Women’s Voices in a Male World: Actions, Bodies, and Spaces Among the Ancient Maya

Feminist archaeology has prompted scholars to reconsider gender roles in ancient Mesoamerica. Current research, however, tends to focus on elite women, classes and sites. Although I do not ignore the potential of these sources, in this paper I am mainly concerned with issues such as the phenomenology of bodies and spaces, subroyal ritual actions, and daily activities such as cooking and weaving. My aim is to offer an overview of the most recent studies on gender in Maya archaeology and to provide ideas for further research by emphasising the need to engender ritual and individuate female discourses in the archaeological record.


Introduction
During the last few decades, the rise of feminist archaeology (Conkey and Gero 1997;Conkey and Spector 1984;Gero and Conkey 1991) has promoted a new sensibility toward themes such as gender, agency, sexuality and the construction of selfhood in antiquity. Studies of the role of women in ancient Mesoamerica have recently increased (e.g. Ardren 2002a; Joyce 2000a; Klein 2001). The aim of this paper is to offer an overview of research on gender in Maya archaeology and to highlight the impact that Maya studies might have on the development of gender theory. My primary concern is not for those areas usually studied by Mayanists: royal architecture, elite burials and hieroglyphic texts. On the contrary, I focus on issues such as the management of the household, the perception of the body and the way in which living spaces were conceived and exploited. My emphasis is on studies which explore how male ideology has concealed women's real position in Maya society. In particular, I note that ideological constructions of body and sexuality were primary means of male domination. I also suggest, however, that a greater focus on female activities such as food processing and weaving can provide valuable information on macro-scale social dynamics, and that further research is needed to identify symbolic and physical 'spaces' in which Maya women were allowed to raise -metaphorically -their voices. Hill (1998), among others, notes that mere hypotheses concerning gender roles in the past have been taken for granted on the basis of ethnographic analogy. She suggests that only a "multivariate approach" based on several lines of evidence may allow archaeologists to avoid the pitfalls which emerge when a single source is employed. I argue that the exceptional richness of the Maya record offers unique opportunities to employ gender theory to investigate both women's conditions in antiquity and macro-scale phenomena such as politics and conflict. Before tackling these issues in the following sections, I offer a critical assessment of the evidence through which Maya gender dynamics are revealed.

Preliminary Methodological Observations: The Sources
Ethnographic and ethnohistorical comparisons can be carefully employed to produce hypotheses subject to testing. General ethnography offers a basic framework to identify gender roles and relations, but fails to take into account the specificity of the context under study. Ethnographic studies on modern Maya provide valuable information and when compared with other lines of evidence, reveal a partial cultural continuity between the past and the present (Joyce 1992: 255). However, Mesoamerican ways of life underwent dramatic changes over time, particularly after European contact. The information provided by the Spaniards must be considered with caution because such information was compromised by religious and cultural biases (Hill 1998: 106-109). Similarly, the hieroglyphic manuscripts painted during the Postclassic period (AD 900-1500) and the accounts written by native individuals after the Conquest do not offer an objective view of local gender dynamics, since they were probably produced by men (Vail and Stone 2002: 203). Similar considerations can be made for Classic Maya written sources (AD 250-900). Although hieroglyphic texts supply exceptional information on political and historical events, writing was a medium employed by elite males to sustain their systems of power.
Iconographic evidence is another important means of understanding gender roles, ideologies and identities. Interpretations based on iconography, however, can be undermined by two main weaknesses. First, they sometimes rely exceedingly on the scholar's viewpoint. This is the case, for instance, of Upper Paleolithic female figurines, variously interpreted as sex-toys produced by men for a male audience, symbols of fertility, religious implements, or female forms of self-representation (McDermott 1996: 233-234). Second, iconography is a medium which retains a strong ideological load and does not offer a neutral view of the social reality. Maya people were depicted on/through three main artistic media: public monuments such as stelae, painted vessels and clay figurines.
Evidence suggests that the artefact's meaning changed according to the producer, the audience and the message conveyed (Joyce 1993: 261-263). Both stelae and painted ceramics were employed by the elites to reinforce their pre-eminent status. We know from the signatures on some works that scribes and artists were generally males, and possibly members of the elite themselves (Inomata 2001: 325). However, whereas stone monuments were more directly related to the royal elite, as testified by the presence of royal names on them, vessels had a wider audience, may have escaped centralised production and were possibly used by non-ruling elites to negotiate their social position (Joyce 1993: 261). This dichotomy between monumental art and ceramics is crucial, since, as I discuss later, the gender identities represented on vessels do not completely overlap with those elaborated on stelae. Compared to other media, figurines show more concern for women's labour. Since there is no clear evidence concerning the artisan's identity, it is possible that women made figurines. This hypothesis has been discussed by Gustafson (2002a: 66-71), who suggests that female figurines may have been evidence of women's agency and spirituality. However, as Cohodas (2002: 43) points out, even female artists may have worked within a framework of socially accepted, male-focused ideological codes. Figurines, therefore, might reveal androcentric views shared by both men and women.

Elisa Perego
In Mesoamerica, space was a medium subject to intense ideological manipulation. Spatial analysis can be considered a viable means to investigate local gender ideologies, but it is open to question whether it can serve as the basis for inferring real social dynamics. Moreover, spaces such as houses, rooms and sanctuaries are rarely attributable to one or the other gender, unless we rely on evidence (e.g. iconography), which could lead to the pitfalls I have already pointed out. Maya mortuary evidence supplies abundant information on local religious beliefs and people's status, gender and wealth. However, as postprocessual and ethnographic studies have demonstrated, the treatment reserved for the dead individual does not necessarily reflect his/her social condition in life (Hodder 1986;Ucko 1969: 265-269). Forms of ideological elaboration of the deceased's identity were common in Mesoamerica. Gender roles and relations, therefore, may have been constructed through the manipulation of the funerary ritual. Osteological analysis offers significant insights into several aspects of ancient lifestyle (e.g. diet, division of labour and health). Nonetheless, the information provided must be carefully evaluated. For example, Cucina and Tiesler (2003) note that at some Mexican sites, the evidence suggests that elite women were more subject to caries than elite males. This pattern, however, may have been caused by different factors, such as minor consumption of meat by women or their more frequent consumption of carbohydrates, or both.
In conclusion, although every line of evidence has its weaknesses, the employment of several of them at the same time should reduce the risk of producing an a priori construction of the past. For example, the quasi-universal assumption that women are in charge of food and textile production seems confirmed for the Maya by ethnographic, ethnohistorical (Vail and Stone 2002: 218), archaeological (Joyce 1992(Joyce , 1993 (Fig.  1) and osteological analysis (Wanner et al. 2007). It is probable, therefore, that Maya women were indeed involved in such activities.

The Data Actions
Ethnographic, archaeological and osteological evidence for the Maya suggests the existence of a sexual division of labour. Men engaged in warfare, hunting, out-field farming and long-distance trade. Women raised children and animals, cultivated the house garden, processed food and produced textiles (Ardren 2002b: 75-81;Joyce 1992Joyce , 1993Neff 2002;Wanner et al. 2007). Recent research suggests that women were possibly involved in farming (Neff 2002: 37-38;Robin 2006). Osteological analysis of groups living in Northern Belize shows that women performed heavy physical work (Saul & Prufer 2005: 311-312). It is unclear if activities such as pottery-making were gender-related. Royalty was generally a man's matter, although women played a significant role in cementing alliances and transmitting royal blood and, rarely, held power on their own (Hewitt 1999;Josserand 2002;Marcus 2001: 324-330). Female involvement in ritual activities is revealed by ethnographic accounts and iconography (Hendon 2004: 313-318;Josserand 2002: 127-139). Hendon (1996Hendon ( , 1997 highlights the economic importance of cloth and food production. The McCaffertys' (1994,1996) works on textile production in Mesoamerica suggest that spinning and weaving acted as metaphors of female identity. The connection of these activities with goddesses involved in childbirth, sexuality and healing symbolised the power women were able to gain in the few sectors open to their agency: reproduction and household technologies. At Ceren, El Salvador, the presence of spindle whorls and artefacts for the processing of different fibres (maguey and cotton), and their peculiar patterns of distribution reveal that women had probably set up a textile workshop. The subdivision of different tasks among different households and the complexity of the labour organisation suggest that the artisans were producing for a market beyond the family needs (Beaudry-Corbett and McCafferty 2002).
Food processing, associated with women, is an activity often devalued both by archaeologists and within numerous societies (Brumfiel 1992). Food, however, is a powerful means of negotiation at different levels, both between political entities and between genders (Appadurai 1981;Brumfiel 1987Brumfiel , 1991Hastorf 1991Hastorf , 2003. A careful evaluation of food processing activities, tools and facilities reveals significant social and symbolic dynamics. At Formative K'axob (c. 800 BC-AD 250), Belize, several sherdlined pits were probably used for cooking (Bobo 2004). The placement of the pits inside or outside the house provides information on the conceptualisation of domestic space; their common location in the courtyard suggests communal and perhaps ritual preparation of food. It also implies that the cooks were not confined to the home and could have contacts with other people. Moreover, the deposition of human remains and offerings inside the pits indicates that there was not a rigid division between sacred/ritualised areas and domestic spaces.

Spaces
Space is the medium through which people substantiate their being-in-the-world (Robin 2002b: 248;Tilley 1994: 7-34). It is also a means to negotiate relations between different social entities and between genders (Hill 1998: 116-117;Hodder 1990). The spatial arrangement is often a way to control women. Numerous forms of female confinement have been elaborated throughout history: segregation at home, segregation in religious buildings or seclusion within limited spaces during menstruation and childbirth (Galloway 1997;Gilchrist 1996). Women, however, are able to create symbolically marked spaces on their own. In his studies on the African Ilchamus, Hodder (1991) suggests that the symbolic decoration made by women on the walls of the hut is a way to assert female power in the domestic sphere.
In Mesoamerica, space was imbued with symbolic meanings referring to local cosmological beliefs (Mathews and Garber 2004). Space was also a powerful means of ideological negotiation. The rulers promoted the creation of space patterns symbolically marked, and manipulated the ways in which these elaborations were "consumed" by the audience (Sanchez 2005). Architecture -for example the presence of buildings that varied in height and were arranged on different levels -and iconography provided a clear sense of hierarchy which people would have begun to internalise in their childhood (Meskell and Joyce 2003: 51-53). At Calakmul, Mexico, the distribution of tools and prestige goods on Structure II reveals that artisans worked on the terraces at the base of the pyramid, while the members of the aristocracy lived in the higher court, over them. This spatial arrangement seems to have marked the elites' authority over the commoners and their right to exploit the production of the entire social body (Robin 2004: 155-156).
Space was also a gendered medium. Joyce (2000a: 74) notes that on public monuments such as stelae women were less frequently represented than men and rarely appeared without a male companion. When men and women were depicted together, their disposition on the monument's surface would have revealed a dichotomy between male and female. On stelae placed at the end of a plaza, women tended to appear on the back of the monument, whereas men were generally represented on the front, facing the open space. Otherwise, women could be depicted standing at ground level, with men placed over them, in elevated positions (Joyce 2000a: 74-75).
Gender ideology and cosmological symbolism permeated symbolically charged spaces such as sweatbaths. Sweatbathing has been practised throughout Mesoamerica since the Preclassic period. Women in childbirth and after menstruation were and are taken to sweatbaths, which are characterised by dark, narrow and womblike spaces. A pervasive symbolism links the sweatbath with sexuality, reproduction and caves, all elements emphasised in Maya ritual. Ethnography shows that today sweatbathing is a practice described by both men and women through hilarious stories about sex and references to the female body (Moyes 2005: 187-191). Although it is risky to extend contemporaneous accounts to past ages, it is possible that this practice was a locus of gender negotiation also among the ancient Maya. On Late Classic vessels (AD 600-900), orgiastic scenes involving old deities attended by young women took place in sweatbaths (Houston 1996: 142;Houston et al. 2006: 197). Houston (1996) suggests that the inner sanctuaries erected in the Cross Group temples at Palenque were perceived as symbolic sweatbaths where the local gods' births were celebrated under the auspices of the ruler.
Research is needed to delineate the role of male rulers as promoters of re-birth and fertility and to understand the meaning of the association between young women and old men in sexualised contexts.
Archaeologists also identify more mundane gendered spaces. In the subroyal residences of Copan, two different kinds of buildings have been discovered. The first kind is elaborate, has traces of ritual activities and contains men's burials; it was probably a male residence. The second one is smaller and less ornate, and contains female equipment, cooking areas and women's tombs. Cohodas (2002: 33-35) considers these data evidence for women's lower status. The existence of houses in which men lived together until their marriage was documented by the Spaniards. Joyce (2000b: 267-269) recognises these specialised palaces in buildings discovered at several Maya sites (e.g. Chichèn Itzà and Uxmal). These residences are characterised by the presence of giant phalli and mural paintings representing aristocratic male activities such as ball-playing, dancing, hunting and warfare. This form of seclusion may have been aimed at promoting male solidarity, male gender identity and homosexuality (Meskell and Joyce 2003: 125-126).
Less hierarchical arrangements probably characterised minor settlements. In her study of the agricultural system at Chan Nòohol, Belize, Robin (2002aRobin ( , 2002bRobin ( , 2004Robin ( : 160-4, 2006 highlights that neither the distribution of the artefacts inside the dwellings, nor the spatial organisation of the site indicates a sexual division of labour. On the contrary, the openness of the pole-walled buildings and the location of the fields near the village would have permitted a great interaction among individuals and women's involvement in farming.

Bodies
The body is the primary means humans posses to interact with the world and to create relationships between individuals (Fisher and Loren 2003;Joyce 2005). Placed on the body, ornaments and clothes become a second skin which mediates the relation between the person and the external world, whilst instruments expand human agency beyond the range of the limbs (Knappett 2006: 241-242). Dress codes are an expression of the numerous axes of the individual's identity: gender, age, economic level and origin (Blum Schevill et al. 1996;Wobst 1977).
The Maya devoted enormous attention to their bodies and their ornamentation. Physical modifications were common: skulls were elongated during childhood; earlobes were enlarged and pierced; noses were adorned with jewels; teeth were modified. Skin was tattooed and hair adorned with complex headdresses. The dead had their faces covered with masks, or their skulls removed and substituted by masks (Meskell and Joyce 2003: 33-43;Tiesler 1999;Williams and White 2006). To describe this pervasive phenomenon, Joyce (2000a: 13) adopts the concept of "beautification": the necessity to become attractive to gain social recognition.

Elisa Perego
In Mesoamerica the construction of the symbolic body (i.e. the body socially controlled and elaborated) was gender related. The creation of 'docile' personalities may have started with the direct manipulation of the physical body. Although the pattern is not general, there is evidence that men enjoyed better nutrition (Ardren, 2002b: 74-75;Cucina and Tiesler 2003: 6;Danforth 1997: 135). As Ardren (2002b: 75) notes, the consequences of this social behaviour were not only physical, but also psychological; since their childhood, girls would have internalised female inferiority through repeated negations of their nutritional needs. It is unclear to what extent body modifications were gendered. At Yaxuna, Mexico, cranial shaping was more common for girls, whereas dental modification, reflecting the exploitation of prestige goods (e.g. incrustation with haematite), was often a male prerogative (Ardren 2002b: 74). Conversely, in her analysis of 1515 skeletons from Honduras, Mexico and Guatemala, Tiesler (1999: 3, 5) does not observe significant differences between sexes, although she notes that dental modification was slightly more common for women. In both cases (nutrition and physical modification) further research is needed to understand why different gendered practices were adopted at different sites, and in social strata and historical phases.
Iconographic studies reveal interesting patterns. The women represented on public monuments are almost completely covered by ornaments, feathers and elaborate clothes. The representations of their bodies do not depict breasts or other sexual characteristics. Conversely, men's bodies are partially naked and, in Joyce's (2002Joyce's ( : 337, 2000b; 264) opinion, highly sexualised and metaphorically offered to the gaze of both female and male spectators. A different treatment of the female body can be detected on ceramic vessels and figurines. Here women are clearly recognisable by the presence of breasts and their bodies are generally less covered (Joyce 1992(Joyce , 1993(Joyce , 2000a (Figs. 1-2). However, painted vessels and figurines do not reveal the same female identities. On ceramics, women appear generally involved in ritual, court scenes and sexual acts with deities and animals (Fig. 2), rarely in productive activities, whereas the focus of figurines is on female labour and reproduction. According to Joyce (1993: 261-263), these different elaborations of women's identities depended on the medium's meaning and audience. Stone monuments, on which royal women are depicted while offering food and textiles to men, would have emphasised the rulers' right to exploit the production of lesser individuals. Figurines and vessels found in non-ruling households may have signalled the commoners' attempts to emphasise their roles as producers. If women made figurines, the attention paid to motherhood and female labour might have been a means to gain social recognition.
The symbolic construction of the body thus had political and economic implications. The products of women's work, food and textiles, were fundamental to maintain the household, to carry out rituals and to sustain men's claims in the political arena (Hendon 1997: 44-45;Joyce 1992Joyce : 66, 1993. The desire of the household elders to control women's production, and the will of the royal elites to exploit the households may have created tensions crystallised in the construction of the socially-shaped body. Living and dead bodies became loci of negotiation between ruling-inferior classes and between genders. According to Joyce (1993Joyce ( , 2000aJoyce ( : 88-89, 2001, the increasing emphasis on ornamentation and heterosexual differentiation would be the evidence of Women's Voices in a Male World: Actions, Bodies, and Spaces Among the Ancient Maya 77 the control which elites and states tried to impose over single households and individuals in order to exploit their production. Sexual control was one of the principal means to achieve power. In the sacred Maya book, The Popol Vuh, the dead god tries to kill his daughter when he discovers her illicit pregnancy (Christenson 2003: 131). In Maya imagery, representations of sexual domination are not rare. Captives are often shown in an aroused state or in sexualised contexts, and a stela from Toninà probably portrays a publicly sanctioned rape (Houston 2001: 213). Sexual degradation may have been a means to achieve control over enemies (Houston et al. 2006: 207-219).
More ambiguous are those vessels presenting women with enormous breasts touched by elderly gods or by animals, or women copulating with monkeys and insects (Houston 2001: 212; Meskell and Joyce 2003: 127) (Fig. 2). Although these ceramics may have had a religious meaning, the fact that women were involved in sexual activities with animals and with males not conforming to proper codes of decorum may be significant in terms of how female sexuality was conceived, and would require further elaboration. Evidence suggests that women's bodies may have been the focus of men's attempts to subsume female prerogatives such as the power of fertility. The ideological exploitation of the feminine principle was crucial to justify male royalty (Gustafson 2002b: 141). It has been hypothesised that rulers wore elements of the female costume to incorporate every nuance of local gender identities (Joyce 1992: 68; see however Houston et al. 2006: 52). The emphasis given to royal male reproduction was shown by the symbolism concerning male activities such as the scaffold sacrifice, which were supposed to enhance fertility (Taube 1988: 350-351). Female reproductive acts such as breastfeeding, childbirth and pregnancy were not revealed on public monuments (Meskell & Joyce 2003: 44), although the roles of women as mothers were indeed emphasised by the inscriptions which accompanied the queens portrayed with their offspring. Since public monuments were erected by male rulers to strengthen their power, this elaboration of motherhood seems a political construction aimed at reasserting the purity of the king's lineage and his entitlement to reign (cf. Marcus 2001: 330;Sanchez 2005: 269), possibly de-emphasising at the same time women's direct control over reproduction.
The lack of sexual attributes of women's royal bodies has its counterpart in Tomb 2 at Yaxuna, where a defeated and sacrificed royal family was buried (Ardren 2002b: 81-87). Here women's royal power appears sexualised and contextualised -but with tragic consequences. One princess was killed not long before giving birth; the other, probably the dynastic heir, had her lower limbs defleshed and was buried with a "La Muneca" figurine, possibly a symbol of sexual appropriation. This emphasis seems to reveal an effort to 'tame' female attempts to gain direct power by 'hurting' them in their femininity.

Suggestions for Further Research
Further research must be dedicated to engendering ritual practices and symbolism. We can consider, for example, an ancient Maya and pre-Maya symbol, the cross, a representation of the world and a symbol of power (Headrick 2004: 370-377). At Formative K'axob, the use of bowls decorated with the cross motif seems reserved for male burials (Headrick 2004: 368-369;McAnany 2004: 416). The exploitation of such symbolism has been read as the attempt by a kin group to gain power in the village (Headrick 2004: 377); the dead were probably male leaders and shamans. Another ritual practice rarely considered in terms of gender is caching (deposition of offerings in the ground). However, an evaluation of caching as a gendered action is possible. An example is offered by the same site of K'axob, where a local cache was composed of four vessels arranged in a cross pattern. The cross, as noted, was probably a symbol of maleness and power. Furthermore, the quadripartite cache and three (probably) male burials containing the bowls with the cross design were covered with fragments of golden-yellow chert bifaces (McAnany 2004a: 416). The bifaces were imported items employed for agricultural work. Heavy agricultural tasks and long-distance trade were more likely men's activities. The colour probably referred to ripened maize (McAnany 2004: 417-418); in ancient Mesoamerica, maize was a pervasive symbol linked to centrality, power, re-birth and prosperity (Taube 2005: 25-30). Male authority, therefore, appears to have been emphasised through different 'layers' of meanings.
Research is needed to identify female forms of self-expression in the archaeological record. Although evidence concerning women's roles in Maya society is not scarce, many of the sources available were products created by men for a male audience. As a consequence, even in those cases in which women are revealed, it is difficult to discern whether we are looking at female discourses and choices or at ideological structures imposed on women by males (Ashmore 2002: 231). However, forms of resistance must have existed. Food and textile production might have been the loci of women's creative responses to a male-dominated system. Research should focus on the evolution of food-processing strategies and technologies to acknowledge women's contributions to macro-scale phenomena such as diet modifications and negotiation between producers and consumers. Textile production was another means women employed to gain social recognition. According to the McCaffertys (1996), women emphasised the importance of their work by linking weaving to sexuality and by using spindle whorls which had been decorated with care. The designs on the spindle whorls would have made these objects symbolically charged and reveal, now as then, the complexity of the female imagery (see however Brumfiel 2001, who notes that the spindles were probably produced by artisans). Joyce (1992: 64-65) notes the symbolic and cosmological meaning of the decoration of the clothes depicted on public monuments. Trevelyan and Forbes (2002: 98) suggest that the unusual architectural decoration of Late Terminal Classic Uxmal (c. AD 680-900), Yucatan, reproduced textile-based designs. Since women produced clothes, it is possible that cloth decoration -perhaps through hidden allusions -conveyed a female-oriented ideology. Although men may have certainly tried to exploit women's work and perhaps compelled women to adopt specific artistic codes, further research may enlighten women's contributions in promoting the creation, manipulation and diffusion of religious and political symbolism through the production of textiles.

Conclusions
In this article I offer an overview of recent attempts to deal with gender issues in Maya archaeology. The works discussed provide examples of the potential that gender studies have in furthering the development of Mesoamerican archaeology. The materials offered by the Maya record provide gender archaeology with a rich dataset that can be used to refine theory and methodology; in turn, a stronger focus on gender dynamics would foster a more holistic understanding of Mesoamerican ancient societies. I hope that my focus on gender and micro-scale phenomena such as caching, weaving and spatial organisation will prove useful in addressing wider issues concerning economy, politics and conflict. Taking into account the weaknesses inherent in the available sources, I argue that the employment of multiple lines of evidence can offer significant insights into Maya gender dynamics. The manipulation of the body, the ideological construction of iconographies and spaces, and the control of (re)productive actions were the loci in which forms of gender negotiation were played out. The physical and symbolic exploitation of women may have been one of the means through which men achieved control over the economic and political system. However, women's importance in Maya society cannot be underestimated. We have evidence that women worked, worshipped, became rulers and contributed to the formation of the archaeological record. A fresher focus on often downplayed female activities such as food and textile production can further expand our knowledge of women's roles in promoting social change and elaborating ideology and symbolism. Although in ancient Mesoamerica men's control over the communicative media was strong, research is needed to allow Maya women to raise their voices -are archaeologists willing to listen?