Looking Beyond Gender Hierarchy: Rethinking Gender at Teotihuacan, Mexico



The absence of gender inequality in art and burials at Teotihuacan, Mexico, poses an interesting problem that has been largely ignored by archaeologists. Archaeologists' lack of attention to gender at Teotihuacan perhaps derives from expectations of gender hierarchies in state societies, resulting in misguided conclusions and unexplored alternative interpretations. In this chapter, I ask how eliminating assumptions about gender hierarchies might help us to gain a better understanding of social relations at Teotihuacan. By incorporating multiple lines of evidence including art, burials, and households, I argue that group identity superseded the importance of the individual, foregrounding “house” identity over gender. Social solidarity may have been particularly necessary at Teotihuacan since a large percentage of male and female residents were foreigners. While socially constructed genders may have been an important component of social identity, I suggest that it was largely an unimportant factor in determining social status.

Archaeologists rarely problematize the notion of gender hierarchy in past civilizations, perhaps because it confronts our basic assumption that gender inequality goes hand in hand with state-level societies. Researchers of ancient states often approach “gender” by asking to what extent a gender hierarchy would have been present, presupposing that males were dominant. Spearheaded by Anne Pyburn (2004), recent critiques argue that in creating rigid dichotomies we are projecting our own cultural conceptions of gender into the past rather than extracting new understandings from the archaeological record. Pyburn writes, “We ask questions that assume the answer, or construct hypotheses that cannot be ruled out with contradictory evidence … Any evidence to the contrary will automatically be treated as exceptional or ‘proving the rule’“ (Pyburn 2004:19). Pyburn urges us to examine our assumptions and to approach gender more critically so that gender can serve as a means to broadening our understanding of social, political, and economic phenomena.

Classic period Teotihuacan was the center of a primary, state-level society located in Central Mexico. The absence of gender inequality in household burials at Teotihuacan poses an interesting problem that has been largely ignored by archaeologists. Possibly because these data challenge our assumptions about the nature of states, Teotihuacan has rarely been the subject of engendered-states research. In this chapter I ask the questions, How might eliminating assumptions about gender hierarchies help us to gain a better understanding of social relations and identity at Teotihuacan? What do we learn about gender and gender roles at Teotihuacan when we challenge our assumptions? How can a study of gender be used to address broader anthropological questions; for example, how (if at all) did the Teotihuacan state function as a “faceless,” corporate society (Blanton et al. 1996)? To address these questions I incorporate multiple lines of evidence including art, burials, and households. I argue that group identity superseded the importance of the individual at Teotihuacan, so that compound identity was emphasized over gender. This kind of social solidarity may have been particularly important for maintaining Teotihuacan's corporate political strategy in light of its large immigrant population and ethnic diversity. While socially constructed genders may nonetheless have been an important component of social identity at Teotihuacan, I suggest that it was largely an unimportant factor in determining social status.

In this chapter I first look at earlier studies of mortuary remains from Teotihuacan to illustrate how assumptions about gender inequality have led to the minimization of women's roles. When these burials are reconsidered and notions of gender hierarchy set aside, it becomes apparent that rather than reflecting gender distinctions, burials were firmly associated with place and group identity. According to Gillespie (2005), people and place were inextricably intertwined at Teotihuacan, so that place was essential to personhood. I suggest that Teotihuacano males and females both may have been economic and political actors within compounds and would have seen their efforts as collaborative rather than dichotomous. I argue that, ultimately, assumptions about gender and gender roles based on modern societies can lead to a detrimental mischaracterization of broader social, political, and economic processes in ancient civilizations.

The Past as a Mirror of the Present

Engels (1972 [1884]) hypothesized that states are detrimental to the status of women because, confined by biological reproduction to the domestic sphere, women are unable to participate in the state's economic and political hierarchies. Traditional Western cultural assumptions similarly dictate that women's roles (whatever they may be) are subordinate to those of men (see Miller, this volume). Such preconceptions about the reproductive role of women result in “public” versus “domestic” dichotomies that uncritically emphasize male dominance and ignore women's agency (Zihlman 1981). Men are associated with action, production, authority, and value, while women are depicted as timeless, static, subordinate, and inconsequential. As a result, gender roles tend to be essentialized rather than problematized.

Stereotypical characterizations of women's and men's social roles often derive from ethnographic and ethnohistoric studies that depict women's roles and statuses as subordinate to those of men. However, feminist researchers warn of uncritically applying ethnohistoric and ethnographic data to the prehistoric past (Brumfiel 2006; Conkey and Gero 1991; Conkey and Spector 1984; Gero and Scattolin 2002; Luedke 2004; Pyburn 2004; Robin 2006; Sørensen 2000; Stahl and Cruz 1998). Pyburn cautions, “Cross-cultural parallels found in the status and treatment of women are more the result of history than of human nature or human biology” (Pyburn 2004:2). The androcentric preconceptions and the assumptions of Western researchers and chroniclers have frequently biased ethnographic and ethnohistorical data by placing greater emphasis and value on male informants, activities, and social roles. Furthermore, archaeological research demonstrates that political and economic changes associated with the market economy and Western cultural values have greatly altered indigenous economies and the roles of men and women within those economies (Pyburn 2004; Robin 2002; Stahl and Cruz 1998). Thus, changes associated with the arrival of Europeans and the introduction of the market economy have greatly altered traditional gender roles and relations, even in the most remote of indigenous societies.

Despite a recent surge in feminist and gender archaeology, misguided assumptions about the essential nature of gender roles persist in archaeological research. Such preconceptions are especially salient in the study of state-level societies where “assumptions about the gendered division of labor and male primacy in political life have led to a body of scholarship which places the rise of the state firmly in the male domain. Women are best considered pawns” (Luedke 2004:47). Even when assumptions about gender are not overtly stated, they are often implied through discussions of patrilineality, emphasis on male social roles, and taking female roles for granted (Luedke 2004). However, by becoming aware of our cultural assumptions, we can begin to deconstruct ethnographic notions of invariable women in a timeless past and to question archaeological projections of present values onto past cultures.

Teotihuacan as a State

This chapter will focus on the height of Teotihuacan from 200–650 c.e. Three phases fall within this period (Rattray 2001): the Tlamimilolpa phase (200–350 c.e.), the Xolalpan phase (350–550 c.e.), and the Metepec phase (550–650 c.e.).

Teotihuacan was a planned, densely populated urban center with a population reaching at least 100 thousand at its height (Millon 1973). The city was arranged according to barrios, some of which held concentrations of craft producers, merchants, or foreigners (Cowgill 1997). Elite residences and ritual structures were concentrated along the Avenue of the Dead, Teotihuacan's four-mile-long north–south axis, according to which the entire city was oriented (Figure 2.1). Two large pyramids lie along this axis, the Pyramid of the Moon, which defines the northern limit, and the Pyramid of the Sun, which lies along the northeast side of the avenue. The pyramids were built early in Teotihuacan's history, around 100 c.e. (Sugiyama 2004). The Teotihuacan Mapping Project identified more than two thousand apartment compounds throughout the city. The compounds were built early in the second and third centuries and appear to have been occupied continuously, remaining relatively unchanged until the city's decline. In general, between 20 and 100 people occupied the compounds and they appear internally diverse (Millon 1973).

Figure 2.1.

Avenue of the Dead, Teotihuacan, Mexico (photograph by the author).

Teotihuacan has been characterized as a society with corporate economic and political strategies (Blanton et al. 1996). In a corporate strategy as described by Blanton and colleagues, “power is shared across different groups and … monopoly control of sources of power is precluded by restrictions on the political behavior of those vested with power” (Blanton et al. 1996:2), suggesting a noncentralized form of governance and minimal social inequality. This is evident in the form of ritual based on collective themes and artwork centered on gods rather than on individuals (Blanton et al. 1996). The political structure at Teotihuacan likely derived from pre-state forms of corporate organization and was governed indirectly through the heads of residential compounds (Cowgill 1997; Pasztory 1997). As there were over two thousand compounds within the city, it would have been difficult to centralize power. While Early Classic leaders may have held considerable power, as evidenced by the mobilization of labor for the construction of monumental architecture such as the Pyramids of the Sun and Moon and the sacrificial burials of the Feathered Serpent Pyramid, Classic period art and monuments center on deities and cosmic renewal imagery that reflect a corporate or collective form of political organization. Further, royal graves have not been located or palaces securely identified, signifying that specific individuals were not idolized or able to centralize wealth and power. The state may have been seen more as a religious than a political entity, possibly ruled by a collection of religious leaders (Sugiyama 2004, 2005).

The Teotihuacan state does not appear to have focused heavily on conquest and tribute extraction, as its administrative control probably extended only slightly beyond the Valley of Mexico. However, Teotihuacan had significant political or economic interaction with various foreign polities, allowing for long-distance exchange of ceramics and exotic raw materials used in craft production (Braswell 2003; Clayton 2005). Craft production may have been at least partially state sponsored and a vital component to Teotihuacan's power (Cowgill 1997). Production appears to have been organized at the level of the community, with apartment compounds functioning as domestic workshops (Sullivan 2006). Workshops specialized in various products including obsidian, ceramics, ground stone, lapidary, and shell (Manzanilla 2004). In sum, given its size, influence, and organization, Teotihuacan can be characterized as a state-level society with corporate political strategies. Maybe it is for this reason that questions of gender have been almost entirely overlooked.

What Do We Know about Gender at Teotihuacan?


One difficulty in interpreting gender at Teotihuacan is that art, such as figurines and murals, rarely represents individuals engaging in gender-specific activities. Portrait figurines (Figure 2.2a), possibly representing male warriors, may be an exception. Generic mold-made faces and androgynous “Gumby-like” bodies that lack clothing characterize these portrait figurines (Berrin and Pasztory 1994). Some researchers have suggested that they are in a spear-throwing stance and may have been clothed in perishable materials (Berrin and Pasztory 1994); however, this interpretation is only speculative. These figurines are nevertheless highly standardized and the mold-made faces lack variability or personalization. Female figurines are sometimes represented carrying infants (Figure 2.3), but little else can be ascertained from figurines in regard to gender roles at Teotihuacan. Biological sex characteristics are usually not depicted or are obscured by clothing, although there are some ambiguous representations of pregnant females, naked female bodies with breasts, and male torsos (Scott 2001). Figurine clothing represents socially constructed genders: males wore loincloths, while females wore loose upper-body garments and long skirts (Figure 2.4a, b). Although it is likely that gender (as well as class and ethnicity) was depicted through clothing styles, headdresses, and stance, to my knowledge these traits have not been systematically tested.

Figure 2.2.

Figure 2.2.

Teotihuacan portrait figurines: (a) Standing figure with headdress, ear spools, and belt, 200 b.c.e.–700 c.e. L83.11.1114, Los Angeles County Museum of Art, on loan by Constance McCormick Fearing (drawing by Santiago Juarez); (b) Figurine heads. Photograph by Joe McDonald, courtesy of the Marian White Anthropology Research Museum at the State University of New York at Buffalo.

Figure 2.2.

Figure 2.2.

Teotihuacan portrait figurines: (a) Standing figure with headdress, ear spools, and belt, 200 b.c.e.–700 c.e. L83.11.1114, Los Angeles County Museum of Art, on loan by Constance McCormick Fearing (drawing by Santiago Juarez); (b) Figurine heads. Photograph by Joe McDonald, courtesy of the Marian White Anthropology Research Museum at the State University of New York at Buffalo.

Figure 2.3.

Small flat-bodied figure with infant, 200 b.c.e.–650 c.e., Teotihuacan. L.83.11.3, Los Angeles County Museum of Art, on loan by Constance McCormick Fearing (drawing by Santiago Juarez).

Figure 2.4.

Figure 2.4.

Gendered figurines from Teotihuacan: (a) Standing female figurine with support, 100–200 c.e. Copyright © Saint Louis Art Museum.

Figure 2.4.

Figure 2.4.

Gendered figurines from Teotihuacan: (a) Standing female figurine with support, 100–200 c.e. Copyright © Saint Louis Art Museum.

While figurines frequently represent humans, Teotihuacan murals primarily focus on the activities of priests and deities. Murals are usually located in elite households and depict ritual activities and processions; thus, they rarely represent everyday social relations and gender interactions. Teotihuacan deities may be gendered—the Great Goddess, for example, has been considered a principal deity, but representations of her are variable and often highly stylized to the extent that an anthropomorphic figure is hardly identifiable (Figure 2.5a). Teotihuacan is thought to have placed a great emphasis on this Great Goddess, whose powers range from fertility to destruction (Berlo 1992); however, her significance and even the attribution of a female gender have recently been questioned (Cowgill 2005; Paulinyi 2006).

Figure 2.5.

Figure 2.5.

Artwork depicting female deities: (a) Fresco fragment of a goddess, circa 650–750 c.e., Teotihuacan. Denver Art Museum Collection: Department of Acquisition Funds, 1965.202, © Denver Art Museum.

Figure 2.5.

Figure 2.5.

Artwork depicting female deities: (a) Fresco fragment of a goddess, circa 650–750 c.e., Teotihuacan. Denver Art Museum Collection: Department of Acquisition Funds, 1965.202, © Denver Art Museum.

The principal Teotihuacan deity may have been female, but murals depict males as primary ritual actors (Berrin and Millon 1988; Berrin and Pasztory 1994). Nonetheless, particular genders are neither glorified in Teotihuacan art nor subordinated. This contrasts with Aztec imperial art that depicts dismembered and mutilated female figures (e.g., the goddess Coyolxauhqui; Figure 2.5b), females in subordinate poses, and the glorification of male warriors (Klein 1993; Umberger 1996). Further, there is no evidence that Teotihuacan art glorifies specific individuals. In fact, “human beings are shown subordinate only to deities, not to other human beings” (Cowgill 1997:136). Thus, females are not represented as inferior to males in Teotihuacan art.

In addition to ignoring gendered activities, Teotihuacan art lacks portraiture and, for that matter, any kind of individuality at all. Figurine faces and masks are standardized with vacant expressions (Figure 2.2a, b); murals primarily depict abstract symbols, animals, and deities; and human figures that are present appear asexual and do not interact (Pasztory 1992). Pasztory (1992, 1997) argues that the abstract nature of Teotihuacan art intentionally sought to de-emphasize the individual to define itself against other Mesoamerican cultures. She argues, “Teotihuacan clearly differentiated itself from its neighbors by negating their naturalistic artistic canons” (Pasztory 1992:289). In sum, the nonspecific nature of facial representation and biological sex in figurines and murals suggests that individual identity was not a primary concern of representational art. Gender-linked activities remain ambiguous at best.


Mortuary analysis is considered by many to be an ideal candidate for studies of gender since biological sex can be determined through osteological analysis and grave goods associated with burials are thought to represent the social roles and identities of the living (Carr 1995). Yet, assumptions about gender roles in mortuary analysis often dismiss women, and the impulse to quantify grave-good “values” may lead to the attribution (or overemphasis) of a gender hierarchy. Objects associated with males are frequently presumed to have greater value and consequently males are attributed higher status because they are associated with high-value items (e.g., Winters 1968). As noted by Gero and Scattolin, “the only way that many societies can be considered to operate as gender hierarchies is by a subtle and presupposed prioritizing of exactly which activities will count as significant indicators of hierarchy” (Gero and Scattolin 2002). Perhaps more important, the preoccupation with determining gender hierarchies often results in nongendered components of burials being overlooked.

The Saxe-Binford approach, which focuses on interpreting social identity, has been commonly employed in mortuary analysis and argues that status differences in life will be reflected by differential burial treatment in death (Binford 1971; Brown 1982, 1995; Goldstein 1982; Peebles and Kus 1977). The relative amount of energy expended on mortuary treatment is argued to be proportional to an individual's former status; thus, individuals with a higher degree of energy expenditure should be individuals of higher rank (Tainter 1978). For this reason there is often a focus on calculating grave-good values and energy expenditure in studies of mortuary ritual. While the Saxe-Binford approach has been insightful in many respects, it nonetheless has its limitations. Because of its focus on status and rank, it often fails to account for variation in “philosophical-religious” factors (Carr 1995), including worldview and understandings of death and the afterlife. For example, a burial recovered by archaeologists may be the result of a long-term process, possibly over generations, rather than an isolated event. Further, a burial may represent relations between individuals or collective social representations rather than a single social identity (Carr 1995; Gillespie 2001). Gillespie argues, “the social persona cannot be considered an essentialized attribute of a single individual—a terminal status—but must take into consideration enacted links to other persons” (Gillespie 2001:78). Every individual has not one, but multiple social identities, which are fluid through time and vary according to different social relations and social roles that may persist even after the physical death of any particular individual.

In a remarkably comprehensive effort, Sempowski and Spence (1994) compiled burial data from excavations at Teotihuacan including five excavated apartment compounds (Tlamimilolpa, Xolalpan, Zacuala Patios and Palace, Tetitla, and La Ventilla) and ritual structures. Using a Saxe-Binford approach, they focused on three main components of mortuary analysis: treatment of the body, techniques of grave preparation, and nature and quantity of artifact offerings in the grave. They evaluated offerings by composing a “complexity score” that accounted for quantity, diversity, and quality of objects. To determine quantity, all objects were assigned a value of 1.0 except for certain small objects (small vessels, obsidian blades, painted slate, miniature vessels, isolated adornos, and clay cylinders), which were assigned smaller values ranging between 0.50 and 0.01. To determine diversity, each different type of object was counted as 1.0. Last, to determine quality, objects were weighted according to complexity of decoration and exotic origin of material. A composite score was then calculated based on all of the above features. This score was then compared with variables such as age, sex, room location, compound location, and time period (Sempowski and Spence 1994).

The most prominent correlation noted by the authors was that residential compounds could be distinguished by overall wealth of offerings. Certain compounds such as Tetitla and Zacuala had greater relative wealth, while others including La Ventilla B, an artisan compound, had less wealth. These distinctions remained steady through time, although there was an increasing disparity through time, indicating a widening gap between the rich and the poor. Additional lines of evidence support these distinctions in wealth: compounds identified as having lower-status burials also had less costly architectural construction and fewer ornate murals (Headrick 1999; Séjourné 1959). This variation in wealth among residential compounds is particularly interesting because it is uncharacteristic of a corporate society.

In terms of gender, the research of Sempowski and Spence (1994) shows no exclusive association of particular types of artifacts with particular genders, making gender attribution through grave goods problematic, if not impossible. Although most artifact types were associated with both males and females in residential burials, Sempowski and Spence found that males were more frequently associated with items of personal adornment than females. This does not include pieces of painted slate that may have been components of headdresses and appear with equal frequency among males and females (Sempowski and Spence 1994).

Ritual contexts, such as the sacrificial burials at the Feathered Serpent Pyramid, do demonstrate greater gender distinctions. Here, males were recovered with slate disks attached to the waist and human maxillae pendants, thought to be a diagnostic of soldiers, while females were exclusively adorned with shell earplugs; however, nobody has attempted to link these earplugs with any particular social position. Both sexes had shell beads and obsidian projectile points (Sugiyama 2004, 2005). It has been argued that these pyramid burials were arranged in relation to cosmic factors and the ritual calendar (Cabrera Castro 2000); thus, gender distinctions may have cosmological significance. Although state-level mortuary ritual appears entirely distinct from the offerings and burial treatments observed in residential contexts, it has likely contributed to notions of gender hierarchy at Teotihuacan. Further research is necessary to fully understand gender distinctions in monumental contexts.

Regarding quantity, Sempowski and Spence state that residential female burials were less likely to be associated with offerings in general, even though the discrepancy was “very slight” (Sempowski and Spence 1994:249).1 Yet, in fact, only 24.1 percent (n= 14) of all females were without grave goods compared to 22.2 percent (n= 12) of males—not a statistically significant difference (χ2= 0.058, df= 1, p= .81). Therefore, the data actually appear to suggest that female burials were equally as likely as male burials to be associated with grave offerings. The investigators also found males to be associated with more complex offerings (primarily within a particular residential subgroup). Table 2.1 shows the complexity scores for residential compounds calculated by Sempowski and demonstrates that males have much higher “maximum complexity scores” than females. Note, however, that the minimum scores are almost identical for males and females (particularly in the earlier phases), indicating that the male gender did not in itself confer a higher status. The large standard deviations for the early male scores indicate that outliers inflated the mean male complexity score. Interestingly, male and female scores converge through time, as there are fewer males with very high complexity scores during the Metepec phase. Exchange relations appear to have become strained during the Metepec phase so that there was less access to luxury goods at this time in Teotihuacan (Sempowski 1992). This change poses an interesting question about changes in gender relations and the roles of men and women during Teotihuacan's decline, an avenue that merits further exploration.

Table 2.1.  Complexity scores of burials from residential compounds (Tlamimilolpa, Xolalpan, Zacuala Patios and Palace, Tetitla, and La Ventilla) by phase
Composite ComplexityTlamimilolpa PhaseXolalpan PhaseMetepec Phase
  1. Source: Data compiled from tables provided in Sempowski and Spence 1994.

Number16201410 6 8
S.D.20.31 6.4518.84 8.5911.7410.49

The differences between the male and female complexity scores calculated by Sempowski and Spence may be partially dependent on the fact that the complexity system is based on subjective interpretation of which objects were more valuable than others. Items more frequently associated with males, for example, items of personal adornment, seem to have been prioritized as objects of higher social value. The assignment of numerical values is also highly subject to interpretation, in which small differences in perspective might translate into profound differences in results. Jade objects (exclusively associated with males) were heavily weighted (10.0) compared to other luxury objects such as conch shells (4.0), shell figurines (3.0), obsidian eccentrics (2.0), and mica, pyrite, hematite, and onyx (1.0). Pieces of painted slate were assigned a low weight (0.20), although the slate may have been components of elaborate headdresses or clothing. Other items such as miniature and small vessels were also assigned a very low weight (0.10), even though these types of objects were exclusively associated with funerary assemblages and would have held special ritual meaning within the mortuary context (Scott 2001). For the sake of comparison, the development of a method that incorporates several dimensions of an artifact's importance (such as complexity), is no doubt necessary and an effective way to approximate value. However, aside from the subjective nature of approximating value, objects also have different social, symbolic, or ritual meanings in any given context that may be independent of economic value (Appadurai 1986; Rodríguez-Alegría 2002). Grave offerings may be incorporated into burials for specific purposes that may or may not be related to an individual's status; thus, it might be equally important to consider burial assemblages as a whole (see below).

Sempowski and Spence conclude that males within a given residential group tended to have higher complexity scores than females within that group. However, because of data limitations this conclusion is based on one phase (Tlamimilolpa) in one compound (La Ventilla B). Exceptions are treated as anomalies. For example, in Cuarto III Cuarto Adobe of La Ventilla B, two females had the most complex offerings, yet this is disregarded because one male had a unique offering—a marine-shell ear spool. Further, in another room (Cuarto Altar III) a young female had the most complex offerings of all other individuals in the room, but this is attributed to flaws in the complexity technique. When we turn to other compounds, including the wealthiest compound, Zacuala Patios, a female had the most complex offering of the Xolalpan phase burials in Zacuala Patios; however, because of the limited data from Zacuala Patios it is difficult to draw any unambiguous conclusions (Sempowski and Spence 1994:260).

Although Sempowski stresses that gender was probably not a primary determinant of social status, from the results described above she nonetheless concludes,

It is not surprising to find that offerings associated with males are generally more complex than those associated with females or that proportionally more females are buried without any offerings whatsoever … Emerging from these observations, then, is the rather predictable implication that males held social positions of relatively greater importance at Teotihuacan than did females … contrary cases stand out as rare exceptions. [Sempowski and Spence 1994:260–261]

Males may have been more likely to hold positions of high status, as suggested by Sempowski and Spence, but the data do not readily lend themselves to this conclusion. Statistically, there is little variation between male and female burials. Moreover, the treatment of complex female burials as “anomalies” rather than as part of the data set overlooks the intricacies of gender roles. The presence of females with elaborate burials indicates that high status (if status is what was being represented) was not defined as a male role. Further, the absence of grave goods from female burials may be a reflection of ritual practices rather than a representation of status. As I discuss below, graves were frequently reopened and objects were removed during ritual ceremonies.

Because systematic mortuary research at Teotihuacan (as at most other sites) has approached gender from the perspective of gender hierarchy, a gender dichotomy and subsequent focus on male dominance is inevitable (even when the data resist this conclusion!). Consequently, assumptions about gender inequality lead to a misrepresentation of social life and minimization of women's roles. By asking the question, Which gender is of higher social status? a hierarchy constraint is immediately created, within which the data must be interpreted. When the data do not conform to the preconstructed dichotomy, they are either made to fit or treated as anomalies. Yet, it is the data that fall outside of our expectations that demand closer scrutiny. As observed by Deetz, “lack of fit serves as a warning that further attempts must be made to bring all of the information into reasonable concordance” (Deetz 1996:198).

Looking Beyond “Gender”

Why is gender so ambiguous at Teotihuacan? Many clues may be overlooked when we focus on gender distinctions and ignore the similarities. To understand mortuary variation at Teotihuacan, we may have to look beyond gender hierarchy. Below I suggest that group identity superseded the importance of the individual at Teotihuacan and, as a result, compound or group identity was emphasized over gender. Social solidarity may have been highlighted at Teotihuacan because a large percentage of male and female residents were immigrants. Isotope studies, archaeological excavation of ethnic compounds, and population estimates suggest that Teotihuacan was sustained by a high influx of foreigners (Price et al. 2000; Storey 1992; White et al. 2004). Group identity may have been reproduced in the mortuary context in response to this influx to serve as a means of boundary maintenance (Barth 1969), even as foreigners became incorporated into the group (which we know they did because they were usually so well integrated that they can only be identified through bone chemistry analysis [Price et al. 2000]). Emphasizing distinctions would probably have been counterproductive.

Teotihuacan compounds demonstrate remarkable continuity through time (Manzanilla 2004) and likely represent what Joyce and Gillespie (2000) refer to as “houses,” in which individuals derive identity and status from their association with a particular compound and its heritage rather than blood ties or kinship. An important component of a “house society” is an emphasis on shared social identity. Sandstrom (2000) notes that among modern Nahuas, house members (often including non-kin) emphasize their shared identity by naming their house compound and by taking this name as a surname. Although the names of individuals may be inaccessible to archaeology, the importance of compound identity in the past can be assessed through continuity, shared labor, and emphasis on place (Gillespie 2005; Joyce and Gillespie 2000).

In the discussion that follows I take a closer look at identity at Teotihuacan through previously published mortuary and household data. Because identity would have been specific to individual compounds, compiling data from multiple compounds would obscure variation and similarity within compounds. Thus, my discussion focuses on variability and patterning within a single residential compound.

Burials—Tlajinga 33

Tlajinga 33 was a craft production compound occupied for almost five centuries between approximately 200 c.e. and 650 c.e. Production focused on lapidary work including greenstone, slate, onyx, and shell crafts; however, in the last two centuries of its occupation, Tlajinga shifted to production of San Martin Orange ware (Widmer and Storey 1993). San Martin Orange ware is commonly found in high-status residences located near the center of the city rather than in low-status areas. Its association with state-sponsored storage (Robertson 1999) suggests that the production of these ceramics was at least partially subsidized by the state. The low status of Tlajinga 33 has been inferred from its poor architectural construction, including insubstantial building materials, poorly implemented and irregular layout, and unimposing size (Widmer and Storey 1993).

Although not incorporated into Sempowski and Spence's mortuary analysis, mortuary data from Tlajinga 33 were included as an appendix to the volume (Sempowski and Spence 1994). Because the site was excavated in the 1980s its archaeological and laboratory methods should be more reliable than those from early Teotihuacan excavations. Sixty-five individuals were recovered from intact contexts, and 141 individuals were recovered from disturbed contexts (N= 206). Fifty-two percent of these individuals are adults and 48 percent subadults. Only primary and secondary interments are considered in the following analysis. Although scattered human remains in fill or burial pits were common, these were not considered intentional burials and therefore are not included in this study. Nineteen males and 13 females were in the sample, an insignificant difference in numbers between the sexes (Storey and Widmer 1999). The burials ranged through three phases (Tlamimilolpa, Xolalpan, and Metepec), although only two burials date to the last phase. For the purpose of this analysis I quantified objects as simple counts, not complexity scores, to avoid thinking of offerings solely in terms of economic value. Instead, I classified ceramic vessels into five categories to consider function and ideological significance: ritual, serving, storage, cooking, and miniature. The other artifact categories included craft production tools, raw materials, obsidian, shell, worked stone, and ground stone.


Burial location at Tlajinga 33 was not determined by gender but by age.2 The majority of burials (including males, females, and subadults) were located in rooms or in patio or courtyard areas. Some burials were also located outside the walls of the compound or under the walls themselves (also not exclusive by age or sex) (Widmer and Storey 1993). However, only subadults were buried in middens.3 Further, of six burials directly associated with the courtyard altar, five were neonates or infants. The single adult was a male with one of the most unusual, if not the most elaborate, offerings. While most adult burials were associated with a handful of ceramics, obsidian, and sometimes a carved stone bead or pendant, this single burial had an elaborate headdress and a garment with four thousand painted shell beads. The burial had been reopened and additional offerings deposited. Perhaps even more interesting, the adult dates to the Early Tlamimilolpa phase, the earliest occupation of the compound, while the infants date to a slightly later period (Late Tlamimilolpa/Early Xolalpan phase). At La Ventilla B, nine infants and only one adult were similarly buried in direct association with an altar—only in this case the adult was a female (Sempowski and Spence 1994). And at the partially excavated compound of Bidasoa, all of the burials in the central patio were neonates located under or in close proximity to the altar (Sánchez Alaniz 2000). Of the 18 burials recovered in the Bidasoa complex, 15 were associated with the central patio and the one adult was a female buried under the floor of an adjoining room. Unlike the other compounds, there was no centrally buried adult (Sánchez Alaniz 2000).

The fact that infants and neonates were buried in spatially segregated areas suggests they possessed social roles that differed from those of adults. Such distinctions in mortuary treatment may indicate that infants lacked a degree of “personhood” and, thus, possessed a liminal role (Gillespie 2001; Senior 1994). In other words, because infants may not have achieved individual personhood, they may have served as an ideal link to the spiritual realm. Among the Maya, for example, social memories and identities were maintained through passing names and heirlooms to the descendants of deceased individuals (Joyce 2000). Similarly, among other indigenous groups infants are perceived as embodiments of reincarnated ancestors (Crass 2001). I argue that at Teotihuacan the pattern of burying infants by an altar that contained the remains of an important ancestor may have been a way of ensuring the continuation of compound identity into the future, particularly in times when this identity might have been threatened.

This interpretation is supported by the presence of “heirlooms” associated strictly with infant burials. Heirlooms, or vessels dating to earlier time periods (at least two hundred years earlier), were found only with the burials of three different infants. Moreover, the importance of ancestor veneration was clearly marked by the fact that several adult burials were repeatedly reopened for the inclusion of new offerings and likely for the removal of old ones. In addition, subadults tend to be overrepresented in Teotihuacan compounds (Storey 1986). Storey (1986) attributes this overrepresentation to an unusually high infant mortality rate, but it is also possible that selective burial practices privileged burying subadults in household contexts. It is known that there is a dearth of human burials overall at Teotihuacan given its population size; thus, it is likely that alternative burial practices for adults also occurred (Headrick 1999).

In sum, the continued placement of new offerings into old burials, the placement of heirlooms into infant burials, and the association of infants with an early important adult burial by the compound altar suggest that mortuary ritual and burial location were closely linked with long-term continuity and group identity. The placement of burials under house floors further reflects the importance of the household as a component of social continuity.


Storey and Widmer's analysis of the Tlajinga 33 burials concludes, “Males tend to predominate in the higher statuses although a few females were accompanied by impressive offerings” (Storey and Widmer 1999:215). To evaluate this conclusion I compared adult male and female offerings (subadult burials cannot be sexed). My analysis did not find a significant difference in the distribution of offerings between males and females. More males (n= 15) than females (n= 9) are associated with offerings; however, there are more males overall. The difference becomes insignificant when proportions are compared: 9 of 13 females (69 percent) have grave goods and 15 of 19 males (79 percent) have grave goods, which is not a statistically significant difference (χ2= 0.389, df= 1, p= .533). Though a higher percentage of females compared to males are without artifacts, three of the four female burials without grave goods had been disturbed or opened, so it is possible that objects were originally included in these graves but were later removed.

Males had a slightly higher median number of objects, while females had a wider variation in number of grave offerings (Figure 2.6). The ranges of male and female burial offerings overlap. The individual with the most objects is a male with 4,017 artifacts (this individual was removed from the plot for the purpose of scale); however, four thousand of these artifacts are beads from one garment, so more accurately this individual was buried with 18 artifacts (17 objects and one fancy garment). This individual is the person buried underneath the altar described earlier; thus, it is likely that this individual was being honored for his association with the compound identity rather than his individual status per se.

Figure 2.6.

Number of artifacts associated with male and female burials at Tlajinga 33.

The normal range of variation is between zero and nine artifacts per individual (Figure 2.7), thus anything above this range (>10 objects) I considered to be “high status.”4 Exactly one-third of males and one-third of females with offerings fall outside the range of normal variation (assuming the altar burial has 18 items). Of the total sample, 23 percent of females and 26 percent of males have “high-status” offerings. Both males and females fall into the outlier and far outlier categories. Thus, males and females appear equally likely to have achieved “high status” (assuming that status is being represented by artifacts). The difference in the frequency of exotic goods for males and females is also insignificant. Exotic, or imported, goods are generally thought to be found more frequently with “high-status” burials, as they are more difficult and costly to procure than local goods. While exotic materials are associated with nine males and only five females, the difference in the frequencies of exotic goods for males and females (47.3 percent of males and 38.5 percent of females) is insignificant (χ2= 0.249, df= 1, p= .618). No exotic materials were exclusively associated with one sex or the other. The sample size is too small to determine whether one sex is more or less likely to be associated with any given material.

Figure 2.7.

Box plot of the number of artifacts associated with all burials at Tlajinga 33. Key: asterisks = outliers; circles = far outliers.

Analysis of skeletal remains confirms that males and females at Tlajinga did not have differential access to resources. The majority of individuals in the compound were unhealthy, with 68 percent of the population exhibiting porotic hyperostosis, linked to chronic undernutrition and illness in childhood (Storey 1992). Dietary differences and offerings were not linked to sex or place of origin. Stable carbon isotope ratios in bone collagen and oxygen isotopes in enamel phosphates indicate that 29 percent of the inhabitants (both males and females) were foreigners (White et al. 2004).

Overall, Tlajinga 33 offerings appear rather standardized. As observed in other Teotihuacan compounds (Sempowski and Spence 1994), variation is mostly in the quantity, not types, of items. The offering complex at Tlajinga generally includes a ceramic vessel (usually a bowl), an obsidian blade, and sometimes an “exotic” item (although these decrease in later periods). The majority of vessels (77 percent) were serving pieces, nine percent were for storage purposes, and eight percent were miniature vessels. Figure 2.8 shows the types of vessels that were associated with the Tlajinga burials. There was one strictly ritual vessel, an incensario associated with a male. Utilitarian cooking wares were completely absent. Not surprisingly, there was no distinction according to sex. Thus, the vessels included with burials were not meant to signify the roles an individual held in life—otherwise, we would expect functional vessels to be present. Instead, these offerings served an important ritual or performance function that was not gender specific.

Figure 2.8.

Vessel types associated with Tlajinga burials. Note that the burial assemblages entirely lack cooking vessels.

Objects used or made in craft production at Tlajinga such as obsidian blades, bone needles, greenstone beads, shell, and slate were all associated with both males and females. The only exception is that mica was associated exclusively with males. In addition, one bone tool was with a female, one “toy” mano and metate was associated with another female, and one ground stone polisher was recovered with a male. The female with worked bone also had a large number (N= 30) of obsidian blades. Many of the items recovered might have been used in craft production; for example, the miniature mano and metate were mostly likely used for grinding pigments and the polisher for working shell or ceramics. The presence of these items provides some indication that both males and females participated in craft production; however, the sample size is too small to tell us whether particular activities were sex specific. Perhaps more important, these burials tell us that an individual's identity as a craft-producer (and hence a Tlajinga resident?) was more important than sex in the mortuary context.

Households and Compounds

Since identity is strongly tied to place (Low and Lawrence-Zúñiga 2003), the household is an essential element for understanding gender and Teotihuacano identity. Houses become socially constructed (take on symbolic meanings) “through people's social exchanges, memories, images, and daily use of the material setting” (Low 1996:861). The “grammar of space,” as argued by Glassie (1975), dictates that the cultural rules of spatial organization can be “read” through the study of architecture and in the organization of activities since they are expressed in a patterned way (Lefebvre 1991; Sutro and Downing 1988). Gender is both implicated in the construction of space and experienced through it (Sørensen 2000). As noted by Sørensen, it is “not just about how artefact distributions or spatial division correspond to gendered activities; it is also always about how through the participation in, and performance of agreed meanings, people are confirmed [or not] in their gendered identities” (Sørensen 2000:166; brackets added).

I consider data from two contemporaneous artisan compounds: Oztoyahualco, extensively excavated by Linda Manzanilla (1993, 1996), and Tlajinga 33, described above (Widmer and Storey 1993). Manzanilla conducted a detailed study of household activity space at the Oztoyahualco compound through the analysis of chemical traces and microartifacts from floor surfaces (Manzanilla 1993, 1996).

Both compounds were demarcated by external walls and were internally divided into households. Each household had its own areas for food production, sleeping, storage, burials, and ritual (Manzanilla 1996; Widmer and Storey 1993). Analysis of activity spaces at Oztoyahualco by Manzanilla demonstrated that kitchens and storage areas were adjacent. Separate rooms were used for lithic production, butchering/hide processing, and breeding animals (Manzanilla 1993). Interestingly, at both sites courtyards were used strictly for compound ritual; no production activities took place within these spaces. At Tlajinga, the ceramic production area and locations for lapidary industry activities were situated on the peripheries of the compound. There were no burials within the craft production areas, suggesting that there was a segregation of production and ritual spaces.

Because houses are the locus of everyday activities and reflect daily behavior patterns, social relations become mapped out in space and can thus be observed archaeologically. Analysis of activity areas, room accessibility, routes of movement, physical isolation, access points, and visual pathways promotes a better understanding of the possibilities of space utilization and the potential consequences for social interactions (Blanton 1994; Gilchrist 1999; Stockett 2005; Sweely 1998). At Teotihuacan, compound organization varied. Hopkins's (1987) study of the spatial organization of six Teotihuacan compounds (Zacuala, Jaguars, Tetitla, Xolalpan, Tlamimilolpa, Yayahuala) determined that while some compounds had limited routes for movement between and among its households, others had multiple circuits to facilitate movement around all parts of the compound with no apparent access hierarchy. Hopkins notes that in Zacuala “one strategically-placed person could have regulated the comings and goings of everyone in the compound” (Hopkins 1987:377). Such spaces organized with few access points and deep entries (requiring passage through other rooms for access) reflect greater social control. However, other compounds had many access points and shallower entries, fostering communication, group interaction, and cooperation. Further research would be required to establish an explanation for these differences, but Zacuala's apparent high status suggests that there was greater social hierarchy among elites.

Hopkins (1987) also found that while some compounds appeared to have been organized around a central patio, others had no single prominent central focus. Yet, even compounds that lacked a central patio appear to have been integrated by broader compound goals and identities. For example, at Oztoyahualco, Household 1 had evidence for rabbit breeding, Household 3 seemed to have engaged in (ritual?) rabbit butchering, while Household 2 had a stone rabbit effigy in its courtyard. To the Aztecs, the rabbit was associated with the alcoholic beverage pulque and lunar symbolism (Sahagún 1979[1959]). Although the missionary Sahagún associated rabbit symbolism with gluttony and malevolence, Anawalt (1993) argues that pulque was used ritually by the Aztecs and may have carried a more auspicious connotation “perhaps relating to a higher state of exalted existence” (Anawalt 1993:35). The rabbits of Oztoyahualco may have held a similar significance and ritual focus for its residents. In addition, plaster polishers were recovered throughout the compound in large numbers, possibly related to an important compound-wide activity (Manzanilla 1993, 1996). Thus, the Oztoyahualco households seem to have been united by an overall compound identity relating to the rabbit and possibly plaster-polishing. At Tlajinga, residents of all three households were also united through compound-wide plastering and craft production activities.

Likewise, household differences may reflect compound integration. For example, at Oztoyahualco the largest courtyard was in Household 1, Household 2 had the largest number of burials, and the most elaborate burial was recovered from Household 3. Each household used its ritual space differently, suggesting that these were shared rather than discrete spaces. A large number of foreign wares in Household 1 may be not a function of status but rather a reflection of group-oriented ritual or feasting occurring within a space large enough to accommodate the compound population. The large number of burials and the rabbit sculpture in Household 2 suggest that this may have been an important location for compound ritual and ancestor veneration. At the Tlajinga site, there was one primary courtyard that Widmer and Storey (1993) argue may have been used for compound-wide ritual. This is the same courtyard with the altar and infant burials mentioned above.

So what can we infer about women's roles and gender relations? In both compounds, courtyards and patios were generally accessible and large enough to accommodate all compound members; thus, spatial organization does not suggest that there were restrictions to ritual participation within the compound (except maybe in elite contexts?). This would provide occasion for contact between men and women in the same and different households. Although one might be tempted to classify courtyards as “public” space in contrast to rooms or “domestic space,” I would disagree with this dichotomy. Courtyards are not public space, as they are within the center of enclosed compounds and are therefore inaccessible to outsiders. If we consider courtyards communal spaces at the level of the compound, then we must conclude (on the basis of burial data and spatial organization) that both men and women would have participated in compound activities. Such spaces emphasized social continuity, ancestor veneration, and, therefore, compound identity, especially since no evidence for production activities occurred within these courtyards.

Further, both men and women were involved in craft production. Ceramics production was a compound-wide activity and likely required pooled labor from all household members to maintain the high temperatures necessary for kilns (Widmer and Storey 1993). The large ceramic production area at Tlajinga suggests that men, women, and children engaged in cooperative household labor. Ethnographic research on the modern Maya demonstrates that large-scale household pottery production requires gender collaboration (Nash 1993), in which women are the producers of pottery and men gather wood, collect clay, and sometimes fire pots. Assuming that craft production was partially sponsored by the state in Teotihuacan (Cowgill 1997), craft producers would not have needed to dedicate substantial time and labor to agriculture, freeing the labor of women, men, and children for craft production.

At Teotihuacan, compounds therefore appear to emphasize group interaction, shared social and ritual space, and collaborative labor. If women and men engaged in collaborative labor and identity was conceptualized at the level of the compound, it seems reasonable to presume that gender would not have been a salient marker of status.


The expectation of gender hierarchy can result in misguided conclusions and unexplored alternative interpretations. Teotihuacan demonstrates that state-level society does not inevitably equate with gender inequality. There is no evidence to suggest that Teotihuacanos devalued women or female roles. Further, within artisan compounds at least, it appears that most craft production involved the collaborative labor of males and females. Both male and female burials were associated with craft production tools and similar ritual objects, while compound organization suggests that households worked together to meet production goals. Looking across compounds we find that both male and female burials had elaborate offerings and were recovered beneath altars and from ritual and domestic contexts. Teotihuacan may be an important example of a state-level society in which not only were gender relations characterized by relative equality but also women had opportunities to gain great authority.

However, a survey of both art and burials at Teotihuacan has provided little insight into gendered divisions of labor. In fact, there are no clear distinctions represented between male and female roles in either iconography or burial offerings. I am not suggesting that there were no differences between male and female activities or social roles; however, I argue that the lack of emphasis placed on gender differences reflects a worldview in which male and female were not necessarily seen as opposed but as collaborative. Our determination to identify gender dichotomies may in fact tell us more about ourselves than about the past.

When we look beyond gender hierarchies at Teotihuacan, we find that compound identity was foregrounded over individual identities. Even though Teotihuacan had a high influx of foreigners, immigrants incorporated into existing compounds are not distinguishable from native Teotihuacanos, except by chemical analysis that can detect change of residence. Compound identity may have been reproduced in response to this influx of outsiders, serving as a means to incorporate newcomers while differentiating resident groups from ethnic enclaves (Barth 1969). This process may have been especially important for maintaining Teotihuacan's corporate political strategy. Gillespie argues that in a house society, “houses are corporate, long-lived units that are organized for specific ends. House members strategically utilize relationships of consanguinity and affinity, real or fictive, in order to legitimate expressions of unity and perpetuity” (Gillespie 2000:468). At Teotihuacan, burial ritual and representational artwork suggest that individual identity was of secondary importance. The integration of households and individuals within compounds is evidenced by organization and activity patterns. Further, in house societies, “ancestral spirits as well as … names should be considered immaterial ‘house’ property that is curated across generations and reproduces the social unit” (Gillespie 2001:95). A focus on ancestor worship, household integration, and faceless representations in artwork offered Teotihuacanos an important sense of collectivity and social continuity despite the change and diversity that no doubt accompanied the growth of a thriving city.

A focus on compound identity did not, however, prevent social hierarchy between compounds. Construction techniques, burials, and household organization suggest that compounds as a whole may have been distinguished according to status, even if there were not great disparities. Nonetheless, Teotihuacan's focus on collective identity at the expense of individuality may have been an important political strategy. From her analysis of art and iconography Pasztory argues, “Teotihuacan was a culture with a utopian view of the world, in which the individual was de-emphasized for the sake of the group, but in which the citizen members enjoyed high status and material benefits as part of the group” (Pasztory 1992:288). Hence, a focus on gender hierarchy at Teotihuacan would be misleading, as it would emphasize distinctions that Teotihuacanos themselves probably regarded as of little consequence. The critical consideration of gender in archaeology thus helps reveal the character of identities that structured social units and interaction in the past, revealing to archaeologists the dynamics of social continuity and change in ancient civilizations.


  • 1

    These data must be interpreted with caution since they derive only from La Ventilla B, the only compound to have adequate data to ensure an accurate count of burials lacking artifacts. Thus, it is unclear whether other compounds would have followed a similar trend.

  • 2

    Sempowski (1992) also notes that differential burial practices are highly determined by age—in particular, infants and fetuses tended to be placed in ceramic vessels and were rarely exposed to fire or cremated.

  • 3

    I considered individuals who were 16 years of age or younger “subadults.”

  • 4

    I put “high status” in quotes because of my earlier discussion about the problems with determining status by quantifying grave goods. I developed this method for the purpose of evaluating prior claims about male and female status. While this method is simplistic, the goal is to get a general characterization of the mortuary assemblage so that I can then move beyond gender characterizations and focus on the assemblage as a whole.