Figurine ontologies, household ritual assemblages, and gendered concerns at Postclassic Xaltocan, Mexico

Historically, most Aztec archaeologists have taken a representationalist approach to figurines, devoting a great deal of attention to discerning whether a figurine represents a person or specific deity, what meanings they had, and what they can tell us about past aesthetics (see Overholtzer 2017). While these are valuable pursuits, there is a growing realization within archaeology that figurines may not have been just representations that signify something else: they may have been efficacious beings or entities themselves (Meskell 2017). The ontological divide particular to modern Western culture—glossed as animate people and inanimate objects—is now recognized as inappropriate for the Aztecs and many other Indigenous peoples of the Americas. This paper takes a different approach by considering the ontology of Aztec figurines with the goal of creating reconstructions that are more in line with pre-Hispanic worlds. Rather than asking what Aztec figurines depicted or represented, it asks what these figurines were and what they did in the home. Drawing on Aztec philosophies recorded in the Colonial period, I argue that Aztec figurines were momentary orderings of sacred and vital energy that came into being in the context of relational ritual household practice with humans. Applying theories of practice and materiality, I consider the agentive role of both humans and other-than-human entities that assembled together and co-created through household ritual practice. Consideration of their depositional context, form, and ethnohistorical mention helps us understand the nature of their embodied engagement with humans, as people solicited their help to ensure household success and health and in response to stressful situations. I conclude that these figurines were vital, vibrant beings or entities that stood, looked back at and spoke with household members, housed powerful spirits or communicated with them on the household’s behalf, received ritual offerings, accompanied and protected human bodies when worn as pendants, and played essential roles within household ritual practice. This paper also takes a different approach in its focus on gender. Most Aztec figurine studies to date have focused on female figurines to the exclusion of other forms, even though Aztec figurines include representations of both men and women (Brumfiel 1996; Overholtzer 2012). This is perhaps because representations of women are common in Mesoamerican figurines, and because their nature as iconographic representations lends themselves well to the study of gender identity and gender ideology. This is also in line with most of the early gender archaeology studies—and indeed most today—which focused on “seeing” women in order to counter the androcentric bias that had long plagued archaeology (Conkey and Spector 1984; Tringham 1991). In contrast, this paper focuses on gender “as a symbolic system that structures social and economic relations within the household and the larger community” (Hendon 1996: 49), and it takes a more inclusive approach to studying gendered daily practices in the household, with explicit attention to men’s roles. It does so by exploring the intensity of household engagement with a sample of 236 Aztec figurine men and women of the flat-backed type recovered from the site of Xaltocan, Mexico. Xaltocan provides an ideal case study context because of 1. its long occupational history spanning the entire Postclassic period, and 2. its large assemblage of figurines recovered in archaeological surface collections, test pits, and excavations conducted across the site by Elizabeth Brumfiel. This paper argues that the intensity of household figurine use—that is, when figurine entities of various kinds were more or less active and ubiquitous in the home—provides insight into changes in household ritual practices and household concerns over time within shifting social, political, and economic contexts (see Halperin 2014 for a Maya figurine study articulating the intersections between household and broader political economies). Assessment of the intensity of household figurine use at Xaltocan reveals the Middle Postclassic to be the period of the most fear, uncertainty, and instability for commoners, more so than the Late Postclassic period of Aztec rule, which is often characterized as the era of greatest household hardship.

Figurine entities

In line with a broader ontological turn in the humanities and social sciences (Descola 2013), archaeologists have recently begun to recognize the existence of alternate ontological frameworks wherein the world is composed of multiple, entangled lively or vital beings, entities, or forces—human and other-than-human. As Baires (2017: ix) explains, archaeological “materials need to be examined not as static markers of broad-scale changes to cultural groups but as embodied players in vibrant past lives of people.” And as Weismantel (2013: 21) argues, pre-Columbian art objects were vital “interlocutors” in a social world, and had an “active, working life” and a materiality that was just as important as their iconography. Within pre-Columbian archaeology, this theoretical turn has resulted in a complete reconceptualization of the role of such varied things as pots, sculptures, mortuary assemblages, and water (Alberti and Marshall 2009; Baltus and Baires 2017; Buchanan and Skousen 2015; Fowler 2013; Harrison-Buck and Hendon 2018; Joyce 2017; Watts 2013; Weismantel 2018; Wilkinson 2013). Within Mesoamerican figurine studies, Formative period figurines have most often been reframed as more active participants in social worlds, perhaps because of previous arguments asserting their intentional breakage. Meissner, South, and Balkansky (2013) argue from a new animist perspective for a “relational metamorphosis” between human and nonhuman beings in the Formative Mixtec world. The efficacy and animacy of these figurine beings can be reconstructed, they contend, based on depositional treatment that parallels that of human beings, especially patterns of intentional breakage and burning that may have served to deactivate their social lives (see Chapman 2000 for figurine fragmentation more broadly). Similarly, Marcus (2019) argues that ancient Oaxacan figurines may have been “animated” and transformed into individuals through ritual utterances, especially in the context of figurine scenes. Finally, Rice (2019) argues that Formative Lowland Maya figurines were social beings that “anthropomorphized the cosmos.” While approaches to vibrant things range from decidedly anthropocentric to decidedly not, I am first and foremost an anthropologist, and I remain fundamentally concerned with the practices, experiences, and ontological frameworks of people in the past. I follow scholars such as Joyce (2012, 2015) and Antczak and Beaudry (2019) in finding promise in the intersection of theories of practice and materiality. Joyce (2015: 188) argues that archaeologists encounter “traces,” which are “the result of continuous human action in assemblages where human/nonhuman distinctions are not particularly helpful.” Similarly, Antczak and Beaudry (2019) offer the conceptual framework of an assemblage of practice, a “dynamic gathering of corresponding things entangled through situated daily and eventful human practice.” Finally, I also find utility in the perspective of relational personhood, articulated by Harrison-Buck and Hendon (2018: 10) in the following way: “humans, plants, spirits, animals, and objects are all potentially persons and are among the many receptable-bodies where the numinous comes to reside, sometimes remaining dormant until an interaction occurs with another relational being, bringing it to life.” In this vein, this paper aims to reconstruct the dynamic gathering of human and figurine entities in the context of household ritual practice. To begin to understand Aztec figurine ontologies—that is, what these figurines were within the household—I turn to Aztec philosophies and metaphysics recorded in the early Colonial period in the form of ethnohistories, pictorial histories, ritual calendars, Nahuatl dictionaries, and so on (León-Portilla 1963; Maffie 2014). Maffie (ibid.) builds on the work of Hill Boone (1989), Burkhart (1989), Read (2000), and others to argue that Aztec philosophy featured ontological monism—the idea that reality consisted of just one thing or kind of stuff: a sacred, life-giving energy or power called teotl. Poorly translated by the Spanish as “god,” teotl infused the entire cosmos with vitality. Deities themselves were “mere momentary ritual constructions,” “momentary arrangements,” just as objects that we today consider to be inanimate were (Maffie 2014: 33). A concept similar to teotl appears in many Mesoamerican languages (Edmonson 1991; Houston and Stuart 1996: 292; Hunt 1977: 234; Joyce 2010: 56; Marcus 1983; Monaghan 2000: 25; Stuart 1996: 162-164). Aztec metaphysics were, therefore, nonhierarchical, “den[ying] any principled metaphysical distinction between transcendent and immanent, higher and lower, or supernatural and natural realities, degrees of being, or kinds of stuff” (Maffie 2014: 12). Teotl is defined not categorically, but processually: reality, existence, and all things are defined in terms of becoming, transforming, and moving. “The cosmos and all its inhabitants are defined by continuous becoming and transformation. As a consequence, reality contains no immutable or permanent entities, structures, or arrangements” (ibid.: 165). All things are energy-in-motion, fleeting configurations of energy, and it is this energy or power that is agential, effecting change and making things happen (ibid.: 24-25). Here, Maffie contrasts an Aztec metaphysics of Becoming with a modern Western metaphysics of Being. Nahua ethnographies (Bassett 2015; Sandstrom 2008, 2009; Sandstrom and Sandstrom 1986) have described the ontological transformation of ordinary paper into animate cut-paper figures that are dressed and receive blood offerings on an altar. By being raised up, dressed, and invoked in prayer, the cut-paper figures “come to embody highly animate deities, totiotzin and Chicomexochitl, a transformation that registers linguistically and ontologically,” and, as a consequence, they bring about rain (Bassett 2015: 23). Interestingly, these totiotzin (“our gods” in Nahuatl) are not only living beings, Bassett (ibid.) argues, but also family members cared for in everyday life. Similar cut-paper figures were found in the exceptionally well-preserved Offering 102 of the Templo Mayor of Tenochtitlan (Barrera Rivera, Gallardo Parrodi, and Montúfar López 2001), suggesting a pre-Hispanic precedent to these god figures. Bringing an ontological perspective into archaeology, Walton (2020: 293-294) argues for the active role of bloodletting implements in ancient Mesoamerican household rituals carried out to communicate with ancestors and petition the help of supernatural forces: “through these physical acts of self-harm, the blood and bloodletting implements themselves were seen as powerful actors that affected, remade, or gave life to people and things in the world.” As Pastrana and Athie (2014: 100) put it, “The same deity penetrates the flesh and is the recipient of the blood offering, since obsidian is the god and the instrument.” Obsidian was understood to have what we consider magico-religious and medicinal qualities (ibid.). Obsidian blades protected unborn children from harm from eclipses (Garibay 1996: 145). Placed in a bowl of water in the home, they protected family members from bad spirits (ibid.: 146). Obsidian in various forms was also used in healing illnesses (ibid.: 151; Hernández 1959: 411-412). It could be a transformative agent. As Heyden (1987: 84-85; cited and translated by Pastrana and Athie 2014: 85) explains,

The mothers who offered their children to Tezcatlipoca made sure they were covered with the paste, “which was the same with which they covered this idol and with which the priests and ministers of this temple covered themselves” (Durán 1967: 48). Just like the priests, with the black paste they felt invincible because this turned them into images of the god.

Thus, for the ancient Aztecs, the volcanic glass we call obsidian was clearly an active, vital substance and produced effects when seen and felt by the human body. Similarly, I argue that Postclassic central Mexicans would have understood both a human being and a figurine to be fundamentally made of the same vital energy, to be constantly transforming and moving, and to have the ability to effect change in the world. That is, I suggest that figurines were momentary orderings of sacred, vital energy. The ubiquity of figurines in Aztec household contexts and the longue durée of their use in household ritual indeed suggest that their users understood these to be efficacious entities, and such a view is supported by descriptions of idols or figures, interpreted by scholars to refer to figurines, in the ethnohistoric record. Like Weismantel (2013: 23), who draws on Mitchell (2004), I am not concerned with whether these figurines had effects that we today would recognize as such—that is, I am not interested in debating whether agency should be attributed to figurines. Rather, I argue that we will misunderstand and erroneously reconstruct human-figurine engagement unless we consider that Postclassic central Mexicans acted as though they did. Fray Diego Durán (1971) and Hernando Ruíz de Alarcón (1984), writing in the late 16th century and early 17th century, respectively, both mention idols or figures in their accounts of native religious practices. We are told that figurines were hung over agricultural fields because of “pagan superstition,” apparently to ensure agricultural fertility (Durán 1971: 419). Ruíz de Alarcón mentions that figurines were inherited and kept in the house on the altar in “little baskets inside boxes for better keeping and veneration” (1984: 51). Alternatively, he mentions that idols would be placed inside granaries and used with incantations for the preservation and longevity of grain (ibid.: 51, 127). Figurines were also placed along the road, and people would make offerings to them and make petitions for “children, wealth, a long life, a family, or health” (ibid.: 57). Ruíz de Alarcón further explains that petitions would be made so “that the deity whom they believe resides there be favorable to them, or that nothing bad happen to them on the voyage they are making, or to have a good harvest, or for similar things, especially the sick on the advice of their seer-doctors” (ibid.: 52). Durán confirms their use for avoiding illness, explaining that figurines were attached to little girls’ wrists so that “they could avoid illness and that no evil would befall them” (1971: 420). Durán (1994) also details daily rituals involving figurines performed by women whose husbands were away at war. After ritual sweeping and cleansing, a wife would enter the household shrine, where she would hang the femurs of her husband’s former captives, and where she would burn incense in a brazier “before all the little household idols, which were many” (ibid.: 162). While the incense was burning, she would perform incantations in which she asked deities to lend her husband life and allow him to return home safely. Iconographic analysis of Aztec figurines supports the argument that Postclassic central Mexican people engaged figurines in their ritual practices designed to avoid harm in household productive and reproductive endeavors. Scholars have argued at length whether Aztec figures represented deities (and if so, which deities), deity impersonators, generalized spirits, or normal humans (Klein and Lona 2009; Millian 1981; Smith 2002). However, from a materiality perspective, the goal of such identification is itself problematic, since those categories were not ontologically distinct in Postclassic central Mexico. Indeed, the very ambiguity of this figurine iconography may be precisely the point, as entities, essences, and forces may have moved between and temporarily resided in bodies made of human flesh, fired clay, copal incense, and amaranth dough (see Klein and Lona 2009 for a discussion of copal figurines). Thus, Aztec figurines may have served as teixiptla, defined by Molina (2004) as an image, substitute, or delegate of a deity, and commonly translated as localized deity embodiment (Bassett 2015; Hvidtfeldt 1958; Townsend 1979). Although scholars more often discuss the human deity impersonators who played a central role in imperial rituals, stone images and figures made of amaranth dough are also examples of teixiptla (Clendinnen 1991:252; see also Dehouve 2020a; Reyes Equiguas 2005; Turner 2020 for archaeological identifications). These were not simply representations, but rather were the sacred substance itself, a material manifestation of the divine energy of teotl. Debate about the figures represented iconographically in Aztec ceramic figurines notwithstanding, there is utility in the identification by these figurine scholars of a prominent health and fertility theme. This includes a concern with the metaphorically linked concepts of human health/fertility/reproduction and agricultural fertility (Kaplan 1958; Parsons 1972). This focus is evident by the presence of iconography associated with wind, rain, or fertility deities in codices, or by the presence of representations of women who are pregnant or carrying children. A concern with household female production is also indicated by representations of cotton and weaving battens in headdresses of women, and what some scholars see as clear deity associations with Xochiquetzal, the patron goddess of weaving and sexuality (see Parsons 1972). At least some of the figurines may have served either as a temporary embodiment of and container for a supernatural essence or as a communicative channel for petitioning deities or spirits for help in contexts over which they presided. While Spanish priests certainly misunderstood much of Indigenous philosophy and ontology, a conceptualization of figurines as temporary bodies for animate potencies is consistent with Ruíz de Alarcón’s (1984) description of ritual practice as petitioning the deity that resided in the idol. López-Austin (1990:139) similarly argues that the gods were forces that resided in their images or relics, considered as vessels or containers. As Looper (2018: 130) explains for the Maya, energy was stored and power was latent in matter, and “through human manipulation, the animate potencies inherent in these materials could be manifested and controlled,” and the materials could be allowed to “speak”—speech, breath, and souls being intimately linked in Mesoamerican thought. A thing could be “activated,” which did not enliven it, per se, but rather summoned the numinous to a physical place where communication and interaction could take place. Etymologically, teixiptla is composed of teotl + ixtli + xiptlahtli. Ixtli is translated by Dehouve (2020a) as eyes, but also, by metonymic extension, all of the organs of the face that enable “breath and word.” Xiptlahtli, or “covering,” referred to the ritual attire provided to bring about the manifestation. As with the ethnographic cut-paper figures, the dressing or attirement, offerings, and invocations of Aztec figurines may have brought about the activation, though I leave open the possibility that human intervention was not always required. I argue that once activated in household ritual practice, Aztec figurines were vibrant, agential participants that worked together with people to bring about a successful outcome for the family. Viewed through the lens of Antczak and Beaudry’s (2019) framework of assemblage of practice, we can understand household ritual figurine use as a dynamic gathering or assembling of vital things—things we would today classify as both human and other-than-human—entangled in the daily and occasional rituals situated in the home and aimed at avoiding harm. The question then becomes: what vibrant things were being assembled or gathered in Postclassic central Mexican household ritual, and how did those assemblages change over time, as the positions, realities, concerns, and stresses of those families shifted in relation to a changing broader social world? To begin to address these questions, I turn to the site of Xaltocan, where extensive archaeological research has recovered excellent contextual evidence for household ritual practice involving figurines.

Household ritual assemblages at Xaltocan

The Postclassic site of Xaltocan is particularly well suited for this task because it has a long history or durée of day-to-day ritual practice: figurines were used at Xaltocan for 600 years before the Spanish conquest. Located on a low, human-made island in Lake Xaltocan in the northern Basin of Mexico (Figure 1), the site was founded by Otomí-speaking peoples according to ethnohistoric sources (Alva Ixtlilxóchitl 1975-1977, I: 293), probably in the mid-10th century CE (Brumfiel 2005: 119). It was an important regional center and capital of the Otomí city-state in the 12th and 13th centuries (Alva Ixtlilxóchitl 1975-1977, II: 17, 18, 51; Carrasco 1950: 260-261; Davies 1980: 144-145; Nazareo de Xaltocan 1940: 124; Anales de Tlatelolco 1948: 28). In 1250, Xaltocan engaged in a war with Cuauhtitlan, then part of the Tepanec empire dominated by Azcapotzalco. Xaltocan fell in 1395, and some residents likely fled, while others stayed and adapted to their new political-economic context (Alva Ixtlilxóchitl 1975-1977, II: 36; Anales de Cuauhtitlan 1992: 54-61; Mata-Míguez et al. 2012; Overholtzer 2013). The Tepanecs became part of the newly formed Aztec Triple Alliance in 1428, and the Aztec ruler is said to have sent taxpayers who were fiercely loyal to Tenochtitlan to live at Xaltocan (Anales de Cuauhtitlan 1992: 76, 104; Díaz del Castillo 1956: 356). A military ruler from Tenochtitlan governed Xaltocan (Nazareo de Xaltocan 1940: 120) until 1521, when Hernán Cortés and his Indigenous allies attacked and burned the city (Cortés 1971: 118). Xaltocan continued to be occupied through the Colonial period and is a thriving community today. Figure 1 – Map of the Basin of Mexico, showing Xaltocan’s location (map by author). Archaeological work directed by Elizabeth Brumfiel began in 1987 with site mapping and extensive surface collections. In 1990 and 1991, twenty 2 × 2 m test pits were excavated, and three of the test pits were expanded to further investigate Early Postclassic houses. The test pits revealed the nature of the stratified deposits and allowed the elaboration of a four-phase chronology based on ceramics and radiocarbon dates (Brumfiel 2005: 125). Extensive household excavations and a significantly larger sample of radiocarbon dates allowed me to update Xaltocan’s chronology (Overholtzer 2014). The Dehe (900-1240 CE) and Hai (1240-1350 CE) phases roughly correspond to the Early and Middle Postclassic periods, when Xaltocan was an independent tribute-receiving capital of the Otomí empire and, later, when it was at war with Cuauhtitlan. The Tlalli phase (1350-1521 CE) spans a few decades of war, as well as Xaltocan’s rule by the Tepanec and later Aztec empires. The Isla phase (1521-1650 CE) represents early Colonial occupation, though unfortunately this period remains outside of the scope of this paper. Brumfiel’s survey, test pits, and excavations produced the assemblage of 588 ceramic figurines that were analyzed for this paper. These figurines were recovered in nearly every excavation unit (Figure 2) and are both spatially and chronologically well distributed, providing contextual data on figurine use and disposal across the site and across its occupational history. They are found in great quantities in every extensive household excavation at the site, suggesting widespread domestic use. Attempts to refit figurine fragments were made; all mended figurines and single non-fitting fragments were counted as single figurines. Figurine densities were calculated in a ratio of the number of figurines per 1,000 rim sherds in the excavated deposits, following Brumfiel’s project methodology. Since rim sherds are a general indicator of basic household activity, this ratio controls for variability in the extent and density of occupation at the site during the periods in question (Brumfiel 1996). Figure 2 – Reconstructed island of Xaltocan showing the locations of Brumfiel’s test pits and excavation units. Excavations that yielded ceramic figures are marked with a red “x” (map by author). Of the variety of types of figurines recovered at Xaltocan (Brumfiel and Overholtzer 2009), this paper addresses only anthropomorphic figurines produced in the Postclassic period, focusing primarily on the solid or flat-backed type, called Type III by Parsons (1972). This type was selected because it includes representations of both men and women, because it is present in all periods of occupation, and because it is the most common figurine form at Xaltocan. Solid, flat-backed figurines (236 examples) account for about 50% of the figurines that are identifiable in their fragmentary state. This large sample size enabled me to compare male and female figurines in terms of iconography and temporal frequencies. One additional anthropomorphic type of figurine, a rattle figurine called Type I by Parsons (ibid.), was used in the Late Postclassic and early Colonial periods at Xaltocan. Rattle figurines are present in low frequencies at Xaltocan, accounting for 3% of the assemblage, and include only female representations. I only briefly consider this type in this paper, since its representational and temporal characteristics preclude comparisons of frequencies over the history of Xaltocan’s pre-Hispanic occupation, and I have published in detail on this type elsewhere (Overholtzer 2012). Additional figurine types present at Xaltocan, including figurines attached to vessels and flutes, animals, and Formative to Classic period figurines that were collected and curated by Postclassic residents, have been studied elsewhere and are not considered here (Brumfiel and Overholtzer 2009; Overholtzer and Stoner 2011). Brumfiel (1996) observed that figurine frequencies were low in Xaltocan’s surface survey collections, likely because of modern collecting. Comparison of all figurine and only excavated figurine frequencies confirmed much higher frequencies in the excavated collections for all phases, but particularly so for the Late Postclassic, probably because this latest occupational phase tends to lie on or nearer to the surface (Table 1). Because an accurate understanding of the relative frequencies of figurines over time is critical in understanding changes in their use in the household, frequency calculations were calculated using only the 192 excavated figurines. All 236 flat-backed figurines were included when studying figurine iconography. Table 1 – Figurine frequencies for all contexts and only excavated contexts at Xaltocan; these frequencies include all figurine types.

Period	Excavated and survey figurines/1,000 rim sherds	Excavated figurines/1,000 rim sherds
Early Postclassic	6.45	12.46
Middle Postclassic	10.92	17.37
Late Postclassic	4.83	13.49

Flat-backed figurines are just that—solid figurines with molded and painted representations on the front and a flat, unpainted back. They often have large bases in the form of either large feet or a support attached to the back that would have enabled them to stand on their own (Figure 3). Based on this form, and in combination with ethnohistoric evidence, Brumfiel and I (2009) argue that these figurines likely would have been used on household altars with their undecorated backs to the wall. In contrast to more rounded, three-dimensional figurines such as the Type I rattles (Parsons 1972), these flat-backed figurines were not designed to be held and manipulated. However, most of these figurines also have small punctation holes, often to the sides of the chest or under the arms, that would have allowed them to be hung above an altar, from trees in the fields or on the side of a road, or to be worn by children as related in the ethnohistoric documents, though some are too large to fit on children’s wrists very comfortably. Figure 3 – Front (a) and top (b) views of a figurine fragment that depicts a standing woman wearing a skirt. The top view shows the large base, which enables the figurine to stand on its own (photograph by author). In some cases, the form of the figurine demands a more particular human engagement. Some of these figurines, all depicting a woman in a kneeling posture, feature small bowls attached to their front, as if the figure is holding the bowl in her lap (Figure 4). These bowls would have encouraged the placement of offerings of paper, cloth, blood, copal, or food. Seven figurines at Xaltocan are identifiable as this sub-type. Figure 4 – Flat-backed figurine women wearing blouses and kneeling with bowls (broken) in front of them (photograph by author). A second sub-type, called pendant heads here, consists of heads of standard scale and size manufactured without bodies and without any iconographic elements associated with particular deities (Figure 5). While they could not have stood on an altar, they feature punctation holes on the sides of the temples that would have allowed them to be suspended. Thirteen of these figurines have been recovered at Xaltocan. It should be noted that these figurines are not masks that would have been worn over other figurines, as seen in figurines at Tlatelolco (Guilliem Arroyo 1997), but rather are solid and morphologically identical to heads of other flat-backed figurines. Nor do these figurines conform to the “jointed” type described by Parsons (1972) with regard to their shape and hole placement. I suggest that these much smaller figurines may have been worn on the body, in a fashion similar to Honduran figurines studied by Joyce (2003), and as described by Durán for Aztec children. Figure 5 – Pendant head variant of the flat-backed figurine. Features two punctation holes that would have allowed suspension. Nearly complete; only a small upper portion of the headdress is broken (photograph by author). Several physical characteristics of flat-backed figurines stand out as especially important for structuring human engagement and enabling some active roles and constraining others. Their small size (typically 10-15 cm tall), suspension holes, and relatively inexpensive construction meant that they were highly portable and could be used in many different settings, including in the fields or on the side of roads, where they might be stolen or where winds might knock them down and break them. The pendant heads could be tied to the wrists of young children, whose abilities to be gentle and careful with fragile things were still developing. The large bases of most flat-backed figurines and their human form enabled them to stand on their own on an altar and look back at the viewer, encouraging a personal engagement between bodies while freeing up human hands to hold incense burners, place other offerings, or play the bone rasps made of human femurs found in the home. As Jackson (2019: 34) argues for the ancient Maya, “to be faced is to have a location in which personhood may lodge, and to have an interface for relational connections that undergird being a person.” Faces, as Jackson points out, include: 1. the head, which was for the Aztecs the human home of tonalli, or the animating life force that came from the sun (López Austin 1984, I: 221-252); 2. the eyes, which were “implicated in active relational work” (Jackson 2019: 33) and which involved a “procreative” power (Houston, Stuart and Taube 2006: 165) in ancient Mesoamerica; and 3. the mouth, which involves ingestion, breath, and speech, the latter two intimately linked and powerful forces for the Aztecs (Furst 1995: 155). Reyes Equigas (2005: 107-109; cited in Dehouve 2020a) argues that amaranth dough figures of deified mountains were provided with faces so that they could see offerings and hear prayers made to them and so that they could speak back to their human interlocutors. It is, then, perhaps not surprising that the mouths of Aztec flat-backed figurines are nearly always depicted open, tongue and/or teeth visible, as if they are mid-speech, speaking back to the human in front of them (Figure 6). The open mouth might also reference the ability of the figurines to consume offerings placed before them. It is notable, and perhaps unsurprising, that the head was the body part deemed essential for inclusion in pendant figurines, while the body could be excluded. Figure 6 – Two flat-backed figurine heads showing large, open mouths with tongues or teeth visible (photograph by author). Excluding the 71 figurines recovered from the surface or in unknown or mixed deposits, the 121 excavated flat-backed figurines at Xaltocan were most often recovered in fragmented form in secondary or tertiary deposits, such as fill contexts placed to raise and level the surface for house construction (61%) and the material used to manufacture mud bricks (2%) or earthen floors (6%). We might suspect that the figurines were accidentally incorporated into these deposits, having been contained within the earth mined for fill, bricks, or floors. However, we cannot exclude the possibility that they reflect materials from ritual practices taking place at the time and in the vicinity of construction, as has been explored elsewhere in Mesoamerica (Newman 2019). Figurine fragments were also found in household refuse deposits (18%) and on top of house floors in what may be in-situ ritual practice either during occupation or at its abandonment or rebuilding (6%). Less frequently, figurines seem to have intentionally been placed in hearths (6%), offerings (1%), and burials (1%). These depositional contexts (Table 2) suggest that many flat-backed figurines were used in the household until broken and were then thrown away with ordinary household refuse. There is no evidence for intentional breakage that would support a shorter period of use before deposition. For the Aztecs, death “is not an ending but a change of status, as that which dies flows into and feeds that which lives. All things are involved in a single, never-ending process of recycling and transformation” (Maffie 2014: 46). Thus, we might view the breakage and deposition of a figurine in the home as a change of status, a de-activation, the figurine re-entering the earth from which it came transformed, perhaps no longer able to channel petitions or house supernatural entities. Table 2 – Archaeological contexts of Xaltocan flat-backed figurines, excluding those recovered on the surface or in unknown and mixed deposits.

Context type	n	Percentage (%)
Fill	74	61
Middens	22	18
Wall matrix	2	2
Floor matrix	7	6
Inside room	7	6
Hearth	7	6
Offering	1	1
Burial	1	1
Total	121	100

The pendant head sub-type seems to have had a different pattern of deposition, never appearing in household refuse deposits. One pendant head figurine, complete but broken into two pieces, was recovered as part of an offering. This offering, which might be associated with two nearby infant burials placed underneath an Early Postclassic house, also contained 15 blowgun pellets, one quartz biface, 15 projectile points, several obsidian blades, a piece of ground stone, two basalt floor polishers, a complete bone needle, two bone spatulas, a jasper punch, an onyx core, two pieces of imported marine shell, and several ceramic sherds (Brumfiel 2010; De Lucia 2014: 386-387). The pendant head included in this offering was the only complete flat-backed figurine recovered at Xaltocan. Several pendant head fragments and nearly complete pendant heads were recovered in hearths. Indeed, of the seven figurines recovered in hearths, four were pendant heads. Thus, even though pendant head figurines constitute only 6% of the flat-backed figurine assemblage, they represent 57% of the flat-backed figurines recovered from hearths. This difference in depositional context, combined with the presence of punctation holes that would have permitted suspension, suggests that pendant heads may have been used (likely worn) for shorter periods of time, perhaps only the length of a child’s illness, before they were interred in a hearth or offering. They were destined for a different transformation through deposition than the other flat-backed figurines. This configuration of teotl may have been particularly fleeting. Scholars working at other Postclassic central Mexican sites have also found ceramic figurines in contexts of types not encountered at Xaltocan: in other kinds of domestic contexts, such as in sweat baths (Evans 1990), and in bodies of water, such as the bottom of ancient canals, springs, lakes, and rivers (Porcayo Michelini 1998: 248, 258-259, 263). These watery depositional contexts are in line with Fray Diego Durán’s observation (1971: 269) that postpartum Aztec women made offerings of clay figurines (“muñecas de tierra”) to river currents. Aztec figurines are also found occasionally as offerings in temples, such as at Temple R at Tlatelolco dedicated to the wind-rain god Ehecatl-Quetzalcoatl (Guilliem Arroyo 1997). All of these contexts share an association with water, rain, and fertility and suggest a different transformation of the figurine entity or change in its status through deposition—perhaps a more active role after its interment as an offering.

Figurine iconography at Xaltocan

If, as been argued here, figurines were active participants, vibrant and vital components of household ritual assemblages aimed at avoiding harm in the household, then trends in iconography and the intensity of figurine use—that is to say, the cumulative choices that people made in engaging with particular kinds of figurine beings—should index changing household concerns. A remarkable degree of continuity in iconography is present in both male and female representations within the flat-backed figurine category. No meaningful temporal trends can be discerned over the 600 years of pre-Hispanic occupation at Xaltocan. There is no iconographic category (e.g. warrior imagery or Xochiquetzal iconography) that is unique to a particular phase, and all of the major iconographic elements are present in all phases. Moreover, figurines are ubiquitous in Aztec household contexts and were likely integral parts of every household assemblage. Flat-backed figurines represent an extraordinarily conservative tradition—that is, they were an incredibly stable presence in the home. However, as will be discussed, there were changes in the intensity of their presence and in the intensity of use of figurines with particular iconography. Solid, flat-backed figurine women (Figure 7) are diverse in iconographic subject matter and style (see Klein and Lona 2009). All are dressed in skirts, and most are bare-chested; breasts are usually depicted even when a blouse is worn. Women most often stand or kneel (Figure 7a and e). Many wear ornamentation typically associated with fertility goddesses in codices, though, as mentioned earlier, figurine representations thwart straightforward identifications of specific, individual deities in many cases, especially in figurine women. Headdresses composed of a band bordered by a single row of beads at the top and bottom, often featuring pleated bark paper fans or amacuexpalli, are often identified as suggesting a relationship with Chalchiuhtlicue (Kaplan 1958; Parsons 1972; Figure 7b-d). The presence of corncobs in the hands of some women depicted in solid, flat-backed figurines may reference the maize goddess Chicomecoatl (see Klein and Lona 2009). Headdresses topped by plumes, especially double plumes, are said to be linked to Xochiquetzal, the patron goddess of weaving (Kaplan 1958; Parsons 1972). These headdresses are present in the flat-backed category (Figure 7c) but are less common than among rattle figurines (Overholtzer 2012). This headdress, however, may simply reflect the adult women’s hairstyle. Some women entirely lack iconographic elements associated with deities (Figure 7a). Some are pregnant or feature a child clinging to their side (Figure 7f) or present a pot or vessel (Figure 4), as previously mentioned. Other scholars are suspicious of deity attributions altogether and suggest that attributes such as paper fans reflect fertility images, broadly defined (Millian 1981; Smith 2002; Klein and Lona 2009). Because of this uncertainty, I do not attempt to categorize figurine women into deity or non-deity iconography. Regardless of whether goddesses, ordinary women, or other types of entities are depicted, the female fertility theme identified in these figurines seems to reflect a concern with general household fecundity, in terms of both biological reproduction and crop fertility (for more information on female iconography, see Brumfiel 1996; Kaplan 1958; Klein and Lona 2009; Millian 1981; and Parsons 1972). Figure 7 – Variability in the iconography of flat-backed figurine women: a. depicts a standing woman wearing a skirt and small headdress but with no identifying deity paraphernalia, and is the most complete flat-backed figurine recovered at Xaltocan; b. and d. depict people wearing a headdress with bands or row of dots and pleated paper fans, commonly interpreted as representing Chalchiutlicue; d. likely held a bowl in front of her; c. depicts a person wearing a two-pronged headdress or hairstyle, sometimes interpreted as representing Xochiquetzal; e. depicts a kneeling woman; f. represents a standing woman with an infant clinging to her side (photograph by author). Two major categories of iconography are present in figurine men: warrior costume or paraphernalia and aspects of particular deities who are sometimes seated on pyramids. Of these categories, warrior and deity imagery are very common and account for nearly a quarter of the figurine men each (Table 3). Parsons (1972: 103) reports having seen complete examples in museums featuring Xipe or Macuilxochitl iconography on the face or headdress and warrior dress on the body. Moreover, the Xaltocan collection includes a man wearing a cut conch shell associated with Ehecatl on his chest and holding a shield in an arm. Thus, these categories are not mutually exclusive. Approximately half of the figurine men could not be assigned a category in their fragmentary state (Figure 8), and some may have had neither warrior nor deity iconography. Interestingly, some of the types of figurine men commonly observed by Parsons in the Teotihuacan Valley collection, such as drummers, are completely absent at Xaltocan. Table 3 – Percentages of iconographic categories in flat-backed figurine men recovered in survey and excavations at Xaltocan.

	n	Percentage (%)
Warriors	13	24
Deities	11	20
Temple models	2	4
Unidentified	28	52
Total	54	100

Figure 8 – Two unidentifiable figurine men, both standing and wearing loincloths, and with possible pectoral or obsidian mirror at right (photograph by author). Warriors can be identified by the presence of paraphernalia such as shields and obsidian clubs or rattle sticks, usually in the left and right hands, respectively (Figure 9a); quilted armor (Figure 9b); or guild costumes, especially eagle warrior (Figure 9c) and jaguar warrior (Figure 9d). Shields are restricted to figurine men, unlike representations in codices, where female deities brandish shields and weaving battens. Representations of warriors are fairly standardized and are similar to those in ceramic sculpture and codices. They stand or kneel on one knee. The clothing worn by warrior figurines, for example, is most similar to that seen in Aztec sculpture, such as the Churubusco Idol exhibited at the Mexican National Museum of Anthropology. The warriors represented in both the ceramic figurines and the Churubusco Idol (Anawalt 1981: fig. 17) don a loincloth and a triangular-shaped hip cloth, worn with the point to the front (Figure 9a and b). Note that, as in representations such as the Churubusco Idol, the hip cloth is worn much lower than a woman’s quechquemitl and so can be distinguished based on its position. Figure 9 – Warrior figurines: a. and b. depict the standardized warrior: standing, holding an atlatl and shield, and wearing a triangular-shaped hip cloth with point toward front, over the loincloth; c. depicts a warrior on one knee wearing a loincloth and cotton armor; d. represents an eagle warrior, while e. represents a jaguar warrior (photograph by author). Scholars are more confident about the identification of male deities, all of whom generally relate to rain, water, and agricultural fertility, but Aztec gods and goddesses alike are of course interrelated. Quetzalcoatl-Ehecatl is commonly identified in Aztec figurines by a buccal mask and/or a spiral pendant made from a Strombus conch; he is associated with the winds that precede the rains in central Mexico (Figure 10a and 10c). Tlaloc, the god of rain, fertility, and water, is also present and is identified by goggle eyes and fangs (Figure 10d). Less common is Xipe-Totec, the life-death-rebirth deity and the god of agriculture, who is identifiable by flayed skin and a conical headdress that is also worn by Quetzalcoatl-Ehecatl and Xochipilli. Macuilxochitl-Xochipilli, who can be identified in the assemblage by the same conical headdress and also in monkey form, is associated with maize and agricultural produce along with dance, song, love, and flowers (Figure 10b). Ruíz de Alarcón (1984: 278) lists some of these deities as those invoked through incantations in the field associated with planting to ensure agricultural success. Quetzalcoatl was petitioned to give the farmers the strength to be able to till the land to plant and later to weed. The tlaloques or helpers of Tlaloc, his manifestations in the four cardinal directions who were responsible for creating rain, were invoked for help in protecting the sown field from small animals (ibid.: 214). Figure 10 – Male deity figurines: a. depicts front and side views of a human wearing a buccal mask, likely Quetzalcoatl-Ehecatl; b. depicts a monkey wearing a buccal mask, possibly a representation of Macuilxochitl-Xochipilli and/or Ehecatl; c. features a spiral conch shell necklace, likely a reference to Ehecatl; d. depicts Tlaloc with his goggle eyes and fangs (photograph by author). Pyramids or temple models are present but not common, with only two examples in the Xaltocan collection (Figure 11; Table 3). It is worth noting here that the temple models recovered were not complete, and we do not know whether deities would have sat on top. Some complete examples of temple models in museum collections feature figures identified as the same deities mentioned above: Quetzalcoatl-Ehecatl, Xipe-Totec, and Macuilxochitl-Xochipilli (Baer 1996: 74-85; Parsons 1972: figs. 19-20), while other temple models do not have deity images (Baer 1996: 64-73). Figure 11 – Temple pyramid model (photograph by author).

Figurine presences across the Postclassic

Leonardo López-Lujan and Salvador Guilliem Arroyo have demonstrated that the intensity or frequency of ritual performance in Aztec Mexico correlates with periods of stress. In his analysis of the offerings of the Templo Mayor of Tenochtitlan, López-Luján (2005: 141) found that in addition to the commemoration of periodic building expansion efforts, ritual offerings were often made in times of stress, such as epidemics, disasters, and illnesses. For example, he argued that Offering 48, a sacrificial offering of 42 subadults, dates to a period of extreme drought between 1450 and 1454 (ibid.: 202-206). Carried out to supplicate Tlaloc, this offering included iconographic representations of the rain god and allusions to water and fertility in the form of blue paint on the sacrificed victims’ bodies and greenstone beads worn as necklaces and placed in their mouths (ibid.: 149). Guilliem Arroyo (1999) similarly argued that a sacrificial offering at Temple R at Tlatelolco containing 37 subadults, mostly under the age of three, and six adults was made during the same drought (see also De La Cruz et al. 2008). This offering was made to Ehecatl-Quetzalcoatl, the god of the winds that precede and help usher in the rains (ibid.). Ethnohistoric documents describe this drought and record an abundance of sacrifices made during that period (Codex Telleriano Remensis 1964: f. 32r); the Templo Mayor and Tlatelolco offerings are likely two of many such ritual offerings made during this time. Ceramic figurines are rarely present in monumental contexts, with the notable exception of an offering likely made in association with the dedication of the circular temple to Ehecatl-Quetzalcoatl at Tlatelolco (Guilliem Arroyo 1997). As archaeologists have argued, household ritual, where most Aztec figurines were likely active, operated outside of state control (Brumfiel 1996; Smith 2002). Nonetheless, some celebrations, such as the New Fire ceremony, had both imperial and domestic components, and household archaeologists have increasingly recognized that many imperial rituals originated in and had long histories within the home in ancient central Mexico and were only later co-opted and redesigned on a grander scale for state purposes. Most importantly, we would expect in this specific historical and cultural context that if anxieties and stresses produced intensified ritual practice in the imperial capital of Tenochtitlan, so too they would have resulted in intensified ritual practice in the homes on the periphery of the empire. On this basis, I argue that the frequencies of figurine beings at Xaltocan—that is, the intensity of their presence in household ritual assemblages—point to the degree of household concerns and anxieties. Because the temporal phases based on archaeological deposits span periods of a century or several centuries, figurine frequencies will mask variability within that span of time and reflect the average intensity over that period. Let us turn now to examine patterns in figurine entity presence within households at Xaltocan across the Postclassic period. When the percentages of figurines by gender are examined across phases (Table 4), we see that figurine men become less popular as compared to women, decreasing from 59% in the Early Postclassic Dehe phase to 42% in the Middle Postclassic Hai phase and to only 25% in the Late Postclassic Tlalli phase. This trend conforms to the pattern observed by Brumfiel (1996). Thus, over time, more and more of the figurines that people at Xaltocan chose to obtain in the markets, and thus a greater proportion of the entities that constituted household ritual assemblages, were women. Averaging across the occupational history of the site, we see that figurine women are slightly more common than men. Table 4 – Relative percentages of flat-backed excavated figurine men and women at Xaltocan.

Period	Men	Women
Early Postclassic	59%	41%
Middle Postclassic	42%	58%
Late Postclassic	25%	75%
Total	44%	56%

However, if we standardize the frequencies of figurines per 1,000 rim sherds, rather than examine relative percentages, we see something slightly different: we see the intensity of figurine presence. The frequency of all figurines for each period reveals the overall scale of figurine use over time, and the frequencies of male and female figurines indicate the extent of use of figurines of each gender (Tables 5 and 6). In this way, the presence and active work of figurine men in the household can be studied independently of changes in those of women. We move now to examine those frequencies. Table 5 – Frequencies of flat-backed figurines per phase excavated at Xaltocan.

Period	n	rim sherds	n/1,000 rim sherds
Early Postclassic	40	11,560	3.46
Middle Postclassic	77	7,830	9.83
Late Postclassic	14	2,298	6.09
Total	131	21,688	6.04

Table 6 – Frequencies of flat-backed figurines per phase by gender excavated at Xaltocan.

Period	Men (n)	Women (n)	Men/1,000 rim sherds	Women/1,000 rim sherds
Early Postclassic	10	7	0.87	0.61
Middle Postclassic	19	26	2.43	3.32
Late Postclassic	2	6	0.87	2.61
Total	31	39	1.43	1.80

In the Early Postclassic Dehe phase, the time period of prosperity and independence, Xaltocamecas (residents of Xaltocan) engaged with slightly more figurine men. Figurine frequencies overall were fairly low, less than one figurine of each gender per 1,000 rim sherds. In the Middle Postclassic Hai phase, when Xaltocan was at war with Cuauhtitlan, residents were using many more figurines: 2.43 and 3.32 figurines, respectively, of each gender per 1,000 rim sherds. Additionally, there was a sharp increase in the presence of figurine men and women. During the Middle Postclassic, people engaged with more figurine men and women than they had during the Early Postclassic. The increase in women was slightly greater than that of men. This observation accords with the trend of an increasing percentage of female figurines throughout the Postclassic noted by Brumfiel. However, masculine figurine entities also increased substantially, nearly three-fold, in the home. In the Late Postclassic, during much of which Xaltocan was under Aztec rule, figurine men were less often active participants in household ritual practice, and overall frequencies returned to Early Postclassic rates. Engagement with feminine figurine beings decreased slightly but remained substantially elevated over Early Postclassic rates. Residents of Xaltocan under Aztec rule engaged with fewer figurine men but continued to engage with figurine women at a high frequency.

Gendered household practices

Having discussed the iconography of flat-backed figurines and temporal trends in the frequencies of these figurines, we move now to a discussion of gendered household practices, the concerns over which people responded to by summoning the help of figurine entities and activating them in their household ritual. The ethnohistoric evidence for figurines described earlier suggests that people engaged with figurines as active participants in household ritual—giving them offerings and verbally invoking them, whether standing face-to-face on household altars, hung in the fields, or worn on the wrist—in order to avoid harm in their everyday lives and to ensure the success of household production and reproduction. Furthermore, the iconography and deity associations suggest that gendered figurines were associated with tasks that were at least ideologically, if not in practice, gendered (see Kaplan 1958; Millian 1981; Parsons 1972). Figurines depicting women feature iconography associated with reproductive tasks such as childbirth, midwifery, and childrearing and with stereotypically feminine productive tasks such as weaving. Figurines depicting men display themes directly related to warfare and agricultural production. There is, however, a degree of overlap in the association of female deities with rain and maize agriculture: Chalchiuhtlicue, for example, was the goddess of water who bathed newborns and treated the sick, but she also partnered with Tlaloc to ensure rains for harvest (Dehouve 2020b). Active warfare was the exclusive domain of men, as only men joined the warrior ranks, and only commoner men were required to provide service in times of war to those who governed their polity (Zorita 1963: 181-188). Women’s labor in service of the ruling elite was directed toward other tasks, such as weaving cloth or working in the palaces of local rulers, although in times of war the preparation of portable foods would have contributed to the war effort indirectly (Brumfiel 1991). While their husbands were away, these “Mexica women on the home front” would have carried out ritual sweeping and offerings with particular care and intensity, thereby providing ideological support (Burkhart 1997: 37). Only in exceptional circumstances did women play a more active role in warfare. Fernando Alvarado Tezozomoc (1975: 392) and Diego Durán (1994: 260) both recount that during the last battle between the cities of Tenochtitlan and Tlatelolco, women actively engaged the enemy in a final effort to defend Tlatelolco’s temple: They exposed their breasts and buttocks; they beat their hands against their abdomens and genitals; they squeezed milk from their breasts; they threw dirt, excrement, and chewed-up tortillas at the Tenochca warriors. They also threw brooms, weaving battens, and warping frames. (Burkhart 1997: 37; see also Klein 1994) As Burkhart (1997) argues, the women in this case did not emulate male warriors. Instead, they “turn[ed] the full force of their womanhood against the invaders,” cleverly flaunting their sexuality and association with Tlazolteotl, goddess of filth and childbirth, and enacting their role as “broom-wielding guardians of the home front” (ibid.: 37-38). Women were clearly not considered warriors in the traditional sense. Ideologically and nearly always in practice, then, active warfare belonged to the male realm, and women most often provided crucial logistical and ideological support. In ideals, if not in practice, farming was also considered men’s work. Ruíz de Alarcón details the incantations performed exclusively by men during the planting, weeding, and harvest seasons (1984: 278). In one incantation, the farmer speaks to the maize kernel, reminding it that his wife has been taking care of it—presumably by guarding the maize bin at the front of the house and making sure that it did not rot or was not eaten by rodents or other animals between agricultural seasons—but that it is now time for him to plant it (ibid.: 278). In reality, women probably assisted with farming tasks beyond guarding corn when help was needed and when they were not occupied with other productive tasks (Evans 1992). Women might also have had more active roles in maguey farming, as suggested by an encounter related by Alva Ixlilxochitl between a prince of Texcoco and a commoner woman collecting aguamiel or maguey sap in the countryside of Chalco (Alva Ixtlilxochitl 1975-1977, I: 346). Women’s participation likely would have been greater in areas where agricultural production was the principal economic base of the community (Brumfiel 1991). Angélica María Medrano Enríquez (2006) employed skeletal markers of stress on a sample of 167 individuals from the chinampa farming community of San Gregorio Atlapulco, Xochimilco, in order to reconstruct occupational activities. Chinampa farming is a labor-intensive way to make swamplands agriculturally productive by creating a system of fields and ditches. Medrano Enríquez’s analysis showed that men, women, and children participated in chinampa agriculture (ibid.: 380-383). Excavations at the Xochimilco site produced little evidence of weaving, suggesting that women were not producing other goods such as cloth and were therefore devoting more of their time to agriculture. This follows the inverse relationship between intensive agricultural and textile production identified by Brumfiel (1991), these two activities representing alternative productive practices of women. However, participation in agricultural tasks was not even: many of the skeletal markers for various agricultural tasks occur at a higher frequency in men (Medrano Enríquez 2006: 380-383). In addition, Medrano found that maintenance of canals used for agriculture, indicated by a specific skeletal marker, auditory exostoses, was a task only performed by men (ibid.: 385). Thus, even in agriculturally intensive areas such as Xochimilco, men’s participation in farming was likely greater and some farming roles would have been exclusive to men. Nonetheless, given the overlap in tasks and deity association, there is no reason to think that women would not have engaged in rituals pertaining to farming and harvest, for example summoning Chalchiuhtlicue or Tlaloc in rituals on their household altar.

Discussion

The data presented here suggest that when Xaltocan was founded in the Early Postclassic, flat-backed figurine and human entities worked together in concert, forming an assemblage that prevented harm to crops and children, the main concerns in human daily lives. Yet the intensity of their presence was rather low during this period, when Xaltocan’s strength and power grew as the capital of the Otomí city-state, and the household was not terribly stressed. These figurine entities became more ubiquitous in Xaltocan homes during the Middle Postclassic, with the need for their assistance growing as the war with Cuauhtitlan dragged on and placed households under considerable strain during what has been called a “troubled era of struggle and conflict” across the Basin of Mexico (Sanders, Parsons, and Santley 1979: 150). The commoner men of Xaltocan would have had to fight in the war, likely more often or for lengthier periods of time than in previous years, because this war was long, and it was against opponents with alliances that afforded them a large population from which to draw warriors. While we cannot know how intensive this 145-year-long war was in terms of the number of battles or the amount of time that commoners would have had to contribute as warriors, it undoubtedly would have taken an emotional and physical toll. The Anales de Cuauhtitlan (1992) suggest that battles were likely episodic or sporadic, and not continuous. Thus, the war with Cuauhtitlan likely consisted of times of relative peace and times of active battle. We might suspect that there were increases and decreases in the intensity of household ritual and figurine use associated with those times of peace and battle, but we cannot assess this variability because of the coarse chronological resolution of archaeological data. On average, however, figurine use was most intense at Xaltocan during this period. I argue that during the Middle Postclassic, households may also have been most invested in the success of their warriors. The war with Cuauhtitlan threatened their everyday lived reality and their ways, standards, and customs of life. Households saw their husbands, brothers, and sons dying in warfare. Their male relatives were no longer consistently able to sow, tend, and harvest crops. Women and children likely worked more in the fields to try to make up for the void. War may have threatened not only loved ones’ lives, but also their ability to feed their family. Xaltocamecas would have understood the consequences of defeat in war for everyday life. Defeated towns would be subjected to increased taxation whether they surrendered or were conquered in battle. Towns that did not surrender were nearly always sacked, the looting serving as one form of compensation to soldiers, and homes and fields were often burned, resulting in utter destruction (Hassig 1988: 112-113). Some residents abandoned their homes once they were conquered, and some even fled pre-emptively (ibid.: 114). While Xaltocamecas fought with Cuauhtitlan, they heard about many other neighboring polities being conquered, often as part of Tepanec expansion. In 1347, Culhuacan was conquered and the city destroyed, and the population is said to have fled to Azcapotzalco, Coatlinchan, Huexotla, and Cuauhtitlan (Anales de Cuauhtitlan 1992: 67, 71; Mengin 1952; Zorita 1941: 268). During Acamapichtli’s reign (1372-1391), in conjunction with Azcapotzalco, the Mexica conquered Xochimilco, Mizquic, Cuitlahuac, Chimalhuacan, and Cuauhhuacan (Hassig 1988: 132). Residents of Xaltocan would have heard about these conquests, the destruction of cities, and the exodus or subjugation of inhabitants. In short, Xaltocamecas saw Tepanec expansion in progress, and their own independent status became threatened. The Anales de Cuauhtitlan recount the successive losses Xaltocan sustained over the years and the concomitant movement of the polity boundaries and the war front closer and closer to the capital and their homes with each loss (Anales de Cuauhtitlan 1992: 59-61). Household members—likely mostly women, according to the ethnohistoric sources—therefore used figurines to try to ensure success in the war, whether that meant the polity winning the war or a loved one returning safely from battle, and to ensure success in agricultural production while the men’s labor was temporarily or permanently lost. According to Burkhart (1997), women were particularly diligent in their offerings and ritual sweeping while their husbands were at war. During this time at war, Xaltocamecas engaged with many more figurine entities—three to five times as many—than they had in previous periods, with each household perhaps placing more figurines on their altars, in their fields, and on their bodies (Table 5). This included greater numbers of figurine men and women, because families were concerned with stereotypically male and female productive and reproductive tasks as well as the fate of their warriors, and they invoked both the male and female deities who oversaw or could influence those activities. In 1395, Xaltocan lost the war to Cuauhtitlan. Ethnohistoric records indicate that the town was deserted (Alva Ixtlilxóchitl 1975-1977, I: 323; Anales de Cuauhtitlan 1992: 104), with nobody living there for 35 years, and a demographic shift of some kind can indeed be seen in ancient DNA at the site (Mata-Míguez et al. 2012). However, recent archaeological evidence indicates that at least some residents stayed and rebuilt their households (Chimonas 2005; Miller 2007; Overholtzer 2013). By 1428, the political circumstances at Xaltocan had changed once again, as the Aztec Triple Alliance was formed, and the town fell under Aztec rule. In this new era, the Late Postclassic, life was different. Xaltocamecas’ homes were not threatened by warfare, looting, and burning. The endemic, constant warfare between neighbors characteristic of the Middle Postclassic had ended with the formation of the Triple Alliance. Imperial military campaigns were focused, directed, strategic affairs—first, at consolidating control and eliminating any resistance within the Basin, and then, conducting singular marches predominantly to the south of the Basin (Hassig 1988: 152). The wars fought by the Triple Alliance were, as time passed, fought farther and farther from the Aztec heartland as more of Central Mexico and beyond fell under Aztec domination. Consequently, wars fought in the Late Postclassic were less visible for residents of Xaltocan and were not as physically present in daily life as was the war at home against neighboring Cuauhtitlan. Since wars were fewer, and the Triple Alliance drew on an ever-larger population of men for its recruits (Hassig 1988), men at Xaltocan may not have had to provide labor for war as often. In addition, the Triple Alliance usually began recruiting in the capital cities—Tenochtitlan, Texcoco, and Tlacopan—before mobilizing troops from elsewhere in the empire (ibid.: 54). Only when a considerable number of troops were needed—that is to say, exceeding the number available in the capital cities—would residents of Xaltocan have been mobilized. The inhabitants may not have been as invested in the wars that were being fought and directed from Tenochtitlan. They were likely more concerned now with making their tribute payments, being able to feed their family, and seeing their sons and daughters born and raised healthy. During this time, Xaltocamecas engaged with fewer figurine men, and frequency rates returned to Early Postclassic norms (Table 6). Warfare no longer took a heavy mental and physical toll on the household, it seems, and the household’s personal investment in wars’ outcome dropped. Men’s lives admittedly would have continued to be affected as they were pulled away from the household, thereby making their agricultural production at home more challenging, and male figurines continued in use, albeit at a lower rate than in the Middle Postclassic. Life under the Triple Alliance would have placed a heavier burden on women’s daily productive and reproductive practices, however. An increase in tribute demands, often imposed in the form of cotton cloth, would have significantly increased their workloads (Brumfiel 1991: 230). Some women would have intensified food production and the production of other goods, at Xaltocan likely salted and roasted fish, and sold them in the market to meet tribute requirements or to obtain other goods to fill household needs (ibid. : 234). As labor drafts imposed by rulers drew labor from the household, women also prepared portable foods such as tortillas that were more labor intensive (ibid.: 243). Therefore, women’s food preparation workload for household consumption was likely higher than Middle Postclassic levels. Life would have been hectic for all household members, but particularly so for women. Xaltocamecas continued engaging with figurine women at an elevated rate, as the Triple Alliance strained women’s productive and reproductive activities. In the Late Postclassic, figurine women remained equally active participants in household ritual designed to avoid harm and to ensure their productive and reproductive success. At the same time, women began engaging with a new type of figurine entity, the rattle figurine or Type I (Parsons 1972), consisting of women in various stages of the reproductive process and likely being used to promote a healthy childbirth (Brumfiel and Overholtzer 2009; Overholtzer 2012). Residents of Xaltocan also used more long-handled censers and braziers in the Late Postclassic period (Brumfiel 2005: 132-137), especially those of a style known as Texcoco-Molded-Filleted (Figure 12; Table 7). These censers featured triangular cut-outs and round molded decoration on the bowl, fertility fans and molded serpents on the handle, and small ceramic balls inside the handle to make a rattling sound. These elements are associated with rain and fertility, lightning and clouds, and “heavenly fire” (Smith 2002: 101), and I suggest they were another material component of the household ritual assemblage, playing a complementary role to the ceramic figurines discussed here. Like the bone rasps and other forms of rattles present at Xaltocan beginning earlier in the Postclassic period (Overholtzer 2016), these musical instruments may have been crucial in summoning deities or activating figurines. Figure 12 – Top and side view of Texcoco-Molded-Filleted censer with serpent figurine on handle end and small ceramic balls inside handle to produce rattling sound (photograph by author). Table 7 – Frequencies of braziers and censers by period per 1,000 rim sherds at Xaltocan.

	Braziers and Censers/1,000 rim sherds
Early Postclassic	15.1
Middle Postclassic	14.6
Late Postclassic	34.4

Conclusion

The contextual analysis of Aztec figurines recovered in Xaltocan households reveals the interconnected labor and engagement of human men, human women, figurine men, and figurine women in ensuring the health of children, a good harvest, the return of men from the war front, and many other issues of household concern. This paper has argued that we must treat all these entities as active agents—all as momentary arrangements of sacred, vital energy—who came together and assembled in the home in relation to concerns that were a consequence of and had consequences for the macro-scale politically, socially, and economically. Pendant heads tied to children’s wrists with thread, then thrown into the home’s hearth; figurine men and women standing on an altar face-to-face with the humans who invoked the spirits they held, their mouths open as if mid-speech; kneeling figurine women receiving and conveying offerings in the bowls held in their outstretched hands: all of these entities were, like humans, fleeting configurations or arrangements of sacred, life-giving energy that had effects in the world for Xaltocan’s household members. This paper has shown how humans did not act alone in their efforts to ensure household success and avoid harm. This paper has also argued that we must restore men to the Aztec household. We do not tend to associate men with the domestic sphere, yet there was no “‘cult of domesticity,’ with the corresponding demarcation of the home as a private domain,” in Postclassic Central Mexico (Burkhart 1997: 26). Burkhart (1997), Brumfiel (1991, 1996), and others have convincingly argued that we must not think of the work women performed in the household as confined to the domestic realm. Fewer archaeologists have examined men’s roles outside of the modern public sphere. Men have been explored as rulers, merchants, priests, and members of elite warrior castes, and perhaps as stone tool producers, but far less often as fathers, sons, farmers, and commoner men conscripted to fight in imperial wars. While much of men’s work was not performed inside the architecture of the house (Robin 2002; Robin and Rothschild 2002), it was certainly domestic in that it contributed directly to household goals. As members of the household interacted with other social spheres, for example through the tribute and market systems, the fruit of both men’s and women’s labor reached beyond the house. Examining men and women’s household work and household anxieties therefore allows us a glimpse into the lived experience of commoners and into the ways in which households articulated with the larger social universe.