7
Ritual Practice at the End of Empire

Archaeological Context

Pacific coastline topography of south coastal Peru’s Ica province has resulted in a drainage pattern for the Río Grande de Nasca Basin that is unique in the Central Andes. Impeded in their journey to the ocean by an ancient coastal range, the basin’s rivers merge on the desert coastal plain to form a single stream before emptying into the ocean (Garayar et al. 2003:127–30). As a result, and unlike the North and Central Coasts where human occupation spreads out along the littoral as well as along the river valleys, settlement in the Río Grande de Nasca Basin is focused well inland of the shoreline and nearer the Andean slope. A general pattern of decreasing discharge from north to south is observed for the rivers of the basin, resulting from smaller headwaters and fewer tributaries for the southern streams (ONERN 1971:265). The northern rivers—Grande, Palpa, Viscas, and Ingenio—tend to flow year-round, whereas the southern Aja, Tierras Blancas, Taruga, and Trancas have low water volumes and intermittent flow. Farming and permanent settlement of the middle stretches of these valleys is made possible by an ancient system of filtration gallery aqueducts, knows as puquios, that were likely first built during the mid- to late Early Intermediate Period (after ca. AD 450) (Schreiber and Lancho Rojas 1995:250). In contrast, river water is generally available in the upper portions of these valleys above about 1,200 meters above sea level (masl).

The Río Grande de Nasca Basin is well-known to archaeologists as the home of the Nasca culture (ca. AD 1–650). Although Nasca was interpreted as a pristine state by early researchers, current evidence suggests a loose confederation of chiefdoms, with regional integration between valleys focused on a shared ideological universe during earlier times that appears to have shifted toward increased competition, strife, and even warfare during the Middle and Late Nasca phases (Schreiber 1998:263; Silverman and Proulx 2002:253; Vaughn 2009:3–5). Nasca religion and ideology is expressed archaeologically in the ruins of important ceremonial sites like Cahuachi, the lines and geoglyphs the Nasca etched onto the desert plain, and their elaborate ceramics—decorated with complex designs executed in a rich color palette made from mineral pigments (Vaughn et al. 2007:18). The latter influenced the development of Wari culture in the highlands, and a close technological and iconographic relationship between Nasca and Wari ceramics has been recognized since the early days of Wari scholarship (Knobloch 1991:248–50; Menzel 1964:9–10).

Pataraya is not the only Wari site in the Nasca drainage. The much-better-known site Pacheco is located along the Nasca River about 30 km to the west and downstream of Pataraya on the coastal plain. Unfortunately, that site, which was excavated in the 1930s and yielded many of the diagnostic ceramics that led to the early identification of Wari styles, has been almost completely destroyed by modern agricultural activity (ibid.:23–25; Schreiber 1999:168–69). We also recorded another site, known as Incawasi, during project surveys of the high-elevation headwaters of the Nasca River’s other main tributary, the Aja—about 20 km northeast of Pataraya (Edwards and Schreiber 2014:226–28; Schreiber and Edwards 2010:161). In addition to this presence in Nasca, Wari architecture and artifacts are found throughout Peru. This observation led the period of Wari influence in the Andes to be demarcated in the standard Andean chronology, one of New World archaeology’s oldest, as the Middle Horizon—even before the site of Huari, the namesake for the stylistic horizon, was identified as the center of diffusion for Middle Horizon cultural markers (Isbell and McEwan 1991:6–7). While Wari sites in the Huamanga Basin of northern Ayacucho, such as Huari itself, have earlier antecedents, Wari sites outside the core are intrusive, unlike anything preceding them in their respective regions, leading most scholars working at these sites to interpret them as administrative outposts of an expansionist state centered in Ayacucho (e.g., ibid.; Lumbreras 1974; Menzel 1964, among others). Such is certainly the case for Pataraya, which was planned and built all at once on previously unoccupied land and inhabited by ethnically foreign residents for virtually the entire span of Wari dominance in Nasca and the Andes, then abandoned as that dominance waned and the Wari system collapsed (see Santley, Yarborough, and Hall 1987 for a discussion of such ethnic enclaves). The events at the moment of this abandonment are the subject of the remainder of this chapter.

Archaeology of Pataraya

Wari Pataraya consists of a main enclosure that measures roughly 20 m on a side and is defined by a double-faced, rubble-filled masonry wall with two additional rectangular enclosures appended to it (see figure 7.2). The square space inscribed by the enclosure wall is divided into four quadrants that themselves are subdivided into patios surrounded on their perimeters by long, narrow rooms called galleries—some of which are further subdivided to form smaller rooms—and corridors. Analysis of wall joining patterns indicates that the site was built according to a preconceived plan in which the square outer enclosure wall was built first, then it was divided into four quadrants, and, finally, each of these quadrants was subdivided into patio groups. Two rectangular constructions were added to the main block and match its orientation: a large enclosure abutting the structure to the east and another, smaller one to the north. Construction technique for the larger eastern addition is as robust as the main enclosure, and it is also incorporated into the site’s overall system of communication, whereas the smaller addition has low walls and is the only sector that can be entered without first passing through the main enclosure’s single entrance.

To evaluate the strength of the inferences made here, it is important to highlight two salient points about site formation: stratigraphic deposition and spatial organization, paying attention to the “taphonomy of ritual evidence” (Rick, this volume). Because of its location on an arid valley bench above the narrow Tierras Blancas River, soil formation at the site is practically nonexistent, as are disturbances from vegetation and burrowing animals. The resulting stratigraphic sequence is therefore relatively simple, consisting of the natural basal layer that underlies the landform, the cultural strata, and a post-abandonment cap of wind-deposited sediments. Except in midden, the cultural strata consist of thin layers of artifacts, features, and, in some places, soot atop prepared floors and sub-floors. A thin layer of sand, which appears to have been taken from the nearby river, covers the occupation level inside Pataraya’s architectural core and marks a clear cessation of Middle Horizon cultural activity at the site. Because all the evidence for cultural activity at the site is compressed into a thin artifact-bearing stratum, inferences about the sequence of events attending abandonment depend largely on Pataraya’s architectural plan and, specifically, the consequences of that plan for communication between sectors of the site.

As mentioned, Pataraya’s architectural plan closely follows the tenets of the Wari architectural canon, and excavations at other Wari sites have demonstrated a profound concern with controlled and restricted access by Wari architects and builders (Spickard 1983:139–40). Because a large percentage of Pataraya’s living surface could be exposed as a result of the enclosure’s relatively small size, we have a complete understanding of access into and communication within the enclosure—a prerequisite for effective access analysis (Cutting 2003:5). The main enclosure at Pataraya has a single entrance located along the north wall (see figure 7.2; table 7.1). Entering the enclosure, the visitor travels south along a walled corridor (Corridor A) toward a blind corner before making a 90° left turn to the east. After this turn an entrance into Patio Group A opens on the right; a little further along, another entrance into Patio Group C opens on the left. From here, Patio Group A serves as a node for access to the rest of the main enclosure. Patio B can be entered using a doorway in the wall that separates the two adjoining sectors. In the southwest corner of Patio Group A is another doorway. This one enters Corridor B, formed by the space between the main enclosure wall and the wall that encloses Patio B. A visitor walking this corridor travels almost 50 m around the perimeter of Patio B, making three blind turns, before entering Patio Group D. For the purposes of this chapter, the most important point of this discussion is that this pattern of access allows extremely limited choices for travel within the site (for a more in-depth discussion, see Edwards 2013:568–74). Specifically, from Patio Group D, in space syntax terms the “deepest” part of the site, one can only exit the enclosure by passing through Corridor B, Patio Group A, and Corridor A (see discussions in Hillier, Hanson, and Graham 1987:364–65; Stockett 2005:389; Vega-Centeno Sara-Lafosse, this volume, for more information on space syntax analyses).

Use of Space at Pataraya during the Middle Horizon

Use of space within the enclosure is highly segregated, matching the hierarchical system of internal communication just described (see figure 7.2). Excavations in Patio A and the galleries of the southeast quadrant revealed the site’s only evidence for extensive cooking and food preparation, as indicated by a thick secondary deposit of domestic and cooking fire refuse in the east gallery, a kitchen and food storage area or granary in the south gallery, and an area for milling grains found in the corner of the patio formed by the two galleries. A large groundstone slab, or batán, found here is the only one recorded at the site, further indicating centralized food production (see Goldstein 2008:39; Nash 2002:51). There is also evidence that guinea pigs, or cuy, a common source of animal protein and fat in the Andes, were raised in Patio Group A as evidenced by apparent pens and an abundance of coprolite in flotation samples (Andrews 1972:129; Rosenfeld 2008:128).

The most substantial and dense site furniture was found in the northern gallery of Patio Group C, located immediately across from Patio Group A in the northeast quadrant of the enclosure. This furniture consists of two freestanding platforms on one end of the gallery and a substantial bench on the other. Artifacts and macrobotanical remains recovered here, while similar to those found in Patio Group A, suggest a different function for Patio Group C and its two galleries. Food remains were recovered here, as were sherds from serving vessels, but evidence of food preparation is absent. Broken but largely complete decorated vessels recovered from wall rubble suggest that they rested in niches. In addition to food remains and culinary equipment, spindle whorls (though not as many as in Patio Group A) and a comparatively high density of lithic debitage found in the patio suggest that domestic activities occurred here as well, but the focus may have been related to other activities including administrative duties, as evidenced by the elaborate furnishings of the north gallery.

The northwest quadrant of the enclosure—Patio Group D and its two galleries—can be considered the most remote area of the site (see figure 7.2). Although somewhat impacted by a brief reoccupation of the space that occurred during the Late Intermediate Period, the site’s best-preserved and most elaborate artifacts were recovered here. Most of these artifacts were recovered from a cache of decorated serving vessels, described in more detail below, found in a small room at the east end of the north gallery. Other richly decorated vessels were recovered from disturbed contexts in the main gallery and from wall fall, again suggesting the presence of niches. Unlike the other patio groups just discussed, this area has no furniture and very little evidence of domestic activity other than spinning, which appears to have been a nearly ubiquitous activity at the site (Edwards, Fernandini Parodi, and Alexandrino Ocaña 2008:91). Other than these materials and the intrusive reoccupation of this sector of the site, the area is devoid of artifacts or features, suggesting a residential area—an interpretation echoed by its extremely private setting within the communicational configuration of the site.

The southwest sector of the site, Patio B, has a unique architectural plan in addition to its specialized function. Organized as a single patio rather than a patio group, the interstitial area between the patio and Pataraya’s perimeter wall that forms the gallery rooms of the other three sectors is instead used to create a narrow and circuitous corridor providing access to Patio Group D. The space also encloses a white, hard plaster floor that, other than the ubiquitous thin layer of sand used to bury the site at abandonment, was left to the archaeological record virtually spotless. A thick layer of clean fill raises and levels the surface of the plaster floor in what appears to have been a low spot in the original construction surface. The cloth-wrapped skeleton of a juvenile camelid was found below the fill in a shallow pit excavated into the subsoil—similar offerings have been documented in other Wari contexts (see, e.g., Cook 2001:147; Finucane 2005:14). These details, alongside the patio’s proximity to the kitchen area of Patio A, suggest an area dedicated to commensal activity in which special ceremonies, rituals, parties, feasts, and other politically important activities—probably involving visitors to the site—occurred (Bray 2003:1–2; Dietler 2001:66–67; Jennings and Bowser 2009:4–9).

Excavations in the two rectangular additions appended to the main square recovered only generalized, secondary deposits of refuse—presumably originating in the main areas of the complex—from the more substantial of the two additions (E). However, a number of unique features were uncovered in the center of sector F, the low-walled rectangular addition that extends to the north. These features consist of two conjoined and well-built cylindrical stone masonry structures and a stone-lined channel to the south. While the channel has not yet been fully explored, this collection of features appears to have formed a drainage or water distribution and storage system. Artifact deposits suggest that the area was also used for various chores and industries, perhaps focused on those that require more space and open air than would have been available inside the enclosure. Most notable is chipped stone tool manufacture, which, while lightly ubiquitous throughout the site, may have been particularly concentrated in this area. A square stone-lined bin filled with obsidian waste flakes was found near the end of the field season, which could have been a receptacle for the hazardous remnant shards of volcanic glass.

Features Related to Pataraya’s Abandonment

Ten features were recorded throughout the main enclosure that are stratigraphically contemporaneous with each other and with the thin cap of clean river sand that represents the last Middle Horizon cultural stratum at the site (see figure 7.2; table 7.2). Two of these are large stones, measuring approximately 0.6 m³ and weighing ca. 50 kg, obstructing Corridors A and B. As discussed, these corridors provide all of the communication between sectors of the main enclosure. Because of the fact that once Corridor B was blocked, Patio Group D could no longer be entered, and once Corridor A was blocked, the entire enclosure was inaccessible, the placement of these obstructions suggests timing and directionality for the inferred events signaled by the remaining eight features. The first of these was also arguably the most dramatic find at the site, a cache of nine whole ceramic vessels carefully placed upside down atop the prepared floor of a small room in the farthest northwest corner of Patio Group D and buried in a matrix of white river sand and Spondylus (sp.) shell artifacts. Eight of the nine vessels are fine Wari or local Middle Horizon (Loro) styles; the remaining vessel is a miniature cooking pot. Spondylus shell is native to the warm Pacific Ocean waters off modern-day Ecuador and is known as a valuable trade item throughout most of Andean prehistory (Paulsen 1974:603–5).

Diagonally opposite the cache, against the wall in the southwest corner of Patio Group D, we found a pair of features consisting of a small charcoal/ash stain and a complete but smashed Wari-style ceramic jar. The ash feature appears to have been created by a single, small, low-intensity fire likely used to burn some kind of small offering, as modern Andeans are known to do historically and ethnographically with coca leaves, llama fat, corn, and similar items (Sillar 2009:374). Another smashed serving vessel/burned offering feature pairing was found alongside the edges of Corridor B. The offering was burned in the extreme southwest corner of the enclosure with the smashed vessel, a fine Wari-style cup, a few meters to the east along the passageway. A third smashed serving vessel/burned offering feature pairing was identified in the southeast corner of Corridor A. Finally, a great deal of smashed decorated pottery was found just outside the main enclosure in sector F, the low-walled rectilinear yard to the left of the entrance into the compound.

A thin layer of clean white river sand was also found covering most of the site’s abandoned surfaces, including the features just described. There is no discernible stratigraphic break between this deposit and the surfaces it covers; nor is there any indication that the sediment was deposited by natural means. A stratum of naturally deposited aeolian sediment that is clearly different in color and composition from the river sand covering the abandoned living surfaces generally caps the entire site, including the modern surface of the landform outside the structure. While areas inside and outside the structure are subject to the same post-abandonment deposition processes, mainly aeolian, the river sand was documented only inside the structure (see Gay 1999:281). The nearest source for the sand is the Tierras Blancas River bed, located 250 m away in a relatively steeply scored channel—suggesting that the sand was carried manually to the site.

Radiocarbon samples recovered from the site demonstrate that Pataraya’s abandonment is roughly co-terminus with generally accepted dates for the collapse of the Wari Empire. The latest radiocarbon sample recovered from Middle Horizon contexts at the site dates to cal AD 922 (median at 2σ, range is cal AD 807–990 [93.4%], μ = cal AD 912).¹ A clear cessation of significant cultural activity is evident after this date. While circumstantial, the close correlation between the latest dates from Pataraya and our best estimates for the collapse of the Wari Empire suggests that Pataraya’s abandonment was directly related to the demise of Wari power in the Andes generally and in Nasca in particular.

Discussion

This pattern of burned offerings, smashed pottery, intentional burial, and obstructed passageways appears to represent the final activities at the site as it was being abandoned, no later than AD 950 or so (see Capriata Estrada and López-Hurtado, this volume, for an interesting discussion of the relationship between terminal rituals involving intentional destruction and site abandonment). Evidence from throughout the Andes suggests that the Wari Empire had begun to show signs of stress by this time and had completely collapsed by the end of the eleventh century (Finucane et al. 2007 :591). Because all communication within the site was afforded by a limited number of corridors and entryways, the placement of the offerings and obstructions suggests a sequential order to the events reported here, possibly indicating a procession of some kind (also see Chicoine et al., this volume). I posit that the hypothetical ritual described in the remainder of this chapter is a good inference from the available data.

In this scenario the performers of this final ritual at Pataraya carefully placed intact ceramic vessels on the floor of the remote room in the northwest corner of the enclosure and buried them in selected sand and Spondylus shell. They then burned an offering and smashed a large decorated jar, possibly containing a ceremonially important liquid such as corn beer, near the doorway into Corridor B before leaving Patio Group D. When they reached the southwest corner of the corridor, another offering was burned and, a few steps away, another vessel, this time a decorated bowl, was sacrificed. The performers then exited the corridor and blocked further entry by placing a large stone to obstruct passage—while too small to effectively prevent reentry, its placement would have communicated the area’s closure to ceremony participants. A final offering was burned in the southeast corner of Corridor A, and yet more vessels were smashed. The performers then left the site for good by way of Corridor A and the enclosure’s main entrance, again obstructing the route as they left. Finally, the last Wari inhabitants of Pataraya may have gathered outside the enclosure for a final round of corn beer served in the smashed vessels found there. Burial of the site’s living surfaces in a thin layer of sand would also have accompanied this ritual performance. Such burial may express an attitude toward cleanliness related to culturally constructed concepts of how living and ritual spaces should be appropriately abandoned (Fernandini and Ruales, this volume). More labor-intensive efforts to bury sites at abandonment have been documented or can be suggested at other Wari sites (e.g., McEwan 2005 :53; Williams and Nash 2002). However, at Pataraya, a smaller resident population may have engaged in an effort that was more symbolic than actual—with the thin layer of sand standing in for a more formal and adequate burial. Similarly, the stones obstructing the corridors, which are not terribly large, were perhaps placed to stand in for the more formal closures of doorways and access routes seen at other Wari sites.²

Conclusion

Returning to the place where this chapter started, these final ritual performances at Pataraya fit into a wider anthropological understanding of ritual generally and the political transformations of Andean prehistory specifically. I conclude by briefly discussing them in terms of their political context, their participants, and the meaning they held for those participants. Although Pataraya was very likely abandoned as a consequence of the Wari Empire’s fading power and influence as it neared political collapse, the occupants of Pataraya do not seem to have been confronting those challenges, whatever they were, directly. While evidence from Wari sites in other regions suggests increased violence, vandalism, and ecological calamities, Pataraya appears to have been carefully abandoned and then quietly forgotten (cf. Arkush 2006:307; Tung 2008:115; Williams 2002:372). Pataraya is not the only Wari site to have been abandoned in this fashion, suggesting that the Wari collapse was not necessarily spectacular or sudden and may have been preceded by a period of strategic retreat (McEwan 2005:53; Moseley et al. 2005:17271; Williams and Nash 2002). Indeed, Pataraya’s Wari residents may have believed they were leaving Nasca for a greater good. Given this wider political context and the nature of the archaeological data, the ceremony described here was probably a largely private affair, conducted by and for the site’s ethnically Wari occupants. Echoes of similar ritual abandonments have been described at other known Wari sites, which suggests there was a Wari style or practice of closing sites. This is perhaps analogous to the lowering and folding of a country’s flag before its embassy is closed today. During such an increasingly uncertain time for their empire, such traditions may have been particularly meaningful to the Wari at Pataraya as they prepared to leave Nasca for the last time.