9
The Demise of the Ruling Elites

Terminal Rituals and Site Abandonment

The authors in this volume are consistent in highlighting the performative nature of rituals. If they can be understood as activities performed in a regular and systematic way, then, as Rick states in this volume, they should be interpreted primarily as a form of communication. We agree with Rick (this volume) that rituals are structured around “customary actions that are effective in obtaining outcomes over which the participants have little or no controlling power” and with Nielsen and his colleagues (this volume) that even though most of the specific actions that take place during rituals are designed to address non-human agents, they suppose the materialization of terrestrial outcomes. In this sense the mobilization of political expectations is a constitutive component of ritual activities in which certain social outcomes are expected as the result of a correctly performed ceremony.

Among the different social outcomes that can be expected as the result of particular ritual performances, the abandonment of public buildings is heavily related to political changes in the organization of past societies. This particular type of ritual is known in the literature as terminal rituals, deliberate and systematic performances that may or may not occur with the end of a site occupation where, per Stanton, artifacts or structures are destroyed, burned, or broken (Stanton, Brown, and Pagliaro 2008:234–36). Stross (1998:37) distinguishes between two types of ritual: dedicatory (animation) and termination (de-animation). The aim of animation rituals is for objects to acquire a soul, while terminal rituals are those associated with the killing of the object, which in some way liberates its soul. The distinction between these rituals can be recognized in the archaeological record by the presence of specific patterns in the disposal of artifacts. In the case of terminal rituals, Mock (1998:5) expresses that “termination actions, although difficult to separate in all instances and often embedded in dedication events, generally include the defacement, mutilation, breaking, burning, or alteration of portable objects (such as pottery, jade or stone tools), sculptures, stellae, or buildings.” Stanton and colleagues (2008:237) also mention the presence of deposits containing destroyed material culture, such as smashed ceramic pot caches, which may be associated with site or structure abandonment and may symbolize reverential ritual activity.

Some terminal rituals can be seen as a part of life cycles, like those cases associated with the end of a construction phase. The ritual destruction of artifacts or architectural features may indicate the end of a building episode and thus mark the beginning of a new phase. However, these events could also mark an end, without implicating any rebirth or regenerating process, as in the case of the abandonment of a particular building. Most of the examples of terminal rituals in Maya sites are presented as a continuum, usually associated with the construction of a new building, where “killing” the old structure was a means of containing its accumulated power (Mock 1998:10).

These rituals have been broadly studied for the Mayas, among whom dedicatory and terminal rituals were very common (Dahlin 2000; Freidel 1998; Harrison-Buck, McAnany, and Storey 2007; Mock 1998; Stanton, Brown, and Pagliaro 2008). Mock mentions several cases for the site Chalcatzingo in the Morelos Valley, where altars and other monuments were partially destroyed and the broken sections were sometimes moved to other areas of the same site. She suggests that these actions were performed to “neutralize [the] supernatural powers contained in the monuments and left uncontrolled by the ruler’s death” (Mock 1998:5).

Intentional destruction constitutes a powerful and symbolic act. However, in what social contexts should we expect to encounter this type of behavior? Although these actions are often related to social, political, or economic factors, in all cases they seem to have a strong ritual component as well. It becomes crucial, then, to determine if these events are related to the partial or total abandonment of the site. Abandonment may occur on different scales; it can comprise entire regions, settlements, or just a few structures within a site. In addition, abandonment can be gradual when it is planned or abrupt when it is unexpected, temporal, or permanent (Joyce and Johannessen 1993; Schiffer 1996). Settlements that have been abandoned abruptly will not usually contain great amounts of trash but may contain particular items that, under other circumstances, would have been transported during evacuation. The opposite occurs during planned abandonments.

The ritual nature of this action is highlighted by the fact that it was deliberate, targeted, and systematic and that its main objective was to prevent the further use of buildings designated specifically for rituals. Settlements can be abandoned for various reasons, and their destruction usually constitutes a secular event (e.g., war, political conflict, drought, invasion). Exploring the relationship between terminal rituals and site abandonment would shed light on the context leading to this particular scenario.

Continuity and Change in Monumental Architecture

The appearance of monumental architecture is strongly related to the increase of social stratification in complex societies (Clark and Martinsson-Wallin 2007; Moore 1996; Stanish 2001; Trigger 1990). The most distinctive attribute of these buildings is that their construction exceeds the pure practical needs of a specific human group, representing the most visible structures in a particular landscape. Furthermore, the visibility of monumental architecture may reflect sociopolitical interaction among different sites, where activities such as rituals may be oriented to large masses or to a select group of people (Chicoine et al., this volume). As Trigger (1990 :126) explains, these buildings “constituted the most public material embodiment of the power of the upper classes,” despite the fact that their meaning may not have been understood by everyone (Moore 1996).

Like many of the other authors in this volume, we suggest that monumental buildings symbolize the materialization of power of certain ruling groups (DeMarrais, Castillo, and Earle 1996). The monumentality of these buildings reflects the ability of local ruling elites to control labor (Abrams and Bolland 1999; Trubitt 2000) and is used to display their power and legitimacy over the population. The buildings’ deterioration or destruction, however, implies a weakening of the religious or political institutions within a particular society. Terminal rituals involving the destruction or burning of structures generally occur within specific buildings that denote political and religious power in a particular community.

In this context terminal events would involve permanent changes to the architectural pattern, in many cases associated with the total or partial abandonment of the structure. Nevertheless, minor changes can take place while the building is still in use. However, some short-term architectural changes could be a consequence of moments of crisis within a society (Driessen 1995). Driessen distinguishes three types of changes: (1) a decrease in the energy used in the construction and maintenance of specific buildings, (2) a change in the original architectural layout of a site, and (3) a change in the use of certain spaces. In addition, in the case of religious buildings these changes seem to be more notorious because of their ideological significance. In both cases mentioned, the alterations of architectural features in monumental buildings reflect sociopolitical changes in a community. Monumental buildings have the ability to represent the power of local leaders; consequently, their destruction has a great impact at the social and political level. It remains to be determined, however, what situation would have triggered these events and who was responsible for these actions. As discussed later, it seems that in most cases some faction within the local population would have taken the lead in the execution of these rituals.

Abandonment and Destruction in Andean Settlements

Terminal rituals, in which the destruction of monumental architecture is associated with the abandonment of at least part of a site, are very common in the Andean region. The level of destruction involved seems directly related to the magnitude of the events leading to the abandonment. Some scholars suggest that site destruction should be seen as part of a long-term process, as opposed to an isolated and unique event (Zuckerman 2007:3), and that it can shed light on some of the activities that were taking place at these sites prior to their final abandonment.

The Moche site of Pampa Grande, located in the Lambayeque Valley, presents evidence of the simultaneous abandonment of different buildings around AD 700, in an event most likely related to the demise of the Moche V polity (Shimada 1994). Archaeological excavations have revealed that a selected set of buildings was intentionally burned and abandoned. The fire caused the floors to harden, and some parts of the plaster from the roof vitrified because of the high temperatures. Moreover, the specific buildings that were burned had architectural styles reflecting some kind of political or religious power. The evidence presented by Shimada suggests that this burning and abandonment might have been part of a terminal ritual. Shimada presents two models that might explain this “systematic attack on and rejection of the material symbols of power and order” (ibid.:252). The first one concerns internal conflicts over political issues, while the second proposes an external factor that led to a shift in political control of the site. In Shimada’s first model, an internal revolt as a result of competing factions within the bureaucratic system would have led to violent actions taken against buildings that symbolized power and the ruling elite. The second model involves a possible conquest of the site by the Wari polity during the Middle Horizon (ibid.).

In the case of the site of Túcume, a settlement in the Lambayeque Valley that dates from the Lambayeque period (AD 250–1000) to the second half of the sixteenth century, two buildings showed evidence of intentional destruction. One of the major buildings known as Huaca Larga exhibits three construction phases. In the central Inka building, dating to the last construction phase (fifteenth century AD), some rooms were intentionally filled while others were destroyed in an event probably dating to the first years of Spanish occupation of the area. The Stone Structure, also dating to the Inka period, suffered two major burning episodes (Narváez 1995).

Another building in the same sector, also destroyed intentionally, was the Temple of the Sacred Stone. The name of this temple derives from the upright stone (huanca) in the middle of the structure that was most likely the central shrine of the temple (ibid.). The building also contained three altars corresponding to different construction phases and associated with this huanca, as well as several offerings. During the abandonment of the temple at the end of the last construction phase, the tops of the walls were knocked over, the central chamber was filled, the huanca was covered, and the columns were burned. This event must have taken place contemporaneous to the destruction episode at Huaca Larga (ibid.). Despite the discussion of the possibility that these acts were caused by weather, local people, or the Spaniards, evidence confirms that they were intentional.

A more elaborate abandonment ceremony seems to have taken place at the site of Cerro Baúl, a Wari settlement in Moquegua. While this site appears to have been abandoned abruptly, evidence demonstrates that several terminal rituals were conducted prior to the final evacuation. This is reflected in the fact that most of the rooms were very clean when found. Nevertheless, researchers have noticed that some spaces were used for rituals during the abandonment. Among these terminal ceremonies, the most elaborate seems to be the one recorded in the brewery, a room containing artifacts associated with the preparation of chicha, an alcoholic beverage known to have been used in ceremonies across the Andean region. Consequently, this would have been a space of great importance related to the performance of rituals and feasting activities. As part of the site’s abandonment, residents would have performed a ritual that would have included chicha. Evidence suggests that once they had consumed the beverage, the room was burned and the keros, used to drink it, were smashed and thrown into the fire (Moseley et al. 2005). A similar “highly ritualized abandonment ceremony” was performed at Pataraya, another Wari settlement in the Nasca region. The ritual included the placement of large rocks obstructing access to some rooms, the caching of entire ceramic vessels, and the intentional burying of some areas (Edwards, this volume).

In the examples given, several structures seem to have been intentionally destroyed. As long as these buildings were functioning, they continued to represent symbols of power, but as soon as they were destroyed or abandoned, they stopped serving this purpose. In the specific case of religious buildings, the connotation of their destruction was even greater since they constituted sacred spaces and were imbued with religious symbolism. In the case of Pampa Grande, the fire was used to attack formal constructions representing an existing religious and sociopolitical order (Shimada 1994:252). The same occurred at Túcume, where selected buildings with the same function were destroyed (Narváez 1995).

It can be debated whether local or foreign people were responsible for these destructions. The question that remains is, how elaborate and selective were these processes of destruction and the consequent abandonment, as it seems in fact that too much effort was put into these actions for them to be blamed on foreigners unaware of the ritual symbolism involved. In the case of Cerro Baúl, it is clear that the actions were part of an elaborate ceremony, probably presided over by the local ruling elite, prior to the final abandonment of the site. All the cases mentioned seem to have been local reactions to adjustments within the sociopolitical structure, especially during periods of regional change. The terminal destruction of a site, however, may also have been a secular action with no ritual components. This seems to be the case with the Omo 10 settlement in Moquegua on the Peruvian South Coast.

The site of Omo 10 is composed of a residential area, a temple, and a large cemetery; it was occupied during AD 725–950. The Tiwanaku polity colonized this settlement during its expansion period (AD 725–1000). Evidence suggests that the site was destroyed shortly after its final abandonment (Goldstein 1993). During the destruction of the site, structures were demolished indiscriminately, and many tombs in the cemetery were opened and looted. In fact, some rock slabs that were originally part of the structures destroyed were reused for sealing the tombs. The event occurred before colonial times, as it appears stratigraphically below a thick layer of ash left by the volcano Huayna Putina, which erupted in February 1600. Goldstein (ibid.) suggests that this destructive event was related to the collapse of the Tiwanaku polity and that a local group rebelling against the Tiwanaku state probably destroyed the site. At Omo 10 the destruction included a series of structures serving different functions. Since this action was not targeted or systematically repeated, it is not considered a terminal ritual. The fact that several tombs were opened and looted further suggests that it was not a ritual action (ibid.).

Burning Pyramids at Panquilma

At Panquilma, a multi-component settlement located in the left bank of the Lurín Valley on the Peruvian Central Coast (figure 9.1), the targeted destruction and consequential abandonment of specific areas within the monumental compounds appear to have been terminal rituals. The site occupies the surface of a dry ravine, extending approximately 150,000 m² at 400 m above sea level. It is 28 km up-valley from Pachacamac, one of the most important religious centers of the ancient Andes (Bueno Mendoza 1983; Eeckhout 1997; Jiménez Borja 1983; Jiménez Borja and Bueno Mendoza 1970; López-Hurtado 2011; Paredes Botoni 1991; Rostworowski de Diez Canseco 2002, 2004; Shimada 1991; Ühle 1991). The site is also part of a group of settlements associated with the Ychsma polity that occupied the Lurín Valley during the Late Intermediate Period (LIP), which extended from AD 1000 to AD 1476 (Menzel 1977; Rowe 1963). Most of these sites continued to be occupied during the Late Horizon (LH) (AD 1476–1535), when the Inkas conquered the area. At Panquilma, evidence recovered from excavations suggests that the site was occupied continuously during the LIP and the LH, but so far no evidence indicates an occupation during the early colonial period.

Figure 9.1. Map indicating location of Panquilma

The settlement is composed of three sectors (López-Hurtado 2011; López-Hurtado and Capriata 2013; Marcone Flores and López-Hurtado 2002). Sector 1 was identified as the site’s public sector and consists of three monumental compounds, each containing a type of building known in this region as a pyramid with ramps (Bueno Mendoza 1983; Eeckhout 1995, 1999a, 1999b, 2000a, 2000b; Franco Jordán 1993; Jiménez Borja and Bueno Mendoza 1970; Paredes Botoni and Franco Jordán 1987; Paredes Botoni et al. 1983) and a series of rooms and plazas associated with the main buildings. Tall perimeter walls surround these compounds whose access ways are restricted by a series of narrow corridors. Sector 2 consists of fifteen habitation complexes composed of a series of domestic structures, including dwellings, storage facilities, and patios that are also surrounded by perimeter walls. Finally, Sector 3 corresponds to the marginal areas of the site and is composed of a series of isolated structures, terraces, and tombs.

Between 2003 and 2013, excavations took place within the domestic and public sectors of the site (López-Hurtado 2010, 2011; López-Hurtado and Capriata 2013; Marcone Flores and López-Hurtado 2002). In the public area, units were placed within pyramid complexes 1 and 3—more specifically, within the plazas, platforms, and adjacent rooms (figure 9.2). During these excavations, we identified two different architectural components in these buildings: a public component that included large plazas and elevated platforms and a residential component formed by a series of adjacent rooms, patios, and storage facilities.

Figure 9.2. (A) Pyramid 1: 1. plaza; 2. intermediate platform; 3. upper platform. (B) Pyramid 3: 4. plaza; 5. intermediate platform; 6. upper platform, Panquilma

Data recovered at the main plaza of Pyramid 1 provided evidence of the ritual nature of these buildings’ public component. Excavations here showed a sequence of occupational floors with a series of pits associated with the different levels. The floors were very clean, and their fine plastered surface differed greatly from ones identified in the site’s domestic compounds. The pits were no bigger than 30 cm in diameter and were originally dug without any particular organization. In fact, some of them intrude into previous ones associated with earlier floors. We also found pits located in the center of the plaza, among all the occupational floors, that contained offerings such as ceramic figurines and Spondylus princeps valves, among others (figures 9.3, 9.4). The pits located in the peripheral areas of the plazas contained only ceramic fragments or were empty.

Figure 9.3. Ceramic figurine

Figure 9.4. Spondylus valves

A similar pattern was identified at Pyramid 3. Here we were able to identify a similar succession of floors and pits. However, in this building the pits containing offerings were disturbed during a later occupation when the plaza was reused as a trash deposit. We have interpreted these findings as evidence of a process in which Panquilma’s ritual areas were de-sacralized.

In addition, the excavation of the main platforms of Pyramids 1 and 3 indicated the abrupt end of the life of these buildings. In both cases these spaces had two access ways: one through the lateral ramps coming from the plazas and one through the adjacent rooms. Also, both had benches that ran parallel to the back walls that were originally covered by a roof. In both cases the roofs, made of canes (Gynerium sagittatum) tied together with ropes and supported by wooden posts, were burned and had consequently collapsed over the floor. The fire had been intense; therefore, some parts of the floors and the plaster over the benches and adjacent walls had turned orange. However, despite the intensity of the fire, it seems it was contained to the platforms without compromising the adjacent structures. No artifact remains associated with these events were found. The burning of these platforms corresponded with the end of the use of at least the public component of these buildings. In both cases the burned roofs were found directly under a layer of collapsed walls and over clean floors, indicating that the burning was probably the last event that took place at the pyramid before final abandonment (figures 9.5, 9.6).

Figure 9.5. Upper platform of Pyramid 1 at Panquilma showing evidence of intense burning over the floor.

Figure 9.6. Drawing of unit placed on Pyramid 3 showing burned room remains over the bench and the floor of the upper platform

Finally, the discovery of walls in the last excavated layers indicates that the pyramid was probably constructed during the last phase of occupation. Although it is not possible to determine the spatial configuration of the first building, we established that there were at least two construction phases. Also, the presence of subsequent floors indicates a series of remodeling events associated with the last construction phase. The presence of large construction fills points to significant effort in its execution, probably to allow the pyramid to gain altitude and the platforms to be built at different levels. Thus, the burning events we recorded marked the end of the life of a building that went through important transformations during its existence.

The excavation of the residential component of Pyramid 3 revealed that in addition to its public ritual function, these buildings were also the residences of Panquilma’s ruling elites. A series of rooms and associated midden deposits excavated in the back part of the building revealed the special status of the people living there. One of the rooms excavated seemed to have served as a dwelling. The size of the room, its plastered walls and floor, and a bench that covered over 70 percent of the room’s area support this hypothesis. The most important artifacts recovered here consisted of fragments of Spondylus beads and a decorated Ychsma head-jar, probably broken intentionally when the pyramid was abandoned. In addition, one storage facility that was excavated next to the upper platform of the pyramid and adjacent to this building’s residential area still contained remains of aji peppers and maize cobs.

Finally, a unit placed behind the wall surrounding one of the inner patios of the residential area of Pyramid 3 showed evidence of large deposits of trash. A statistical analysis of the artifacts revealed a drastic change in the relationship between fine ceramics and serving vessels. Our analysis revealed that during the first occupational phase, in the public sector the number of fine ceramics and serving vessels was almost equal, while during the second phase, right before the pyramids’ abandonment, there were almost four times as many serving vessels as fine ceramics. This drastic change in the proportion of fine ceramics to serving vessels has been interpreted as an indicator of an increase in feasting activities in the elite’s residential area before its abandonment (López-Hurtado 2011).

Some scholars suggest that feasting is a type of “commensal politics” or a specialized form of political competition in which a relationship of obligation, retribution, and loyalty is established between host and guest (Dietler 1996). Feasting can also be seen as a type of ritual practice conducted systematically or spontaneously by a society as a way to diffuse social tension. Based on the higher proportion of serving vessels and preferred food remains among the elites, such as aji peppers and maize, we suggest that competitive feasting took place on this patio (López-Hurtado 2011). The size of this space and its location within the private area of the pyramid, however, indicates that the feasts were performed for a small, select audience.

Political Landscape in the Lurín Valley

Pyramids with ramps have been identified in several sites along the lower and middle Lurín Valley. Eeckhout (2003a) mentions the presence of at least forty-three pyramids, with the vast majority located at the sites of Pachacamac and Pampa de Flores, both in the lower valley. While most of the pyramids share similar features, such as superimposed platforms connected by a ramp, a main plaza area, and several rooms associated with the platforms, they vary in size and complexity. In addition, there is variation in the position of the ramps that connect the platforms to the main plaza. At Panquilma, the three pyramids located in Sector 1 have what Eeckhout calls “offset ramps.” As opposed to the majority of pyramids at Pachacamac and Pampa de Flores, which have a central ramp, pyramids at Panquilma have ramps located on one side of the main plaza (see figure 9.2).

Based on excavations at Pachacamac, Eeckhout suggests that central ramp pyramids were mainly used as palaces or elite residences, serving primarily administrative and residential functions (Eeckhout 2003a, 2003b). The occupants of the pyramids would have offered banquets to and performed ceremonies for a select audience. He indicates that most plazas contained storage vessels, hearths, and food remains, suggesting that these spaces were used primarily for processing food intended for the elite residents and sporadically for a greater number of people attending the feasts (Eeckhout 2003a). The buildings also had rooms dedicated to the production of crafts, such as textiles and ceramics, and several storage facilities, probably for the accumulation of residents’ wealth (Eeckhout 2003b). Pyramids would have been continuously occupied while the main resident was alive and abandoned and sometimes ritually closed when he died. This approach implies a dynastic succession among rulers, in which the pyramids served as palaces used for a brief amount of time and did not function simultaneously (ibid.).

In addition, Eeckhout was able to identify events related to changes in the life use of ramp pyramids at Pachacamac. He discovered elite burials in three platforms excavated in the pyramids. These data are consistent with his theory of the use of pyramids by residential elites and a dynastic succession, which would explain the sudden abandonment of these palaces when the main resident died. Eeckhout also mentions that the abandonment of these pyramids could have been voluntary, as evidenced by the presence of ritual activity, which consists of the blockage of entrances, the covering of structures with fine sand, and the burial of offerings containing Spondylus princeps shells, among others (Eeckhout 2003a). Moreover, a new building was sometimes constructed on top of the old one (López-Hurtado 2011). As mentioned, this implies that most of the buildings in Pachacamac would have been abandoned before the Inka arrival, which, according to Eeckhout’s model, also suggests that only one pyramid would probably have been functioning at that time: the one with the current ruler. It also implies that these events were not necessarily related to drastic disruptions of the sociopolitical order. On the contrary, the construction of new pyramids on top of the abandoned one can be interpreted as indicating continuity of the existing power structures.

However, the evidence found at Panquilma contrasts with Eeckhout’s theories. First, excavations in Pyramids 1 and 3 suggest that even if the construction of Pyramid 3 might have been slightly older, the two were probably contemporaneous. In both buildings ceramic fragments associated with Late Ychsma A and B styles were found, which suggests they were used at the same time (ibid.). A small proportion of Inka-style ceramics was also associated with these occupations, indicating that the buildings were at some point also used during the Late Horizon, although this does not necessarily imply an Inka presence at the site.

Second, no evidence of public feasting was found in either of the excavated plazas in pyramid complexes 1 and 3. As opposed to what Eeckhout found at Pachacamac, these spaces in Panquilma seem to have served different functions. While both cases demonstrate that plazas represented the largest inner precincts in the pyramids and would have held a greater number of people, at Panquilma these areas were used for ritual practices that involved the placement of offerings during ceremonies probably presided over by ruling elites (ibid.; López-Hurtado and Capriata 2013). These inconsistencies may be a result of the different spatial configurations of the buildings, indicating that pyramids with an offset ramp served different purposes than those with a central ramp.

Finally, the abandonment of the buildings seems to have been more abrupt at Panquilma than at Pachacamac. In Pachacamac, the pyramids were used to bury the main occupant, which implies continuity as these spaces transformed from residential palaces to mausoleums (Eeckhout 2003a). The burning episodes at Panquilma, however, demonstrate that specific areas of the pyramids were targeted for destruction prior to final abandonment. Moreover, this destruction appears to have been systematic and deliberate, as we only found evidence of burning in the upper platforms of Pyramids 1 and 3. Given that this burning was targeted to specific areas of the pyramids, the destruction was most likely performed advisedly.

During the LIP, the Lurín Valley was populated by a series of groups forming the Ychsma polity, with the religious center of Pachacamac located in the lower section of this valley. Based on research carried out at Panquilma, it appears that Pachacamac did not have a strong influence on the political configuration of the valley (López-Hurtado 2011). Each settlement seems to have had its own political administration run by local ruling elites. Still, Pachacamac played a leading role in the ideological landscape of the valley, as it represented the largest religious center in the area (Astuhuamán Gonzáles 2000; Bueno Mendoza 1982; Cieza de León 1985; De La Calancha 1975; Guamán Poma de Ayala 2002; Rostworowski de Diez Canseco 2002).

The arrival of the Inkas to the Central Coast brought significant changes to the political landscape. Pachacamac was probably one of the first sites occupied. Inka authorities built a number of administrative and residential structures at the site, including the Temple of the Sun, the Acllahuasi, and the Tauri Chumpi Palace. Evidence indicates that most ramp pyramids were no longer used as residences for the elite but instead became temporary dwellings for pilgrims visiting the site (Eeckhout 2010; López-Hurtado and Nesbitt 2010). As Pachacamac gained importance during the Late Horizon as one of the empire’s major religious and administrative centers, the political organization of some settlements in the valley was radically affected, perhaps as a consequence of these changes.

In this context, evidence indicates that Panquilma continued to be occupied during the LH, but its political importance may have diminished, as explained later. Evidence recovered in the domestic and private sectors of the site shows the presence of cultural material related stylistically to the Inka polity, appearing in contexts contemporaneous with the local Late Ychsma style. In addition, excavations at one of the isolated structures in Sector 3 displayed a vast amount of Inka-style ceramic fragments, more than found in the domestic sectors or the pyramid complexes. Further studies will determine whether this isolated structure was built during the LH or reoccupied at that time.

At Huaycán de Cieneguilla, another Ychsma settlement 2 km northeast of Panquilma, a series of architectural modifications with clear Inka influence took place during the LH, including the construction of a public building known as the “room of four windows” (López-Hurtado and Nesbitt 2010). Huaycán de Cieneguilla is the only site in the Lurín Valley that presents a monumental Inka-style architecture (Marcone Flores 2004). Marcone suggests that this evidence reflects indirect control by the Inkas, where they would have acquired control in the area through local rulers. In the case of Huaycán, this would have been the point at which the site gained hierarchical superiority over previously more important sites such as Panquilma (ibid.).

Given this evidence of Inka expansion in the Lurín Valley, it is possible to say that the arrival of the Inkas had an impact on Ychsma local populations. However, how did local ruling elites at Panquilma deal with the foreigner presence in the valley in terms of political control of the settlement? Prior to the arrival of the Inkas, Panquilma seems to have been ruled by different elite groups. These factions would have coexisted while competing against each other for prestige and, consequently, for control over the population. Evidence of feasting activities in the privately located patios of the pyramidal complexes demonstrates the competition between various elite groups. At the same time, ruling elites would have interacted with local populations by conducting rituals involving the placement of offerings in the main plazas of the pyramids. During these ceremonies the elite groups would have sat on the benches of the upper platforms in the pyramidal complexes, probably with their backs to the audience while this select entourage deposited offerings.

The evidence presented suggests that the political organization at Panquilma differed from the one at Pachacamac and that this may be reflected in the different configuration of the buildings as well as in their different use. At Panquilma, local ruling elites did not host banquets or feasts for commoners. As it has been established that the two pyramids may have functioned simultaneously, we propose that ruling groups had two ways of interacting with the rest of the population. In the first way, relationships with commoners occurred in the plazas where the rulers conducted ceremonies for a select audience, while the relationship among elite groups would have occurred through feasting activities carried out in private patio areas inside the pyramids. In this sense, although a ritual component of these banquets cannot be ruled out, given the evidence recorded in these spaces, the primarily mundane nature of these activities became very clear. Presuming the ritual nature of every communal activity homogenizes the varied ways in which mundane and ideological factors interacted in the history of Andean societies. In the second way of interacting, the relationship between Panquilma’s elite and the rest of the population was the opposite of the first. Evidence recorded in the plazas and platforms on top of the three pyramids at the site highlighted the primarily ritual nature of these ceremonies.

Conclusion

The events leading to the abandonment of the pyramids at Panquilma occurred during the Late Horizon, when social and political changes were developing in the valley; consequently, they may be attributed to the arrival of the new Inka polity. But how did this affect the political power of local ruling groups at Panquilma? Did they refuse to align with the new political order? Those questions remain for further investigations.

As argued for the cases of Pampa Grande, Túcume, Cerro Baúl, and Panquilma, terminal rituals seem to have been a common practice in the Andean world. In all cases the events revealed different levels of complexity, but they always reflected important changes in the sociopolitical landscape at a regional level. Furthermore, the local ruling class seems to have played an important role in these processes. Nevertheless, not all intentional destructions and abandonments in the Andes would have had a ritual component, as has been demonstrated for the Omo 10 site in Moquegua.

While terminal rituals can result from endogenous conflict in Andean society, it seems probable that a violent and extensive ritual response reflects the collective trauma caused by large-scale exogenous factors (e.g., conquest, coercion). All of the events mentioned above were triggered by violent events such as invasions or threats of invasion. In this context terminal rituals were caused by external factors and the need to resist them. Consequently, they were probably performed by local elites, the original administrators of rituals in these precincts. It seems clear that in addition to the ritual implications of these actions, the common goal was to destroy things either to prevent their future illegitimate use by others or to “kill” them in a metaphoric way, thus undermining and denying legitimacy to imposed authority.

While it has been stated that Panquilma continued to be occupied during the Late Horizon, its political scenario suffered, directly or indirectly, from the impact of the Inka conquest. As other settlements such as Huaycán de Cieneguilla gained importance with this new political situation, Panquilma lost its place in the valley hierarchy. The evidence of sudden destruction of the platforms in the pyramidal complexes points to an abrupt end of this polity. However, this intentional destruction does not appear to have been executed by foreigners. The fact that the burning events were confined to areas used by ruling elites to preside over ceremonies highlights the preeminent ritual nature and the probable existence of preconceived “culturally constructed concepts of how living and ritual spaces should be appropriately abandoned” (Edwards, this volume). Moreover, we found no evidence that suggests that outsiders were participants in these events. If this was a consequence of an invasion, the burning and destruction would have been more extensive. On the contrary, the burning events followed certain guidelines more in accordance with those proposed for terminal rituals. Within this context, the selection of spaces, the recurrence, and evidence indicating that these actions took place in spaces linked to ritual activities inside a monumental building support this argument. In addition, the effort and precision that went into these actions may suggest that local ruling groups were involved in this burning event, which suggests that they still had control over ritual practices at the site, even near the demise. Finally, the collapse of burned roofs over clean floors and the presence of these very clean spaces in the surrounding areas may also indicate a planned evacuation. In this scenario the ruling elites would have known that the end was coming and would have packed their belongings and ritually closed their sacred spaces.