4
Complexity without Centralization

Political Economic Strategies and Political Organization

Constant conflict and warfare reigned at the dawn of Spanish incursion in the Tequila valleys. Polities built temporary alliances among themselves to protect their territories, but they never unified into a single political structure; instead, they retained their political autonomy. Caxcanes and Tarascans had surrounded the valleys and, as the early documents suggest, they made several attempts to take control over this highly coveted region. How did these societies manage to stand off their enemies while at the same time preserving their autonomy rather than merging into a single polity?

In response to external military and political pressure, the political architects of the Tequila valleys could have deployed network strategies to recruit the followers necessary to maintain their independence (Blanton et al. 1996). In network strategies, individuals seek to concentrate power, authority, and riches in particular families by manipulating wealth generated through long-distance exchange (commercial, political, etc.). Wealth accrued through such trade is used to recruit clients locally via elaborate consumption, gifting, and feasting, accompanied by patrimonial rhetoric. An aura of tension and competition holds sway in these cases since there are multiple competing groups fostered on personal relationships and gift-giving. In this scenario, alliances are highly volatile between people or groups; in addition, ties among the population within these polities are weak, because only a portion of the population benefits from such relationships. The overall populace has little or no social mobility and their demands can be ignored because the power and authority of leaders are grounded elsewhere or depend on external revenues (see Blanton and Fargher 2008 for a discussion of the impact of external revenues on political organization). This results in little incentive for elites to provide for voice and public goods in an effort to obtain compliance from a broad constituency (e.g., free peasants). Rituals emphasizing community integration are also absent, hence large and accessible plazas that may suggest community-wide gatherings do not figure into the overall configurations of towns and cities (see Fargher, Heredia Espinoza, and Uriarte Torres 2014). The result is a politically fragmented landscape that is in some ways analogous to the Tequila valleys during the Postclassic, as ethnohistoric sources suggest (see below). Yet, evidence of prestige goods and patrimonial rhetoric in the Tequila valleys during the Postclassic is fleeting, suggesting that network strategies were not the basis of polity-building during this period.

Conversely, people in the Tequila valleys could have responded to the external threats by using collective strategies to unify the polities in a single powerful state with the capacity to mobilize the entire population for military defense, like their contemporaries in Tlaxcallan (Fargher et al. 2010; Fargher, Heredia Espinoza, and Blanton 2011). In Tlaxcallan (the prehispanic name of the state of Tlaxcala, Mexico), an ethnically diverse group responded to external military pressure (from the Aztec Triple Alliance) by building an inclusive state grounded in collective action. In previous publications, my colleagues and I (Fargher et al. 2010; Fargher, Heredia Espinoza, and Blanton 2011) argued that Tlaxcallan was built around cooperation in which the state bureaucratized to accommodate taxpayer voice and offer public goods, especially military leadership, in exchange for revenue payments, especially military corvée, by commoners. However, the state went further in order to build the trust and confidence among Tlaxcaltecans necessary to cultivate compliance with their revenue demands. The state’s political architects instituted a powerful egalitarian or corporate ideology that rewarded service to the state, especially in warfare, with promotion to “noble” status and political offices (including, possibly, a seat on the ruling council), as well as implementing a strict and austere moral code that governed the behavior of officials (cf. Fargher, chapter 15, this volume). Religion was also employed to justify and reinforce the merit-based egalitarian ethos, emphasizing Tezcatlipoca, a deity associated with merit, judicial equality, and self-sacrifice regardless of social position (Fargher 2012; Fargher et al. 2010). Building on this corporate ideology, the Tlaxcaltecas created a unified political structure based on a ruling council that consisted of somewhere between 50 and 250 officials, who made consensus-based decisions and were promoted on merit (Fargher et al. 2010:238).

Much like Tlaxcallan, wealth and prestige goods were deemphasized in the political strategies employed in the Tequila valleys during the Postclassic; yet, the Tequila valleys remained decentralized and the political landscape was fragmented, which was very much unlike Tlaxcallan. Thus, I posit that an additional pathway to complexity was followed in Tequila that differed from both network (or external revenue) strategies and collective action based on internal revenues, bureaucratization, voice, and public goods. Here, the multiple multiethnic groups maintained their political autonomy, yet they created short-term alliances to defend their territories. Specifically, I contend, based on the archaeological record and ethnohistoric information, that Late Postclassic political architects employed a form of sequential hierarchies and powerful corporate or egalitarian ideologies to build strong polities able to withstand common foes, retain political autonomy, and limit personalized power based on monopolies over long-distance exchange and ideological resources. They built a horizontal and highly collective political structure, but unlike other states high in collectivity, they did not assemble large amounts of revenue and invest in major public goods and/or a complex bureaucratic apparatus.

Corporate Political Strategies

In the original formation, corporate strategies are those where aggrandizing behaviors are suppressed in favor of more egalitarian or group-oriented dimensions (Blanton et al. 1996). In these strategies, images of individuals are obscured and/or shown in groups or crowds. Depictions of specific personages that could be identified with a name are absent and instead groups or individuals are depicted, if at all, wearing masks. Writing, where present, tends to be dedicated to keeping records of transactions, rituals, and other important events and it is not used to document the deeds of specific individuals. Corporate strategies used by political actors are implemented to create community-wide integration and cooperation, and can be recognized archaeologically by investment in public ceremonial and civic architecture, an emphasis on fertility rituals, community feasting, and the minimization of individualizing behaviors (Blanton et al. 1996; see also Blanton 1998a). Thus, palaces, personal monuments, elaborate tombs, and control over prestige goods are masked or absent.

As Fargher points out (chapter 15, this volume), corporate strategies are one tool that may be used in building collective states, but the institutional plan varies among polities. In the particular case of the Tequila valleys, corporate strategies involved power-sharing (probably in the form of consensus decision-making) and an egalitarian ideology but not heavy investment in the development of institutional infrastructure. Instead, I posit that consensual decision-making was achieved through the development of sequential hierarchies and an ideology that downplayed the role of wealth accumulation in gaining political prestige and power.

Sequential Hierarchies and Corporate Strategies

Johnson (1982, 1983) pointed out that the information stress created by the nucleation of large numbers of individuals within a single site or polity could be solved through heterarchical strategies, such as sequential hierarchies, instead of developing more centralized and hierarchical structures through either network strategies or bureaucratization. The development of sequential hierarchies involves grouping individuals or households into larger and inclusive decision-making units (e.g., complex households, clans, moieties, etc.) thereby diminishing their number and allowing for more horizontal decision-making (e.g., power-sharing or consensus decision-making). As opposed to dealing with a myriad of small units, the reduced number of larger aggregations decreases the number of individuals with whom consensus must be reached and, thus, decision-making stress is scaled down. At the same time, these larger units are horizontally constructed with unranked representatives who possess equivalent decision-making authority. They are also responsible for ensuring that their unit conforms and cooperates with decisions reached in council.

Sequential hierarchies have been documented in a range of societies from simple to complex (Bargatzky 1995; Johnson 1982, 1983; see also Kowalewski 2006 for similar patterns). For example, the Huron and Iroquois used sequential hierarchies to build confederacies. Huron councils originated informally from extended families. At the village scale, Huron ruling councils were built of the heads of the long houses (extended families) and larger political agglomerations (e.g., communities and peoples) were formed through supravillage councils made up of representatives from each member village. In both village councils and supravillage councils, decisions were reached through voting and consensus (Trigger 2002:89–91). The position of chief was hereditary and within a community he was “recognized as the principal chief and spokesman for the entire community” (Trigger 2002:87–88). Two types of chiefs existed among the Huron: civil or peace chiefs and war chiefs. Only the former were members of the council and hence of the national and confederacy levels. Civil chiefs were selected from their own lineages, and their positions were inheritable within a lineage. Their main duties were to act as speakers or chairmen: they assisted their group to reach agreements and represented them in foreign affairs (Trigger 2002:81). Although some chiefs were more prominent than others, there is no indication that they had more authority than the rest. In fact, chiefs could not coerce or force their will on anyone (Trigger 2002:84), their duties resided in being the spokesmen of their people. Given this nested political structure, all clan segments had representative chiefs at the Huron confederacy level.

Among the Iroquois, the structure of the confederacy of 50 sachems representing the six nations was formed through “nested hierarchies” that strongly resemble sequential hierarchies. At the very base of this political structure were households that made up tribes. A group of tribes (e.g., Onondaga, Seneca, Mohawk) constituted a nation. Tribes selected sachems, which then represented their nation and group decisions at the confederacy level. Each sachem possessed the same powers and shared equal rank and authority as all the other members of the confederacy despite the uneven number of representatives for each nation. The sachem office was hereditary within the tribe, but elective among the male members of the tribe itself (Morgan 1901).

Finally, the Cheyenne are another example of government by council where the council of 44 peace chiefs was established to prevent internal disagreement and to solve matters of war (Hoebel 1988). Those chiefs selected for the council possessed certain qualities, which included “good temper, generosity, energy, courage, altruism, wisdom” (Hoebel 1988:43). Chiefs selected for the council were experienced warriors who upon selection into the council became peace chiefs, losing their position (but not their memberships) as war chiefs (Hoebel 1988:43). Hereditary factors did not play an important part as a requisite in the selection process, but frequently, after a 10-year term in office, a peace chief could choose a son as his successor (Hoebel 1988:51). A chief was both a representative of his extended family and band; first and foremost, however, as member of the council he acted as protector of each member of the tribe (Hoebel 1988:43). Above all, the council made decisions such as “camp moving and tribal war polity” and in some cases (e.g., criminal acts) it acted as a judicial body (Hoebel 188:52). Decision-making was reached by prolonged debate until agreement was attained. All peace chiefs were equally ranked and as a council their authority was invested with supernatural skills, making it the supreme authority (Hoebel 1988:49).

The aforementioned forms of organization are quintessential examples of successful sequential hierarchies. Most important, this form of organization distributes authority and representation of the population in such a manner that people have a voice in the political decision-making process. Nonetheless, the distribution of decision-making among various segments is not sufficient for maintaining order and cohesion.

Creating a sequential hierarchy does not by itself prevent fissioning, internal conflict, or prestige-good consumption and wealth competition among groups (see for example the societies of the Northwest Coast [e.g., Ames 1995]). As Stark points out (chapter 5, this volume), there is always tension between individuals seeking personal gain (network strategies) versus those who favor group interests and welfare (corporate strategies). Accordingly, individuals seeking personal power may attempt to derail or obstruct consensus-building to thwart collective decision-making. As Blanton and Fargher (2016) point out, achieving cooperation in human societies is extremely difficult and designing functional governing structures based on collective action may be the most difficult challenged faced by humans as a species.

In order to control such behaviors, political architects must support sequential hierarchies with corporate cognitive codes. Such codes function to reinforce group identity and an egalitarian ideology that suppresses prestige goods and network strategies. However, once codes are in place, ceremonialism and ritual activity become essential for group maintenance and cohesion. In situations where leaders seek to integrate multiethnic groups into larger aggregations, such as was the case in the Tequila valleys, a common symbolic language that would cross-cut linguistic barriers and allow integration and a shared identity must be developed (see Kowalewski 2006). Ritual, in these situations, plays a significant role in creating group cohesion and allowing corporate cognitive codes to support a “corporate political structure” (Blanton et al. 1996:2).

Corporate cognitive codes also emphasize reciprocal obligations among groups of distinct origin through the enactment of rituals that “transcend local belief systems” (Blanton et al. 1996:4). Integration of different units requires the creation of new myths and spaces for the enactment of these rituals. Sizable plazas capable of bringing together large numbers of people—or conversely, multiple plazas—are indicative of such collective events (for examples, see Fargher et al. 2011; Fargher et al. 2014). The sizes of plazas or the number of them may indicate such collective codes, and public architecture, such as certain monuments, may also function as an integrative mechanism. Connectivity between different architectural groupings may also be indicative of such codes (e.g., Blanton and Fargher 2012; Fargher et al. 2014). Emphasis on shared identity through these communal events would have served as a key ingredient for the construction of strong polities.

A corporate cognitive code is also vital to restrain the emergence of exclusionary strategies such as aggrandizement, massive and luxurious tombs or burials, or monuments in honor of particular individuals or dynasties (Blanton et al. 1996:6). Such codes allow power-sharing strategies “across different groups and sectors of society” (ibid. 1996:2; Blanton 1998a:158, 159), thereby inhibiting the power over decision-making by a particular group (see also Fargher, chapter 15, this volume). Representatives of the different units, and hence the people represented, would have voice in the decision-making processes. Reduced wealth disparities are also consistent with a corporate cognitive code, where significant hoarding of material riches and their use as political currency is controlled. A corporate cognitive code and egalitarian ideology thus create unity in groups that cross-cuts ethnic and linguistic boundaries. This must have been extremely advantageous in times of warfare. Mobilization for military defense would have profited from these egalitarian strategies and ideologies as they effectively functioned to congregate an army when needed (cf. Kowalewski 2006).

Data Collection: Ethnohistory and Archaeology

A full-coverage systematic survey north of the Tequila volcano provides important data complementing the ethnohistoric information available on political organization. Using a modified version of the regional survey developed in Central Mexico and Oaxaca highlands (Kowalewski et al. 1989; Kowalewski et al. 2009; Sanders et al. 1979), the Tequila survey recorded all evidence of human occupation from small artifact scatters to sites covering hundreds of hectares. Our methodology consisted of walking every field with a distance of 20–40 m between surveyors. Site limits, identified by the fall-off frequency of artifacts and architectural limits, were drawn on the aerial photo. Our collections were of two different kinds: grab bags, or general collections, and specific collections. General collections were usually taken at small sites that we recognized as single-period occupations. Specific collections were taken at exact locations within a site such as structures, terraces, and other prominent features in order to provide accurate information on site components.

Results

Ethnohistoric Data

The cultural history of the Tequila valleys of Jalisco can hardly be framed in simplistic diachronic terminology where early (or simple) societies gave way to more complex social formations. Here we see periods of stability and continuity, highlighted by rupture, coalescence, and diversity in the way that societies formed and were created at different points in time. This is the case for the Late Postclassic, when multiple independent, ethnically and linguistically distinct polities occupied the Tequila valleys (Baus de Czitrom 1982; Soto de Arechavaleta 1994; Weigand 1993:136–137; Weigand and García de Weigand 1996:31–32). Ethnohistoric documents mention Cocas, Tecuexes, and Cazcanes among those ethnic groups (Acuña 1988; Baus de Czitrom 1982). Linguistically the region was also diverse and Cora, Coano, Cazcán (closely related to Nahuatl), Coca, Vitzurita or Wixárika, Tecozquín, and Nahuatl were spoken in the region (Yáñez Rosales 1994, 2004:85–86). In addition, Caxcanes (and other Chichimec groups) and Tarascans were encroaching on the valleys and, as the early documents suggest, both groups made several attempts to take control over this highly coveted region (Coria 1937:558; Weigand 2013:59–61).

Ethnohistoric data on the polities north of the Tequila volcano are paltry and dispersed in various documents. The Spanish understated the political organization of the region and descriptions are superficial, mixing terms such as señor, calpixque, and cacique, among others. Nonetheless, it is evident that not all towns shared a single organizational structure, and the documents also mention the existence of several polities, which indicate political fragmentation, decentralization, and the presence of many rulers. In some cases, sources (Coria 1937:559) document the presence of two calpixque. Calpixque, however, were not heads of towns such as a tlatoani; yet the presence of two in one town, each from a distinct ethnic group, points to some sort of power-sharing, where each ethnic group had a representative in government affairs. This information indicates that each barrio or small settlement had a representative who ruled together and/or formed short alliances with his peers. Etzatlán, in the southwestern Tequila valleys, was a province (provincia) headed by a señor and organized “a barrios,” each with its own representative (Weigand 2013:33–34). Weigand (2013:31) considers significant the fact that Etzatlán was organized “a barrios” and not “de barrios,” the difference being that the use of a barrios signifies that the barrios were not continuous, but rather dispersed and de barrios refers to a single settlement subdivided in neighborhoods (Weigand 2013:31). A recent regional survey in Etzatlán recovered settlement-pattern data, indicating a discontinuous settlement rather than a centralized capital (Heredia Espinoza 2014).

Finally, a ceramic stamp, from Etzatlán, portrays two defleshed faces or skulls that may represent regional variants of Xipe Totec and Tezcatlipoca, respectively (Weigand 2013:28). Tezcatlipoca has been suggested as an important deity in the region (Weigand and García de Weigand 1995:23, 65–66), perhaps originating in the Classic period. He is closely identified with the desert north and nomadic or seminomadic groups, such as the Caxcanes, who were encroaching on the Tequila valleys sometime before the Conquest. Caxcanes spoke some sort of corrupted Nahuatl that was intelligible with Nahuatl from the Basin of Mexico (Yáñez Rosales 1994:60). Nahuatl has been identified as one of the many languages spoken among the resident populations in the Tequila valleys (Baus de Czitrom 1982:25). As I noted above, Tezcatlipoca is associated with merit, judicial equality, self-sacrifice regardless of social position, and egalitarianism. He is associated with Atecpanecatl, the “destroyer of royal privilege,” in opposition to Quetzalcoatl, who is associated with royal lineages and privileges (Fargher et al. 2010:241; Ringle et al. 1998). As such, the ideological association with Tezcatlipoca favored social mobility over injustice, aggrandizement, and strong hereditary power. Therefore, the promotion of his cult and associated ideology would have proved the basis for the type of egalitarian corporate cognitive code that I suggest was a key aspect of the region’s political organization.

Archaeological Data

Settlement Patterns

In an area of 463 km², we recorded 242 sites that date to the Postclassic (figure 4.1), and 119 sites with standing architecture. Although population densities may have been lower here than in Central Mexico or Oaxaca, a conservative population estimate suggests that tens of thousands of people occupied the Tequila valleys at the Conquest (Weigand 1993:130) and nearly 15,000 lived in the area surveyed (including Santa María but not Etzatlán) based on a conservative figure of 17.5 persons per hectare, but debate exists on how densely populated these valleys were at the time (Weigand 1993:128–130). An increase in number of settlements and in the overall size of sites from previous occupation phases along with data on residential architecture and domestic debris are in line with these observations.

Figure 4.1. The central valleys of Jalisco, showing the area surveyed, settlement patterns for the Postclassic, as well as various localities mentioned in the text.

Late Postclassic sites are located on hills, lomas, and along the Río Santiago canyon (figure 4.1), but the largest centers and political seats tend to cluster around the natural entryways into the valleys and toward the northern piedmont of the Tequila volcano. The canyon, which limits the valleys to the north, was occupied with small sites on hilltops and lomas, but no large site was recorded in this subregion. To the northwest is Santa María and to the southeast and into the Atemajac Valley is a cluster of four large sites. Etzatlán and Tala are mentioned in ethnohistoric documents as important centers in the Late Postclassic (Coria 1937; Tello 1968). The Tarascans ravaged Tala just 50 years before the Conquest (Tello 1968), and it was located on the southeastern entry. Etzatlán on the other hand, was located on the southwestern passage. Recent archaeological survey identified a large and sprawling site on a piedmont ridge above (south and west of) the modern town of Etzatlán that was probably the settlement of the same name mentioned by the Spanish. All of these are strategic locations for vigilance and to control access, themes of primary importance during this period.

Based on site size as a proxy for the political administrative hierarchy, I have identified at least four settlement tiers (figure 4.2). The first tier consists of one large site (Santa María) at the northwestern corner of the survey area (212 ha). The second tier is composed of six sites that range from 17 ha to 75 ha, located in the southern half of the survey region, including four clusters of sites in the southeastern corner. A rough population estimate of 10 to 25 persons per hectare indicates that together these centers housed approximately 7,000 individuals. The third tier consists of secondary centers ranging in size from 4 ha to 16 ha (10 people per hectare). The lowest level consists of dispersed rural settlement, artifact scatters, and isolated residences (5–10 people per hectare). Political seats are surrounded by secondary centers and rural settlement in compact clusters separated by shatter zones, or scantily occupied areas (figure 4.2).

Figure 4.2. Postclassic settlement tiers, based on site size.

The settlement pattern, evident in the map, suggests a highly decentralized or fragmented landscape (figure 4.2). There are two shatter zones, suggesting the presence of multiple polities (possibly three). In the southeast, four sites are clustered together, suggesting one polity, then a shatter zone of 6.5 km separates this cluster from another with two large sites at its center that may also represent another polity. The latter cluster is located near the center of the survey area. In the northwest, Santa María stands as a third polity. Unlike the two other polities, settlement at Santa María is highly nucleated, nearly everyone lived in the capital and there was little or no rural settlement. This site corresponds with the settlement of Xochitepec (home of Guaxicar), now called Magdalena, described in ethnohistoric documents (Tello 1968:128). South of Santa María at the southern end of the Magdalena Basin, the archaeological site of Etzatlán forms the seat of yet another autonomous polity. Thus, I interpret these settlement clusters, divided by shatter zones, as politically independent units, which concurs with the fragmented political landscape described in early colonial documents (summarized in Soto de Arechavaleta 1994:346).

At 212 ha, Santa María is the largest Late Postclassic site. Based on survey data, it functioned as a political seat in the northern Magdalena basin. Only two small sites (< .4 ha) northeast of Santa María were recorded in this portion of the valleys, indicating a strongly nucleated, primate settlement pattern for the polity. This cluster is separated from the nearest cluster to the east by a significant linear distance of 25 km.

Of the six remaining large sites located north of the volcano, three of them do not have standing architecture due to their location in agave fields and toward the center of the valleys. It is my contention that these places were similar to Santa María, but mechanized agriculture has done away with extant architectural remains. This conclusion is based on our field observations, which recorded both construction stone and domestic debris on these sites. Besides Santa María, the only other primary site with standing architecture is located in the southeastern corner of the survey area on a smoothly sloping loma. Only two large residential platforms, superstructures, and stone alignments were visible on the surface, significantly less impressive than Santa María (see below). The loma has been affected by the contemporary urban growth of town of El Arenal (figure 4.1) and most of the site is currently planted with agave. Therefore, the possibility that many more structures existed is very likely. All of these primary sites are located within four to eight linear kilometers of one another, indicating dense populations in a politically decentralized landscape.

Intrasite Architectural Arrangements

Previous studies in various regions of Mesoamerica have tested the degree of centralization and differing political strategies through architectural spatial configurations at the site level (Blanton 1989; Blanton and Fargher 2012; Fargher et al. 2011; Fargher et al. 2014; Heredia Espinoza 2007). Generally speaking, network strategies (Blanton et al. 1996) are correlated with the presence of large palatial edifices, as well as small and enclosed (private) plazas that are, at times, directly associated with civic-ceremonial spaces. Conversely, corporate strategies are often linked with open and accessible civic-ceremonial architecture; the presence of multiple, equally ranked public spaces or civic-ceremonial architecture point to community integration at multiple levels and power-sharing strategies (see Beekman, chapter 3, this volume; Fargher et al. 2011).

For major sites in the Tequila valleys north of the Tequila volcano, it is important to identify those potential spaces where public events would have taken place as well as their location within a site, since ritual is an important aspect in the organization of sequential hierarchies. Plaza sizes are also important because they indicate the potential number of people who could have participated in various gatherings (e.g., large for public vs. small for private events). At Santa María, plaza sizes range from about 125 m² to 835 m². Some plazas are not formal, meaning they are open spaces that do not have walls nor are they delimited by buildings on all or some sides.

For the purposes of this section, I concentrate mostly on Santa María, which, I argue is exemplary of the architectural arrangement of polity capitals, and where archaeological preservation is optimal. A recurrent residential architectural form, known as corrales, prevails in most Postclassic settlements. As yet, there are no excavated corrales, but we know that they had a limited distribution following the canyons to the north (Weigand and García de Weigand 1996:51). During salvage work at Bugambilias, in the Atemajac valley to the east, Galván Villegas (1983) recorded several of these structures; and Hers (1977) reports on them in the Sierra del Nayar in Nayarit.

At first glance, corrales seem to conform to a regular construction style; two rows of stone blocks parallel to each other (between 40 and 60 cm apart), forming a rectangular C; but the survey recorded more information on their varied morphology. For example, corrales can be built directly on the surface or atop platforms and terraces. Most have a single entrance, but others have two or three entrances and staircases/steps, and therefore are of substantial size. Weigand and García de Weigand (1996:61) had already noted somewhat of a hierarchy of corrales in this area, yet they seem to be built following general guidelines. We think they were either special-function architecture or the residences of principals within a settlement (as defined by Blanton and Fargher [2008:19], principals are “the chief decision-makers and policy-makers of a polity). Sizes of corrales range from 12.5 m² to 772.8 m², following a lognormal distribution. The geometric standard deviation (1.852 m²) indicates significant variation, and given the geometric mean (56.023 m²), there are some very large corrales that appear to be outliers. The largest corrales are found in primary centers. Differences in size may also be indicative of stratification within the site such that larger corrales housed important, perhaps extended, families, while smaller structures served as residences for commoner households, possibly attached to the larger corrales. The importance of corrales to this discussion lies in the fact that this type of architecture is ubiquitous throughout the survey region, and it is to some extent a diagnostic feature of the Late Postclassic period.

In order to understand the architectural spatial patterning at Santa María, I drew Thiessen polygons between all standing residential architecture greater than 8m², which allowed the easy identification of architectural clusters (figure 4.3). The polygons indicate that architecture clustered in groups consisting of several structures, small plazas, mounds, large corrales, and other features of unknown function. These architectural groups tend to be spread evenly across the site, pointing to a decentralized patterning within the site where several formal arrangements can be detected. These architectural clusters are the material representation of larger units (e.g., such as groups of households or neighborhoods).

Figure 4.3. Santa María architectural settlement clusters.

Next, using structure areas as a proxy for labor investment (person days), I identified those places within the site that point to high amounts of labor (figure 4.3). Cluster analysis shows several locations where labor investment is highest, indicating that labor-intensive architecture is located in multiple locations within Santa María (a heterarchical pattern in the architecture). High amounts of labor investment concentrate at the highest point of the hill, where I mapped a three-mound group surrounding a small patio atop a high platform (figure 4.4). Terraces and structures run northeast and southwest of this complex. To the west, there is an abrupt slope and from there the view of Laguna Magdalena is impressive. While we collected artifacts on the slopes around this three-mound group, we did not find a single artifact associated with it that suggested a nonresidential or civic-ceremonial function. This architectural group is located on one of the highest points of the mountain and is located in a cluster, which includes a number of large corrales, terraces, and structures. Its location at the center of a neighborhood cluster probably indicates that it functioned as the neighborhood civic-ceremonial space, as well as the central architectural complex of the site. Hence, this architectural group can be interpreted as a unit formed by both residential and nonresidential architecture (e.g., civic-ceremonial).

Figure 4.4. Three sections of Santa María, showing various residential groups: highest point of site.

In the middle section of the site, spatial arrangement of the largest corrales and other residential architecture follows a pattern where open spaces or plazas are surrounded by groups of three or four large corrales, smaller corrales, residential structures, and terraces (figure 4.5). These groupings are separated by natural topography such as drainages, constituting discrete or separate residential zones. In the northern portion of the site, several discrete groups of residences, divided by natural drainages, are arranged in circular fashion around a large plaza. The plaza is located on an open area high on the hill and at the very center of the above-mentioned residential aggregations. Given that these groups are divided by natural topography, they suggest discrete units (groups of households); thus, it appears that these units shared a large plaza as a common public space. However, no single group was significantly larger or more impressive than any other, suggesting relatively equal ranking among the groups that participated in public events at the central plaza.

Figure 4.5. Three sections of Santa María, showing various residential groups: hill.

On the piedmont, the arrangement varies somewhat, and groupings are not as evident, but again clusters are divided by natural drainages (figure 4.6). In this section of the site, some residences and corrales are arranged in groups of two or three. Spatial proximity to other residences is high, so the pattern is more clustered than on the highest portion of Santa María. A mound associated with a plaza and/or large corrales repeats in various places on the piedmont. Mounds have few to no artifacts on the surface, but fragments of ritual vessels (braseros, sahumadors, etc.) are found in association with these structures; thus I interpret these features as civic-ceremonial architecture. Mounds may also function to separate residential groups or create independent residential groups. In one case, we recorded a large corral and other residences associated with a mound and a small open plaza (figure 4.6). The mound and the plaza do not have any associated artifacts that indicate a nonresidential function, whereas the large corral and the surrounding residential architecture display a high density of domestic debris such as manos, metates, comals, and jars. On a different hillside, a large corral or platform is directly associated with a mound, and a “double U” shaped structure that may have functioned as a public space (figure 4.6). Finally, in another section, a large corral is associated with one of these “double U” structures.

Figure 4.6. Three sections of Santa María, showing various residential groups: piedmont.

Thus, the entire site displays multiple clusters of equally ranked civic-ceremonial and residential architecture, indicating distinct units based on spatial breaks between clusters. The ancient settlement at Etzatlán suggests variation on this pattern where there is not a single settlement focus and instead people lived in multiple separate clusters of settlement distributed along the piedmont. However, this pattern may be similar to Santa María in that these distinct groups or sites may be equivalent to the clusters. In addition, all large Postclassic sites feature debris patterns that suggest the presence of public architecture, a pattern not seen at smaller sites. Besides the three-mound group mentioned above, there are two large open plazas surrounded by residences, resulting in small groups of structures analogous to neighborhoods. Further down on the piedmont, the spatial structure of architectural arrangements follows a similar pattern; houses are situated in groups of various sizes where large corral structures were recorded, some of which are grouped around patios and yet others were constructed on terraces in the sloping terrain. Thus, it appears that the site consists of unranked and relatively architecturally similar neighbors characterized by a group of one or a few large houses, small houses, and public spaces.

Spatial Distribution of Artifacts

Our field methods did not implement systematic, controlled collection points of a specific size, but our survey methodology was systematic. We made collections in specific features and in small artifact concentrations within the site in order to have better control over chronological components and over intrasite artifact distributions. Collections included only diagnostic artifacts such as rims, decorated vessel bodies, and figurine fragments. In the absence of diagnostics, and where ceramic densities were too low, we collected undecorated sherds and dated them based on paste attributes. Hence, although a statistical sampling method was not implemented, some generalizations on the distribution of artifacts for the entire survey region and within Santa María as well can be made based on these collections.

Differences in the quality of artifacts in a spatial dimension are one way to identify possible variations in wealth and status. Artifacts, especially ceramic artifacts that display high amounts of labor investment (Feinman et al. 1981; Garraty 2000; Hagstrum 1988) and specialized goods that require complex technologies or secret procedures, can be used as proxies to study differential access to wealth within and among sites. Differential distributions of these goods within settlements as well as among sites (Fargher et al. 2010; Heredia Espinoza 2007; Kowalewski et al. 2008; Uriarte Torres 2011) can effectively identify the degree of difference in access patterns to a diversity of goods and resources as well as to possible discrete functional spaces, and they shed light on the variety of strategies used by individuals or groups.

Postclassic diagnostic ceramics can be reduced, but not limited to, three main forms according to their frequency in our collections: jars, molcajetes, and tablets (figure 4.7a, b, c, respectively). Jars vary in size, but in terms of form they do not differ significantly. Typically, they display a distinctive, thickened rim (Beekman and Weigand 2000:68–69) and they tend to be painted in red and highly burnished, although decoration in red and white vertical bands also occurs. Molcajetes are a recurrent form. They are striated on the interior, have supports, and display diverse decorations (monochrome, bichrome, or polychrome) and vary in ceramic fabric texture. Molcajetes are very common and they do not seem to be limited to domestic contexts, as they have also been identified in interments (Glassow 1967; Smith Marquez and Herrejón Villicaña 2004). Finally, tablets are flat and rectangular with raking or striations on the upper surface; their function is unknown. They do not show wear patterns and, given their shape, they were not used for food serving or food preparation. Again, they do not seem to have restricted distribution and they are not confined to a specific context. Surprisingly, bowls and plates, among the most common forms in other parts of Late Postclassic Mesoamerica (Brumfiel 1987, 1991; Garraty 2000), occur rarely in the Tequila valleys’ Late Postclassic ceramic assemblage (Beekman personal communication 2012; Heredia Espinoza 2011).

Figure 4.7. Late Postclassic ceramic diagnostic forms: (a) Atemajac jars; (b) Huistla molcajetes; and (c) tablets.

Overall, frequencies in ceramic forms and ceramic types do not show significant differences across the surveyed region or in areas within sites, suggesting a highly homogeneous distribution. Two forms dominate the entire collection: jars and molcajetes. Together, these two make up 85 percent of the entire assemblage. The proportion of jars to molcajetes is 2:1. Other forms that head the list, but that occur in significant lower proportions, are comals and, still less frequently, bowls/plates.

Competitive and aggrandizing feasting behavior is usually inferred by the presence of fancy serving ware and a wide variety of ceramic forms, which suggest consumption of diverse and elaborate foods, in a few (elite) households (Garraty 2000; Levine 2011; Smith et al. 2003:251). Specifically, the ratio of jars to bowls (11:1) indicates that serving for aggrandizement and political legitimation was not an important pursuit at this time; furthermore there is no evidence that bowls were concentrated in high-status residences (e.g., in large corrales). Cooking and storage were more important and, in the absence of highly decorated or luxury serving forms, the ceramic assemblage is more indicative of communal types of food-sharing behavior. Thus, the near absence of fancy serving vessels, the lack of diversity in vessel forms, and the low frequency of serving vessels, suggests that competitive feasting was limited. I think that such archaeological patterns indicate that cognitive codes downplayed the importance of personal wealth aggrandizement and gifting as a pathway to political power. This pattern fits with the ideological focus on Tezcatlipoca and his cult, mentioned above.

Moreover, two ceramic types dominate the assemblage: Atemajac and Huistla that together represent over 93 percent of the entire Late Postclassic collection. Both types are ubiquitous and represent the bulk of the domestic pottery in all households. Furthermore, these types do not show high amounts of labor investment in decoration, but the Atemajac type tends to be well burnished. Various pastes were used in the production of these two types, but on average pastes seem midway between a fine and a coarse clay (Beekman and Weigand 2000:68), suggesting the use of piedmont clays with little additional labor investment (e.g., they did not levigate and carefully temper clays). Accordingly, Atemajac vessels for the most part are plain, but examples of polychromes appear in small quantities. On the other hand some Huistla type vessels may have fine pastes and bichrome or polychrome decoration (Nance 2013:104–107), indicating the presence of some more finely made dishes. However, in the survey the highly decorated sherds do not constitute a large enough sample to be significant, and they were not disproportionately concentrated in and around large corrales.

Several conclusions can be drawn from the ceramic evidence. First, there are no major differences in types, suggesting a collective form of material culture and perhaps the presence of exchange mechanisms that cross-cut political boundaries, such as a regional market system (tiangues are mentioned for several towns in Coria 1937). The similarity in local types with limited decoration also indicates a lack of competitive behavior such as in feasting events, and it also concurs with the presence of markets. The limited ceramic forms indicate standardized food-consumption patterns and again no major differences that may indicate considerable distinctions in behavior (consumption and wealth). Finally, the data described above show that big residential units (e.g., corrales) did not monopolize access to better and fancier pottery or consume significantly larger amounts of pottery, especially serving vessels, than more modest households. The architectural clusters attest to a greater investment in architecture, but not in portable wealth.

Sequential Hierarchies, Communal Rituals, and Egalitarian Ideologies: A Summary of the Evidence

At the regional scale, the spatial distribution of sites indicates a fragmented political landscape, with small-scale, relatively autonomous polities divided by shatter zones. Each polity apparently consisted of a dominant center surrounded by lower settlement tiers. Each center was internally organized around multiple semiautonomous and unranked residential groups or neighborhoods. Each neighborhood seems to have focused on a small plaza and to have shared access to a large central plaza at the center of the site. Thus, the heterarchical organization of residential groups gives the impression that they were integrated through mechanisms analogous to sequential hierarchies and not simultaneous hierarchies. Each residential group in turn had its own civic-ceremonial architecture, where important public ceremonies may have taken place, yet they shared a common plaza accessible to all residents of the site.

Consistent with a sequential strategy, evidence indicates that corporate ideologies downplayed the importance of personal wealth accumulation and networks of power. These corporate ideologies were personified in the worship of Tezcatlipoca and the adoption of his cult, which are evidenced in figures and ethnohistoric documents (in Weigand 2013:28). Furthermore, these corporate ideologies were materialized in artifact inventories, especially ceramics, which were highly homogeneous within sites and across the region. Moreover, elaborate serving vessels comparable to Tlaxcallan polychromes or Aztec III Black-on-Orange occur only rarely in the artifact assemblage and were not disproportionately concentrated in large corrales. Given the political situation in the Late Postclassic, the societies occupying the Tequila valleys adopted corporate strategies and a corporate cognitive code that provided for the effective mobilization of the population for defense against attacks from outside forces (e.g., Tarascans), but limited the degree to which warrior-leaders could monopolize material and ideological resources for their personal aggrandizement and power (see Earle, chapter 14, this volume).

Conclusions

The introduction of dual-processual theory into archaeological theory has opened a wide door for the explanation of the evolution of societies along multiple paths toward complexity. Research in Jalisco has benefited from these new insights and is already producing important evidence for the coexistence of diverse political strategies at certain points in time (e.g., Beekman 2008; Beekman, chapter 3, this volume). Whereas information was limited in the recent past, we now have assembled an important corpus of data that is throwing additional light onto the ways people organized themselves in the Tequila valleys during the late prehispanic era. Here, I have made a first attempt to characterize the Late Postclassic political organization through regional and intrasite spatial organization, architecture, and portable material culture. Rather than simply fitting these societies into known societal types or political forms elsewhere in Mesoamerica, I ventured to think about how these different ethnic groups organized effectively without centralizing power into a single polity or lineage in response to military threats from the Tarascans and the Caxcanes. The information available indicates an emphasis on differences in organizational strategies, implying an alternative pathway to complexity. In this case, sequential hierarchies proved a useful tool for disentangling these strategies, which combined corporate political strategies and collective action to achieve complexity without centralization. In addition to providing an alternative model for a more egalitarian political structure, sequential hierarchies enrich corporate theory because they help operationalize this theory for archaeological research.

Acknowledgments. This project was supported financially by El Colegio de Michoacán, A.C., the Foundation for the Advancement of Mesoamerican Studies, Inc. (FAMSI, #07012), National Geographic Society (#8711–09), National Science Foundation (BCS-1219619), Consejo Nacional de Ciencia y Tecnología (#131056), Secretaría de Cultura del Estado de Jalisco, and three municipios in Jalisco (El Arenal, Amatitán and Tequila) provided small funds for the project. I would also like to thank a very dedicated group of archaeologists and students of archaeology from different institutions in Mexico and abroad. Thanks to Lane F. Fargher and Christopher S. Beekman, who have been involved at various stages of the project throughout the years and have provided helpful advice in the development of the survey and laboratory stages. Stephen A. Kowalewski and Beekman read an earlier version of this chapter and provided useful comments. However, all omissions and errors are my responsibility.