reached sexual maturity, but have already begun puberty. This process of becoming
gendered through the course of ones lifecyle is also evident among the Aztecs (Joyce

2000b). Joyce (2000b:474) suggests the Aztecs believed humans are born without a distinct
gender. By ritually emphasizing their physical differences during their pre-pubescent years
children become socialized into fully gendered individuals by the time they reach puberty.
Prior to this transition children would only possess bodily modifications suggestive of what
their gender would become, but would not obtain a full set of distinctive markers until an adult
status was achieved.
In addition to this, figures of children among the West Mexican pairs are smaller in size
than the other figures, and are often held by one of the larger individuals. Figures that can be
confidently identified as children are found with eleven (11) of the pairs (see appendix A,
Figures A.3, A.16, A.17, A.42, A.82, A.116, A.122, A.141, A.145, A.146, A.162) in this dataset.
Interestingly, within this dataset children tend to be found primarily within the Ixtlan del Rio
and Ameca-Etzatlan substyles. Only one (1) child is found within the context of the pair in
each of the Zacatecas and Comala substyles. In four (4) other Ameca-Etzatlan pairs (see
appendix A, Figures A.10, A.35, A.94, A.145 (146B)), and one (1) Ixtlan del Rio pair (a third
figure not included in testing, but part of a numbered pair), it is possible that one of the two
figures is an individual who is approaching a fully gendered status. This is, again, evidenced
by their slightly smaller size and lack of multiple status markers. However, they possess more
markers than the infant and toddler staged figures, and have achieved an adult-like
appearance, making them more identical to their counterpart and less like a child.
Recognizable children were excluded from statistical testing in this project, however, those
figures that were almost fully gendered were included.
On the other end of the spectrum, it is worth noting that five (5) of the pairs (see appendix
A, Figure A.9, A.90, A. 108, A. 129, A. 157) are clearly of an elderly age, seen in their aged and
wrinkled faces, hunched over body posture, and visible ribcage. Among the Ixtlan del Rio
substyle, which included all but one of the elderly pairs, female figures are always shown
holding some type of food over a bowl on the ground, while males always hold a tube to their
mouth that extends into a pot held at their feet. It is likely that these males are partaking of an
alcoholic beverage, such as pulque, which among the Aztecs is a privilege given only to the
elderly who were allowed to become intoxicated. Additionally, both males and females are
repeatedly associated with similar body tattooing emphasizing their ribcages. The only other
elderly couple in this dataset belongs to the Lagunillas substyle (see appendix A, Figure
A. 108). This pair shares the primary indicators of aged individuals listed above, however,

they differ in that it is the male who holds the bowl rather than the female, and the female in
this case, is holding no object at all. All elderly couples were included in statistical testing.
A second, possible age transition visible in this dataset, is that of individuals who are past
their reproductive years, but are not yet considered elderly. Three (3) figures (see appendix
A, Figures A.48, A.83, and possibly A.112), all from the Ixtlan del Rio substyle, may be
reflecting couples in this part of their lifecycle. This is evidenced by their possession of
attributes of both reproductive adults and elderly. These couples are adorned and carry
objects associated with the former, and are posed in positions characteristic of the latter. All
three of these couples were included in statistical testing.
Digital images were captured of all but seven pairs (160A&B, 161A&B, 162A&B, 167A&B,
168A&B, 169A&B, 175A&B) and incorporated into a visual database. Images not already in
electronic format from the internet and digital photographs, were scanned and converted into
digital images (i.e. photographs in books and film photographs taken of DMNS and DAM
figures). All digital images were closely examined in Adobe Photoshop 6.0. Specific photo
viewing capabilities of this software, such as zooming, contrast and color adjustment, etc.,
allowed for maximum identification of selected attributes.
Ideally, every known male-female pair should be included in this analysis, however, there
were several factors that made this impossible. Time and money were important concerns,
but access to the figures was the most controlling factor. Not all collections were available for
viewing, as many belonged to private owners and were usable at their discretion; some
collections remained undiscovered; and even those collections available to the public required
time and money to examine. While male-female pairs are housed in many other locations, it
is believed that those sources of data were effectively tapped through the use of photographs
in books and via the internet. It must be stated though, that this approach was not without its
problems. Access to various collections through photographs in books and online were
limited not only in their availability to the investigator, but also in the information they provided.
Angles at which pictures were taken, use of detail shots, and the quality of the photographs
varied based on the technology available at the time and the experience of the photographer.
Pairs were either included or excluded from the final dataset based on their accessibility and
ease in clearly identifying a number of key attributes. Even though photographs posed
limitations on what parts of the figure were visible, it was still valuable to include figures in
which only some attribute data could be collected. For example, figures in which tattooing

and clothing designs could not be determined, provided information on other attributes such
as biological sex, head adornment, objects held, and so on. Therefore, while it was not
possible to collect all attribute data on every figure, the data that was gained on the visible
attributes was still of great value to the analysis. Examining only attribute data available from
a number of different pairs did not affect the statistical analysis negatively because the focus
was on the patterning seen among the attributes and not the individual figures themselves.
Knowing that a one-hundred percent collection of attribute data from each figure was not
always realistically possible, coding included a notation for missing data. Missing data were
defined as attributes that were undeterminable, unidentifiable, rubbed off, broken, or for
whatever reason were missing.
General Figure Information
For this project it was necessary to assign new figure numbers to all pairs in order to
create uniformity in the dataset because specimens were compiled from multiple collections.
A figure number was given to the pair, and each figure within the set was arbitrarily
designated a and b. The location or published source from which the specimens came
from was also noted, along with the original numerical designations given to each figure. This
allowed for easier identification of specimens during examination. Besides recording specific
attribute information, to be described later, information on the figures associated style and
substyle, relative date range, figure type, figure dimensions, figure authenticity, and pair
attachment also were noted for each specimen.
Style and Substyle
Both style and substyle were recorded for each figure, however, it was the substyle
designation that was of most value for this project. Not all substyles were equally represented
in the sample, and since substyles are to some degree tied to the larger state defined style
regions, this pattern was repeated in the styles as well. Tables 4.1 and 4.2 show a
breakdown of the styles and substyles for the total sample. Many factors contributed to the
uneven representation of figures. For instance, not all areas in which figures are found have
been extensively explored (e.g. not much human traffic through the area, or not much
archaeological work done in the area); certain substyles may have been better constructed
than others and they simply lasted longer (e.g. better preservation of some more than others);

certain substyles may have been produced in larger quantities than others; collectors may
prefer one substyle over another for their aesthetic and/or monetary value and it is those that
show up more often in private collections, museums, and dealer inventories; or it may simply
be sampling error, such that the investigator only had access to sources in which there was a
heavier representation of certain substyles. These are just some speculations, but why
certain substyles are more abundant than others in the overall sample is not known for
certain. There is however, a definite under representation of certain substyles to which the
public has access. In addition, some of the figures style and/or substyle designations were
impossible to determine, or were unknown. Both the source location and this investigator
were unable to place the figures into one of the style and/or substyle types. These figures
were included in the Unknown count of the sample in Tables 4.1 and 4.2.
Table 4.1. Number of samples by style, n = 374*
Nayarit 206
Jalisco 148
Colima 16
Michoacan 2
Unknown 2
* Sample counts are based on individual figures and not on pairs
The implications of uneven figure representation on statistical analysis, in general, are
that sample sizes too small not only yield less reliable data, as they are not an accurate
representation of the population, but they often cannot be tested and must be eliminated from
analyses. Although the sample size for some substyles was too small for the more refined
analyses, it was possible to incorporate them into those analyses that lumped several
substyles together.

Table 4.2. Number of samples by substyle, n = 374*
Ameca-Etzatlan 62
El Arenal Brown 2
Coahuayana 2
Comala 10
Ixtlan del Rio 162
Lagunillas 10
San Juanito 18
San Sebastian 30
Tala-Tonala 18
T uxcacuesco 14
Zacatecas 30
Unknown 16
* Sample counts are based on individual figures and not on pairs
Time Period
Dating of the male-female pairs is somewhat problematic as they are not often recovered
in their original context. The majority of figures that have been collected were looted and sold
on the black market, and not systematically excavated by archaeologists. As a result, most
figures have not been directly dated, but are rather assumed to date to a certain period based
on the regional style or substyle they belong to. The dates for these periods have been
determined by art historians, and are not based on archaeological evidence (Christopher
Beekman, personal communication 2006). Some of the figures found in context have been
associated with radiocarbon dates (Bell 1974), and those dates tend to correspond with the
ones given to the different substyles by art historians (Christopher Beekman, personal
communication 2006). While the figures only roughly date to a 400-500 year time span during
the local Early and Middles Tabachines phases (also known as Tequila II and III phases)
(Beekman and Weigand in press), which correspond to the Late Formative and Early Classic
periods of the surrounding areas, the general contemporaneity of the substyles allows them to
be successfully used to address sociopolitical organization during this period.

For data collection, time period was simply the relative date that the source location or
publication associated with the piece. This date is usually based on the current chronology of
the time for the region at the date of publication or incorporation into a collection. Dates for
some of the figures in books came from references in the associated text or from timelines
published within them. Figure 3.3, in the previous chapter, shows the variety of date ranges
given for the figures used in this project.
Figure Type
Figure type is a general measure of the relative size and construction of each figure.
Hollow refers to figures of an overall larger size, greater than 13 cm (5.07 inches) in height,
and of hollow construction. Solid refers to figures of an overall smaller size, less than 25 cm
(9.75 inches) in height, and of solid construction. There is some dimensional overlap to these
designations, however, stylistic differences between hollow and solid figures is so great that it
is usually very obvious to which category they belong. There were only a few instances in
which specimens fell within this overlap and were difficult to determine whether they were of
hollow or solid manufacture. Photos did not allow for this information to be collected on the
figures, so decisions were made as to which category they belonged based on what was
typical for figures of their particular substyle.
Figure Dimensions
Figure dimensions of height, width, and depth were recorded when possible, in both
centimeters (cm) and inches (in.). In some instances, figure dimensions were only attainable
in inches and required a conversion into metric. Therefore, providing both metric and imperial
dimensions allowed the actual figure measurements to be retained within the data set. In
addition, having dimensions in centimeters and inches provided a frame of reference from
which to grasp the relative size of the figures regardless of which form of measurement one
was familiar with. For some figures, all dimensions were recordable, as they were listed by
the source location, or were physically measured by the investigator. For others, however,
only height, and sometimes, width measurements were able to be recorded. Height
measurements were most commonly listed for figures. There were only a few figures for
which no dimensions were noted at all.
For all figures found in photographs, the dimensions were the ones recorded by their

original source location or publication. For the several specimens which were observed
hands-on, these measurements were taken by the analyst of this project using a ruler. Ruler
measurements, while less exact in nature, were sufficient enough to make figure distinctions
for this project. While it was recognized that additional and more precise measurements
could have been taken, only approximate measurements of the figure dimensions were
important for addressing the questions of this analysis.
Figure Authenticity
Figure authenticity was determined either by the original location or publication source or
through observations made by the investigator of this project. Possible responses for figure
authenticity included the following: Authentic {A), Questionable Authenticity (QA), Suspected
Forgery (SF), No Data Available (NDA), Believed to be Authentic (BA), Reconstruction (R),
and Question Pairing (QP). The first five responses were considered to be required (i.e. A,
QA, SF, NDA, BA). That is, it was necessary to determine and note one of these responses
for each figure during data collection. The latter two responses, Reconstruction and Question
Pairing, were optional, and only noted if information was provided. As such, these responses,
when present, supplied additional information on a figures authenticity. Authentic means the
figure has been evaluated specifically, by others and not the analyst of this project, using
some method of analysis to determine its authenticity. Method of analysis refers to
something as simple as visual observation of the figure by a scholarly individual familiar with
the styles of the region, all the way to examination of figure construction materials, and
mineral and insect pupa analysis like that performed by Robert Pickering and Ephram Cuevas
(2003) on the West Mexican collections at the Denver Museum of Nature and Science.
Questionable Authenticity means that while the figure was evaluated, either by others or the
analyst of this project, its authenticity was difficult to determine. Suspected Forgery means
that based on visual observation, by the analyst of this project, the figure seems out of place
from what is typically characteristic of the substyle. Visually these figures do not conform with
what was often observed to be the norm for that particular substyle. No Data Available means
that authenticity was not addressed by the source location and that visually, when observed
by the analyst of this project, the object appears to be authentic, but the analyst of the project
did not have access to the specimen to observe it further. The assumption here with the No
Data Available response is that authenticity was NOT addressed because the figure was

simply believed to be authentic by its original source location. The difference between
Authentic and No Data Available was that Authentic was proven and information was stated
that the figure was tested and found to be authentic. No Data Available means no information
was put forth as to whether or not the figure was tested, as such the figure is assumed to be
authentic because its lack of authenticity was not stated. All evaluations for the above were
made by someone with scholarly background. Claims of authenticity by art and antiquity
dealers were not considered to be scholarly in nature. As such, an additional authenticity
response was added, Believed to be Authentic. This notation only applies to figures found
within the context of an art or antiquity dealer, auction house, etc., that required the figures
authenticity to be judged solely by the analyst of this project and determined based on their
experience with the West Mexican figures.
Due to some contextual issues, the No Data Available response was broken down into
two responses. The first, retained the No Data Available title, but was defined differently than
before. By this new definition, the No Data Available response only applies to figures found
within the context of any accredited scholarly institution (i.e. museums, universities, etc.), or
publication funded by one of them. The second response, Believed to be Authentic, was
selected for use when figures were found within the context of an art or antiquity dealer, or
auction house.
These two responses are similar in that authenticity is not addressed by the source
location, however, when visually observed by the analyst of this project the figure appears to
be authentic. More extensive analysis of these figures was impossible as the analyst did not
have access to the specimens. The responses differ, however, in how authenticity is defined.
The assumption here with the No Data Available response is that authenticity was not
addressed because the figure was determined to be authentic by individuals associated with
its original source location. While each institution has its own mission statement and
acquisition policies, they would not knowingly permit forgeries to be displayed to the public or
placed in a publication associated with their institution. Should such an instance arise in
which a forgery was placed on public view, or determined to be an imitation, only after it had
been displayed, either in an onsite exhibit at the institution, or in an online collection, a
notation regarding the objects inauthenticity would accompany it, or it would be removed. As
such, authenticity for the No Data Available response is assumed to exist based on the
general goal and nature of such scholarly institutions to further public knowledge and protect

cultural resources. The Believed to be Authentic response was developed because claims of
authenticity by art and antiquity dealers were not considered to be scholarly in nature, and
were therefore, unreliable determinations of authenticity as defined by this project. Dealer
assessments of authenticity cannot be relied upon as their goal is to make money, often
without regard to resource protection or public education. Figures found in antiquity dealers
inventories and on the auction block can only, at best, be believed to be authentic based on
the visual inspection and experience of the analyst of this project. While the Believed to be
Authentic response was used primarily for figures in dealer collections, there was one
exception to this use for figures found with in a specific scholarly context in which neither of
the authors of the various essays nor the editor of the catalogue notes [had] seen the objects
in reality. The descriptions [were] based on photographic material and...therefore [they gave]
no guarantee of [an objects] authenticity" (Holsbeke and Arnaut 1998:61). However, they did
state that they had full confidence in [those], who...made the preliminary selection of the
pieces for the publication (Holsbeke and Arnaut 1998:61). It was due to these conditions that
the Believed to be Authentic (BA) response was selected to replace the Authentic (A) one in
this particular case. In the dataset, authenticity was broken down into two columns. One
reflecting the required authenticity notation, and the other, an additional optional authenticity
notation. This was done so the required notation could be included in statistical tests, while
excluding the optional notation from the analyses, but not the dataset.
Reconstruction is simply another notation used in addition to the general authenticity
notation which acknowledged that some part of the figure had been reconstructed. Question
Pairing is also an additional figure notation tied to authenticity which means that based on the
visual observation, by the analyst of this project, it is believed that the pairing of the two
figures may be artificial, or that the figures visually appear to belong with one another and
have been placed as such, but the analyst of this project was unable to verify the authenticity
of the pairing. Artificial" pairing was done after the fact, either in the past, or today, by others
who were not part of the initial conception and construction of the figures. As opposed to
intentional pairing, desired for this analysis, which is the purposeful design and construction
of the figures together by their original artisan. This notation is most valuable for pairs in
which the figures are not attached. It is assumed that figures of an attached pair were
intentionally placed together during design and construction.
Most figures were considered to be authentic either under the Authentic, Believed to be

Authentic, or No Data Available notation, however, a few were deemed to be of questionable
or unknown authenticity. My own personal observations lead me to suspect that some sets of
figures were either forgeries or artificially paired. As a result, certain pieces were flagged as
a means of marking highly suspicious specimens that may potentially skew statistical results.
Atypical appearance and/or pair context made these pieces standout as especially suspect
with respect to the rest of the dataset. Flagging made for easy re-identification and selective
exclusion of specific objects during statistical analyses.
Pair Attachment
The Pair Attachment category makes note of whether or not the two figures of the pair are
physically attached to one another (i.e. built as a single piece). Attachment of figures is an
indication as to the intentionality of the pairing as described in the previous section. The more
assurance that the figures were meant to be a pair, the more confident one can be that the
relationship between the two figures is an emically perceived one. As such, we can be
relatively confident of our interpretations of gender ideology and how it is viewed in context
with other gender forms.
Symbols of Identity and Status
The appearance cues used to mark individuals for gender, as well as other forms of
social status and identity within ancient Mesoamerica, are visible in a wide range of sources,
including text and pictorial documentation of sixteenth century Spanish conquistadors and
missionaries (Joyce 2002:81), and the material culture and assemblages associated with
burials (McCafferty and McCafferty 1994, 2003). However, they are most apparent within
indigenous artistic depictions of human imagery in figural sculptures (Joyce 1993, 2002),
monuments (Joyce 1996), and codices (Anawalt 1981, 1998; Troike 1982).
Within these different contexts, males and females, who can often be identified by there
exposed genitalia or gendered dress, are shown engaged in variety of activities (usually in the
company of others). Objects held by figures, or found in association with them are believed to
reflect specific gender roles, and other social statuses and positions held by the individuals
(Joyce 1993:26; McCafferty and McCafferty 1994, 2003). Figures identified as male tend to
be associated with weapons and religious paraphernalia, such as shields, axes, staffs, and
musical instruments (Joyce 1996:179; McCafferty and McCafferty 1994:145), that are believed

to reflect their roles in warfare, hunting, and ceremonial activities (Joyce 1993:260, 261).
Female figures, however, are shown involved in either productive or reproductive tasks (Joyce
1993:260). Women are depicted pregnant, nursing, or caring for small children or infants
(Joyce 1993:259, 261, 1996:188, 2002:89). They are also associated with spinning and
weaving implements, namely weaving battens, combs, picks, spindle whorls, and spinning
bowls (McCafferty and McCafferty 1994:148, 2003:42), as well as food processing tools,
cookware, and containers of food-like materials (Joyce 1993:261, 1996:188). Women are
commonly depicted in a kneeling position which emphasizes their role as producers of food
and cloth, because both weaving and grinding corn, and even tending the hearth, are
performed in this position (Brumfiel 1996:158-159; von Winning 1974:53). The kneeling
position is also believed to be indicative of pregnancy, as this is a position that women
assume to give birth (von Winning 1974:66). Females have also been shown to assist males
in ritual activities (Joyce 1993:261).
Symbols of identity and status are visible on the figures of the ceramic pairs as attributes
which reflect the social roles and positions held by the individuals (Butterwick 1998:96). The
interconnection between how figures are posed, how they are adorned, and what they are
holding indicate the various identities held by each gender. Displayed in the material culture,
these positions and roles depict the stereotypical ideals embodied by different genders.
Additionally, the depiction of the two genders together in one context, reveals idealized
relationships between genders.
To guarantee this information would be interpreted properly it was crucial that it be shown
obviously and distinctly from all other attributes within the artistic context (Troike 1982:177-
178). Therefore, the communication aspect of [art must] always be the artists prime
consideration, with all other expressions of art being of necessity subordinate to it (Troike
1982:178). For proper interpretations to be is essential that all the important
features be large enough and have sufficient detail to be seen clearly by all viewers (Troike
1982:178). With the main purpose of communicating information, attributes found on artistic
depictions will sometimes override...considerations of realism and require that the size or
details be enlarged or exaggerated until all the necessary information was plainly displayed
to viewers (Troike 1982:178). Since consistency is crucial to...communication (Troike
1982:185), repeated associations between certain attributes and biological sex would likely
indicate stereotypical perceptions of the roles and positions afforded to different genders.

Body Poses and Gestures
Body poses and gestures are often not included in examinations of identity because they
cannot be attained from the archaeological record. Mesoamerica, however, offers a unique
situation in which the preservation of human images in artistic contexts not only makes them
visible but amenable to examination and study as well. Observation of body poses and
gestures within their artistic context is not only crucial to determining their meanings, but is
essential to understanding the relationship between figures depicted together (Troike 1982).
In her analyses of Mixtec codices, Troike (1982) has demonstrated how small scale body
variations, in the form of posture and gesture, can be used to recognize and interpret
relationships between individuals of different roles and statuses. While Joyce (1996) has
shown that the relative placement of the whole body, or macro scale body variations, within a
single artistic context in Classic period Maya monuments can indicate not only idealized
gender relationships, but also the presence of gendered spaces.
If gender is viewed as complementary, this will be evident in the context of the two figures
Variation in gesture between paired images places the emphasis on the action,
and suggests the necessary interdependence of the individuals of distinct gender.
Paired figures with different gestures illuminate complementary gendered actions
(Joyce 1996:179).
The placement of males and females on Classic period Maya monuments divide space into
gendered settings along a variety of axes consistent with the spatial symbolism of male and
female costume as vertical axis and horizontal plane (Joyce 1996:172). Females are
believed to be associated with the left and lower elevation", and males with the right and
higher elevation (Joyce 1996:174).
Body poses and gestures, as well as the placement of the body itself, maximizes the
amount of information that is conveyed in an artistic form without adding additional decoration
or materials to an artistic space (Troike 1982:180). This is particularly important for figurine
construction as the canvas is highly limited. The use of this technique is evident in many of
the figurine substyles of pre-Columbian West Mexico in which women are consistently
depicted in one of two different kneeling positions. The seated kneel in which the feet are
placed directly under the bottom, and the collapsed kneel, in which the figure is in a sitting
position with their legs off to one side (see Figure 4.1). The absence of these poses used with
figures believed to be males, greatly supports the argument that body poses and gestures

may have been used as a means of signaling gender identity. Betty Bell (1974:152-157) has
found that among the Zacatecas substyle, the overall body poses of the figures are unique to
males and females. Additionally, the depiction females directly behind or slightly offset and to
the back of males, especially in attached pairs may suggest that the placement of the female
body with respect to that of the male body is indicative of a gendered space.
To define the units for this analysis, I first studied the various body poses and gestures
displayed by the figures not only in the context of the pairs, but also those found on individual
figures, as well as those seen in the dioramas. Combining these observations with Nancy
Troikes (1982) approach to postures and gestures in artistic depictions of the pre-Columbian
Mixtec, I decided to break down the human body form into a number of different units for
examination. The largest unit was the body as a whole, or the pose of the figure (Troike
1982:179). From here, the body could be further subdivided into any number of parts in which
different gestures could be identified (Troike 1982). For the purpose of this analysis, these
were the head, the upper body consisting of the torso, arms, and hands, and the lower body
emphasizing the legs and feet. The head was addressed in terms of its relative position, as
well as by expressions on the face. The placement and positioning of the arms and hands,
and the legs and feet of the figure were examined for the upper and lower body, respectively.
Each of the units can be examined and compared individually, however, each part must also
be examined with respect to the larger unit to which it belongs (i.e. hand and arm as part of
the upper body, and the upper body as a part of the larger whole body form).

Taken from Anawalt 1998, Figure 1; Townsend 1998c, Figure 22.
Data were collected on poses and gestures using the following five categories, Body
Pose, Position of Legs & Feet, Position of Arms & Hands, Position Head, and Facial
Expression. Body Pose refers to the overall position of the body of the figure, such as is the
figure standing, kneeling, or sitting. The Position of Legs & Feet is a more specific notation of
body position, focusing on the placement of the lower body, such as are the legs and feet in a
crossed position, a straight position, and so on. The Position of Arms & Hands is another
notation of body position that specifically addressed the placement of the upper body, such as
the relative location of the arms and hands to the rest of the body. If the arm and hand were
not functioning to express information...[, they were shown] hanging down at the figures
side... and...empty with the fingers straight or slightly curved as though relaxed (Troike
1982:193-194). This too is how Troike (1982:193-194) found neutral gestures depicted in the
Mixtec codices. The Position of Head specifically notes the placement of the head with
respect to the rest of the body, such as is the head positioned forward, backward, cocked or
turned to the side, or rotated differently about its axial attachment to the body. Finally, Facial

Expression is the most specific breakdown of body position focusing on gestures displayed on
the face. Primarily this category makes note of the expression or position of the eyes and
mouth, such as are they open or closed, and are they relaxed or intentionally placed into a
specific expression.
Body poses and gestures of the figures were examined individually, and then as a pair
due to the importance of evaluating poses and gestures in context for their meaning (Troike
1982:181-183). The spatial-posture relationship between the pairs (i.e. what the pose and/or
gesture indicates about the relationship between the two) proved to be more difficult to
determine on unattached pairs where their poses or positions with respect to one another
were more ambiguous. However, the fact that the figures belong together, even if unattached,
created a context in which more than one gender was depicted. This offered more
information about idealized gender relationships than a single figure would have.
Poses and gestures repeatedly associated with males or females were closely examined
and their possible meanings determined. Meanings were based on associations with other
forms of material culture, such as depictions of poses and gestures within diorama and
individual figure contexts, as well as meanings determined by others for similar poses and
Body Adornment
Body adornment refers to various types of modifications made to the body itself, as well
as supplements or items added to it (Eicher and Roach-Higgins 1992:16). Defined in this way
there are many forms of body adornment. Some are permanent, such as tattooing or
scarring, while others like garments, body paint, and jewelry are less permanent (Joyce 2002).
All of the forms, however different in nature they may be, have one thing in common, they are
a means by which people manipulate their bodies to create a social skin," that is used to
convey information to others about their identities, including their status, their social roles, and
even their gender (Joyce 2000a:186). In this way body adornment is used
to visually reinforce the status quo, leaving no doubt in any viewers mind of the
place of an individual with the social network, and hence, the correct mode of
interaction with that person. Clothing and ornaments thus are important visual
clues to a social system (Bruhns 1988:105).
It was not always easy to determine the difference between the various forms of body
adornment depicted on the figures, therefore, each was defined differently as an attempt to

somewhat standardize data collection. For the adornment categories specific in nature,
counts were recorded as part of the response and were represented as -#. However, if a
count could not be determined it was simply left off of the response, and only the type was
noted. Thus, a B notation could mean either a single adornment of type B, or the presence
of multiple type "B adornments where the actual number could not be identified.
Clothing provides humans with much more than a mere functional quality, it is also a
symbolic medium through which they can convey information to others about themselves.
Most commonly, the type of garment worn by an individual indicates the gender group they
are associated with. Throughout Mesoamerica certain forms of clothing are consistently
associated with men and women, and can be used as a means of identifying the different
sexes (Anawalt 1982) (see Figure 4.2). The skirt and a shoulder garment called the
quechquemitl are believed to be the quintessential pan-Mesoamerican garments for females
(Anawalt 1981:211, 213). At times, the alternative shoulder garment called the huipil has also
been included in this repertory (Anawalt 1981:214; Cordry and Cordry 1968:9) because in
central Mexico it is the primary garment worn in place of the quechquemitl (Gillespie
1983:816). The principal garments worn by men in Mesoamerica were the loincloth (maxtlatl),
hip-cloth, and cape (tilmatli) (Anawalt 1981:209, 210, 214, 215 ; Cordry and Cordry 1968:8).
Other garments cited by Patricia Anawalt (1981:209-211) to have been exclusive to males
were found among only certain groups and were limited in their use. With the exception of
armor, which was the essential martial garment, the other costumes were special purpose
garments used only in ritual contexts (Anawalt 1981:209-211). Others, like Rosemary Joyce
(1992, 2000a, 2000b, 2002), who have examined different Precolumbian populations in
Mesoamerica, have recognized the use of these specific garment types to identify human
images as males, females, or a combination of both. Cherri Pancake (1996:49-50) who has
examined modern Indian communities of Guatemala has found this pattern to continue to
exist today. Although some of these models of gender depiction have been applied to West
Mexico (e.g. Anawalt 1981, 1998), most have not, and there have been no attempts to use
these conventions to study gender relations.
Based on the clothing worn by the people depicted in the Relacion de Michoacan, a
document recorded by a Spanish missionary on the Tarascan culture, Anawalt (1981:84) has
suggested that similar clothing-biological sex patterning existed in this region of Mesoamerica

as well. The Tarascans, who were not contemporaneous with the populations producing the
shaft and chamber tombs, appeared much later in western Mexico (ca. AD 1400 -1525).
However, the relatively close proximity between the ancient populations and the later
Tarascans has led Anawalt (1998:237) to believe that the clothing worn by both cultures was
similar. Tarascan males covered their lower torsos with garments resembling short breeches
and. ..wore sleeveless, tuniclike shirts (Anawalt 1998:236). Females of the region not only
wore...short...skirts, but often went about topless, either completely bare above the waist or
wearing small quechquemitls that covered little more than the neck and one shoulder
(Anawalt 1998:236). Others have also recognized that, among the West Mexican figures,
females typically wear skirts and quechquemitls, while males are clad in shorts and tunics
(Butterwick 2004:28-29; von Winning 1974:23-25). Therefore, it is expected that there are
statistically significant associations between these garment types and males and females
among the West Mexican figures in this analysis.
Different aspects of identity also can be signaled through the design elements of
clothing, such as construction, color, design, materials, and even the manner in which
individual garments are worn (Pancake 1996:46). Even the patterning of the designs and
their placement on garments have been found to be important indicators of not only gender,
but other social identities as well (Pancake 1996:47). Louise Allison Cort (1992) notes, in her
study of Japanese robes, that it is not the type of clothing that makes gender distinctions, but
rather the design or pattern that is used on them that does.

Rounded Triangular
Quechquemid QucchquemitJ Huipil Skirt Maxtlatl Hip-cloth Cape
Jfl: - J V t- JUf if]
"^jll $ Lr* s tV 1 £ 0" V £
ff ||f f $
% 1 § pi tf. ^ "V, ^ ar
l 1
Taken from Anawalt 1981, Chart 22 and Chart 23.
Cherri Pancake (1996) argues that garment imagery in modern Guatemalan dress
communicates different messages about the wearers identity at multiple levels. These levels,
ranging from regional to local, convey locational, social, and personal information,
respectively, about the individual (Pancake 1996). Locational symbolism can provide
information on ethnicity, linguistic region, and the community that an individual is part of
(Pancake 1996:54). Social symbolism details an individuals gender, marital status, age
group, socioeconomic position, group membership, and ceremonial roles (Pancake 1996:54).
Symbols reflective of personal information display family affiliation, personal aesthetics,
technical skills, and self-expressiveness of individuals (Pancake 1996:54). In ancient West
Mexico, Butterwick (2004:30) has suggested that the designs depicted on the garments of

ceramic figures reflect the social order and provide clues to a person's village, ethnic identity,
and lineage. Identical or similar patterns might suggest a closer relation between the figures,
such as siblings or cousins, or possibly an endogamous marriage (i.e. individuals paired
within the same lineage), while very different patterns may indicate a exogamous marriage
pair in which both retain membership after marriage (i.e. individuals paired from different
For this analysis clothing is defined as applique or painted on decoration that conceals
primary and secondary sex characteristics. The types of clothing worn by figures were noted,
as well as the specific patterns on the garments, and the matching of patterns between
figures. Specific notation of whether or not the patterns matched in entirety or only in portions
was essential for making interpretations about degree of relatedness between the figures.
Seven categories were developed for the data collection of clothing attributes. They were
Clothing Type, Placement of Quechquemitl, Protective Piece, Clothing Motifs, Clothing Design
Component Schemata, Clothing Spatial Arrangement Schemata, and Matching Clothing
Schemata. Clothing Type is a notation recording the specific garments worn by each figure.
Placement of Quechquemitl is a specific notation regarding the precise placement of a certain
type of clothing. This category was added after viewing a number of figures in which there
was a clear and intentional placement of this particular piece of clothing in different locations
on the body. The assumed importance of the distinction between different placement
locations of the quechquemitl was further supported by finding this different location distinction
within the context of attached pairs. That is, within pairs, one figure is seen with the
quechquemitl in one location while the other has it in another.
Clothing Motifs is a notation of the general overall motifs represented in the clothing worn
by each figure. This more general recording of motifs was based on the classification of
design motifs found in ancient Mexico by Jorge Enciso (1953). Through an evaluation of the
different designs evident in the material culture, Enciso (1953) was able to identify nine
general geometric motifs that were repeatedly used throughout ancient Mexico. These were
zigzags, triangles, squares, circles, spirals, steps, stepped-fret patterns, crosses, and various
geometric patterns (i.e. other designs that could not be easily placed into one of the previously
defined eight pattern types), all of which he showed various examples for, as well as made
notations as to the cultures they were associated with and their possible meanings (Enciso
1953). Responses used for Clothing Motifs in this analysis included most of Encisos
categories and incorporate several new ones that he did not identify (see Figure 4.3). The

more general nature of his distinctions made the motifs comparable across substyles.
Responses for this category are comprised of a collection of clothing motif drawings which are
coded so responses could be recorded.

Wavy lines
Clothing Design Component Schemata and Clothing Spatial Arrangement Schemata are
notations more specific in nature designed to identify important biological sex/gender-attribute
patterning within substyles rather than between them, as designs were often unique to
substyles. The idea of breaking down the decoration found on the clothing of the figures into
its most basic design units (i.e. design component schemata) and design unit orientations (i.e.

spatial arrangement schemata) is based on the approach taken by EW Jernigan (1986) for
stylistic analysis of ceramics in the Southwest. The objective of this approach is to discover
the units of design as they were conceived by the makers (Jernigan 1986:9). This is done by
identifying the basic units of design, or schemata, which consist of a configuration or pattern
of configurations that retain its distinct identity across an number contexts in a
particular style repertory" (Jernigan 1986:9, 10). Style, then, refers to the repertory of
distinct schemata used by an individual, a social group, or any other distinguishable socio-
cultural entity (Jernigan 1986:9, 10). This approach not only emphasizes the importance of
the units that are arranged to make designs but also the important aspect of the nature of
the arrangements for stylistic analysis (Jernigan 1986:17).
Three spatial arrangements for designs were found to be universal among the West
Mexican figures in this dataset. Design schemata were either arranged randomly, or in a
rectangular or triangular pattern (see Figure 4.4). These alignments were found on all
locations where design component schemata were used. There were no limitations to the use
of the spatial arrangement schemata based on the object on which the design was placed.
Whether it was a decorated pot, piece of clothing, or some form of head adornment, the
universal alignments used were the same. The only exceptions were body tattooing and
scarification which had their own unique spatial arrangement schemata. While the location of
general body tattooing and torso tattooing is significant, its overall placement on the body was
more random in nature, as it did not appear in the patterned triangular and rectangular
alignments seen in clothing, headdresses, and object designs. Face design schemata were
broken down into two foci, one centered around the eyes and the other around the mouth.
Only rectangular schemata alignments were seen in designs on some of the ear adornment
forms, but due to the repetition of the same schemata in this alignment, specific ear
adornment forms were recorded rather than adding additional categories for design
component schemata and spatial arrangement schemata for ear adornment. By evaluating
variation in this way, a more emic perspective of design and patterning is achieved in
comparison to other approaches that artificially place an etic categorization on the design
motifs conforming them to the analysts typology (Jernigan 1986:3-9).
Responses for the two clothing schemata categories are comprised of a collection of two
sets of drawings one for the design component schemata, and the other for the spatial
arrangement schemata both of which are coded so responses could be recorded (see
Appendix B, Clothing Design Component Schemata and Spatial Arrangement Schemata).

The final clothing category of Matching Clothing Schemata was originally designed to be
used for accessing the similarity between the basic design components (units) of the two
figures of each pair. However, through data collection it was determined that this category
was of little value to the analysis, as it was often highly ambiguous, could be responded to in a
number of different ways, and overall appeared to yield little if any real data to the analysis.
Triangular Design Alignment Rectangular Design Alignment
Body Painting, Tattooing, Scarification
Body painting, tattooing, and scarification mark differences between genders, as well as
other identities (Joyce 2002). For the Aztecs scarring marked sexual status (Joyce 2002:82).
Among the West Mexican figures tattooing is believed by some to indicate social position
(Butterwick 2004:30). Similarities in patterns on the skin may also indicate degree of
relatedness between the figures in a pair, much like the designs found on clothes.
Joanne B. Eicher and Mary Ellen Roach-Higgins (1992:11) cite Paul Bohannans (1956)
observations of body modifications among the Tiv of Nigeria, as an example of the use of
scarification to distinguish between different genders. Specifically, they note that while the
designs of scarification...[were]...a requisite for beauty among both females and males, [they]

varied by sex (Eicher and Roach-Higgins 1992:11).
In West Mexico the use of tattooing to emphasize the breasts of females is a common
theme despite the varied use of patterns. It is possible that this practice is used as a means
of signaling gender or perhaps emphasizing something that was already obvious.
Additionally, the depiction of specific schemata may be used to convey information about
other forms of identity. Bell (1974:152-157) argues that for the Zacatecas substyle face
tattooing is associated with biological sex.
Body painting and tattooing were lumped together under tattooing for this analysis
because it was difficult to make a distinction between the two forms of skin modification on the
figures. Tattooing" is identified on the figures as painted on decoration in which primary and
secondary sex characteristics are left exposed. Incised slits and applique welts, two types of
scarification characteristic of the region, were more easily distinguished from other forms of
skin modification. As a result, they were recorded separately.
For data collection the following seven attribute categories were used to record paint and
tattoo markings on the figures, Body Tattooing General (Simplified) Schemata, Face Tattoo
Spatial and Design Schemata, Matching Face Schemata, Torso Design Component
Schemata, Matching Torso Schemata, Breast Emphasis Tattooing, and Schemata
Emphasizing Breasts. The Body Tattoo General (Simplified) Schemata records attributes on
the figures derived from a combination of Jernigan (1986) and Encisos (1953) approaches.
By doing this an attempt is made to determine what the basic design units were on the
figures, the emic aspect, but added an etic aspect by trying to keep these basic design units in
a more general motif form that is easily comparable between substyles. This approach is so
little like Jernigans (1986) that it may not even be appropriate to refer to the units as
schemata because they were more etically defined responses. Etic responses were
selected for this attribute category because derived emic schemata within each substyle were
not comparable across different substyles. Responses for this category are comprised of a
collection of listed and coded general schemata possibilities. This category is designed to
compare the very basic shapes used in tattooing between the groups.
The Face Tattoo Spatial and Design Schemata category was also designed to record data
based on Jernigans (1986) emic approach to stylistic analysis. The same approach applied
to clothing designs was again used for tattoo designs on the face. The eye and mouth
distinction used as the response for this category appeared to be the universal way the face
was broken down in terms of design. Eyes tended to be the focus of one design

arrangement, while the mouth tended to be the center of another. This same theme also
appears in the face painting of the Huichol Indians (Lumholz 1900, 1904). Therefore, the eye-
mouth arrangement was the primary spatial arrangement schemata. Only two exceptions
were found in which the entire face was covered by a single design and no eye-mouth
distinctions were made. As such, it was the design component schemata that were of
greatest importance for this category.
The Matching Face Schemata category which was originally designed to access the
similarity between the basic design components of the two figures was found to be of little
value to the analysis, as it was often highly ambiguous, could be responded to in a number of
different ways, and overall appeared to yield little if any real data to the actual analysis.
The Torso Design Component Schemata again records attribute data based on Jernigans
(1986) emic approach to stylistic analysis. However, spatial arrangement schemata could not
be identified and appeared to be unimportant on this portion of the body as all schemata
arrangements were random in nature. Therefore, the spatial arrangement schemata aspect
was eliminated for this specific design analysis. Again, the Matching Torso Schemata
category was originally designed to access the similarity between the basic design
components of the two figures. However, through data collection this category too has been
determined to be of little value to the analysis, as it is often highly ambiguous and can be
responded to in a number of different ways.
The Breast Emphasis Tattooing category was designed to compare the general nature of
breast emphasis tattooing across the substyles. Jernigans (1986) approach was used here
as well, but because of the nature of the location of the design all of the spatial arrangement
schemata were similar. Therefore, this aspect again was eliminated from the analysis, and
only design component schemata were identified. This category was designed to compare
the more specific designs and examine the relationship of their use between the different
substyles throughout the region. This category provides notation for a more specific type of
stylistic analysis that goes beyond the mere presence-absence analysis of breast tattooing.
Scarification was only addressed in one category during data collection. Responses
specifically made note as to whether scarring existed or not, and if it did where it was located
and what its nature was. Scarification on the shoulders was in the form of round welts, while
face scarring was usually limited to slits in the cheeks, but also included slits across other
parts of the face as well. Also recorded in this category were the presence of paint markings
on the shoulders that may be indicative of scarring (Butterwick 2004:30).

As with clothing, jewelry marks the roles and statuses held by different individuals
throughout Mesoamerica (Anawalt 1998). In a discussion on gender and adornment
associations, Joyce (2002:85) notes that through the examination of several burials at La
Venta, Mexico, she discovered ear adornments were used only by fully gendered adults.
Based on their examination of associations between ornamentation and biological sex in the
Mixtec codices, the McCaffertys (1994:149) have suggested that pre-Columbian jewelry was
not gender specific. Von Winning (1974:25-27) has also found the same to be true among the
West Mexican figures. Due to the contradictory results in whether or not jewelry can be
associated with gender, and the fact that jewelry is most often indicative of some form of
social status or identity, I have chosen to evaluate this category in this analysis with the hopes
that it might shed some light on this issue.
Jewelry is defined as both relief and painted on ornamentation that is located on the neck
and extremities. For data collection six categories were designed to evaluate the different
areas of the body that were typically adorned. They were Ear Adornment-General, Ear
Adornment-Specific, Nose Adornment, Neck Adornment, Arm Adornment, and Leg
Adornment. This break down made the recording of adornment data much easier.
The Ear Adornment-General category was designed to obtain a more general response of
the types of ear adornment present so cross substyle comparison could be made easier.
While the Ear Adornment-Specific category, which was also designed to look at ear
adornment, aimed at getting more specific responses to be used to look at the specific
patterning between biological sex, gender, and ear adornment within each substyle.
Responses for the ear adornment categories were composed of a collection of listed and
sketched ear adornment possibilities (see appendix B, Ear Adornment-General and -Specific).
It must be noted that the responses between these two categories were relatively similar and
did not need to be broken down into general and specific responses.
The Nose, Neck, Arm, and Leg Adornment categories simply made note of what forms of
each were present. Adornment forms for these categories appeared to be shared by the
substyles for each so more general response categories were not helpful. Responses for
these categories are composed of a collection of sketches of the different types of adornment
possibilities which are coded so responses could be recorded (see appendix B, Nose, Neck,
Arm, and Leg Adornment).

Headdresses and Hairstyles
The head, as a central focus of the body, provides an important canvas upon which to
convey social information. Within Mesoamerica many forms of head adornment are used to
show different roles and statuses of the wearers. Rosemary Joyce (2002:82) notes that
among the Aztecs hair was the focus of an exceptionally elaborate set of treatments
providing the richest communication of dimensions of social status[, including gender,] a
single kind of appearance cue. Headdress and hairstyles as indicators of gender still persist
today among the modern descendants of ancestral populations of Mesoamerica (Pancake
1996:49). Pancake (1996:49) has shown how headdresses among the Sacapulas in
Guatemala are used to distinguish between different gender statuses of women, specifically
indicating when they are single and when they are married. Cranial deformation or intentional
head shaping has also been associated with varying social identities in Mesoamerica,
specifically that of biological sex and gender (Joyce 2001:115). The elongated head shapes
of some of the West Mexican figures is evidence that intentional head shaping was occurring
(Butterwick 2004:30), however, its ability to distinguish between different genders or indicate
status has yet to be determined for this region. For the Zacatecas substyle, it has been
suggested that hair adornment was used to distinguish between different genders (Bell
The head was addressed in this analysis through the collection of six different forms of
attribute data. The categories used in data collection were the following, Head Shape, Hair,
Head Adornment-General, Head Adornment-Specific, Head Adornment Design Component
Schemata, and Head Adornment Spatial Arrangement Schemata. Head Shape is composed
of two possible responses, Natural (N) or Modified (Mod). Natuml referred to an unmodified
state of the skull, while Modified referred to any form of modification or reshaping of the skull
from slight flatening to outright changing of over all skull shape (i.e. elongation). The Hair
category simply notes whether or not the figures own hair appeared to be incorporated into
the overall head adornment. Responses were noted as either present or absent.
The Head Adornment-General category is of a more general nature that allowed for cross-
substyle comparison of forms of head adornment. Responses for this category are a list of
general head adornment type possibilities. The Head Adornment-Specific category is fairly
specific in nature and is designed more for examining within substyle patterning rather than
for substyle comparisons. Responses for this category were comprised of a collection of head

adornment drawings which were coded so responses could be recorded (see appendix B,
Head Adornment-General and -Specific). Visual images of responses reduced the ambiguity,
inconsistency, and misinterpretation often associated with mere textual responses. The Head
Adornment Design Component Schemata category was again a category based on Jernigans
(1986) approach to stylistic analysis in which basic design components were derived emically.
This approach was identical to the one used on clothing design, only it focused on a different
form of adornment. Responses were depicted as visual representations and were coded so
they could be recorded. The Head Adornment Spatial Arrangement Schemata category was
simply the other aspect of Jernigans (1986) approach used to determine the basic spatial
arrangement of schemata in designs. Responses for this category were comprised of
drawings of what appeared to be a universal set of spatial arrangement schemata shared by
the substyles which were coded so responses could be recorded (see appendix B, Head
Adornment Spatial Arrangement Schemata).
The distinction between headpieces and hairstyles was quite problematic. Occasionally,
the detailing of individual strands of hair on the head was used to indicate a hairstyle, as
opposed to a headpiece. However, on some figures (e.g. Ameca-Etzatlan style figures)
smooth cap-like paint or applique could easily be interpreted as hair or a headpiece. In other
cases, the figures hair adornment reflected the hair being combined with decorative
elements, such as fabrics, ribbons, metals, or organic materials and worked into a headdress-
like piece (Cordry and Cordry 1968:115-134; Flannery and Marcus 1998:31-38). This further
complicated attempts at distinguishing the different types of head adornment. As a result, no
distinction was made between headpieces and hairstyles for this analysis. Head adornment"
was used to define all decoration that occurred on the head regardless of whether or not the
adornment was predominantly hair or added ornamentation (i.e. hats, headbands, etc.).
Objects Held
Objects held by figures, or found in association with them, in artistic contexts in
Mesoamerica are often found to define gender specific roles, as well as other social statuses
and positions held by individuals (Joyce 1993:261). Knowing different status and tasks are
often gender specific allows objects held by figures to not only be invaluable for understanding
the kind of social roles and statuses that existed, but also which ones may have been gender
unique. Determining which items are consistently displayed by certain genders and

examining their meaning gives insight into idealized roles and positions assumed by each
The meanings and functions of such objects to the ancient populations of West Mexico
may be obtainable through comparative analyses of material culture with that of other groups
who have resided in the region in the past and today (i.e. Tarascans and Huichol).
Additionally, similarities with other non-local items and their associated designs, specifically
those of Ecuador, which is believed to have either directly or indirectly influenced ancient
West Mexican populations (Anawalt 1998), are invaluable for determining the possible
significance of items depicted with the figures.
Objects held included any items that were grasped by the hands, attached to the upper
extremities or torso, placed in the lap, or were being used by the figure. Such objects
included, but were not limited to religious paraphernalia, balls, containers, children,
subsistence items, and stone tools.
Identifying the difference between a plate, bowl, cup and pot was a bit ambiguous. For
this reason distinctions were made based on the diameters of the mouth and base of the
vessel and the height of the vessel walls. First, it was assumed that these vessels fell on a
continuum where changes in their dimensions produced changes in their overall shape. On
one end was a pot and on the other a plate. Taller walls tended more towards a cup, while
shorter walls were more like a plate. More closed mouth (i.e. base and opening
approximately equal in diameter) more toward a cup, more open mouth (i.e. base smaller
diameter than vessel opening) more toward a bowl. A very small diameter opening was a pot.
Usually when the mouth diameter was smaller than the base diameter the vessel was
considered a pot. Mainly the pot, cup, bowl, plate distinction was based on the diameter of
mouth and base, and wall height, however, the vessels overall size and the context within
which they were found also played a role in the final determination.
For data collection an Objects Held category allowed for recording of specific objects held
by the figures. Responses for this category were coded drawings of objects (see appendix B,
Objects Held). Associating responses with visual images rather than text only allowed for the
recording of responses that contained less variability especially for complicated designs and
adornments. Visual representations reduced the amount of possible error and bias by clearly
depicting the responses rather than trying to determine what was being referred to in a textual

In addition to collecting data on the specific type of object held by each figure, the object
design component and spatial arrangement schemata were also recorded. These categories
were again based off of Jernigans (1986) approach to stylistic analysis and were identical to
those used for clothing and head adornment designs. Responses were coded drawings of
basic design component and spatial arrangement schemata (see appendix B, Object Design
Component Schemata and Spatial Arrangement Schemata). Coded drawings reduced much
of the error and ambiguity associated with the use of textual responses. They also allowed for
a more accurate recollection and continued collection of data based on the same visual
attributes as opposed to textual descriptions that greatly varied in how they could be
interpreted by a different analyst.
Determining Sex and Gender
The biological sex of each figure was determined by the presence of primary and
secondary sex characteristics. Additionally, associations between biological sex and
protective piece (for males) and swollen belly (for females) were evaluated to determine if
these two attributes would be reliable indicators of biological sex. While these latter two
measures of biological sex were used mainly to support biological determinations for figures
with identifiable sex characteristics, they were however, valuable in situations where a figures
primary and secondary sex characteristics were not visible.
Biological sex is identifiable on many of the figures, however, there are a number for
which it is not. This is likely due to the presence of clothing obscuring the sex characteristics.
The absence of any identifying sex characteristics on a figure is as significant as it is for those
on whom these can be seen. The lack of sex characteristics on a figure could potentially
indicate the presence of a third gender.
Priority was given to primary and secondary sex characteristics as they are a direct
indication of ones biological sex. Sexing a figure by their exposed genitalia is a fairly reliable
means of determining biological sex. Unfortunately, these features are not always visible. As
a result, identification was also dependent on secondary sex characteristics, such as a mother
nursing a baby, or developed breasts depicted on a figure as opposed to a flat chest, for
making the distinctions.
If neither primary nor secondary sex characteristics were visible, the presence of a
swollen belly or a protective piece, was used to determine biological sex. These attributes

were considered only secondarily to sex characteristics, as they are a more indirect means of
identification, and can, in some cases, be a somewhat ambiguous indication of biological sex.
While these attributes are not primary or secondary sex characteristics, they can indicate the
presence of them.
The presence of a protective piece would suggest the figure is biologically male,
especially if it is lacking the secondary sex characteristics of females and a swollen belly.
While the primary male sex characteristic is not visible, the existence of an object necessary
for its protection likely indicates the presence of one behind it. Others have also speculated
that the protective piece, depicted on the West Mexican figurines, may have been used for
genital protection (Anawalt 1998:238). The frequent portrayal of only male figures with a
protective piece in this region (Anawalt 1998:238, von Winning 1974:24-25) suggests that it
may have been used to signal biological sex or gender.
The presence of a swollen belly would likely suggest the figure is biologically female
(Joyce 2002:89), especially when primary and secondary female sex characteristics are also
present. It is assumed that a swollen belly is due to pregnancy, which is only biologically
possible for females, rather than the result of being overweight, because the enlarged portion
of the figure is limited to the abdomen or lower belly.
Once biological sex was determined, chi-square tests were performed on each key
attribute to determine which, if any, were consistently associated with biological sex.
Attributes repeatedly associated with figures determined to be male or female were deemed
to be characteristic of the two basic genders based on biological sex, man and woman
The dataset was then recoded using the attributes determined to be characteristic of men
and women (Munson 2000:133). Figures that possessed one or more traits characteristic of
only one gender or the other were considered to be of that gender (Munson 2000:133). If a
figure had no attributes associated with either gender, but had visible sex characteristics, they
were assigned to the gender which corresponded to their biological sex (Munson 2000:133).
There are two possible ways of identifying the presence of other genders in the dataset.
A figure could either have traits characteristic of both men and women (i.e. wearing an article
of clothing associated with females as well as one associated with males); or a figures
biological sex is contradicted by the presence of traits characteristic of the gender defined by
the opposite biological sex (Munson 2000:133).

Using the following gender definitions, six and, possibly seven, different gender types can
be identified. The woman gender is defined by figures that are biologically female with only
characteristically female associated traits. The man gender is defined by figures that are
biologically male with only characteristically male associated traits. Figures deemed to be
either biologically male or female and possesses only traits characteristic of the opposite
biological sex are gendered, man-woman and woman-man respectively. Figures deemed to
be either biologically male or female and possesses traits characteristic to both males and
females are gendered, man2 and woman2 respectively. Figures in which there is no
indication of biological sex, but possesses traits characteristic of both males and females are
gendered unknown alternate, indicating they potentially represent another gender type not tied
to biological sex. Table 4.3 lists out these specific gender type definitions. Additionally,
figures that did not possess any traits characteristic of the male and female genders, but were
depicted with male and female sex characteristics, were defined as male and female genders,
Table 4.3. Gender type definitions
Gender Type Biological Sex Traits Characteristic of Gender
Woman female only female
Man male only male
Woman-Man female only male
Man-Woman male only female
Woman2 female both male and female
Man2 male both male and female
Unknown Alternate unknown both male and female
The results from the statistical testing of these attributes are presented in the following
chapter. Additionally, the implications of the association of different attributes with gender are
discussed, as well as the identified gender types at the local and regional levels.

Statistical testing of the dataset mainly addresses the microscale issues of biological sex-
attribute associations and gender marking within each substyle. Although the substyles are
evaluated together for regional associations, these results only define relative regional
patterning, and were not considered in final gender definitions. Macroscale analysis focuses
on comparisons between the substyles in the core and periphery regions once gender
ideologies are established.
Marking is evaluated by performing chi-square tests on the dataset to determine if any of
the attributes are consistently associated with biological sex. If either males, or females, are
explicitly marked, then many attributes will be associated with one, while few, if any, with the
other. Additionally, the attributes associated with the marked biological sex, should not occur
with the other. If neither biological sex is explicitly marked, then there should either be no
attribute associations, or the number of attributes consistently associated with males and
females should be roughly the same; the latter of which could also suggest both sexes are
equally marked. Should the relationship between the two figures be complementary, then
each would associate with a unique set of traits.
A comparative analysis performed across the substyles was used to determine the extent
to which perceptions of gender were shared. The idea of a shared ideology (i.e. similar
gender definitions) does not suggest that specific identity markers (e.g. headbands or
obsidian necklaces) need to be the same, but it is rather the use of the same kinds of
attributes (e g. head adornment or neck adornment) to signal the identity that is important.
For an ideology to be considered shared, all of the attributes signaling gender must be
shared. For substyles that have less physical attributes, the attributes that are used to signal
gender must be the same, even if they do not use all of them. However, perceptions of
gender can still be shared, despite finding that different substyles use different attributes to
make the distinctions.
Data Preparation
The original dataset was recoded and downloaded into the SPSS program for statistical

testing (see Table 5.1). Recoding not only created a more uniform dataset, but was also
necessary to make the responses compatible with SPSS. All recoding during this step was
done manually. Textual responses were recoded and given a numerical value. Multiple letter
responses were first divided into separate columns, and then each column was recoded using
numerical values. This was done because the individual responses could potentially be
valuable in their own right. That is, important patterning in the dataset might be missed if
these responses were kept together and not separated. Once the dataset was recoded, it
was hand checked against the original dataset in an attempt to reduce error by ensuring the
responses were the same using the new numerical codes.
It is important to note that the groupings I have chosen not only for the original columns,
but the recoded ones as well, directly affect the relationships and patterns revealed through
statistical analysis. Had the responses been grouped differently, different relationships may
become apparent that were missed in this analysis simply because of the way that I have
chosen to group them.
Groupings used were designed to get at the most minute associations present between
the physical attributes and biological sex of the figures. Attributes were not simply grouped
into larger categories of body adornment and ornamentation, but were rather broken down
into particular types of bodily decoration (i.e. clothing, headdresses, jewelry, tattooing,
scarification, etc.) that accounted for specific locations of those different types (i.e. ear
adornment, nose adornment, leg adornment, torso tattooing, etc). Matching categories, which
were initially hoped to assess similarities between the figures in each pair were found to be of
little value to the analysis, and were eliminated from statistical testing.

Table 5.1. Original data collection categories with associated recoded categories for SPSS
Original Data Categories Recoded Data Categories
Authenticity Authenticity, Authenticity-Additional Notation
Position Hands & Arms Recoded Position Hands/Arms, Position Right Arm/Hand, Position Left Arm/Hand
Facial Expression Recoded Facial Expression, Eye Expression, Mouth Expression
Clothing Motif(s) Recoded Clothing Motif(s), Clothing Motif 0 through Clothing Motif 12
Clothing Design Component Schemata Recoded Clothing Design Component Schemata, Clothing Schemata 0 through Clothing Schemata 46
Body Tattoo General Schemata Recoded Body Tattoo General Schemata, Tattoo Schemata 0 through Tattoo Schemata 18
Face Tattoo Spatial Arrangement and Design Component Schemata Recoded Face Tattoo Spatial Arrangement and Design Component Schemata, Face Tattoo Spatial Schemata, Eye Tattoo Design Schemata, Mouth Tattoo Design Schemata
Ear Adornment-Specific Recoded Ear Adornment Specific, Ear Adornment Specific Minus Count
Nose Adornment Recoded Nose Adornment, Nose Adornment Minus Count
Neck Adornment Recoded Neck Adornment, Neck Adornment Minus Count, Neck Adorn(ment) 0 through Neck Adorn(ment) 21
Arm Adornment Recoded Arm Adornment, Arm Adornment Minus Count, Arm Adorn(ment) 0 through Arm Adorn(ment) 28
Leg Adornment Recoded Leg Adornment, Leg Adornment Minus Count
Head Adornment Design Component Schemata Recoded Head Adornment Design Component Schemata, Head Adorn(ment) Schemata 0 through Head Adorn(ment) Schemata 23
Object Design Component Schemata Recoded Object Design Component Schemata, Object Schemata 0 through Object Schemata 23
Testing the Data
Chi-square tests were run, first on the whole dataset, and then on each substyle, using
the SPSS statistical software. The value for missing data 999" was coded as missing data in

SPSS and was not included in the chi-square tests. Results were deemed to be statistically
significant if the chi-square p-value was .059 or less with 20% or less of the cells having an
expected count of less than 5. Cramers V values were also calculated and used to determine
the strength of the associations, where 0 to .3 was weak, .3 to .6 was moderate, and above .6
was strong. Results with p-values above .059 were recorded, but eliminated from further
statistical testing. Of the remaining variables with p-values reflecting a significant association
with biological sex, those having expected count percentages greater than 20, were either
eliminated if they already existed as a presence-absence response, or if specific attribute
types were initially grouped into a single variable they were recoded into presence-absence
responses using Microsoft Excel. Recoding reduced the table size of the chi-square and
resulted in the reduction of the number of cells with high expected counts less than 5. Chi-
squares were then rerun on the biological sex indicators and new variable columns.
In an effort to detect any possible patterning, I had inadvertently broken down biological
sex too far in these initial tests, so that instead of associating males and females with each
attribute, I was evaluating the associations between the attributes and specific sex
characteristics. As a result, these first rounds of testing and recoding proved to be an
excellent means of establishing if existing categories lumped the variables enough to make
their results pertinent for later discussions. Additionally, these initial tests provided a general
view of attribute patterning and allowed refinements to be made to attribute categories to
determine if potential significant associations could be accepted at a more refined level of
Chi-square tests were rerun a third time on the entire dataset and each substyle using
only the Male/Female biological sex category in which figures were either determined to be
male or female based on the presence of the collection of corresponding primary and/or
secondary sex characteristics. This time, not only the 999" response for missing data was
eliminated from the tests, but also the 0" response of the Male/Female biological sex
category, which represented figures with no sex characteristics present. All substyles were
included in statistical testing of the entire dataset, however, only substyles with sample sizes
of 25 or greater were evaluated individually for specific substyle associations. Twenty-five
was chosen as the minimal sample size required for statistical testing. Sample sizes below
this value were believed to be too small to yield representative results, and were not included
in the individual substyle analyses. The substyles with a sample size of 25 or larger that were
a part of this round of testing were substyle 1 Ameca-Etzatlan, substyle 5 Ixtlan del Rio,

substyle 8 San Sebastian, and substyle 11 Zacatecas. Significance was evaluated as
described above, with the exception of the value for the percentage of cells having an
expected count less than 5. A new value of 25% was chosen for the remainder of the tests
because the previous value of 20%, which was more arbitrarily selected, did not reflect a
whole cell given the sample sizes used. Tables were created for the results showing which
attributes associated with males and females. Attributes in these tables were arranged based
on the strength of the associations.
This round of tests actually evaluated the associations between biological males and
females and specific attributes, however, in excluding figures without visible sex
characteristics at least 50%, and often more, of the dataset was eliminated in both the
evaluation of the entire dataset, as well as by individual substyle. This proved problematic to
me in three respects. First, if some figures without visible sex characteristics were actually
males (i.e. primary sex characteristic covered by garment), then males were under
represented. Second, substyles in which sex characteristics were not often present on one of
the individuals (i.e. Ixtlan del Rio and Ameca-Etzatlan) in the pairs were completely excluded
from tests on the entire dataset. Finally, the results from the tests on the entire dataset
seemed to reflect associations of the substyles which were most heavily represented in the
Pan-Regional Results
Quite a number of attributes were associated with males and females when all of the
substyles were evaluated together (see Table 5.2). Most associations were weak, however
several were moderate. Although females did have a few more associations than the males,
the relative number of associations between males and females was roughly the same,
suggesting that pan-regionally gender, or at least biological sex, was not being marked
Finding females moderately associated with breast tattooing and a swollen belly pan-
regionally was not surprising, especially because both of these attributes were tied closely to
biological sex. Weak associations with several specific designs of breast tattooing were found
for the female figures as well. These include, zigzag ladder, solid line, solid triangle line, 3-bar
line, and spiral patterns (see appendix B, Design Component Schemata Emphasizing
Breasts). Presumably all breast designs would associate with females if they occurred
frequently enough. Resting a hand on the belly or abdomen region, also associated weakly

with females. This gesture is often depicted on pregnant female figures, and is indicative of
fertility and child bearing (Goldstein 1988:54, 56). Correspondingly, males were moderately
associated with the absence of tattooing on their breasts, and there was no statistically
significant relationship between male figures and a swollen belly.
Also, quite expected was a moderate association between female figures and the skirt,
cup, and kneeling body pose, which included both the seated and collapsed kneel. It has
been suggested that females within this region are afforded a lower social status because
they are found primarily in a kneeling position, which is a submissive posture (Goldstein
1988:54). Although this is possible, it is more likely the kneeling body pose is tied to female
reproductive imagery, specifically childbirth, rather than portraying a subordinate female social
position. With the exception of the cup, all of these attributes have been recognized as being
present only among females in the West Mexican imagery (Anawalt 1998:236-237; von
Winning 1974:23,25). Even the cup, which sometimes is used by males, is considered to be a
predominantly female possession in this region (Gallagher 1983; von Winning 1974).
Moderately (and also weakly) associated with female garments was the use of a single color
for decoration displayed in a rectangular arrangement (see Figure 4.4). Designs on objects
held by women were weakly associated with rectangular pattern arrangements (see Figure

Table 5.2. Pan-regional biological sex-attribute associations, n = 374
Associations with Females
Weak Moderate Strong
No Head Adornment Breast Emphasis Tattooing
Plain Headband Swollen Belly
Twisted Headband Kneeling Body Pose
Earring Fan Skirt
Single Loop Nose-ring Cup
Neck Adornment: Dotted Line(s) Clothing Design(s): Solid Color
Neck Adornment: Single Obsidian Plate Clothing Design(s) Arrangement: Rectangular
Arm Adornment: Dotted Line(s)
Arm Adornment: Solid Color Limbs
Disk Armband
No Leg Adornment
Tic Line Tattooing
Dotted Line Tattooing
Circle Tattooing
Triangle Tattooing
Spiral Tattooing
Zigzag Ladder Breast Tattooing
Solid Line Breast Tattooing
Solid Triangle Line Breast Tattooing
3-Bar Line Breast Tattooing

Table 5.2. (Cont.)
Associations with Females
Weak Moderate Strong
Spiral Breast Tattooing
No Scarring
Clothing Motif(s): Solid Color
Wide Open Eyes
Bearing Teeth/Smile
Both Hands on Hips
Holding Object to Chest
Hand on Abdomen/Belly
Seated Kneel
Collapsed Kneel
Seated with Legs Straight Out in Front
Design Arrangement on Object Held: Rectangular

Associations with Males
Weak Moderate Strong
Head Adornment Design: Solid Color Head Adornment: Two-Horn
Earspools Leg Adornment: Wavy/Zigzag Line(s)
Generic Type Earpiece No Breast Emphasis Tattooing
No Nose Adornment Not Wearing Clothes
Neck Adornment: Solid Line(s) Arms Crossed

Table 5.2. (Cont.)
Associations with Males
Weak Moderate Strong
Leg Adornment: Solid Color Limbs A-Frame Knees
Wavy Line Tattooing
Slits on Cheeks
Head Tilted Down Facing Front
Eyes Squint/Slit Open
O-Shaped Open Mouth
Right Hand/Arm Holding Object in Air
Left Hand on Knee
Sitting on Ground/Squatting
No Object Held
Finding males moderately associated with wearing no clothing pan-regionally, at first
appeared peculiar, but when given some thought this was a quite logical association. The
only way the biological sex of male figures could be identified was if their primary sex
characteristics were exposed, which meant they could not be wearing any lower body
garments. Among the figures in this dataset, males who were nude from the waist down were
typically unclothed from the waist up as well. There were only thirteen instances in which
males, whose genitalia were exposed, wore a garment covering their torso, such as a shirt,
quechquemitl, or body armor (see appendix A, Figures A. 11, A.40, A.41, A.43, A.91, A.98,
A. 110, A. 125, A. 133, A. 142, A. 165, A. 168, A. 172). For the most part, figures that were
identified as male were completely nude. However, in roughly 50% of the dataset females
were found to be paired with figures with no visible sex characteristics, often the result of
wearing shorts or loincloths. If these figures had been included as males, then it is likely that
statistical significant associations between males and other items of clothing, such as shirts,
shorts, and loincloths would have been present in the results.

Finding that males are both moderately and strongly associated with not wearing clothes
pan-regionally is possibly due to the number of valid cases that were used in the different chi-
square tests. The Cramers V for the Clothing Schemata determination (n=51) of no clothing
worn is .596 which approaches the cutoff of .6 for the difference between a moderate and
strong association between the variables. Clothing Type (n=54), Clothing Motif (n=49), and
Clothing Schemata (n=51) consider the association between males and no clothing worn to be
strong. The fact that Clothing Type, which is a direct measure of the type of clothing that is
worn, if any, considers the association to be strong is probably a good indication that the
strength of the association is in fact strong. With three other variables indicating that males
are strongly associated with not wearing clothes and the one variable considering the
association is moderate is only five hundredths (.005) from being considered a strong
association, the association between males and no clothing worn could easily be interpreted
as strong. Elimination of figures that had no visible sex characteristics, but were usually
interpreted as male based on the clothing they wear (Anawalt 1998; Butterwick 1998:93; von
Wnning 1974), was also believed to be the reason other male-attribute associations were not
apparent in the statistical testing.
Pan-regionally, males were moderately associated with the two-horn head adornment (i.e.
which is used here to refer to figures with two horns protruding from their head), wavy or
zigzag line markings on their legs, crossed arms, and A-frame knees. Sitting with both knees
raised into an A-frame position, has been argued to be a high status posture, especially when
associated with males, because it allows them to display their genitalia (Goldstein 1988:55),
and in essence their power and masculinity. The horned headdress alone (Furst 1998:ISO-
181, 1978:30; Goldstein 1988:55), and in combination with crossed arms resting on A-frame
knees has been argued to be characteristic of shamans and shamanistic practices (Holsbeke
and Arnaut 1998:73). Additionally, horned headdresses have also been suggested to indicate
ruling elite, specifically marking their superior political rank (Graham 1998). It is possible that
these moderate attribute associations, two of which have a strong statistical association with
males of the Zacatecas substyle (i.e. two-horn head adornment and A-frame knees), and the
others (i.e. wavy or zigzag leg markings and crossed arms), have been visually determined to
be primary male traits of this substyle (Bell 1974; Gallagher 1983; von Winning 1974), may
have been influenced by the large number of males present in the sample from the Zacatecas
substyle because they are always depicted nude, and were therefore, easily identified as
male. Again, this is an example of how eliminating potential male figures based on the lack of

visible sex characteristics may have negatively affected statistical results.
The identification of females based on primary and secondary sex characteristics is fairly
reliable, because there are always a number of possible indications on the body to suggest
that the figure is female. However, more problematic is the assessment of males by sex
characteristics, especially because there is only one attribute indication. If this attribute is
covered by clothing there is no positive way to indicate that the figure is in fact male. This
proved to be quite a dilemma for this project, as 50% of the figures paired with females were
of this latter type. Without visible sex characteristics it is impossible to say whether clothing is
enough to justify these figures as male, or if they are better described as representative of a
third gender.
Only females were weakly associated with particular forms of head and arm adornment
pan-regionally. For head adornment these included, not wearing head adornment at all, and
wearing either a plain or twisted headband. Dotted lines, solid color limbs, and disk armbands
were the three forms of arm adornment associated with female figures (see appendix B, Arm
Adornment, #8,#10,#11). While males were moderately associated with the two-horn head
adornment, they were only weakly associated with the use of a solid color on the different
Ear, nose, neck, and leg adornments, however, were found to weakly associate with both
males and females pan-regionally. Female figures were associated with only the earring fan
(refer to appendix B, Ear Adornment Specific, for similar examples, #5, #6, #7), while males
were associated with either earspools or generic earpieces. Men tended to be depicted
without nose adornments, and women had single loop nose-rings (see appendix B, Nose
Adornment, #2). Females were associated with single obsidian plate necklaces and dotted
line paint or tattooing around their necks (see appendix B, Neck Adornment, #2, #3). Male
figures associated with straight line paint and tattoo designs on their necks (see appendix B,
Neck Adornment, #7). Female figures showed a weak association with not wearing leg
adornment at all, while males were weakly associated with solid patches of colored paint or
tattooing on their legs.
Body paint or tattooing appeared to weakly associate with both male and female figures
pan-regionally. However, females showed associations with several different designs, while
males were only associated with one. Female patterns included tic lines, dotted lines, circles,
triangles, and spirals (refer to appendix B, Torso Tattoo Design Component Schemata, for
similar examples, #4, #7, #12, #14, #23, #29). Males designs were limited to wavy lines

(refer to appendix B, Torso Tattoo Design Component Schemata, for similar examples, #8,
Several different variations of scarring were depicted on the figures in the dataset (see
appendix B, Scarring). However, males only are associated weakly with one form, slits in
cheeks, and females are weakly associated with no scarring at all when testing all the
substyles together.
Body poses and facial gestures appear to weakly associate differently with males and
females pan-regionally. Female figures were associated with wide open eyes and mouths
that were either displaying a smile or bearing their teeth. Facial expressions for male figures
included squinty eyes and O-shaped open mouths. A weak association suggested the heads
of males were typically tilted down and facing front. In addition to being associated with the
collapsed and seated kneeling positions, female figures were also weakly associated with the
seated position in which their legs were straight out in front of them. If female figures were not
placing a hand on their belly, weak associations suggested that it would either be holding an
object to their chest or both hands would be on their hips. Moderately associated with males
were A-frame knees, however, they were also weakly associated with either sitting on the
ground or squatting. While males were weakly associated with holding no object at all, those
figures depicted with objects, held them high in the air with their right hands. Male figures
were also weakly associated with placing their left hand on their knee.
A-frame knees are moderately associated with males pan-regionally, however, the mates
to many of the figures identified as female are also seated cross-legged. These figures were
typically wearing shorts and therefore were excluded from the analysis because they had no
visible sex characteristics. Again these figures are often argued to be males because they
are wearing shorts which are believed to be a garment associated only with males in
Mesoamerica, including West Mexico. It is therefore, likely that sitting cross-legged may also
be significantly associated with males as well.
Substyle Results
Of the individual substyles tested, only those of the periphery were found to have
statistically significant associations between biological sex and the attributes (see Table 5.3).
Both the El Arenal Brown and Coahuayana substyles had to be eliminated from the individual
substyle analysis as their sample size of two was too small. Statistically significant
associations were found to exist when the chi-square tests were run on the Ixtlan del Rio, San