ELEVEN

Fluctuating Community Organization

Formation and Dissolution of Multifamily Corporate
Groups at La Joya, Veracruz, Mexico

VALERIE J. McCORMACK

INTRODUCTION

In many Formative Mesoamerican communities multifamily corporate groups emerged following the transition to sedentism (Flannery 2002). When surface remains of artifacts and architecture are detectable, aspects of social organization can be inferred. However, in some regions, recovering evidence of household and community organization is difficult because of the perishable nature of houses and the fact that archaeological deposits are deeply buried. For this reason we know little about household or community organization of Formative southern Gulf Coast societies. This chapter investigates community organization at the Formative village of La Joya, Veracruz, Mexico, by determining the conditions that produced nuclear family households and those that produced multi-family corporate groups. Several changes in household organization are detected for La Joya and these changes are placed within the regional environmental and social landscape. While this community-level focus is unable to reveal fine-scaled analysis at the level of individual households, the community approach reveals significant changes in village organization over time. The intrasite spatial patterns at La Joya illustrate that community organization fluctuated between independent nuclear family households and multifamily corporate groups. Comparisons of the activities based in the multifamily corporate groups reveal that organization differs through time, but the presence of multifamily corporate groups is strongly correlated with periods of land scarcity.

La Joya is a twenty-five-hectare village situated along the Río Catemaco, a major drainage and transportation route of the Sierra de los Tuxtlas (Figure 11.1), which is a 4,500-square-kilometer active volcanic mountain range located along Mexico’s southern Gulf Coast. While the southern Gulf Coast is best known for the emergence of Olmec societies by 1200 BC (uncalibrated), social organization varies widely throughout the region during the Formative period (Stark and Arnold 1997). For example, societies in the Tuxtlas engaged in similar ideological and ceremonial traditions, shared ceramic technologies, and participated in similar long-distance exchange routes as the San Lorenzo Olmec, but a chiefly society did not emerge until 400 BC in the Río Catemaco drainage.

Over the past decade, research of Formative societies in the Tuxtlas has helped establish broad cultural patterns and chronological sequence for the region (Table 11.1). The earliest evidence for human occupation derives from pollen data, which indicate that around 2250 BC, populations were clearing forests and growing corn in the Tuxtlas (Goman 1992). However, no sites have been identified for this time, suggesting populations moved frequently and did not occupy any single location for an extended length of time. By 1300 BC, hamlets and villages occupied the best agricultural lands within the Tuxtlas, a pattern that continues throughout prehispanic times (Santley and Arnold 1996; Santley, Arnold, and Barrett 1997). La Joya was first settled around 1300 BC and occupation continued through AD 100, when Cerro Puntiagudo erupted and blanketed the region with a layer of thick volcanic ash (Arnold et al. 1996; Arnold and McCormack 2002; Reinhardt 1991). This 1,400-year occupation from the Early Formative through the Terminal Formative makes La Joya an ideal location for studying changes in community organization as social complexity increases.

HOUSEHOLD ORGANIZATION AND THE MULTIFAMILY CORPORATE GROUP

The internal organization of communities reflects the social, economic, and political relationships and boundaries defined and maintained by individuals and households. Household organization varies along a continuum from independent nuclear family households to multifamily corporate groups. Corporate groups consist of several families that are linked together through lineage, kin, or clan affiliation, and members share residence or live in close proximity to one another (Hayden and Cannon 1982). Anthropologists have linked corporate group organization to scheduling conflicts, resource control, inheritance, and the size of the labor group necessary to extract key subsistence resources (Hayden and Cannon 1982; Nimkoff and Middleton 1960; Pasternak, Ember, and Ember 1976). Cross-cultural studies of household composition and size suggest that the natural tendency is for households to resemble nuclear family units. When independent families are able to carry out subsistence activities on their own, household organization tends to favor nuclear families (Hayden 1995; Pasternak, Ember, and Ember 1976; Wolf 1966).

Figure 11.1. Location of La Joya within Sierra de las Tuxtlas, Veracruz, Mexico

Table 11.1. Chronological phases of La Joya

In reality, this continuum is more complex than it first appears, because independent nuclear family households can be nested within corporate groups (Gillespie 2000; Santley 1993; Smith 1993). Households may form the basic unit of production, but political, social, and religious roles within a community may be coordinated by larger entities. Cross-cultural comparisons suggest that as the number of activities shared by members of a corporate group increases, so does the closeness of the group’s organization (Hayden and Cannon 1982).

In other regions of Mesoamerica, multifamily corporate groups formed shortly after the transition to sedentism and during periods of emerging hereditary social inequality. This type of group ranged from nucleated apartment compounds of La Coyotera in the Cuitcatlan Cañada (Spencer 1982) to the residential wards marked by iconographic motifs at San José Mogote, Oaxaca (Flannery and Marcus 1994; Parry 1987; Pyne 1976) and the loosely knit neighborhoods composed of the economically independent but politically linked households of Cerros, Belize (Cliff 1988).

One strategy for creating hereditary status differences lies in the formation of multifamily corporate groups. Hayden (1995) proposes that aggrandizers—individuals who seek wealth, power, and prestige—attempt to form these types of groups to increase a wealth-producing base. This process involves aggrandizers who draw on existing kinship ties and obligations and convince their kin to increase production and ultimately the group’s wealth. An aggrandizer’s kins-people also seek to cement their relationships with this individual to elevate their own status positions through association and to ensure recompense. As wealth, power, and prestige become entrenched in the corporate group, the ability to transfer status from one generation to the next becomes permissible. In many Formative Mesoamerican communities, the timing of corporate group emergence may indicate that similar processes played a role in the emergence of complex societies (Blanton et al. 1999:36–42).

While the formation of some corporate groups correlates with critical transitions in social evolution, the range of household organization may also vary within a single community. When household organization varies from nuclear family households to corporate groups within communities, the corporate groups are strongly correlated with wealthy families (Netting 1979, 1982). From the Late Formative period on, households with the largest architecture, elaborate burials, and the most specialized craft items are organized along corporate descent lines in many Mesoamerican communities (see, e.g., Collier 1975; McAnany 1994). The curvilinear model of household organization suggests that the amount of land available for newly married couples to establish an independent household is a particularly influential factor of household composition (Collier 1975). When land is highly available, newly wedded couples will move to unoccupied land to establish a new household. In contrast, when land is scarce, young couples will remain part of the parental unit, which forms a multifamily corporate group. In this case, the personal interests of the household head and the subordinate generation act in a complementary manner to keep the corporate group perpetuating as a unit. The younger generation desires to retain membership to access corporate resources and the corporate group headman delays inheritance to retain labor in the household, thereby increasing production and minimizing corporate group fissioning (Collier 1975; McAnany 1994). When land becomes too scarce to support a large group, however, the trend reverses. Consequently, when land is extremely scarce, the progenous generation will leave the household to establish its own household. In this situation, agriculture is no longer a viable option, and the new household will likely shift to craft production (Collier 1975). This relationship between household organization and land availability provides an interpretive framework for the fluctuating community organization documented at La Joya.

DOCUMENTING COMMUNITY ORGANIZATION AT LA JOYA

Natural and cultural depositional processes have buried Formative cultural deposits up to four meters below today’s ground surface at La Joya and within the Tuxtlas. Furthermore, there is little correlation between surface and subsurface artifact densities (Arnold et al. 1996; Arnold and McCormack 2002), limiting the ability to target subsurface deposits through excavation within the twenty-five-hectare settlement. Therefore, unlike many regions of Mesoamerica where community organization can be investigated through collecting surface material, mapping house-mound locations, or exposing architectural remains that are buried near the surface, the study of community organization at La Joya requires alternative methods.

To model community organization along the lines of independent household groups to multifamily corporate groups and to trace changes through time, an extensive subsurface sample was collected with a bucket auger from La Joya. Data from the auger test probes were used to create density maps in much the same way surface collection data are used. The auger test probes were spaced 10 to 12.5 meters apart on a staggered grid. Diagnostic cultural material and stratigraphic information were used to assign the temporal period to each ten-centimeter arbitrary level for each auger test probe. Finally, the cultural material within each period was plotted to form distribution maps, which define areas of high and low densities. Research in the Tuxtlas suggests that the remains of prehispanic houselots encompass areas of about 0.5 hectare (approximately 70 × 70 meters) and include a number of common elements such as structures, clear areas, intermediate areas, garden orchards, and middens (Arnold and McCormack 2002; Killion 1987; Pool 1997). In this case, the houselot is taken to represent an archaeological signature of past households (Wilk and Ashmore 1988). In this study, I use the high-density areas as a proxy measure of houselot locations.

While artifact distributions were examined using density maps created through a variety of methods (McCormack 2002), the maps presented in this chapter are based on the mean and standard deviation of log (weight/m³). These statistics define contour intervals illustrating high-density areas unique to each phase. I prefer to use weight per volume, rather than counts, because it is less influenced by differential depositional and postdepositional processes than the number of sherds. For example, sherds dropped in open areas may become fragmented through trampling, whereas material immediately deposited into a midden is less likely to fragment. The overall weight of sherds per volume of soil will be greater in a midden than for open areas. Since the goal of the density maps is to identify the approximate locations of houselots and general areal extent, weight/volume is the best measurement to use. In addition, this approach defines the high- and low-density areas based on the characteristics of deposits unique to each phase.

The comparisons are akin to statistical probability tests such as the t-test or analysis of variance, which require single peaked and symmetrical batches (Drennan 1996). Archaeological samples are commonly upwardly skewed and require a transformation prior to performing probability tests. The samples of sherd weight/m³ are no exception. Therefore, the batches were logarithmically transformed; which produced single peaked and symmetrical batches (McCormack 2002). Finally, the mean and standard deviations of the transformed batches were used to create the contour maps and to identify the approximate locations of past houselots for each phase. The artifacts from auger probes within each high-density area are grouped, allowing comparisons of occupational zones within the settlement. The contour maps suggest that community organization fluctuates between independent nuclear family households and corporate groups.

COMMUNITY ORGANIZATION AT LA JOYA

Below, I discuss the nature of community organization at La Joya by time period.

Tulipan Phase (1300–1150 BC)

The initial occupation of La Joya dates to the Tulipan phase, when small villages and hamlets were dispersed throughout the Río Catemaco drainage. Early Formative occupants of La Joya were farming maize and probably practiced an extensive slash-and burn-farming strategy. In addition, they also relied on hunting and fishing to supplement the staple crop (VanDerwarker 2003). In the Tuxtlas, the locations with the best agricultural land typically do not adjoin the ecozones rich with aquatic resources. Therefore, a degree of mobility was likely required to grow corn and extract wild resources (McCormack 2002).

Figure 11.2, the distribution map based on the log (weight/m³), illustrates numerous high-density areas, more or less evenly distributed throughout La Joya. If we use the high-density areas as a proxy measure of houselot location, it appears that many nuclear family households occupy the site. The archaeological deposits (Arnold 2000), tool design (Arnold 1999; McCormack 2002), and nature of site structure (McCormack 2002) suggest that the initial Tulipan phase settlement of La Joya was occupied by groups who regularly visited the site, moved their houses frequently, or did both (ibid.). Given the mobility associated with the Tulipan phase, these high-density areas more likely reflect incidental reoccupation of the same locations rather than permanent houselot locations (ibid.), making it difficult to ascertain Tulipan phase household composition. Population estimates are difficult to calculate for this phase because of the mobile nature of the residence patterns but have been tenuously estimated between fifty-five and ninety-one persons (ibid.). In other areas of Mesoamerica, households tend to be organized in independent nuclear family groups during periods of residential mobility and early stages of village organization (Flannery 2002), and the same is likely true for the Tulipan phase occupation at La Joya.

Coyame Phase (1150–850 BC)

Around 1150 BC, Cerro Mono Blanco erupted and blanketed the upper Río Catemaco drainage with a layer of tephra. Fires, floods, and mudslides likely followed the eruption, causing long-term, wide-scale environmental degradation for two centuries or more (Chase 1981; Reinhardt 1991). Settlement-pattern research documents large-scale abandonment of the upper Río Catemaco drainage in the Middle Formative (850–400 BC; see Santley and Arnold 1996; Santley, Arnold, and Barrett 1997), and this abandonment likely reflects a long-term response to the Cerro Mono Blanco eruption. Occupation at La Joya, located beyond the perimeter of the major impact zone, continued into the Coyame phase. While the amount of cultivable land was severely reduced in the upper Río Catemaco drainage by this eruption, the population at La Joya increased from an estimated 55–91 persons to 168–272 persons during the Coyame phase (McCormack 2002). In addition, an increase in the productivity of maize around 1000 BC (Borstein 2001) may have allowed a larger population to support itself on the land available near La Joya. An intensified emphasis on maize in the diet and fewer options to move swidden fields appear to have encouraged sedentism in the Coyame phase (McCormack 2002).

Figure 11.2. Distribution of ceramics based on log (weight/m³) for the Tulipan phase of La Joya

Figure 11.3. Distribution of ceramics based on log (weight/m³) for the Coyame phase of La Joya

Table 11.2. Population estimates for Coyame phase residential clusters for La Joya

The increased occupation at La Joya is visible in Figure 11.3, the ceramic distribution map of log (weight/m³), which illustrates the presence of three large nucleated artifact clusters within La Joya. Given the areal extent, between 0.5 hectare and 2 hectare, and the associated population estimates (Table 11.2), these clusters suggest the presence of multifamily corporate groups, similar in size to residential wards or neighborhoods identified for other Mesoamerican Formative villages. Analysis of ceramic production techniques also suggests that these Coyame households were a continuation of the Tulipan phase populations (Arnold 2003). Settlement in the Coyame phase is not uniform, and Figure 11.3 also illustrates a number of small high-density clusters dispersed among the large clusters. Artifact densities and the structure of space in these clusters are similar to the Tulipan phase deposits (ibid). Therefore, I interpret these clusters as the result of short-term occupation. The transition to settled life is often discussed at the scale of the society; however, it is more likely that the decision to become sedentary was based within the household. Appropriately, a variety of residential strategies should be expected during the transition to sedentism.

Regional settlement patterns in the Río Catemaco drainage suggest that social organization in the Tuxtlas remained undifferentiated and unstratified at this time (Santley and Arnold 1996). Comparison of intrasite residential groups reveals some ranking differences among the residential clusters and suggests distinct differences between mobile and sedentary populations.

During the Early Formative period, individual households and corporate groups obtained obsidian through independent long-distance trade connections (Clark and Lee 1984; Pires-Ferreira 1975; Santley et al. 2001). The comparisons of surface collections from La Joya and other contemporary sites indicate that La Joya has a significantly higher obsidian to rim-sherd ratio (Santley, Arnold, and Barrett 1997), suggesting that the occupants of La Joya were consuming obsidian at a higher rate than the residents of other sites within the region. Santley, Arnold, and Barrett (1997) also propose that the occupants of La Joya were specializing in an economic activity requiring obsidian at this time. The focus on intrasite residential group differences suggests that groups acquired obsidian at different rates, which also influenced how obsidian was used and reduced within each residential group.

Table 11.3. Coyame phase obsidian-to-sherd ratios for La Joya

Table 11.3 presents the ratio of obsidian to sherds for the Coyame phase and illustrates that corporate groups consumed different amounts of obsidian. This ratio is the highest for residential group C, suggesting that the obsidian consumption rate is two to four times the consumption rate of residential clusters A and B. Also, the small clusters have an obsidian-to-sherd ratio greater than residential clusters A and B but less than residential cluster C. In addition to the relative amount of obsidian acquired in each group, how it was reduced differs. Technologically, reducing the small imported spalls and nodules of obsidian to usable flakes is not a difficult or a skilled process (Clark and Lee 1984; McCormack 1996). Flakes are first removed through percussion, and when the nodules become too small to effectively remove more flakes, the expended cores and flakes are reduced though bipolar reduction. This process creates small flakes and shatter. Given the Early Formative reduction technology, attempts to maximize obsidian might involve a greater effort to reduce every last piece of obsidian, resulting in small fragments produced through bipolar reduction. Most of the obsidian chips recovered from La Joya are extremely small and were probably mounted onto a wooden or bone handle and used as a knife (see Andrefsky 1998), inserted into grater boards (see Lewenstein and Walker 1984), or used with some other type of composite tool.

Table 11.4 presents the categories of obsidian flakes and shatter size and the reduction technology that likely produced the pieces (Boksenbaum 1980; Boksenbaum et al. 1987; Jeske and Lurie 1993). The difference in the obsidian chip sizes among two residential clusters, B and C, and the small clusters is not very significant (X² = 5.85; 0.50 > p > 0.20, V = 0.16). The sample of obsidian from residential cluster A was not included in the X² statistic because of the small sample size.

Although we can not have a great deal of confidence that there are differences in the proportions of chip sizes among the entire sample, comparisons of the proportions and error ranges of the fifteen-millimeter subset indicate that there is greater than 95 percent confidence that the proportion of fifteen-millimeter obsidian chips is greater in residential cluster C and the small clusters than in residential cluster B (McCormack 2002). This statistic suggests that there are some differences in how obsidian was reduced within the different residential groups.

Table 11.4. Coyame phase obsidian flake sizes at La Joya

The proportions of flakes produced through bipolar smashing also support these differences. Table 11.5 presents the proportions of obsidian produced through core-flake and bipolar production associated with each residential cluster. Overall, the difference in the technology that produced the flakes among the residential clusters is not significant (X² = 2.25; p < 0.50, V = 0.15). However, when focus is shifted to the subset of residential clusters B and C, there is a greater than 80 percent chance (X² = 1.76; 0.20 < p < 0.10, V = 0.15) that the differences observed between these two samples actually do reflect differences in core-flake and bipolar technology. Taken together, with the information on the ultimate size of obsidian chips, these patterns suggest that residential cluster B was obtaining less obsidian and maximizing the obsidian more so than residential cluster C.

In contrast to obsidian acquisition and reduction strategies, the comparison of ceramic assemblages reveals more similarities than differences among the Coyame phase residential clusters. While a substantial number of sherds have been recovered from each residential group, vessel type is identifiable for only a small percentage of the sherds. To document domestic activities the proportions of vessel types are compared. Analysis focuses on two major categories, serving vessels (bowls and plates) and food-preparation vessels (tecomates and jars) (after Santley 1992).

Table 11.6 presents the proportions of vessel forms associated with each residential cluster. Comparison of the vessel forms indicates that there is no statistical difference (X² = 1.25; p < 0.50, V = 0.11) in proportions of vessel forms among the four residential groups, which suggests that all residential groups were engaging in similar food-consumption and preparation activities. This pattern conforms to observations at the regional level where “Formative sites in the Tuxtlas exhibit little intra-assemblage variability in types present within sites” (Santley, Arnold, and Barrett 1997:193), implying that the occupants of sites performed similar food-preparation and serving activities in similar amounts.

Table 11.5. Coyame phase stone-tool reduction technology at La Joya

Table 11.6. Coyame phase vessel types at La Joya

In addition to serving and cooking, households and corporate groups sometimes contain different amounts of fancy service wares, which are frequently used as an indicator of high status (Pool and Britt 2000; Smith 1987). While there are no ceramic wares that are distinctly obvious indicators of status at La Joya during the Coyame phase, the proportions of decorated and undecorated ceramics reflect differences in the labor and craftsmanship invested in the production of ceramics (Table 11.7). Decorated sherds include all sherds with evidence of surface modification such as incisions, punctation, and slips, while undecorated ceramics exhibit no surface treatment. The difference in the proportions of decorated and undecorated sherds is significant (X² = 17.5; p > 0.001, V = 0.06) among the four residential clusters. Examination of the proportions associated with each residential cluster illustrates that the small clusters have the lowest proportion of decorated sherds. However, it is important to note that the significance associated with this X² statistic is influenced by the large sample size. In addition, Cramer’s V indicates only a 6 percent difference in the overall proportions of decorated and undecorated sherds, which is an overall weak finding despite the significant results of the X². Given the general patterns of ceramics, it appears that all residential clusters were engaging in similar amounts of cooking and serving activities, but the tendency is for the residentially mobile subpopulation to have fewer decorated vessels than the sedentary folks.

Table 11.7. Coyame phase decorated sherds at La Joya

Table 11.8. Coyame phase figurine-to-ceramic ratios at La Joya

Figurines are generally considered as objects that all Formative Mesoamerican households used in rituals (Follensbee 2000; Marcus 1996) and were material points of reference used in the negotiation and reproduction of social relations (Joyce 1993). However, figurines are sometimes more commonly found associated with large structures than small structures (Lesure 1999), suggesting that large households used figurines in public rituals, and small households did not.

Table 11.8 presents the ratios of figurine fragments to sherds in each group. As with the comparison of obsidian above, these ratios show that residential cluster C is associated with the greatest number of figurines. While figurines were used for household-level rituals, they may also have been used in community-level rituals. The greater number of figurines recovered from residential cluster C may indicate that this residential cluster performed rituals more frequently than the other residential clusters. If figurines were used in ancestor rituals, as proposed by Marcus (1996), perhaps ancestors and the link to the past were of greater importance to this group than the other residential clusters. These comparisons of figurine-to-sherd ratios also indicate that the small clusters, the residentially mobile subpopulations, have the fewest figurines. Overall, this distribution may hint at subtle activity ranking among the residential groups. Also, the limited number of figurines associated with the residentially mobile subpopulation suggests that they were not participating in rituals involving figurines as frequently as the sedentary residential groups.

Table 11.9. Summary of Coyame phase residential clusters La Joya

The comparison of artifact assemblages associated with the residential clusters illustrates minor differences among the Coyame phase corporate groups. Table 11.9 presents the generalized pattern, and when there are differences among the groups, residential cluster C stands apart from the others. To recapitulate, residential cluster C acquired relatively more obsidian, possibly reducing the need to conserve the resource in the manner necessary for the occupants of residential clusters A and B. Residential cluster C also contains the greatest number of figurines, suggesting that it performed rituals more frequently than the others. These patterns suggest that residential cluster C had independent contacts beyond the village and that they may have held a higher social position above the other residential groups.

The small clusters differ from the large artifact clusters in that they have fewer figurines, fewer decorated sherds, and an intermediate amount of obsidian. The relative lack of figurines and decorated sherds suggests that it was less important to express membership or individuality in a permanent medium for the members of the small clusters than for the members of La Joya corporate groups. Ritual uses of figurines, whether they occurred in the household or at the community level, may serve the function of bonding individuals within a group. It may have been less important for the part-time residents to participate in, host, or display materials within their houses during rituals and feasts.

Gordita Phase (850–400 BC)

During the Gordita phase, settlement patterns in the Río Catemaco were concentrated in the lower reaches of the drainage, which may have been a continued response to the 1150 BC eruption of Cerro Mono Blanco and the associated environmental degradation. Similar to the Coyame phase, La Joya is one of the few sites in the upper Río Catemaco drainage that was occupied during the Gordita phase. However, Figure 11.4, the distribution map of log (weight/m³), indicates some significant shifts in community organization. Instead of the relatively nucleated areas, a number of relatively small clusters are distributed throughout the site of La Joya. Population estimates for La Joya also suggest an overall population decrease from 163–272 persons to 61–102 persons (McCormack 2002). The shift toward dispersed settlement suggests a breakdown of multifamily corporate groups. While no major settlements are located between La Joya and Cerro Mono Blanco, a few hamlets are present (Santley and Arnold 1996:fig. 3). In addition, excavations at Bezuapan, an 8.5-hectare Late Formative settlement located five kilometers east of La Joya, reveal that this location was first occupied at the end of the Middle Formative period (Pool and Britt 2000). While data from the regional settlement-pattern study suggest continued abandonment of the area most adversely affected by Mono Blanco’s eruption, it is likely that some of the land had recovered to support vegetation and agricultural crops. Consequently, small groups could begin reoccupying this area, a point I return to below.

Figure 11.4. Distribution of ceramics based on log (weight/m³) for the Gordita phase of La Joya

Table 11.10. Population estimates for Bezuapan phase residential clusters at La Joya

Bezuapan Phase (400 BC–AD 100)

Around 400 BC, Chuniapan de Abajo emerged as a center of a four-tiered settlement hierarchy in the Río Catemaco drainage, marking the emergence of chiefly society in the Tuxtlas. La Joya is one of the largest villages and may have served as a secondary center. In addition, numerous settlements were established in the upper Río Catemaco drainage in locations of good agricultural land (Santley and Arnold 1996). Figure 11.5, the distribution map of log (weight/m³), illustrates a significant increase in the intrasite occupation from the previous period. Settlement is distributed throughout the site, including the presence of three high-density areas. While the boundaries of these clusters are not as distinct as the Coyame phase clusters, low-density areas within the site do separate the nucleated residential clusters. The population sizes associated with these clusters also suggest corporate group organization (Table 11.10).

Table 11.11, which presents the ratios of obsidian to sherds for each residential cluster of the Bezuapan phase, illustrates that no one residential cluster obtained markedly more obsidian than the other residential clusters. With the emergence of a chiefly center, all trade into the upper Río Catemaco drainage may have been filtered through Chuniapan de Abajo, rather than directly to the consumer, a change that may have altered and limited the amount of obsidian available to all residents at La Joya.

In addition, Table 11.12 presents the proportions of obsidian flake sizes for each residential cluster. Comparison of the sizes (X² = 0.06, p < 0.50, V = 0.01) show that residential groups were reducing obsidian in a similar manner, which further supports the notion that each residential group acquired obsidian in similar amounts. Furthermore, comparison of reduction technology presented in Table 11.13 reveals no significant difference (X² = 3.17, 0.50 > p < 0.20, V = 0.12) among residential clusters A and B. Area C was excluded from these calculations because the sample is small.

Figure 11.5. Distribution of ceramics based on log (weight/m³) for the Bezuapan phase of La Joya

The one exception to this pattern is the proportions of obsidian blades in the assemblage of each residential cluster. Blade technology increased during the Late Formative period, eventually becoming the major technological reduction strategy throughout Mesoamerica. Blades are rare at La Joya, and they may have served as a prestige item. While there is little confidence that the overall technological reduction strategies differed among the groups, we can have greater confidence (X² = 2.973; 0.1 > p < 0.05, V = 0.12) that the difference between the proportions of blades associated with residential clusters A and B reflects that residential cluster A obtained greater numbers of blades. Because of the single blade and overall small sample of obsidian from residential cluster C, it was not factored into this analysis.

Table 11.11. Bezuapan phase obsidian-to-sherd ratios at La Joya

Table 11.12. Bezuapan phase obsidian flake size at La Joya

Table 11.13. Bezuapan phase reduction technology at La Joya

The comparison of the ceramic assemblages presented in Table 11.14 reveals similar proportions of cooking and serving vessels (X² = 1.6, 0.5 > p < 0.2, V = 0.08). Therefore, we can conclude that each residential group engaged in similar food-consumption and preparation activities, and no particular group appears to have been serving food more frequently, an activity associated with hosting feasts. However, the comparison of the proportions of decorated versus undeco-rated sherds reveals a significant difference among these groups. The inspection of the proportions in Table 11.15 (X² = 10.522; 0.01 > p < 0.001, V = 0.03) indicates that residential cluster C contains the fewest decorated sherds, pointing to a lower social position than the other residential clusters.

Table 11.14. Bezuapan phase vessel types at La Joya

Table 11.15. Bezuapan phase decorated sherds at La Joya

Specialized serving vessels appear during the Bezuapan phase with the emergence of a chiefly society in the Río Catemaco drainage. Fine-paste vessels are made without temper and are on average thinner than vessels of the same size with temper, attributes that would convey group affiliation and social status (Pool and Britt 2000:154). As reflected in Table 11.16, no fine pastes were recovered from residential cluster C, a detail further suggesting that residential cluster C held a lower position among the residential clusters at La Joya. But the lack of fine-paste sherds also prevents including it in the X² analysis. The X² reveals that the higher proportion of fine-paste sherds in residential cluster A is statistically significant (X²= 200.82; p > 0.001, V = 0.13), suggesting that residential cluster A has a higher status than B.

To summarize, the Bezuapan phase residential clusters still appear to be engaging in similar domestic activities. However, as Table 11.17 illustrates, the differences observed for artifacts associated with status (e.g., blades and fine-paste ceramics) are consistently greater for residential cluster A. This pattern suggests that at this time, when regional hereditary social differences emerge, status differences are also endemic at La Joya. While these differences are not great, their presence suggests graded social differences within the region.

Table 11.16. Bezuapan phase fine-paste proportions at La Joya

Table 11.17. Summary of Bezuapan phase La Joya residential clusters

SUMMARY AND CONCLUSIONS

The review of intrasite artifact distributions suggests that community organization at La Joya fluctuated between independent, nuclear family–sized households and multifamily corporate groups. The emergence of corporate groups at La Joya is neither static nor a unidirectional development. This fluctuating pattern therefore raises the question, why did corporate groups emerge when they did at La Joya? The examination of available land resources reveals a correlation between land scarcity and multifamily corporate group organization. To quantify land scarcity, detailed information regarding regional and local populations, diet, subsistence, agricultural techniques, land under cultivation, agricultural yields, storage technology, and a host of other factors is necessary. Formative period research in the Tuxtlas is a relatively recent endeavor, with most studies dating to the last couple of decades. Therefore, much of the data needed to quantify land scarcity is not yet available. A number of changes in the regional social and environmental landscape do suggest, on a broad scale, that the relative availability of land fluctuated through time.

During the Tulipan phase, the combination of low population levels and mobile residence patterns suggests that land was highly available, allowing residential groups to shift houselot and, presumably, milpa locations fairly frequently. The direct and consequential effects of the eruption of Cerro Mono Blanco would have dramatically changed that relationship. The blanket of tephra near Mono Blanco rendered soil infertile, probably for centuries. When ash deposits are thin, however, they act as mulch, increasing soil fertility (Reinhardt 1991). The thin ash deposit at La Joya and continued occupation suggest that agriculture was possible in this location, but the ability to move about the landscape and relocate milpas would have been greatly reduced.

Moreover, La Joya may also have served as a refugee zone, absorbing some of the populations displaced by Cerro Mono Blanco’s eruption. The increased locational stability and reduced agricultural land might influence the young generations to remain within the parental households as no other land suitable for agriculture was available. While there are some minor wealth differences, these Coyame phase corporate groups were not vehicles enabling the emergence of hereditary social inequality. Therefore, it appears that corporate groups of the Coyame phase were adaptive responses to a decrease in the available agricultural land.

By 800 BC, approximately 300 years following the eruption of Mono Blanco, the tephra deposits likely degraded to a point where they would have formed highly productive agricultural soils. The inheriting generation would therefore have an option to establish a new household on unoccupied land. This option could very well lead to the dissolution of multifamily corporate groups.

Finally, following a period of nuclear family–sized household organization, multifamily corporate groups formed again in the Bezuapan phase at La Joya. While the natural landscape was not dramatically altered as it was in the Coyame phase, the social landscape changed between the Gordita and Bezuapan phases. Populations infilled the locations with the best agricultural soil, and the presence of a chiefly center may have further restricted a household’s ability to settle in an unoccupied location with good agricultural land. Therefore, when households were presented with the option to fission and move to locations without land or remain part of the parental household and become a multifamily corporate group, the option to remain part of the original household may have been the more attractive choice. Comparison of La Joya and Bezuapan subsistence strategies suggests that the residents of La Joya incorporated a combination of farming, hunting, and fishing, while farming was dominant at Bezuapan (VanDerwarker 2003). This strategy may be a result of differences in household organization and the ability of larger corporate households to diversify production (see Henderson 1998, this volume).

This longitudinal study of community organization with a focus on corporate-group presence within the social and environmental landscape reveals a relationship between nuclear family households and land abundance, and corporate group formation and land scarcity. While the broad nature of these data prevents a detailed discussion of the specific household activities and composition, it does reveal dynamic changes within one community over time as individual households adjusted to alterations in the natural and social landscape.

Acknowledgments. The National Science Foundation, the Alcoa Foundation and a predoctoral Mellon fellowship from the University of Pittsburgh funded this research at La Joya. Fieldwork was carried out under a permit from the Instituto Nacional de Antropología e Historía and with the cooperation of the community of Comaopan, Veracruz, Mexico. I thank Philip J. Arnold III for his encouragement and enthusiasm from the time when the initial ideas were formulated through final production. The clarity of the arguments presented in this chapter greatly benefited from discussions with Olivier de Montmollin, Kathleen Allen, and Robert Drennan. I thank John Douglass and Nan Gonlin for organizing the original symposium at the 66th Annual Meeting of the Society for American Archaeology and for dedicating their time to develop the original symposium into the current volume. John’s and Nan’s time and comments on earlier drafts, as well as two peer reviewers, are greatly appreciated, although all errors and omissions, of course, remain my own.

REFERENCES CITED