chapter four

Site Structure and Community Organization

Blanton’s (1978) intensive survey of Monte Albán revealed patterns in the distribution of Xoo phase mounds that suggested that the ancient urban center may have been organized into fifteen barrios, each with its own local marketplace, temple, and civic center. Determining whether smaller and less populous district centers, such as Lambityeco, were similarly configured or manifested different patterns of community organization requires an investigation of the structure of the site (Bawden 1982).

Lambityeco is located 2 km west of the present-day market town of Tlacolula (see Fig. 2.3). The Pan-American Highway, built in the 1940s, passes through the northern part of the ancient settlement, and several dirt roads, used principally by oxcarts, also cross the site. A drainage, dry except during heavy rains, traverses Lambityeco in a north-to-south direction and terminates in the south in a salt marsh. Today Tlacolula Zapotecs call the drainage yahuish and the salt marsh zidi naana. Near the eastern edge of the site along the south side of the Pan-American Highway is a small mountainous spur called Cerro Yegüih (Fig. 4.1)

Lambityeco was first officially recorded as an archaeological site in 1953 by Ignacio Bernal, who was making a survey of valley sites. He recorded the site name as Yegüih, although informants gave John Paddock, who accompanied Bernal on his visit to the site, the name Lambityeco. Yegüih, according to Tlacolula Zapotecs, means cerrito (small hill or mountain) and is the name given to the mountainous spur at the eastern edge of the site (Fig. 4.2a). Its summit has remnants of a plaster floor with smoothed boulders that may represent an ancient shrine (Fig. 4.2b).¹

4.1. The archaeological site of Lambityeco

The etymology of Lambityeco, which Tlacolula Zapotecs insist is the name of the entire archaeological site and which was the name given to David Peterson (personal communication, 1979) by the salt workers who lived there in the 1940s, is not easy to decipher. Pityec means “mound” in Zapotec. Tlacolula Zapotecs translate Lambityeco as lugar de muchos mogotes, or “the place of many mounds.” Lam- or Lambi-, however, has no meaning for them in Zapotec. It does not mean “place” or “many.”

Paddock first pointed out in 1967 that the Lam- or Lambi- in Lambityeco is almost certainly a Zapotec corruption of the Spanish word alambique, which itself is derived from Arabic and means “distillery.” Zapotecs frequently shorten Spanish words and names: Eligio becomes Lig, Antonio Ton, and María Li (Parsons 1936:81). By shortening alambique, which in Zapotec is pronounced lambik, and adding it to pityec (mound) the result is Lambikpityec. Since bik and pit next to one another are phonologically redundant, this could explain how the name Lambityeco came into being.

4.2. Cerro Yegüih

Paddock believed that the meaning of Lambityeco was “distillery mounds.” He pointed out that his excavations on top of Mound 195 in 1961 had revealed a lot of ash as if a distillery or “still” had been located on top of it. At the time he was thinking of the distillation of alcoholic beverages, because Tlacolula is famous for its mezcal distilleries. The work of Peterson, however, later made it clear that the distilleries referred to are salt distillation areas and that the “mounds” are mounds of filtered earth, by-products of salt production. The name Lambityeco, then, is a combination of Spanish and Zapotec words and might best be translated as “salt distillery mounds.” The ancient name of the site is lost in antiquity.

THE SURFACE SURVEY

Beginning in 1970, personnel of the Institute of Oaxaca Studies conducted a surface survey that eventually resulted in the location and mapping of 206 mounds within the 1.17 km² area covered by Lambityeco. When excavations at the site began in 1961, a designation for the mound that was the target of the exploration was needed. Paddock chose the number 195 in the momentarily mistaken belief that the Pan-American Highway, which passes just north of the mound, was Route 195; it is Route 190. The numbering sequence at Lambityeco begins with Mound 10 and runs through Mound 223, with some numbers in between not used (Fig. 4.1; see Appendix 2 for data on mounds illustrated in Fig. 4.1).

The mounds were located with the aid of air photos and by clearing brush from some fallow fields to enhance visibility. Mound heights were measured to the nearest meter and mound bases were paced off to determine their approximate north-south and east-west dimensions in meters. Diagnostic artifacts were collected from the surface of each mound. In addition, all artifacts were collected from a 1 m by 1 m control square on each mound. Peterson (1976:82) also surveyed areas between mounds in a series of randomly located transects placed across the site. From these data, it was possible to determine that 147 of the 206 mounds at Lambityeco show evidence of Xoo phase ceramics on their surfaces and that Xoo phase Lambityeco covered an area of nearly sixty-four hectares (Fig. 4.3).²

XOO PHASE LAMBITYECO

Determining the nature of Xoo phase Lambityeco is no easy task for several reasons. First, there is no guarantee that all the mounds with Xoo phase sherds on their surfaces were actually constructions dating to that phase. At least two large mounds (14 and 21) in the extreme southeastern sector of the site were almost certainly Pe phase constructions atop which some later Xoo phase trash was dumped. On the other hand, some mounds not showing Xoo phase occupation on their surfaces may have been Xoo phase constructions. This became clear when a stratigraphic test pit was dug into Mound 57—a 10 m high mound whose surface manifested only Chila phase materials. It turned out that Mound 57 was a huge Xoo phase construction in its entirety from top to bottom. The Chila phase materials came from salt-boiling activities carried out on top of this abandoned Xoo phase structure and resulted in covering the surface of the mound with a veneer of later materials. Therefore, mounds thought to be Xoo phase may represent structures built in earlier times and some apparently non–Xoo phase mounds may be Xoo phase constructions. However, we assume that very few mounds have been misidentified as to phase.

4.3. Xoo phase Lambityeco

A second and more serious problem concerns the growth and development of Xoo phase Lambityeco through time. Even if all the mounds have been identified correctly as Xoo phase constructions, there is no way available at this time to determine from surface remains that all these structures were in use at the same time during the 200 years of the Xoo phase. Some may have been built and used in the earlier years of the Xoo phase and they remained as abandoned and ruined buildings later in the same phase. Others may have been built only during the later years of the Xoo phase and never existed during the earlier years of that phase. It is even possible that whole sections of the town were occupied early in the Xoo phase and abandoned later with a shift in population to other sectors. Such a population shift, of course, is offered only as a hypothetical example of what may have happened and our inability to detect it from surface surveys, which rely on a sequential segregation approach. Analysis of excavated features employing a sequential integration approach, on the other hand, could detect this type of spatial shift in population.

Assuming that no population shift occurred, assuming that all the mounds have been correctly identified as Xoo phase constructions, and assuming that all these constructions were in use at the same time during that phase, then, presumably the layout of Lambityeco as it existed on the eve of its abandonment is depicted on the map in Fig. 4.3. An examination of the distribution of the Xoo phase mounds throughout the site reveals that Lambityeco was a nucleated community in ca. 850 CE. The mounds are concentrated within the sixty-four hectares of the site and few, if any, isolated mounds indicative of homesteads occur dispersed in the surrounding countryside. This suggests that the Xoo phase inhabitants clustered together in Lambityeco, living in contiguous house plots within the community, and traveled on foot to any farmlands they may have worked in the surrounding countryside instead of living in isolated homesteads on their farmlands.

Most present-day Zapotec communities in the Valley of Oaxaca are also nucleated. People live on contiguous house plots within the community and travel from their houses in town to their farmlands in the surrounding countryside. Homesteads on farmlands are rare. An understanding of the nature of an ancient nucleated community like Lambityeco may best be achieved by beginning with an analysis of a present-day nucleated Zapotec community in the Tlacolula arm of the Valley of Oaxaca, Mitla.

MODEL OF A NUCLEATED ZAPOTEC COMMUNITY

To establish a reliable body of quantifiable data on a nucleated Zapotec community to serve as a model against which the archaeological data from Lambityeco can be compared, information was compiled and analyzed from Elsie Clews Parsons’s ethnography of Mitla. Around 1930, Parsons (1936:10) had Eligio Santiago, a native Mitla Zapotec, draw a map of the town (Fig. 4.4). Although it may be argued that Spanish colonial policies affected community layout in Mitla, there are several reasons to suggest that Mitla was not severely impacted by post-contact regulations.

4.4. Map of Mitla, ca. 1930 (after Parsons 1936:map II)

First, unlike Tlacolula, which was established early in the Colonial period when the Spaniards moved the people of Yagul to the new Spanish settlement of Tlacolula, Mitla remains in its Prehispanic location. The town is built over the Postclassic community (Robles 1986:18). Large mounds occur within the town and the Postclassic Mitla palaces as well (Robles 1986:18–21). Even the major church, San Pablo, built in 1590 CE, is in Patio C of one of the Mitla palaces (Grupo del Establecimiento Católico) and the other two patios of this palace (A and B) include the remains of long rooms with Prehispanic murals decorating the lintels above their doors (Robles 1986:18–21; Pohl 1999).

Second, Mitla did not even have a resident Spanish corregidor (mayor or magistrate) in 1580 CE. The corregidor, Alonso de Canseco, lived in Tlacolula. Although a Spanish priest, Cristóbal Ruíz Maldonado, is mentioned in the Relación de Tlacolula y Mitla (Canseco 1580:153) as having Mitla as part of his curado (parish), it does not state that he lived there; he apparently lived in Tlacolula and visited Mitla from time to time. It appears, then, that no Spanish administrative or religious authorities lived in Mitla as late as 1580 CE (Robles 1986:17).

Third, an examination of the map of Mitla does not show the town conforming to a Spanish grid pattern. It does appear that the Spaniards may have attempted to impose a grid pattern on the town by first establishing a plaza in 1575 CE with a church, which no longer exists, in the center of the town (Robles 1986:21). An east-west street passes along the south side of the plaza and continues in a straight line across the town in the east, but its western extent quickly breaks down and curves north to dead end. A north-south street passes the plaza along its west side but snakes across the town from north to south. It appears that this is the extent to which the Spaniards tried to establish a grid. Outside this “grid,” most other streets are short and narrow and do not follow a straight line but curve around and frequently dead-end, which appears to reflect a Prehispanic pattern. The Spanish impact on community layout in Mitla, then, appears minimal.

Santiago’s map can be seen by comparison with air photos taken in the 1960s to be remarkably accurate, although not to scale. The map correctly portrays the course of the Río Mitla (also known as the Río Grande) and its feeder streams and the locations of the Mitla ruins and Church of San Pablo; the Calvario chapel and nearby soccer field; the plaza with its municipal building, jail, school, and market; and stores, streets, and barrio boundaries. Most importantly, however, the map locates each “square block” of Mitla (or, more properly, “block” bounded by streets since few blocks in Mitla are square) and identifies each plot within each block and the structure or structures built on it.

From air photos to scale (taken in the 1960s), it is easy to identify the blocks drawn on Santiago’s map and determine the surface area of each block in square meters or hectares. Using this method, it was possible to determine that Mitla covered about fifty-five hectares in 1930. Parsons (1936:10) places Mitla’s population in that year at 2,500 persons, which would be about forty-five persons per hectare. Lambityeco covered about sixty-four hectares in ca. 850 CE and, if it was as densely populated as Mitla, would have had a population of 2,880 persons, which is remarkably close to Kowalewski’s independently derived estimate of 2,702 persons for Xoo phase Lambityeco (Kowalewski et al. 1989:287, table 9.6).

TOWN PLOTS

From an analysis of Santiago’s map, it was determined that 599 plots existed in Mitla in 1930. Plot boundaries were easily defined within each block because most plots either had walls built around them or, more commonly, were bounded by a fence of organ cactus into which an opening was left and fit with a cane gate. Of the 599 plots, nearly four-fifths (475, or 79 percent) had structures built on them, whereas slightly more than a fifth (124, or 21 percent) were vacant plots. It should be noted that vacant plots occurred in nearly every block and served as convenient areas for the disposal of trash.

To date, no evidence exists that house plots within Xoo phase Lambityeco were bounded by adobe walls or fences of organ cactus. However, in an intensive survey and sampling of 68 of the 147 Xoo phase mounds at Lambityeco, Peterson (1974a, 1974b, 1976, 1979) found physical evidence (floors, walls, wattle and daub) that fifty-five, or 81 percent, were structures and thirteen, or 19 percent, were refuse heaps (middens) containing no structures. If refuse heaps were deposited on vacant plots, then these mounds mark vacant plots, and Lambityeco, like Mitla, had about 20 percent of its plots vacant and for trash disposal and about 80 percent with structures built on them.

Peterson’s (1974a, 1974b, 1976, 1979) intensive survey and sampling of mounds involved an examination of the mound surface for evidence of a structure (floor, wall, or wattle-and-daub fragments), an analysis of artifacts from the surface of the mound and, in the absence of an exposed structural feature, subsurface sampling of the mound through full or partial excavation or by probing with a posthole digger to see if any structure occurred within. We conducted a comparison to determine if the mounds Peterson identified as structures could be distinguished from mounds containing refuse.

Refuse mounds could not be distinguished from structure-bearing mounds with regard to shape. They manifested neither more elongated, more circular, nor more irregular-shaped bases than structure-bearing mounds. Nor did refuse mounds have gentler or steeper slopes than structure-bearing mounds. Furthermore, no clear distinction obtained between refuse mounds and structure-bearing mounds with regard to the surface area covered by the mound base. Only by the fact that all mounds 3 m and higher represented structures, and not refuse, does any distinction occur (Table 4.1). Therefore, the identification of Xoo phase refuse mounds is necessarily limited to those identified by Peterson and does not represent all the extant refuse mounds or, by extension, all the probable vacant plots in Lambityeco at the time of its abandonment.

TABLE 4.1. A comparison of types of Xoo phase mounds by heights

Mitla had 599 plots within its fifty-five-hectare area, or about 11 plots per hectare. Only 475 plots, or about 9 plots per hectare, had structures built on them. Lambityeco, if it were similarly divided into plots, would be expected to have had around 700 plots with 20 percent (or 140 plots) vacant and 80 percent (or 560 plots) with structures built on them. Yet only 147 Xoo phase mounds, or about 2.3 per hectare, exist at Lambityeco. Furthermore, only 117 of these mounds, or 1.8 per hectare, may represent structures. This means that the Xoo phase community of Lambityeco either had one-fifth as many plots with structures per hectare as Mitla or that only one out of every five structures at Lambityeco left a mound or escaped obliteration over the past 1,200 years.

It is conceivable that 79 percent, or 443, of the possible 560 structure-bearing plots had all superficial traces of structures built upon them obliterated. In the nearly 1,200 years since its abandonment, only one mound need have been eradicated about every three years to account for the loss. Certainly, the construction of the Pan-American Highway through the northern part of the site resulted in the annihilation of all mounds in its way. Likewise, the dirt roads crossing the site contributed to the destruction of mounds, as can be seen by some partially dissected mounds along their paths (Fig. 4.1). Finally, plow cultivation of fields within the archaeological site for an undetermined number of years since the Conquest inevitably resulted in the elimination of many mounds, especially smaller ones of 1 m or less in height.

Clear evidence that many Xoo phase structures at Lambityeco are not represented by mounds comes from excavations. At least eighteen Xoo phase structures have been discovered by pure chance in areas where no mounds occurred. These moundless structures were encountered accidentally during excavations in areas specifically chosen for their apparent lack of structures (e.g., Fowler’s stratigraphic test pit project) and by farmers who chanced upon them in featureless fields. Each of these moundless structures is located on the map of Xoo phase Lambityeco. Their presence makes it apparent that many additional structures may be encountered in areas of the site that appear to be devoid of structures (Fig. 4.3; see Appendix 3 for data on moundless structures).

It is not possible to demonstrate that Lambityeco was as densely populated and as highly nucleated as Mitla on the basis of the extant Xoo phase mounds. Nevertheless, the weight of 1,200 years of abandonment combined with the known history of mound-destruction processes provides for the possibility that as many as 443 Xoo phase mounds may have been destroyed at the rate of one about every three years. Furthermore, the purely accidental discovery of eighteen moundless Xoo phase structures conclusively demonstrates that many Xoo phase structures are no longer visible on the surface of the site. Finally, Peterson’s sample of 46 percent (68 of 147) of the extant Xoo phase mounds demonstrates that Lambityeco, like twentieth-century Mitla, had about 20 percent of its plots vacant and 80 percent with structures built on them. This strengthens the possibility that Lambityeco was at least as densely populated and as highly nucleated as Mitla.

RESIDENTIAL PLOTS

Of the 475 plots with structures in Mitla in 1930, about 94 percent, or 448, had houses built on them, while only 6 percent, or 27, had nonresidential structures. Nearly two-thirds of the 448 residential plots contained a single house usually occupied by a nuclear family (Parsons 1936:66), whereas slightly more than one-third contained two—or rarely more—houses occupied by related nuclear families forming a joint family household (Table 4.2). For example, Parsons (1936:66) notes that four houses in one residential plot were occupied by a married couple and their three married sons, each of whom lived in one of the houses with his wife and children.

Mitla residential plots range in size from 340 to 2,800 m² with a mean of 915 m² and a mode of 800 m². The plots are typically rectangular, about twice as long as wide, usually measuring around 20 m wide by 40 m long. As noted above, nearly all plots are enclosed by blind fences, generally a fence of organ cactus with an opening fitted with a cane gate. Because of the fences, the interiors of the residential compounds remain hidden from the outside, providing a great deal of privacy.

The components of Mitla residential plots provide a model against which the archaeological remains from Lambityeco can be compared. The main components of the fenced plot include the house or living quarters, kitchen, courtyard, well, and toilet area (latrines or outhouses are not generally used; instead, a designated area of the plot is used as an open toilet). Additionally, some residential plots have a sweatbath structure. Unlike certain other areas of Mexico, however, residential plots do not include granaries or special structures to store corn. Instead, corn and beans are stored on the floor in a corner of the house or living quarters (Fig. 4.5).

TABLE 4.2. Number of houses per residential plot in Mitla

The house or living quarters represents the largest structure built on the plot, although it occupies only a small section of the plot, which is dominated by a large open-space courtyard. The house consists of a single rectangular, generally windowless room about 4 m wide and 10 m long with a single doorway at the center of its long side. The walls are usually built of adobe blocks set in mud mortar atop a stone foundation and elevated to a height of about 3 m. The roof is generally made of thatching over a cane surface supported by pine roof poles. The living quarters have earthen, cement, or flagstone floors and invariably have an altar at one end with images or pictures of saints upon it (Parsons 1936:27–28). At the beginning of November on All Souls’ Day (Día de los Muertos), the altar is filled with flowers, fruits, and fancy loaves of bread (pan de muerto) and becomes a focal point of the single-room house.³

The arrangement of more than one house in a residential plot varies. However, about 77 percent of the 144 plots with two houses on them have the houses built at right angles to one another, whereas 14 percent face one another across the common courtyard, and only 9 percent are built next to one another along the same side of the courtyard. Although three houses occur infrequently on a single plot, when they do occur they are most frequently arranged around three sides of the common courtyard.

Each of the houses (or living quarters) built in a residential plot has associated with it a kitchen, which is built as a separate structure sometimes, but not always, at a right angle to the living quarters. As Kearney (1972:13) points out with regard to Ixtepeji, a Zapotec town in the Sierra Juárez (see Fig. 10.1 for location), kitchens are usually attached to a wall of the living quarters. No doorway connects the kitchen directly to the house. The kitchen has its own doorway and must be entered from the courtyard. The kitchen is a small structure, usually 2 m wide and 3 m long. It may be made of adobe or wattle and daub (cane plastered with mud) and has a thatched roof and an earthen, cement, or flagstone floor.

4.5. Model of a residential plot in Mitla

When adobe houses—living quarters and kitchen—like those in Mitla are abandoned, they easily deteriorate into mounds like many of the mounds at Lambityeco. First, the roof collapses over the floor as the roof poles rot. With the roof gone, the adobe walls are exposed to rain and begin disintegrating into dirt—eventually collapsing over the floor and stone foundation of the house. The result is a low mound of earth covering the floor and foundation. When archaeologists excavate the mounds, generally only the floor and foundation stones of the house remain. Excavations in mounds at Lambityeco have revealed Xoo phase residences that may be compared to the model of Zapotec residences in Mitla.

XOO PHASE RESIDENCES

At least nine separate mounds containing at least twenty-six houses corresponding to the Xoo phase have been either fully or partially excavated at Lambityeco. Elite residences (thirteen in four mounds) and the ordinary houses of commoners (thirteen in five mounds) were excavated.⁴ Several excavated elite residences were enough preserved to provide complete floor plans (as will be seen in later chapters), but the floor plans of commoner houses were difficult to determine because few had any of the rooms around their patios preserved or fully excavated enough to provide a complete floor plan. Because the basic components of Xoo phase residences appear to be the same for elite and commoner houses—differing only in size and degree of elaboration—a model of a Xoo phase residence serves to illustrate these basic components (Fig. 4.6).

4.6. Model of Xoo phase residence

The main components of a Xoo phase house included four rooms arranged around the sides of a central patio with four smaller rooms in the corners and a household tomb. The rooms were all built with the entrances opening onto the patio. The latter was generally a small area, about 4 m on a side, and was usually paved with a smooth layer of white plaster over a cobblestone base. The patio was usually encircled by a walkway about 1 m wide and raised about 20 cm above it. Although nearly all houses excavated had a patio paved with plaster, the patio of one house was paved with crushed ignimbrite gravel, a white chalk-like substance that occurs naturally at Lambityeco, and another had only a compact earthen surface. Often a drain in the form of a stone-lined conduit (or in elite houses, a tubular ceramic drainpipe) was located at one corner of the patio to facilitate the runoff of rainwater.

The rooms built along the sides of the patio were between 2 m and 3 m wide and about 4 m long, with walls on three sides. The walls were built of adobe blocks set in mud mortar atop a stone foundation. The open side of the rooms, facing the patio, may have had curtains or other perishable coverings that functioned to close them off during cool weather or for privacy, but as yet no evidence for such coverings, such as curtain holders, has been discovered in excavations. The floors of the side rooms were elevated a step (20 cm–30 cm) above the level of the walkway around the patio floor and were paved with a smooth layer of white plaster over a cobblestone base. Often, in elite houses a thick-walled ceramic pan was set flush in the center of most of the rooms.

Corner rooms are small, usually about 2 m by 3 m, with walls on four sides and a single doorway about 1 m wide. Two of their walls were shared with the side rooms adjacent to them. The other two walls were generally made of adobe blocks set in mud mortar atop a stone foundation, although in some cases these walls may have been made of cane with or without mud plaster covering them. Most floors of corner rooms were elevated a step above the walkway around the patio and paved with smooth white plaster over a cobblestone base. Some floors, however, were flush with the walkway and some were compact earthen surfaces. In elite houses, at least, a ceramic pan was set in the floors of most of the corner rooms. Evidence from excavations indicates that the roof of at least one corner room was thatched and supported by pine roof poles. One of the four corner rooms was an entryway to the house and had an additional doorway opening to the outside. This corner room served as an entry vestibule to the house and did not have a ceramic pan set in its floor.

We are unaware of any material evidence for the types of activities conducted in these rooms. Generally, plaster room floors were swept clean of any trash, which was then deposited in middens. No samples of plaster have ever been taken from the floors of Xoo phase Zapotec houses to subject them to chemical analysis. The practice of sweeping patios and rooms clean and the lack of chemical analysis of floors make it difficult to interpret specific uses of these domestic spaces.

We assume that the open patios, aside from allowing for constant traffic between the surrounding rooms, served multiple purposes, especially when weather conditions permitted people to carry out different chores and tasks. We also posit, based on the larger size of side rooms and on the assumption that the ceramic pans embedded in the floors of some of them (in the case of elite houses) most likely functioned primarily as hearths to take the chill off the rooms in cold weather, that most of these enclosures were living quarters where families rested, visited, slept, and stored their possessions. The fourth side room had no hearth and was the ancestral shrine with the household tomb beneath it.

Setting aside the corner room that served as entry vestibule into the house, each of the other three corner rooms seem to be associated with one or another of the three side rooms adjacent to it. Yet, we disagree on their function. Lind postulates that the hearths in these rooms were actually used for cooking. Since the ceramic pan hearths (in both corner rooms and living quarters) contained a fine powdery gray ash, he suggests that they had held glowing chunks of charcoal, which produce little smoke or soot that might have blackened the room floors or walls. Given their diameter (about 30 cm) and depth (10 cm), they could easily accommodate patojos (shoe-shaped cooking vessels) (see Fig. 7.20a). The only such vessel thus far documented at Lambityeco measures 11 cm in height and is 21 cm long. Then, based on ethnographic analogy from the Mixteca of Oaxaca, where a nuclear family in a joint family household maintains its own living quarters and kitchen within the residential plot (Romney and Romney 1966:12), he interprets the corner rooms as kitchens associated with each of the living quarters. Since one of the corner rooms in the houses was seemingly used as a vestibule, the presence of three additional corner rooms would imply that the family groups could consist of up to 3 nuclear families, and that food preparation and consumption was not a shared social practice.

Urcid disagrees and assumes that cooking was done in nonmasonry structures built outside the houses but within the residential plot and that although the ceramic pans could have been used at times to heat meals prepared in the outside kitchens, their primary function was to warm the rooms. His assumption that cooking was done outside the houses rests on the seeming lack of blackened areas or traces of soot on the floors, especially around the hearths, or on the lower portions of the walls. If corner rooms had thatched roofs, such a feature may have solved the problem of ventilation to avoid the accumulation of smoke generated by cooking. But only one room so far discovered at Lambityeco, whose roof had burned, has yielded material evidence that it was roofed with thatch, and this corner room has other unique features that sets it apart (Urcid’s interpretation of this room is presented in Chapter 6). As to patojos, their shape suggests to him that at least some cooking activities required some type of semi-enclosed hearths, so that the elongation of the pot could be inserted in a small area where the heat was concentrated, implying that such cooking vessels could not have been used in the ceramic pan hearths embedded in the floors. Thus, the use of the side and corner rooms would have been more flexible, depending on a household’s particular cycle of growth and specific needs (e.g., that of joint families or, in the case of the elite, polygynous arrangements). Activities within these rooms might have included sharing meals; manufacturing certain items such as textiles; processing foodstuffs before long-term storage or prior to cooking; carrying out other daily activities if weather conditions like extreme heat, extreme cold, or rain and wind prevented doing them in open spaces like the patios or areas in the residential plots; hosting close kin visitors; and at times accommodating for gender-specific and/or age-specific seclusions in rites of passage or sickness.

We both concur, however, that because of the contingencies of life, the household groups may have included at times widows or widowers, aged relatives, and, in the case of elite houses, polygynous arrangements. This is supported by evidence of adult burials of both sexes interred beneath house floors or beyond the confines of the house and not in the household tomb where the married couple who headed the household were eventually buried.

Because the household tomb is always located beneath one of the side rooms, most often the one on the east side of the patio, we believe such an enclosure to have been used as an ancestral shrine. Like the altars in present-day Zapotec houses, the shrine room of Xoo phase residences was probably the locus of household religious activities and may have housed images or representations of Zapotec deities. Also, just as the altar in present-day Zapotec houses becomes a focal point of the household on the Day of the Dead in honor of deceased family members, the shrine room was certainly the focal point for ancient Zapotec rituals aimed at evoking and invoking the ancestors buried in the household tomb centrally placed beneath its floor.

Tombs, like the ancestral shrine rooms above them, vary from elaborate structures in elite houses to simple structures in the houses of commoners. All tombs, however, had entrances sealed with stones and most contained multiple burials. From analyses of the skeletal remains in Xoo phase tombs from both Lambityeco and Monte Albán, it is known that virtually all individuals buried in those tombs were adults of both sexes. Furthermore, the disturbed nature of most of the skeletal remains indicates that burials were placed in tombs at different times and that earlier burials were disturbed either to remove certain bones or to make room for later burials. As commented in Chapter 3, it seems clear that the tombs beneath shrine rooms in Xoo phase houses were the burial places for married couples who headed the household and controlled the household estate during their lives and who were especially venerated following their deaths as ancestors and also the ones who bequeathed the household estate to their legitimate heirs. Other members of the household—infants, children, adolescents, and adults who did not head the household—were buried under the floors of rooms and patios or directly beyond the confines of the house structure in the residential plot.

4.7. The center of Xoo phase Lambityeco

The presence of some burials, salt-boiling activities, and sweatbaths near but beyond the confines of houses indicates that residences were surrounded by residential plots. As Winter (1972, 1976) has pointed out, residential plots can be identified archaeologically by household clusters in which various features reflecting household activities occur in areas outside and adjacent to the confines of the house. Cira Martínez (personal communication, 2008) reports that recent salvage excavations at Lambityeco in conjunction with a widening of the Pan-American Highway have shed light on the probable limits of the large residential plot within which the elite houses of Mound 195 and Mound 190, the sweatbath north of Mound 195, and the houses of commoners (the houses of Tombs 3 and 4) were built (Fig. 4.7). What remains unknown, however, is the range of variation in Xoo phase residential plots and whether these, like Mitla residential plots, were enclosed behind blind fences that hid the house from outside view.

Like Mitla in 1930, Xoo phase Lambityeco had one or more living quarters around each patio that probably housed nuclear or joint families. There is no reason, therefore, to assume that each Xoo phase “house mound” represents a single house occupied by a nuclear family. For the time being, it is not possible to determine if Xoo phase Lambityeco, like Mitla in 1930, had about two-thirds of its residential plots occupied by nuclear families residing in single houses and one-third occupied by joint family households residing in two or more houses built in a single residential plot. However, from the excavated examples at hand, it seems that joint family households were much more common at Xoo phase Lambityeco than in Mitla in the 1930s.

HOUSEHOLD COMPOSITION

Although Parsons (1936) presents indirect data on nuclear and joint family households in Mitla via the number of houses in a residential plot, neither hers nor other ethnographic studies of Valley of Oaxaca communities have provided detailed statistics on household composition. In the absence of published evidence from the Valley of Oaxaca, then, complementary data from Carrasco’s (1964:185) analysis of a very detailed sixteenth-century (1530–1540 CE) census of households in the Tlacatecpan barrio of Tepoztlán, Morelos, and the Romneys’ ethnographic study of the Santo Domingo barrio of Juxtlahuaca in the Mixteca Baja of Oaxaca must suffice.

Carrasco (1964:190–191) defined four types of households in the Tlacatecpan barrio of Tepoztlán: (1) nonfamily households, composed of unrelated persons; (2) consanguineal households, composed of persons connected by purely consanguineal ties; (3) nuclear family households, composed of a married couple and usually including their children and/or other unmarried relatives; and (4) joint family households, composed of two or more married couples, usually two or more married brothers or a father and his married son or sons.

Nuclear and joint family households constituted 97 percent (531) of the 549 households in the Tlacatecpan barrio (Carrasco 1964:191). This suggests that Xoo phase Lambityeco may have had at least 97 percent of its households occupied by either nuclear or joint families instead of the rather unstable nonfamily or consanguineal arrangements. Of the 201 households subject to the cacique of Tepoztlán, nearly 72 percent lived in joint family households whereas only 26 percent lived in nuclear family households. Comparative ethnographic data from the Santo Domingo barrio of Juxtlahuaca are quite compatible with Carrasco’s data and indicate that joint family households account for about 77 percent of the total and nuclear family households 23 percent (Romney and Romney 1966:xxi, chart IV; 43). These figures are probably fairly compatible with Xoo phase Lambityeco based on samples of excavated houses.

Within these two dominant household types—joint and nuclear family households—household size was about 4.4 persons (296 households with 1,298 persons) per nuclear family household and about 7.3 persons (235 households with 1,727 persons) per joint family household in Tepoztlán. However, other analyses led Carrasco (1964:209) to point out that household size in the Tlacatecpan barrio was well below average for Central Mexico. In the Santo Domingo barrio of Juxtlahuaca, average nuclear family size was 6.5 persons, whereas joint family household size averaged 12.6 persons (Romney and Romney 1966:xx–xxi, charts III–IV). These data at least provide a possible range of the number of persons per household for the households of commoners at Xoo phase Lambityeco.

The cacique’s household was the largest in Tepoztlán and included twenty-three people (Carrasco 1964:189–190). Among these were the cacique and his wife, his thirteen children, and six female slaves who were probably concubines given the fact that his eldest child was only six years old (Carrasco 1964:205). Again, these data suggest that Zapotec elite households of nobles (xoana) and rulers (coqui) probably contained many more persons than households of commoners at Xoo phase Lambityeco.

From his analysis, Carrasco defined a cycle whereby nuclear family households became joint family households with the passage of time. A nuclear family household composed of a married couple and their children became a joint family household of the “head-with-sons” type when the sons married and remained in their father’s compound. With the death of the father, the eldest son became the household head and together with his younger married brothers formed a joint family household of the “head-with-brothers” type. “We see then the main cycle in family development from nuclear toward joint families of the head-with-sons and later head-with-brothers types, only occasionally growing to larger types” (Carrasco 1964:208).

A similar situation obtains in the Santo Domingo barrio of Juxtlahuaca and Mitla. A young married couple will build their house within the husband’s father’s compound (Parsons 1936:66; Romney and Romney 1966:42–43). In Juxtlahuaca, “the major decision as to whether or not a newly married couple will reside with the husband’s family is the availability of space in the prospective compound” (Romney and Romney 1966:43). If space is not available to build a house, the couple will seek a place as near as possible to the family compound. Unlike Tepoztlán, the present-day Juxtlahuaca household groups have no institutionalized household head. “Rather, the adult men who are still in the prime of life all maintain more or less equal status” (Romney and Romney 1966:50). Even when they live in joint family households in Juxtlahuaca, each nuclear family forming the joint family household within the residential compound maintains its own kitchen for cooking and its own living quarters for sleeping, storage, and visiting (Romney and Romney 1966:12).

Carrasco’s cycle of household development is bolstered by ethnographic data from Juxtlahuaca and Mitla. However, with one possible exception, none of the house remains from Lambityeco can be shown to conform to Carrasco’s cycle of household development. A sequential integration approach to the sequence of houses of commoners within Mound 92 at Lambityeco appears to show an initial nuclear family household, indicated by two side rooms (one a living quarters and the other the ancestral shrine with a household tomb beneath it) with a corner room (which Lind considers a kitchen) between them flanking two sides of the patio. This is followed by a joint family household composed of two nuclear families, indicated by three side rooms (the ancestral shrine room and two living quarters facing one another across the patio) and two corner rooms (which Lind considers kitchens) between them flanking three sides of the patio. The sequence of houses ends with a joint family household with three nuclear families, indicated by four side rooms (the ancestral shrine room and three living quarters) and four corner rooms (one entry vestibule and three that Lind considers kitchens) surrounding the patio (Lind 2009). Nevertheless, the excavation data on which this analysis was based are too limited to fully substantiate it.

HOUSES OF NOBLES AND COMMONERS

Determining what proportion of the ancient Xoo phase population of Lambityeco was composed of nobles and what proportion was represented by commoners requires examining the mounds. Defining which mounds may have represented the houses of nobles is extremely difficult. Setting aside the two very large mounds, Mound 155 at 12 m tall and Mound 57 at 10 m tall—which were most likely temples—leaves 145 mounds, most of which were probably houses of commoners. Apart from the civic residential complex (PPA) in Mound 195, two other mounds representing houses of nobles have been excavated. Mound 185, a 4 m high mound representing several superimposed elite houses, was partially excavated by Winter and colleagues (1979), and Mound 190, a 3 m high mound also representing several superimposed elite houses, was completely excavated under the direction of John Paddock.

Based on the above known examples, it might be assumed that mounds in the 3–4 m height range were houses of nobles. Because there are nine mounds in the 3–4 m height range at ancient Xoo phase Lambityeco, 9 of the 145 mounds may be the houses of nobles. This works out to be about 6 percent of the mounds, and it could be assumed that this percentage of the population of ancient Lambityeco was composed of nobles, or about 173 adults and children based on a population of 2,880 persons. Also, in this case, mounds 5 m tall or taller would be temples, making for nine temples at Lambityeco, and most mounds less than 3 m high would include houses of commoners and some (19 percent) refuse mounds. However, other factors complicate this reasoning.

One 5 m high mound (Mound 125) may be an elite house. In 1967, Lind examined a looter’s pit on top of it. The pit had exposed part of a patio floor with a raised walkway along its east side like those known from elite houses. This suggests, but does not prove, that Mound 125 might have had the house of a noble atop it. If houses of nobles included mounds in the 3–5 m height range, then 14 of the 145 mounds at Lambityeco were houses of nobles and accounted for about 9.7 percent of the population, or about 276 adults and children based on a population of 2,880. In this case, only three mounds 6 m tall or taller would have been temples, but the number of houses of commoners and refuse mounds would remain unchanged.

Another complicating factor was a 1 m high mound (Mound 148) that was partially excavated and possibly could be an elite house. Not enough of the mound was excavated to prove that it represented the house of a noble. No house floors or patios were exposed. However, a 30 cm high stone talud (sloping wall) was partly exposed and may have formed the wall of a low platform on which the house was built (Urcid 1983:117). Because an elite house in Mound 195 was built atop a similar low platform with a talud, it appears possible that Mound 148 was the house of a noble. Mound 148 had a basal area of 675 m². An examination of the basal areas of mounds known to be houses of commoners revealed that they ranged between 80 m² and 300 m², whereas known elite houses ranged between 650 m² and 1850 m².

With this in mind, 1 m and 2 m high mounds were examined to determine how many of them had basal areas 650 m² or greater and, therefore, might have been houses of nobles. A total of nineteen mounds 1 m high and five mounds 2 m high had basal areas 650 m² or greater, making a total of twenty-four mounds that may have been elite houses. Added to the fourteen mounds 3–5 m tall, this totals thirty-eight houses of nobles representing 26 percent of the 145 mounds and making nobles account for about 748 adults and children based on a population of 2,880 persons for ancient Xoo phase Lambityeco. In this case, the number of possible temples remains at three, but the number of houses of commoners decreases.

From the data at hand, it appears that anywhere from 6 to 26 percent of the population of ancient Xoo phase Lambityeco was composed of nobles. This is certainly an unacceptable range, although it should be kept in mind that it is based on the extant Xoo phase mounds and that many mounds were probably destroyed over the past 1,200 years. If Lambityeco were as densely populated as Mitla in the 1930s, it would be expected to have as many as 560 Xoo phase mounds with structures. This means that as many as 443 mounds may have been destroyed and most, if not all, of these mounds were probably 1 m tall or less and represented houses of commoners. It is a certainty that mounds corresponding to the houses of nobles are disproportionately represented relative to mounds corresponding to the houses of commoners because the latter are small and more easily subject to destruction whereas the former are much larger and less easily eradicated.

COMMUNITY ORGANIZATION AT LAMBITYECO

Insofar as it is possible to determine community organization from site structure, it is evident that Lambityeco was a nucleated community with a population of nearly 3,000 persons on the eve of its abandonment at the end of the Xoo phase. Like the plaza in Mitla in the 1930s and the Main Plaza at Xoo phase Monte Albán, the community of Lambityeco had a center, the North Marketplace—a plaza measuring 50 m east-west by 65 m north-south. A major thoroughfare, the West Avenue about 25 m wide and nearly 135 m long, entered the community from the west and led directly to this plaza (Fig. 4.7).

The community of Lambityeco was organized around this plaza. Its principal government building, which housed the residence of the ruler of the Lambityeco district, was a PPA (Mound 195) located in a large residential compound on the northeast side of the plaza. Its principal temple, Mound 155, the tallest structure within Lambityeco, formed a TPA located along the south side of the plaza, and a probable 5 m high temple (or possibly elite house, Mound 175) occurred at its extreme northwestern corner.

Most of the mounds at Lambityeco in the areas surrounding and beyond the plaza are probably the remains of houses. These houses are not bounded by streets, as in 1930 Mitla. Prehispanic societies had neither wheeled vehicles nor beasts of burden and therefore did not require streets. Instead, footpaths probably wended their way throughout the community. This, however, does not mean that houses were haphazardly scattered about Lambityeco. All excavated structures at Lambityeco were oriented about 16° 30’ east of north, or what came to be called “construction north.” From his research at Lambityeco, Peterson (1991) concluded that this alignment was determined based on the point of sunrise during the winter solstice and the point of sunset during the summer solstice (Fig. 4.8). Because all known structures at Lambityeco follow this alignment, even the houses of commoners, it is clear that community planning went into the building of structures at Lambityeco.

4.8. Sunrise over Mound 195 during the winter solstice, December 22, 1987 (photo courtesy of David Peterson)

Unlike Monte Albán, there is no evidence at Lambityeco for “barrios,” each with its marketplace, civic center, and temple. Lambityeco was too small to have separate marketplaces for any barrios it might have had. Furthermore, mounds that might represent small temples are not distributed as would be expected if they were barrio temples. Likewise, mounds that might represent the elite houses of nobles do not show any pattern of distribution that would indicate that they were associated with particular barrios. This, however, does not mean that barrios did not exist at Lambityeco. Mitla had six barrios in 1930, but archaeologists would be hard put to identify them from any material remains (Fig. 4.4).

Most of Lambityeco’s Xoo phase population of nearly 3,000 persons were commoners. Probably over half of them were involved in nonagricultural pursuits. As much as one quarter of the population carried out salt-production activities. Another 15 percent may have been potters and an additional 15 percent may have produced textiles, but there is no solid archaeological evidence for these latter two percentages. A number of commoners may have cultivated the fields around Lambityeco, perhaps some planting corn and some planting maguey and beans, which require less rainfall than corn.

Nobles may have constituted at least 6 percent of Lambityeco’s population. There is no evidence they lived together in a single barrio at Lambityeco, but in some cases, their houses appear near one another, suggesting that closely related nobles may have lived next to one another (Lind and Urcid 1983). The ethnohistoric model, presented in Chapter 1, suggests some of the roles these nobles might have occupied at Lambityeco.

The ruler of Lambityeco (the coqui) probably occupied the elite house atop Mound 195. Also, the elite house seemingly occupied by a priest of Cociyo (the Zapotec deity of lightning, thunder, and rain) was located in Mound 190, directly south of Mound 195 (Lind and Urcid 1983) within the same residential compound. Other nobles (xoana) residing in Lambityeco probably included additional priests, political advisers, captains who led military units, and administrators, some of whom oversaw each of the eight villages within the Lambityeco district. Whether the latter resided in Lambityeco or in the village they administered is uncertain.

NOTES

1. Bernal customarily named previously unnamed sites after a prominent nearby topographical feature, which is why he named Lambityeco Yegüih. Sr. Juan López, a Tlacolula Zapotec, provided Lind with the translation of Yegüih. He also related that Yegüih is enchanted (encantado). At times a very tiny rooster emerges from the solid rock and, crowing, encircles the base of the hill. The trip around the hill takes the little rooster five minutes to complete. At other times, a giant of a man dressed as a charro, emerges from the solid rock, encircles the hill, and disappears like the rooster. Anyone who has spent time in villages in Oaxaca knows that every nearby cerro is enchanted. The rooster is probably the zanate de oro (bird of gold) and the charro is undoubtedly the charro diabólico (diabolical horseman), both of which are common in Zapotec folklore (Cruz 1946).

2. In two earlier publications (Lind and Urcid 1983, 1990) the total number of mounds was given as 197 and Xoo phase mounds as 141. In a more recent publication, Lind (2001) cited 213 total mounds and 169 Xoo phase mounds. This inflated number resulted from including moundless structures excavated at Lambityeco as mounds (Appendix 3). After a careful and complete revision, the above numbers of 206 total mounds and 147 Xoo phase mounds stands as the most accurate count (Appendix 2).

3. Lucha Sosa, a Mitla Zapotec, informed Lind in 1982 that deceased members of the household come back on the Day of the Dead to partake of the food and drink, and therefore their favorite foods and drink are placed on the altar for them and sometimes their favorite clothing as well.

4. In a recent salvage excavation along the Pan-American Highway, at least five other houses of commoners, two of them moundless, were excavated at Lambityeco to the east of Mound 195 (Cira Martínez, personal communication, 2008).