THREE
Production and Consumption in the Countryside
A Case Study from the Late Classic Maya Rural
Commoner Households at Copán, Honduras
INTRODUCTION
Producer households are the backbone of agrarian societies and make up the bulk of the domestic economy, an observation that holds through time and space.1 Anthropologists routinely investigate the nature of production, its organization, what goods or services are produced and by whom, and whether the domestic income is supplemented with extra-household production. These questions reflect a cross-cultural interest in what Hirth (2009b) calls “housework” and can be answered in both archaeological and ethnographic contexts (e.g., Robin 2003; Wilk 1991). The perspective of household archaeology offers a way to explore these issues through the material expressions of cultural practices.
This chapter presents an archaeological case study drawing from eight “Type 1” sites2 (Gonlin 1993, 1994, 1996; Webster and Gonlin 1988) that were inhabited by Maya people who lived in the hinterland of the Copán kingdom during the Late Classic period (AD 650–900) (Figure 3.1). Most, but not all, people who lived in rural areas of Classic Maya polities were commoners (Lohse and Valdez 2004), as reflected in the particular material signature seen throughout Mesoamerica (e.g., small, simple structures; utilitarian artifacts; lack of hieroglyphics; and lack of elaborate burials; see Lohse and Gonlin 2007 for a discussion of the commoner concept in Mesoamerica). While it is presumed that farming was the primary occupation of rural commoners, a wide variety of activities took place, some of which have been traditionally assigned exclusively to elites.
Figure 3.1. Map showing the locations of eight rural sites that were excavated during 1985–1986, Copán Valley, Honduras; site names are indicated (from Gonlin 1993:77)
These eight rural sites3 offer a particularly rich database to explore issues of household production and consumption for a number of reasons. They are well contextualized since they were excavated as a continuation of Pennsylvania State University’s urban zone excavations and rural survey and testing program, co-directed by William T. Sanders and David Webster.4 Like the urban excavations, this project strove for complete horizontal exposure of all architecture, clearing of structure floors, excavation of extramural space (behind buildings and in the courtyard), excavation to sterile soil in many cases, and testing of the toft area (see Varien, this volume, for a discussion of the toft area in the American Southwest and Stanton, Brown, and Pagliaro 2008 for an informative discourse on refuse disposal). This strategy ensured a large sample of artifacts, features, architecture, and extramural space (average = 80%) where activities were likely to have taken place (Hendon 1987, 1996). As increasing numbers of sites in the hinterlands are investigated, an emerging view of the complexity of such settlement is coming into focus for the Classic Maya (Dixon 2011; Douglass 2002; Iannone 2005; Iannone and Connell 2003; Yaeger 2000; inter alia), which will undoubtedly refine our reconstructions of these kinds of ancient households.
Furthermore, comparisons between rural and urban areas, and commoners and elites, are possible at Copán because of the various types of research and the long history of investigation that has taken place over the past centuries. Western explorers in the 1800s (Galindo 1836; Gordon 1896; Stephens 1841) recorded preliminary information about the city and its surroundings. Since then, many projects have focused on Copán’s grand temples, palaces, sculptures, iconography, and hieroglyphics (Agurcia 1996; Andrews and Bill 2005; Baudez 1994; Doonan 1996; Fash 2001; Sharer et al. 1999; Traxler 2001; inter alia), urban neighborhoods (Ashmore 1991; Diamanti 2000; Gerstle and Webster 1990; Hendon, Fash, and Aguilar P. 1990; Maca et al. 2011; Sheehy 1991; Webster 1991; Webster et al. 1998, Widmer 2009; Willey et al. 1994; inter alia), outlying occupation of the valley (Baudez 1983; Canuto 2002; Diez 2011; Fash 1983a; Freter 1998; Gonlin 1993; Saturno 2000; Webster 1985; inter alia), and projects that incorporate technological advances (e.g., Aoyama 1999; Bill 1997; Richards-Rissetto and von Schwein 2011; Whittington and Reed 1997), all adding immensely to our understanding of the kingdom’s history. Several recent publications nicely summarize this vast database (Andrews and Fash 2005; Bell, Canuto, and Sharer 2004; Fash 2001; Webster, Freter, and Gonlin 2000).
THE HOUSEHOLD CONCEPT
Production and consumption within society can be investigated by focusing on the household. Anthropologists see these two functions as intrinsically connected and often analyze them together (e.g., Kramer 1998). As the title of this volume suggests, what households do is of fundamental importance to the economy of a society. As archaeologists we must first identify “the household” from material remains. We look for, and expect to find, the physical embodiment of social, economic, ideological, and political processes, and ideally, all in one location. Households in ancient Mesoamerica are most often identified by clusters of buildings found together in courtyard groupings. These groupings can range from the closed courtyards of Teotihuacan to the more open pattern of associated buildings often found in the lowland Maya region. The buildings within any particular cluster may have been residences or ancillaries, the deduction of which can be problematic, even in the best of conditions (e.g., El Cerén, El Salvador [Sheets 1992, 2002]). Structure function figures importantly in the task of assessing production and consumption but is addressed elsewhere (Gonlin 2004). While the excavations in rural Copán are of houses, not households, there is sufficient evidence to make the theoretical leap from material remains to socioeconomic unit (see Varien, this volume, for further discussion). In rural areas especially, the Classic Maya have neatly packaged themselves into discreet settlements. Thus, for the purposes of this chapter, it is assumed that the spatially isolated groupings of houses found on the rural landscape throughout the Copán Valley are indeed indicative of ancient households and the activities and processes that occurred within them during their lifetimes. While archaeologists realize that this conclusion is not entirely correct, we can productively employ it with caution. The relative geographical isolation of groups combined with the close proximity of building remains within groups, as well as the nature of the artifact assemblages, points to self-contained units of operation (Lucero 2001, 2002). This statement of course does not indicate complete social, economic, ideological, or political isolation or self-sufficiency, however, as a household is a part of a larger community. (The value of community studies is well-known in archaeology [Canuto and Yaeger 2000; Stomper 2001] but is not emphasized in this chapter.) As we attempt to understand the household and its operations, we must recreate institutions, relationships, and ways of life from the artifacts, features, and architecture of the past. Maya household archaeologists (Gonlin 1993; Haviland 1985; Hendon 1987, 2010; Lohse 2001; Tourtellot 1983) have long focused on socioeconomic organization, with more recent work specifically considering the criterion of gender (Beaudry-Corbett and McCafferty 2002; Gustafson and Trevelyan 2002; Haviland 1997; Hendon 1997; Neff 2002; Robin 2006; Robin and Brumfiel 2008; Sweely 1999). This chapter too shall consider gender, a variable of household consumption and production (other chapters in this volume that consider gender include those by Douglass and Heckman, Gougeon, Henderson, Neff, Snow, and Wiewall, among others). The division of labor in every society breaks down along the lines of age, gender, ability, and status. Hence, consideration of these variables will enrich our understanding of the past and paint a more complex picture.
A standard definition of a household, one that is used by many authors in this volume, is that it is a coresidential activity group (Ashmore and Wilk 1988), which includes the women and girls, and men and boys, who produce, consume, distribute, and perform other activities. If we assume that there is a universal gendered division of labor, that this division is expressed in household activities, and that the evidence for these activities is recoverable through artifacts, features, and reconstructable patterns, then it may be possible to determine who the performers were in such activities, the locations of these activities, and the contribution of these activities to the domestic economy. If particular artifacts are associated with certain activities, then theoretically, it may be possible to determine which gender was more likely to have used certain tools. Ideally, production of goods and services may be assigned to one gender or another and household space may be partitioned along the lines of gender. In reality tools are multipurpose, tasks may be interchangeable between genders, there may be overlap in the use of space, there may be more than two genders (Storey 2005), many productive activities leave behind little trace, children may contribute substantially to productive efforts, and cultural and natural transformations obscure patterns.
GENDER: A CONSIDERATION OF CHRONOLOGY, GEOGRAPHY, STATUS, CLASS, AND AGE
Additional criteria should be evaluated to understand the connection between gender and production and consumption for the Classic Maya: chronology, geography, status, class, and age. The Late Classic period at Copán spanned hundreds of years, from AD 650 through AD 900. Given the dynamic nature of culture, patterns may very well have changed over this time period, which encompassed the dramatic decline of the kingdom. Earlier patterns during the Early Classic (AD 100–400) and Middle Classic (AD 400–650) may have differed from household strategies employed later in Copán’s history. This chapter looks at small rural households occupied throughout the Late Classic period. In addition to chronological variation, geographical differences need to be considered in an area as diverse as the Maya lowlands. Recent work (Ardren 2002a; Gustafson and Trevelyan 2002; McAnany 2010) highlights both chronological and geographical variability, particularly with respect to gender roles. This chapter focuses on the southernmost expression of the grand style of Classic Maya culture, as expressed at Copán, Honduras.
Gender in a complex society encompasses many different classes and statuses. For the Classic period Maya of Mexico and Central America, archaeologists have been fortunate to have depictions of women and men from pottery vessels, figurines, jadeites, murals, sculptures, and architectural features (Freidel and Schele 1993; Schele and Miller 1986). While epigraphy and iconography offer snapshots primarily of elite Late Classic men and women, their value for understanding non-elites may lie in the symbolism that we derive from these portrayals (Lucero 2010). For example, from these depictions, a concept of gender complementarity has been used to explain the differences perceived in male and female roles. Rosemary Joyce (1992, 1993, 1996), Christopher Fung (1995), Joel Palka (1999), and others subscribe to this reconstruction of complementarity. However, these depictions are primarily of royal and elite women and men. It cannot be assumed that people of low status would have had similar roles and have been similarly portrayed or that the relationships between the genders or age-specific roles were identical from one class or status to another. As Vail and Stone (2002:204) have stated, “Classic Maya imagery, dating to c. 250–900 C.E., says little about the lives of the commoners because it is narrowly focused on the elite who commissioned such artworks.” This situation is perceived by others as well. In referring to Maya women, Ashmore (2002:234) has aptly noted, “If the material traces of women’s lives are generally subtle relative to other categories of evidence, the traces of commoner women are subtler yet.”
When examining productive activities in an agrarian economy, one must account for age and the early recognition of adulthood. Kramer (1998) studied intensively both age and gender in the division of labor in a modern agrarian context and found both characteristics to be of vital importance. Netting (1993:71) used cross-cultural information to conclude that in many societies, by the time a child reached adolescence, he or she was already performing at the adult rate of work. Furthermore, biological aspects must also be considered. Vail and Stone (2002:204) note, “Our analysis suggests that Maya women were broadly divided into two categories according to an age-based dichotomy: premenopausal women, whose sexuality was seen as a threat to men and in need of control, and grandmotherly figures associated with the security of the home.” Many questions need to be addressed. Did gender complementarity, a concept largely derived from elite portraiture, exist among the rural commoners of Late Classic Maya Copán? Were productive tasks carried out at farmsteads in a complementary fashion and were goods consumed likewise? Is there archaeological evidence that supports this model or another? And where did children fit into the picture of production and consumption (Ardren and Hutson 2006)?
To answer these questions and many others, an agrarian model of production is examined to shed light on the types of activities performed by swidden and intensive agriculturalists in a gendered context. Three main lines of evidence from the archaeological record of rural Copán, consisting of artifacts, architecture, and analyses of human remains, will be drawn upon. The presence of some artifacts that are often viewed as gender-specific, such as grinding stones, spindle whorls, celts, and projectile points, will be examined to analyze their distribution. The style, function, and layout of architecture can provide insights about the distribution of production and consumption since buildings provide physical places for harboring activities. Osteological studies involving isotopes and paleopathological analyses are useful in determining patterns of consumption within the household.
AN AGRARIAN MODEL OF PRODUCTION: CHILDREN AND ADULTS
As members of an agrarian society, ancient rural Copán householders performed innumerable agricultural tasks, and the household figured prominently in the implementation and completion of such tasks. The chores on a farm are never-ending and the production of food is of prime importance while supplemental income may be derived from intermittent crafting and multicrafting (Hirth 2009a). All able-bodied members, regardless of age, may have contributed in some fashion, as a wide range of activities occurred on a daily basis. As others who have addressed the topic of household labor and gender have done, particularly in relation to Maya farmsteads (Neff 2002), a listing of associated activities is in order. There are several studies of agriculturalists that provide information on productive tasks, and here I refer to a couple of these works (Kramer 1998; Wilken 1987). Kramer focuses on the productivity of children, since the additional criterion of age should be added to gender studies, and Wilken examines resource management of traditional agriculture in Central America and Mexico.
According to Kramer (1998), children do a variety of productive tasks on a regular basis as members of a subsistence agricultural society, such as the one she studied ethnographically in Xculoc in the Puuc region of Mexico. Children generally are neglected in anthropological studies (Hirschfeld 2002) and probably more so in archaeological contexts (Ardren 2006; Kamp 2001), where their presence is even less visible than in ethnographic situations. What we can learn from Kramer’s work is not what Maya children specifically contributed to the household in terms of production 1,500 years ago, but the general pattern of contribution of children living in agrarian societies throughout the world. Kramer (1998:appendix B) provides a list of tasks that were likely performed by children anywhere, given the similarity of subsistence strategies: household activities; maintenance and manufacture; food processing and preparation; tending animals; resource acquisition; child care; personal maintenance; social activities; garden work; other labor; education; milpa work; and ritual activity. The types of artifacts associated with these activities that children perform would differ little, if at all, from the artifacts used by adults. In other words, it is highly likely that it will not be possible to determine the age of the user of such implements, such as a spindle whorl, a metate, or a knife. It also may not be possible to determine the gender of the users either, especially when tools, such as obsidian blades, are general purpose in nature.
Kramer has inferred the following patterns from her work. Not surprisingly, “[d]omestic work is largely sex-patterned and, although young boys may allocate some time to domestic work, girls’ investment far surpasses that of boys” and “[f]emale children 12 and older spend between 49% and 52% of their time in domestic work, which is comparable to their mother’s work effort” (Kramer 1998:131). These quotes and list of activities allow us to envision the work of children, as well as women and men, on the farmstead. Our Western twenty-first-century ideas of “adulthood” and “childhood” and productivity are a poor model for ancient agrarian societies, or even modern agrarian ones. We cannot assume that a person in Classic Maya society would had to have reached twenty-one years of age to be considered an adult and to be a productive member of the household. Both production and physical reproduction probably happened much earlier in life, and adult patterns of consumption of goods and services may have also occurred. Whittington’s (1989) extensive osteological work at Copán on remains of Type 1 and Type 2 inhabitants has shown the mean age at death for individuals ages 20 and above to be 31.67 (95 percent confidence interval is 20.00 to 48.97). A cautious interpretation of this figure (keeping in mind the osteological paradox [Wood et al. 1992]) means that lifespans were short for the Classic Maya and cultural recognition of adulthood may have been accelerated. In Storey’s (1992) analysis of the Copán sample, she considers subadults to range in age from newborns to fourteen years.
Wilken (1987:22), in reference to agriculture, makes an important observation for the purposes of this study, stating that “generally, traditional technology consists of ways of doing things rather than equipment for doing them,” and that most traditional tools are general purpose. Field preparation first involves clearing, which consists of tree felling, clearing shrubs and grass, and bush felling (Wilken 1987:13), all of which are physically demanding chores. In prehispanic Mesoamerica, the lack of beasts of burden precluded field preparation involving the animal-drawn plow. The most likely implement for field preparation may have been a stone hoe, the remains of which are found at some Classic Maya sites (especially in Belize; see, e.g., McAnany 1992; Neff 2002). Digging sticks, usually made of perishable materials, would have been the tool of choice for planting. There is limited evidence for intensive forms of agriculture in rural Copán (see Davis-Salazar 2006 for a discussion of flood control) involving terracing, raised fields, or permanent irrigation canals, but according to Wilken (1987), there can be other forms of intensification, such as maize mounds and manual irrigation using water jugs (as in Oaxaca), basins, or gourds for watering crops. These types of intensification would leave little trace in the archaeological record. So although the major productive activity for Classic Maya farmers was food production, there may be little direct evidence of the related activities (i.e., weeding, planting, harvesting).
In addition to food production, other chores of farming families include cloth production, toolmaking, woodworking, house building and maintenance, the chopping and gathering of firewood, hunting, trapping, gathering, beekeeping, cooking, and craft production. Ritual and religious practices may also be carried out by individuals within households (Blackmore 2011; Gonlin 2007). Typical activities assigned to pre-Columbian Mesoamerican women and girls are cloth production (Hendon 2006), including spinning by females of most ages (Ashmore 2002:240) and weaving by women (McCafferty and McCafferty 1991); food preparation and cooking (Brumfiel 1991); tending to animals; child care; gathering; aiding in agricultural tasks; and gardening. Hunting, trapping, warfare, construction activities, toolmaking, ritual activity, resource procurement, beekeeping (Vail and Stone 2002:216), and heavy agricultural tasks are usually activities assigned to men and boys. One may argue that such a division of labor represents a stereotypical view, but we may ask if there is archaeological evidence to support these reconstructions. The study of figurines by Joyce (1993) shows women employed in maize-grinding tasks. The linguistic survey by Clark and Houston (1998) indicates that all craft activities (except for cloth production) contained male prefixes, indicating that such crafts were the products of men (however, see Ardren 2002b:77–78). A discussion of the archaeological evidence from rural Copán follows.
AN ANALYSIS OF ARTIFACTS
Most often, particular types of artifacts are tied in with particular types of activities, and archaeologists rely upon the correlation between tool and action to reconstruct production. It may be tempting in gender studies to assign certain artifacts to certain genders, but Karen Bruhns strongly advises against this practice. The ubiquitous metate illustrates this idea well.
[E]very household from the period of incipient agriculture to the present had its metate or metates. This is because the metate is your pre-electricity blender, food processor, mouli julien and grinder: in the kitchen one grinds spices, herbs, chocolate, nuts, vegetables and fruits, makes baby food, mashes cheese for papas a la huancaina, grinds up dried fish, mashes meat, and so on ad infinitum. Everything that needs to be pulverized, flattened, mashed, squashed or ground is processed on that metate. (Bruhns 1991:422)
Aside from food preparation, a list of non-culinary uses of metates, including industrial and pharmaceutical applications, is given (see also Hayden 1987). It is also noted that women or men may perform these activities. Specific attention to details such as class, status, cultural situation (Bruhns 1991), and chronology (Haviland 1997) may identify the users and the context. With Bruhns’s cautionary note in mind, the artifact distributions in rural Copán can be examined.
While some archaeologists have used monuments, murals, and figurines to successfully extract information about ancient Maya gender and economic activities (e.g., Hendon 2003; Joyce 1993, 1996; and many others), a recent find sheds clear light on what may be one of the very few depictions of commoners. A literal picture of Classic Maya production and consumption was found in the remains of a mural at Calakmul, Mexico (Vargas, Vásquez López, and Martin 2009). Drawings of women, men, and a child portray activities and glyphs name various roles, such as the “maize-gruel person,” “maize-bread person,” “maize-grain person,” “salt person,” “tobacco person,” “clay-vessel person” (Vargas, Vásquez López, and Martin 2009:19248).
The basic artifact assemblage found in rural Copán is also found in urban Copán, particularly in the high-ranking neighborhood of Sepulturas (Diamanti 1991; Hendon 1987, 1997; Willey et al. 1994). From ceramics to lithics to ground stone, these items represent essential household equipment, regardless of class or status and regardless of who used them. What differs from rural to urban contexts is the diversity of artifacts (sometimes referred to as richness; Leonard and Jones 1990). The urbanites of Sepulturas utilized more specialized tools and a greater range in style for a particular artifact. Besides higher status, a greater diversity of artifacts may correlate with a longer occupation. Overall, however, it is clear that activities carried out were of a similar nature in all areas of the Copán polity, based on the similarity of tool assemblages (Gonlin 1993).
The task now is to examine the rural Copán homesteads to see if their assemblages include the evidence for the above activities and if artifact distributions within a site reveal anything about a gendered use of space. Nearly all artifacts were recovered from secondary deposits making interpretations about a gendered use of space tenuous at best. For example, in rural Copán, at all residences in this sample, ground-stone artifacts are ubiquitous, with metate fragments slightly outnumbering mano fragments (136 to 105, from Gonlin 1993:table 4.17). Shown in Figure 3.2 is the artifact distribution of Site 7D-6-2, displaying fragmentary manos and metates. If only one gender used these objects in particular areas of the house, it cannot be determined from this distribution map. Cultural and natural processes have so transformed the record of small habitation sites to the extent that few in situ remains are recoverable. What is really being mapped is the distribution of refuse behavior and postoccupational events rather than the behavior of food preparation or industrial or pharmaceutical production, since most remains are fragmentary and are found in secondary deposit types (LaMotta and Schiffer 1999). At Aguateca, Guatemala, where in situ materials have been recovered in the site core, it is clear that women utilized certain parts of elite structures for their domestic tasks, including food grinding and textile production (Inomata et al. 2002). Inomata and colleagues (2002:325) make a good point when they observe, “The apparent lack of divisions of male and female spaces at the Copán residential compounds, then, probably reflects the lack of resolution of the data rather than the reality of past practices,” in reference to Sepulturas activities. Likewise, at Cerén, domestic activities seem to be compartmentalized (Sheets 1992).
The pattern evident in Figure 3.2 does highlight, however, that every structure is associated with the remains of grinding stones, possibly indicating that women and girls performed food preparation in each of the buildings and then disposed of their broken implements just outside of their houses. This statement involves another assumption, that is, that the refuse associated with a building came from activities that occurred within and around that building. The presence of manos and metate fragments in relatively great numbers indicates the prevalence of grinding activities. Grinding maize by hand was a laborious chore. In addition to other tasks associated with maize, such as processing, food preparation would have consumed much of a woman’s waking hours, especially since maize was the staple food source. It has been documented through isotopic means that the Copaneco diet consisted of between 62 percent and 78 percent maize (Reed 1998). In addition, almost 60 percent of ceramic remains are associated with cooking and storage activities in rural Copán (Gonlin 1993:table 4.8). The symbolism of manos and metates, as well as many other mundane artifacts, are meaningfully discussed by Hendon (2010:88), who states that metates “index the labor of women and girls.”
Figure 3.2. Spatial distribution of grinding-stone fragments at Site 7D-6-2, Río Amarillo, Copán, Honduras; note that each structure is associated with this type of artifact (from Gonlin 1993:807)
Evidence for spinning was found in five of seven rural commoner residences in the form of ceramic spindle whorls, worked sherds, or discs (Gonlin 1993:389). Although the overall count is extremely low, it is on par with the percentages of these types of artifacts found in the densely populated Sepulturas by both the Harvard (Willey et al. 1994) and Penn State excavations (Diamanti 1991). There is no distinguishable intrasite distribution that shows the activity of spinning being performed in particular areas of buildings, terraces, or outdoor areas. Had such a spatial arrangement existed in the past, it has long been blurred by the elements of time and nature. Hendon’s (1997) data for Sepulturas show a greater frequency of textile production tools in larger residences, perhaps indicating a larger population of women and girls. Though present, the frequency of textile manufacturing tools found in rural areas is extremely low. It is possible that textile production was not a “defining activity of nonelite females” (McAnany and Plank 2001:96), although they did engage in it. Alternately, based on the recovery of organic spindles and whorls at Cerén (Beaudry-Corbett and McCafferty 2002:59–60), where some types are made of wood and coyol palm endocarp, spinning tools may be underrepresented in archaeological contexts. Needles, which indicate weaving, were not recovered in the rural area; however, bone needles and awls were recovered from Sepulturas (Diamanti 1991:236). According to Vail and Stone (2002:211), weaving likely took place in the dry season months of November and April, which correlates with the agricultural off-season, a form of production that Hirth (2009a:20) calls “seasonal crafting.” Weaving is an activity that would have required more skill than spinning and may have been practiced by adults rather than children.
There is artifactual evidence from rural Copán, in the form of grinding stones and spinning implements, to support the activities stereotypically assigned to Classic Maya women and girls. The presence of “male” activities may be examined in a similar fashion, by looking for artifactual evidence of farming, hunting, house building, and warfare. McAnany and Plank (2001:95) suggest that a monolithic role did not exist for males or females, with status or class being an important element for consideration. They cite the fact that Classic Maya rulers have never been portrayed as farmers, although some have been portrayed in deer hunts (McAnany and Plank 2001:93), indicating that the farming role most likely was confined to the non-elite sector of society. Ethnographical and ethnohistorical observations support the reconstruction of men as farmers; in rural Copán, however, there is little direct evidence in the way of artifacts to confirm this activity. The standard biface, ubiquitous at other Classic Maya sites (McAnany 1992) for land-clearing activities, has not been recovered from rural or urban Copán (Willey et al. 1994:264) or in areas just outside the Maya region, such as the Naco Valley, Honduras (Douglass 2002). In the absence of direct artifactual evidence, it pays to keep in mind Wilken’s (1987:22) observation that traditional technology may not involve equipment but rather knowledge.
Another way to approach the reconstruction of agricultural production is through an analysis of ecofacts. When we look at the particular agricultural regime at Copán, and specifically in rural Copán, paleoethnobotanical evidence, being site-specific, can be used to further our knowledge. Lentz (1989, 1990, 1991) examined remains from both urban and rural areas of Copán to gain insight into commoner and elite diets and to determine plants that were cultivated and the ones gathered as wild species. By using this information, we can determine the kinds of crops that were grown and consumed by the populace.
Not surprisingly, Lentz (1991:272–273) found the remains of Zea mays at Copán from sites of all statuses and time periods, and estimates that corn was likely the main staple. In particular, grains and cupules were recovered from rural sites 7D-6-2, 7D-3-1, 34A-12-1, 34C-4-2, and 99A-18-2. Given the ubiquity of grinding stones and isotopic analysis cited above, it is clear that maize comprised the major portion of the diet for people of all statuses. The diet was supplemented with beans (Phaseolus vulgaris), squash (Cucurbita moschata), chayote (Sechium edule), and tree fruits like avocado (Persea americana) and nance (Byrsonima crassifolia) (see Lentz 1991:table 1 for a complete listing of macro-remains from Copán).
In addition to these cultigens, there were numerous tree species that were either cultivated or their products gathered from wild-growing ones. More than any other plant remains at Copán, the carbonized seeds or endocarps of the coyol palm were recovered in great abundance. Lentz (1990) has studied the coyol palm (Acrocomia mexicana) and Heather McKillop (1996) has examined it along with other palm species from archaeobotanical and ethnobotanical sources. At Copán, coyol was found in numerous contexts, such as middens, house floors, vessels, platforms, and patios. Since the seeds are extremely durable, this abundance may result from preservation factors rather than their use as a staple food source.
The coyol can be used as a food source in a variety of ways (Lentz 1990). The kernels or endosperms can be eaten or made into a beverage and it is possible to extract the oil from the kernels for use in cooking. McKillop (1996:280, 288) reports that the trunks of palm trees can be burned to produce salt, and that the sap may be extracted for wine-making. The seeds have a high fat and caloric content (Lentz 1990:189) and it is estimated that they may have provided up to 7 percent of the calories in the diet. Though this percentage is small, the coyol may have supplemented the diet in important ways, especially during the latter half of the Late Classic period, when ecological stress and population pressure put harsh demands on the food supply.
Apart from the direct evidence of coyol, the presence of particular artifacts indicates that coyol may have been processed within the household. Stone slabs with small circular depressions the size of a coyol nut were found at two larger-sized rural Type 1 sites, 11D-11-2 and 34C-4-2, in front of structures in the courtyard. Additionally, nutting stones were found at several rural (11D-11-2, 34A-12-2, and 32B-16-1) and urban sites and may have been used to process coyol nuts. The meat is particularly difficult to extract. Ethnographically, the coyol is found in kitchen gardens, as reported by McKillop (1996:288). An infield-outfield system of agriculture is suggested by the types of plant remains recovered. At one site, 7D-6-2, the remains of “chich” mounds may be indicative of arboriculture (Kepecs and Boucher 1996, in Ashmore 2004), providing additional evidence for the existence this type of agricultural system. Women may have been responsible for the kitchen gardens, while the milpa may have been the primary responsibility of the men of the household. These conclusions, however, are not based on archaeological evidence, nor are they new insights (Netting 1977).
The presence of wild species at some sites indicates that gathering may have played a subsidiary role in the diet. Wild beans and wild grapes (Lentz 1991) may have been collected and then brought home to be consumed. Weedy species were also recovered, indicating that they may have been intentionally brought in and used as bedding or matting material (Lentz 1991:278). Alternatively, they may have been inadvertently carried or blown in.
One wild creature that was captured and brought home was a species of freshwater snail, called jutes. The gathering of jutes was notable at some homesteads (386 jutes at 34A-12-2 and 676 jutes at 99A-18-2) where abundant remains of these snail shells were recovered (Gonlin 1993:table 4.24). The shells are intact, save for a hole at the distal end of the shell where the creature was extracted, a form known as “spire-lopped” (Halperin et al. 2003:208). Lawrence Feldman analyzed mollusks from the Harvard Sepulturas excavations at Copán and notes that the jutes belong to the freshwater genus of Pachychilus. According to Feldman (1994:478), P. corvinus is today used as a snack food and consumed raw or boiled in its shell. The shells of this species, as well as those of P. largillierti, are useful as a source of lime, a substance valuable for maize processing, plaster making, and pottery temper. Healy, Emery, and Wright (1990:178–179) have discussed at length the domestic uses in both modern and ancient Maya contexts, noting that these types of mollusks are most often prepared as a thick, spicy soup. According to Halperin and colleagues (2003), jutes were not only used for subsistence but figured importantly in cave rituals. After consumption in the household, their remains were gathered and deposited in caves, returning them to Mother Earth. This activity may explain why some ancient Copán households have abundant remains of shells (34A-12-2, 99A-18-2), while other households have little (two jutes at 11D-11-2, three jutes at 34A-12-1) or none (7D-6-2, 7D-3-1, 32B-16-1, 34C-4-2). Healy, Emery, and Wright (1990:178) also note that snail harvesting today is most often a male activity, one that can be conducted as men go to and from their fields or on the return trip from a hunt.
Hunting and warfare, supposedly male activities (cf. Reese-Taylor et al. 2009), may be connoted by the presence of arrowheads, projectile points, darts for blowguns (Ventura 2003), or other points (Aoyama 2005), or hunting may be recognized through the identification of hunting shrines (Brown and Emery 2008). In rural Copán, all sites in this sample had these types of artifacts (Gonlin 1993:409), albeit in low numbers. To what degree commoners participated in military activities is unknown, and as McAnany and Plank (2001:93) state, “[i]n the absence of warrior burials, it is difficult to characterize the frequency of military activity among Classic Maya males of either the royal court or households.” McAnany and Plank (2001:94) further state, “Additional evidence for a military role for males from households has been garnered from an analysis of the seasonality of recorded military events that revealed dates of engagement coinciding with agricultural downtime (Child 1999).” Conscripted labor requirements for commoner males have long been proposed (Abrams 1994, 2010; Abrams and Bolland 1999), so it is not inconceivable that military labor may also have been required. Aoyama (2004:291) has noted the notoriously difficult task for recovering archaeological remains of conflict, but sufficient evidence from across the Copán polity suggests that the participation was valley wide. All eight sites in this study produced low numbers of obsidian or chert points (Gonlin 1993:tables 4.22 and 4.23), which could have been used in conflict or hunting. The subject of the role of hunting is addressed below in a discussion of the Copán diet.
Another typical activity for Classic period Maya men and boys probably included construction, such as building of houses and monuments. Celts, hammerstones, and polishers are artifacts commonly attributed to these activities (Fung 1995; G. McCafferty 2001; Pohl 1991). Celts are found at all rural sites, but hammerstones and polishers are less common (Gonlin 1993:table 4.18). Just as grinding stones were common to each structure within a site, a similar pattern holds for supposedly male artifacts as well. Shown here is the distribution of hammerstone and polishing stone fragments (Figure 3.3, from Site 7D-6-2 in Río Amarillo). Hammerstones are multifunctional tools that serve equally well for cracking open nuts as for pounding construction elements into place. Likewise, polishers have more than one purpose and are often assigned a function in ceramic production. Freter (2004) hypothesized that it is possible Copán rural ceramic production was performed by community cooperatives, along the lines of the Chorti sian otot, per Wisdom (1940). Recently, Landau (2011) has tested several models of social interactions in the Copán Valley, of which the sian otot is one of several different possibilities.
The social organization of construction activities likely varied with the size of the task at hand. One can envision work groups of related or unrelated males organized by bureaucratic administrators for the tasks involved in monument building. Elliot Abrams (1994) has broken such tasks into quantifiable energetic costs and their associated organizational correlates. The family or lineage (or sian otot, per Fash 1983b and Freter 2004) likely played a part in construction activities in both rural and urban areas and for small and large projects. Organization ranged from various forms of familial recruitment (familial reciprocal, familial contractual, and community contractual) to custodial recruitment (including festive custodial and corvée) (Abrams 1994:97–101). This pattern is likely common to other areas of Mesoamerica as well (see Carmean, McAnany, and Sabloff 2011).
Figure 3.3. Spatial distribution of hammerstone/polisher fragments at Site 7D-6-2, Río Amarillo, Copán, Honduras; each structure is associated with this type of artifact
In what other productive activities did rural commoners engage? There is evidence of papermaking in the form of bark beaters. Bark beaters may have been used to produce cloth and paper (McAnany and Plank 2001:96), but bark beaters are not ubiquitous in rural commoner households at Copán and do not seem to be an essential tool in the domestic assemblage. From our sample of eight sites, three contained one bark beater each (Gonlin 1993:table 4.18). Bark beaters are found in Sepulturas (Diamanti 1991; Willey et al. 1994) but in small quantities. Whether paper or cloth was the end product of these implements, the ultimate consumer of these products is not known, although depictions of royalty in bloodletting rituals are shown to employ paper. Just as likely, paper could have been manufactured and used by the inhabitants themselves in their own rituals (Gonlin 2007).
The prolific amount of sherds recovered from Lowland Classic Maya sites provides a general picture of activities. Hendon (1987) organized the wide variety of vessel forms for Copán and, based on ethnographic analogy, assigned meaningful functional classifications to various forms. Her broad categories of “cooking/storage,” “serving/eating,” “ritual consumption/ceremonial,” and “other” are useful for understanding past activities in a general way, especially when the original locations of such activities no longer exist in the archaeological record. After sherds, obsidian prismatic blades are the most numerous artifacts at Copán sites, rural or urban, low or high status. Blades are truly multipurpose tools used for all sorts of cutting activities, from slicing open ears of corn (Neff 2002:36), to bloodletting (Marcus 1996:288), to weapons (Clark 2001:553). These tools are easily tailored to specific requirements but on archaeological sites are most often found broken, and most often use-wear analysis is not performed on the majority of broken blades. It would be illogical to assign this particular artifact to one gender or to one task.
AN ANALYSIS OF ARCHITECTURE
An examination of architecture within a site shows trends that may lend clues to gender relations. Houses, like other artifacts, have styles that can be identified. At some rural sites, it is evident that the same style was followed in house construction, at least for the intact remains of the substructures that were built of cobbles. At Site 7D-3-1, located in Río Amarillo (Figure 3.4), it is evident that the three buildings, one of which served primarily as a kitchen, were built in a similar style. Most likely they were probably not all built at the same time over the multiple decades of occupation of the site (Gonlin 1993:472). The length of occupation means that different generations inhabited the site and maintained the style of building. If men were the primary house builders, then it may be tenuously hypothesized that patrilocal residence was practiced at this homestead, with generations of men passing on their knowledge and style to their kinsmen. Alternatively, regardless of residence, family members contributed to maintaining the farmstead.
Figure 3.4. Plan map of Site 7D-3-1, Río Amarillo, Copán, Honduras; note the similarity in construction of each building (from Gonlin 1993:166; original by David Webster)
Lentz provides some guidance as to the species that may have been selected by ancient Copanecos for house building. In his analysis of charcoal remains, “[p]ine predominates throughout all phases in middens, cooking areas, hearths, activity surfaces, and in construction collapse. It may have been a preferred wood for building construction as well, since it was the most common charcoal in post molds and construction collapse” (Lentz 1991:280). Hardwoods were also used for firewood and construction, but not as often as pine. The remains of burnt clay, some of which is pole or grass impressed, also lend clues to construction (Gonlin 1993). We can glean from this information the types of firewood collected by the rural inhabitants.
Kitchens and hearths are considered to be the domain of women, yet as Graham (1991:474) has observed, neither has been the focus of intensive study in the Maya region. Identifying structure function is difficult because functions change over time, refuse may not be related to structures, and architecture may be multifunctional. Just as kitchens are difficult to identify archaeologically, hearths too provide a challenge to the excavator, a point addressed elsewhere (Gonlin 2004). Nevertheless, kitchen signatures may be identified if long-term use prevailed and if chemical analysis of soils has been undertaken (Manzanilla 1987; Manzanilla and Barba 1990). Both Diamanti (1991) and Hendon (1987, 1997) have identified kitchens at Sepulturas, and it has been possible to identify them at rural sites, too. For rural commoners, kitchens are almost always located to the magnetic north or east, not to the south or west, as one might expect from concepts of gender ideology derived from ethnographic observations (Palka 1999). Nevertheless, the identification of women’s space is important in understanding household organization, since it reveals the physical location of their activities.
Identification of the number of contemporaneous kitchens or hearths within a group may reveal more than the location of women’s space. According to Sabrina Chase (1991), who studied polygyny in a cross-cultural fashion with relation to architecture, it is common for each wife to be provided with her own hearth, if not her own living space. The existence of polygyny has not been given serious consideration for the commoners of Classic Maya society, but it is a pattern taken for granted among the royalty and elite. For example, it has long been proposed that polygyny existed, given the layout of architecture and the predominance of female burials at the elite Type 4 site in Sepulturas (Sanders 1989:96). With increasing research on commoners (Lohse and Valdez 2004), it may be possible to determine just how widespread this cultural practice was.
In terms of overall size, the variation within rural Type 1 sites warrants comment with relation to agricultural production. In this sample of residences, Type 1 structures may number from two to five around a plaza, and vary in substructure area from a low of around 13 square meters to a high of 66 square meters (Gonlin 1993:table 5.5). According to Netting (1993), larger households may signal bigger farms that are more productive. The larger Type 1 sites (11D-11-2 and 34C-4-2) are also located on better agricultural land that is well drained and relatively flat. A strong correlation between length of occupation and type of habitat was found by Paine and Freter (1996; Paine 1992) for the Copán Valley. The soils of the valley (Wingard 1992, 1996) have played a dynamic role in the evolution of the Copán kingdom and its households. In particular, the Copán pocket soils have supported relatively large populations since these ones are the most fertile in the region. As population densities increased during the Classic period, farmers moved to less desirable soils and cropped more frequently. Perhaps they grew root crops as well as maize, as suggested by the recent finds at Cerén, El Salvador (Dixon 2011). By the very end of the Late Classic, Copán farmers were probably working harder than their counterparts did during the beginning of this time period.
HUMAN REMAINS: ANALYSIS OF CONSUMPTION
Another approach to gender studies is through the analysis of human remains. Brian Hayden (1992) views the field of skeletal and mortuary studies as a fruitful approach to understand prehistoric gender. To accurately interpret skeletal or isotopic data and to perceive pan-Maya (Gerry and Krueger 1997) or pan-Mesoamerican trends, one must ideally have large sample sizes from single sites rather than few samples from many sites. If we examine the work carried out at Copán, where one of the largest and best samples of burials exists, there are a number of different approaches that researchers have taken. First, Stephen Whittington (1989, 1992, 1999) has studied many osteological characteristics, including dental caries and tooth loss. In one of his studies, the sample included forty-two low-status males and forty-five low-status females (Whittington 1999). He concludes that the difference in the presence of caries, 14.3 percent for males verses 26.0 percent for females, is statistically significant and represents different behavior (ibid.:158). The formation of caries is positively correlated with consumption of carbohydrates in the diet, but it is the frequency, rather than the amount consumed, that is important. In other words, women had more opportunities to eat probably maize-based foods throughout the day than did men. Perhaps women’s snacking behavior is the result of their activities centering on the residence and kitchen rather than farther afield.
David M. Reed (1998; Whittington and Reed 1997) has examined carbon and nitrogen isotopes on Copán skeletal data. Paraphrasing from Reed (personal communication, March 15, 2001), it is accurate to state that “[a]dult women and men had similar diets, but women’s diets appear more varied than men’s diets.” This conclusion accords well with the results from Whittington stated above. It is likely that women had access to a greater variety of foods in the preparation and cooking of them, and also perhaps in the gathering of them. Recall that paleoethnobotanical evidence indicates the use of wild species and perhaps the existence of kitchen gardens.
The role of protein in the Classic Maya diet has been approached in a variety of ways. If we rely on isotopic data, Reed (personal communication, March 15, 2001) states that another difference in men’s and women’s diets is that “women’s diets appear to be lower in protein relative to men’s.” If foods with protein were distributed unevenly according to gender, this pattern would not be unexpected. These conclusions indicate that perhaps an unequal rather than an egalitarian distribution of more prestigious foods was the norm, presuming that meat was a more prestigious food (White 2005).
The role of deer in the ancient Maya diet has been debated by archaeologists (Carr 1996). Ever since Bishop Landa was quoted as observing Maya women raising deer in sixteenth-century Yucatán (Tozzer 1941:127), archaeologists and others have wondered if Classic Maya women reared deer as well. In Ardren’s (2002b:78) analysis, she states that “[a] number of lines of evidence suggest that deer and women may have been linked in an economic and symbolic capacity.” These lines of evidence include portrayals of women and deer together on ceramic vessels, figurines of women and deer, and ethnographic and ethnohistoric observations of women raising deer.
Direct evidence for deer consumption in the form of bones is not found in rural Copán, although the remains of deer (Odocoileus virginianus) are quite frequent in the Harvard excavations in the Sepulturas urban zone (Pohl 1994). It is premature to comment on the few faunal remains from the rural contexts (six bones at 11D-11-2 and two bones from 34A-12-2) since the species have not been identified. The following comments are derived from Pohl’s (1994) analysis of faunal remains from a Type 1 site (9N-5 or CV-16), a Type 2 site (9M-27 or CV-20), and a Type 3 site (9M-18 or CV-43), all located within Copán’s urban core. She states that “the primary animal consumed at all social levels was white-tailed deer. This finding reflects a strong food preference among residents at Copán, who may have raised deer in their house compounds to meet demand for venison, as well as for deer sacrifices, dance headdresses, skins, and bone tool blanks” (Pohl 1994:459). Close behind deer bones in terms of numbers were the remains of dogs, which were used for food, sacrifice, and hunting (Pohl 1994:459). There was a clear association between the status and frequency of large animal remains, with sites of higher status containing higher numbers of bones and greater species diversity. Other species utilized at Copán were tapirs, peccaries, pacas, coatis, ocelots, opossum, birds, and turtles.
Given the paucity of non-human bones found in rural contexts, it is difficult to say with certainty that women raised deer within their compounds. If deer were penned up and habitually kept in one area, chemical analysis might show the existence of such an area, especially given the lack of faunal and architectural evidence. From random testing at some rural sites for pH levels, with values ranging from slightly acidic to alkaline, there is nothing to suggest from these limited tests that animals were penned and kept for extended periods of time on any of these farmsteads (Gonlin 1993). Ardren (2002b:80) realizes that there is no morphological evidence to suggest this practice, even though the possibility of women opportunistically raising deer in the household compound should be considered.
Isotopic evidence by Reed (1998:183) contradicts the prevalence of deer remains at elite and urban sites. According to his data, if deer were a major food source, then the C3 signature for humans should be much stronger than observed. Also, it is quite possible that the large numbers of floral and faunal remains at elite sites are indicative of their longer occupations and the use of particular species in ritual activities. Of fundamental note is that “[i]nferences from the isotopic evidence lead to the conclusion of equivalent diets across social strata and reliance on maize with little faunal supplements” (Reed 1998:183–184).
However, when other criteria are used to examine gender differences of consumption using paleopathological evidence, Storey (1998, 1999) concludes that status may be a more important determinant than gender, with commoners suffering more than elites. She examined human skeletal remains that fell into different socioeconomic categories for evidence of childhood stress. Dental enamel hypoplasias, markers of anemia (porotic hyperostosis/cribra orbitalia), and adult stature were used as measures of this stress. Her findings show that “there is little evidence that males were preferred and females neglected during childhood” (Storey 1998:146) in any class or status at Copán, perhaps because of the value of both males and females within the household and their complementary roles.
SUMMARY AND CONCLUSIONS
In summary, there are various lines of evidence to support gender complementarity among non-elite Copán Maya. The distribution of artifacts shows that households, to function properly, had to contain essential tools that consisted of typically male and female artifacts in addition to numerous multipurpose tools. However, the layouts of buildings within a site, especially the kitchens or kitchen areas, do not conform to expectations based on gender ideology. Some osteological data suggest gender complementarity while other data support a different relationship. McCafferty’s (2001) suggestion that probably both gender hierarchy and gender complementarity existed simultaneously and was adopted situationally may provide a realistic picture of the complicated past.
The criteria used in this exercise—artifact distributions, architectural function and style, and bioarchaeological data—are only a few of the ways in which gender can be explored at the household level. This exercise has shed light on who the ancient inhabitants were and their daily lifestyles during Late Classic times in the kingdom of Copán. The information presented here also illuminates the annual pattern of activities for Classic Maya farmers. Some tasks, such as cooking, gathering firewood, or hauling water, would have been performed on a daily basis, while other tasks, such as military conscription, construction, or textile production, may have occurred at certain times during the year.
Although it is not likely that production and consumption were identical from household to household, given the similarity in artifact assemblages from rural site to rural site and from the urban to rural areas of Copán, it may be logical to conclude that similar activities were repeated across the polity. Robin (2002:26) has observed similarly, “At Chan Nòohol [Belize], artifact inventories from each farmstead indicate that the same set of basic domestic and agricultural tasks was conducted at each farmstead.” With archaeological data from ancient households of Mesoamerica growing by the day, we can hope to see patterns of production and consumption in a broader comparative fashion, lending insight into the similarities and differences across time and space of the domestic economy.
Acknowledgments. My gratitude to the Instituto Hondureño de Antropología e Historia for allowing me to present material on Copán. I have benefited from the expert technical advice of Frances J. Peppard and from Benayah Israel’s interlibrary loan acumen, both of Bellevue College. The Social Science Division of BC has supported this work. Thanks to David Reed, Scott Zeleznik, John Douglass, and K. Viswanathan for commenting on earlier drafts of this chapter, as well as the two anonymous reviewers who helped to refine the arguments presented here.
NOTES
1. Shorter versions of this chapter were published in Yaxkin (Gonlin 2000) and presented at the 66th Annual Meeting for the Society for American Archaeology (Gonlin 2001).
2. Copán’s site hierarchy, detailed in Willey and Leventhal (1979), is given briefly as follows. The “Main Group” in the Copán pocket, where the monumental remains of palaces, temples, ballcourts, and grand plazas are located, is referred to as a Type 5 site. There is only one such site in the settlement hierarchy. The Type 4 sites are those that have several courtyard groupings of structures, some of which may have a height of ten meters or more; remains of sculpture may be present as well as corbelling stone and dressed stone. Type 4 sites are primarily located near the Main Group in the Copán pocket, the largest alluvial deposit in the valley. Type 3 sites may have six or more structures arranged around one to three courtyards, with a mound height of less than five meters, and dressed stone and vault stones are likely present. Type 2 sites typically have six to eight mounds with one to two courtyards. Mound height is less than five meters and construction may include dressed stone with cobbles. Type 1 sites are by far the most numerous and typically have three to five mounds around a single courtyard. Remains are less than one meter in height and cobble construction predominates, although dressed stone may be present. Freter (1988:75) added the category of “aggregate site” to connote Type 1 structures that have no formal courtyard arrangement and no presence of dressed stone. Single mound sites may be the remains of large tall temple buildings or small isolated fieldhuts. In addition, artifact scatters are labeled non-mound sites.
3. The sites or groups included in the present analysis are 11D-11-2 in El Jaral; 7D-6-2 and 7D-3-1 in Río Amarillo; 32B-16-1, 34A-12-1, 34A-12-2, and 34C-4-2 in the Sesesmil Valley; and 99A-18-2 in the Río Gila Valley.
4. As principal investigator, David Webster was awarded National Science Foundation grants (BNS 84-19922 and BNS 82-19421) to fund the survey, testing, and rural excavation projects, along with grants from Wenner-Gren and the Honduran government with co-investigator William T. Sanders. I am indebted to Webster and Sanders for making me an integral part of these projects.