Dulanto Place Of Death.pdf

The Archaeological Study ofAncestor Cult Practices: The Case ofPampa Chica, a Late Initial Period and Early Horizon Site on the Central Coast ofPeru Jalh Dulanto Pontificia Universidad Catolica del Peru, Lima ABSTRACT This chapter explores the spatial and material dimensions of ancestor cult practices for one specific archaeological case: the late Initial Period and Early Horizon (ca. 700—200 B.C.) site of Pampa Chica in the Lurin Valley on the central coast of Peru. It combines the analysis of ethnohistoric accounts, used to formulate an ideal-typical image of the spatial and material organization of ancestor worship in the Central Andes, with the analysis of several independent lines of archaeological evidence, in order to explore how the specific archaeological case resembles and/or deviates from the ideal-typical ethnohistoric image. Despite regional and temporal differences, Pampa Chica seems to have been a site designed, built, and used in ways that closely resemble ancestor worship sites described in ethnohistoric documents.

A

ncestor worship was the most important component of collective rituals in the Central Andes. Early Colonial accounts describe practices of ancestor worship in a variety of collective rituals undertaken at different levels of demographic, social, and/or political aggregation—from small agrarian communities such as the ayllus of Caxatambo and Huarochiri, to entire ethnic groups such as the Bolivian Macha, to multiethnic states such as the Inca. From a sociological perspective the ubiquitous character of ancestor worship in Andean ritual life is not surprising. In societies in which social relations are mainly framed in an idiom of kinship, such as in the Andes, ancestors serve as points of collective reference in the construction and reconstruction of social relations. Ancestor worship is a way of reasoning and arguing about social relations and, because of this, it is a means by which the collectivity invents and reinvents itself. Elucidation and comparison of ancestor worship practices among Andean societies is worthwhile for the light it can shed on their organization and transformation. Yet ancestor worship is one of the most elusive practices when

it comes to inferring it from the archaeological record. This chapter is a contribution to the archaeological study of ancestor worship. It shows how a holistic approach permits the identification and characterization of ancestor worship practices in a specific archaeological case. This approach combines the analysis of ethnohistoric information on ancestor cults with the analysis of archaeological information regarding location of sites in relation to each other and to natural resources, internal organization of architecture and activity areas, and, in particular, the formation and transformation of various distinct types of contexts containing human remains. In this study, ethnohistoric information is used to formulate an idealtypical image of the spatial and material organization of ancestor worship in the Central Andes. This construct should be useful in studying these practices elsewhere in the Andes and for periods during which the only available information is archaeological. The three lines of archaeological evidence mentioned above are, in turn, used to explore how the specific archaeological case resembles or deviates from such an ideal-typical ethnohistoric image.

98

Jalh Dulanto

The specific case analyzed is the site of Pampa Chica, a late Initial Period and Early Horizon site located on the central coast of Peru (ca. 700-200 B.C.).1 This study shows that, despite regional and temporal differences, Pampa Chica was a site designed, built, and used in ways that closely resemble ancestor worship sites described in the documents. Yet, this study also permits us to discover important particularities in the ways that Pampa Chica was designed, built, and used that could be interpreted as the result of specificities in the organization of ancestor worship practices represented in the site.

The Ethnohistoric Record The ethnohistoric record on ancestor worship in the Andes is fragmentary and highly problematic. Nonetheless, if compared and complemented with the more complete ethnographic record, ethnohistory can be used to formulate an ideal-typical image of the organization of ancestor worship that could be useful in the study of particular cases from different regions and periods of Andean culture history (Salomon 1995:315). If, in addition, such an ideal-typical image emphasizes the spatial and material dimensions of the organization of these practices it becomes particularly useful in the study of cases for which only archaeological information is available. Largely on the basis of the comparative ethnohistoric study of Andean ancestor worship presented by Frank Salomon (1995; see also Doyle 1988; Dulanto 1992; Duviols 1979. 1986; Huertas 1981; Isbell 1997:69-100; Zuidema 1973), it is possible to formulate such an idealtypical image of the organization of Andean ancestor worship in terms of three spatial and material dimensions: (1) the landscape that was imagined as a complex of sacred objects, spaces, and natural forces originating in the actions of ancestral heroes; (2) the public spaces within this complex where the communities gathered periodically to add their recent dead to the heroic dead and to venerate their ancestors; and (3) the mummified bodies of the recent dead, which were intentionally preserved and frequently manipulated and moved within and between such public spaces.

son as the source of entitlement among a group of people who shared rights or identity." However, a basic important difference can be noted between lower and higher levels of aggregation regarding the type of material referents identified with the ancestors. At lower levels, such as small rural communities composed of a few households, the tendency was to identify the group's immediate ancestors with the mummified bodies of the group's recent dead. At higher levels, however, the tendency was to identify the common parents of several small rural communities with landscape features or natural forces. Landscape features such as monoliths, mountains, and islands were often considered to be ancestral superhuman heroes (i.e., huacas) who turned or were turned into stone, or places where the heroes originated or engaged in violent confrontations with each other once upon a time. In contrast, natural forces like lightning, fire, rain, or hail were usually regarded as the means by which some of these ancestral heroes confronted others. Ultimately, however, the whole complex of mummified bodies, sacred objects, spaces, and even natural forces was seen as a unit, as "the final product of a mythohistorical transit from huaca origins, to huaca actions..and finally to a stable order in which the living continually aggregate the recent dead to the heroic dead" (Salomon 1995:322). Human and superhuman ancestors were imagined as related to each other within a complex network of nested genealogies that seem to have served, among other things, to reckon group membership and to legitimize the transmission of goods and rights at different levels of demographic, social, and political aggregation. Although the exact levels of aggregation at which human genealogies were replaced by superhuman ones seem to have varied among different Andean regions, it is clear that the two were normally integrated within a single system. Because of this, not only the mummified bodies of the recent dead and the public places in which they were continually added, but also almost any feature of the landscape and even the cosmos were seen as tangible points of articulation in a complex web that integrated social, geographic, and cosmological orders. In the Andes, ancestry was literally mapped onto the whole world, and the whole world onto ancestry (Salomon 1995:320).

Landscape

Public Spaces The ethnohistoric record leaves few doubts that ancestor worship was a central feature of Andean social and political life at different levels of demographic and/ or political aggregation. As Salomon (1995:320) has noted, at each of these levels "the minimal element of ancestor cult was the veneration of at least one dead per-

At least in seventeenth-century Caxatambo and Huarochiri, the ethnohistoric documents indicate that one or more small rural communities were usually affiliated to a local ceremonial center that housed the mummified bodies of their recent dead and also served to accommo-

Ancestor Cult Practices at Pampa Chica

date the large numbers of people who frequently gathered there to venerate their ancestors—not only during mortuary rituals but also during rituals of ancestor enshrinement and cyclical rituals of ancestor veneration.2 According to Salomon (1995:321), the typical local ceremonial center of Caxatambo and Huarochiri "had a small plaza close to, and often overlooking, an area of houses for the living. The plaza would be bordered by small chambers or cells described as 'tiny lodgings,' something 'like storehouses,'...These contained the preserved bodies. It was at this plaza, often called cayan, that the village's ayllu(s) jointly feted their 'founders.' The small chambers in which mummies dwelled usually also contained collections of lesser holy objects, such as conopas or fertility 'idols.'" In contrast to cemeteries, in which the recent dead usually remain buried after interment, these local ceremonial centers consisted of permanent, clearly bounded, and highly visible buildings that were especially designed and built to provide easy access to the remains of the recent dead during the gatherings of large numbers of people. Their construction involved a significant investment by the group of small communities in a program of ancestor veneration in which mortuary rituals, the enshrinement of ancestors, and the cyclical veneration of ancestors were closely integrated with each other. However, the documents indicate that these local ceremonial centers were not the only type of public space at which mummified bodies were stored and/or people gathered to venerate their ancestors. Almost any place that was seen as significant in the origins and actions of ancestral heroes was likely to be ritually marked, to be the object of public gatherings and pilgrimages, and even to be used for the secondary disposal and/or temporal dwelling of mummified ancestors—especially in the case of the enshrinement of mummified ancestors. Although these local ceremonial centers were important sacred places in which the recent dead were usually disposed for the first time, they were not the only type of sacred place in which the integration of human and superhuman ancestors was celebrated. Mummified Bodies The ethnohistoric record contains abundant descriptions of the manipulation of the mummified bodies of the recent dead during different types of public, collective rituals of ancestor veneration. The documents say that during these rituals, the participants used to extract the mummified bodies from their communal sepulchers, put them on display, sacrifice llamas and guinea pigs in

99

front of them, and sprinkle them with the blood of these animals. There are also accounts of how the living used to dance carrying the mummified bodies on their backs, and take them to distant sacred places on pilgrimages that sometimes lasted several days. Preservation of the bodies and their transformation into ritual objects that could be repeatedly manipulated and transported seems to have been accomplished by means of different types of techniques and procedures (see Lastres 1953:70-73; Moodie 1931:46; Tello 1929:131-35; Vreeland and Cockburn 1980; see also Buikstra 1995; Rivera 1995). One of these techniques was the wrapping of the flexed corpse in several layers of textiles to transform it into a mummy bundle. This usually took place during the mortuary rituals, before the first interment of the dead person. Mummy bundles, however, continued to be frequently treated long after their first interment in order to either preserve or transform their physical appearance. They were frequently cleaned, refilled, rewrapped, redressed, and perhaps even painted. Unfortunately, it is not clear in the documents whether preservation or transformation of the mummy bundles' appearance was used to mark continuity or change in their genealogical position and status or the expansion of the groups of descendants they encompassed. What is clear, in any case, is that mummified bodies were frequently incorporated in public spaces within and between which they were frequently manipulated and, in a sense, used as an important piece of ritual paraphernalia.3

The Late Initial Period and Early Horizon Landscape of Ancestor Veneration on the Central Coast It is difficult to study the landscape of the late Initial Period and Early Horizon of the central coast. Severe alluvial events and modern agricultural practices have destroyed almost all valley bottom sites (see. e.g.. Agurto and Watanabe 1974; Dulanto 1994; Earle 1972; Engel 1983; Feltham 1983, 1984; Palacios 1988; Patterson 1966; Silva 1996; Silva et al. 1982, 1983). Except for sites located in areas that were not affected by alluvial processes, such as shore areas (e.g., Ancon and Curayacu) or slopes and hilltops on the sides of the valleys (e.g., Quebrada dc Jicamarca), only sites with monumental architecture have survived. Because of poor site preservation it is very difficult to fully understand the relation between mortuary rituals and ancestor cults and the landscape. Nonetheless, it is still possible to analyze the location of individual sites in relation to other sites and natural resources and make

ion

Jalh Dulanto

comparisons with the location of sites described in the documents as having been used for mortuary rituals and or rituals of ancestor veneration. Pampa Chica is a late Initial Period and Early Horizon site on the central coast, whose location resembles that of at least one specific type of site described in ethnohistoric documents as a place used in mortuary rituals and or rituals of ancestor veneration. Like the local ceremonial centers described in the documents, Pampa Chica shows a relatively marginal location in relation to habitation sites and a strateuic location in relation to important complementary resources. The site is located on the bottom of a small, hidden ravine on the north side of the Lurin Yallev It overlooks a valley bottom area where habitation sites were probably originally located.

and it is located in between the agricultural fields in the valley bottom and the area of the Lomas de Atocongo, which was an area rich in hunting and gathering resources (Figure 6.1). Pampa Chica is the only known site of the Initial Period and Early Horizon that shows a location like this. Interestingly, a number of cemeteries of the later Early Intermediate period have been found in similar locations (e.g , Tablada de Lurin and Limay).

Pampa Chica as a Space of Public Veneration of Ancestors The location of Pampa Chica as well as the organization of space within the site and the types of activities carried out in it resemble, to some degree, those of the

Chillon * ^M River l U f l V

IK

Rimac J River

^^^M ^ H ^

c

• V ?//) Pacific 1

Pampa

Chica

10Km

^'^

V

Ocean ^ 0

/

Lurin ^ ^ ^ \

Figure 6.1. I oc at ion of Pampa Chica in i elation to the bottom section of the Lurin Valley and the Lomas de Atocongo area.

Ancestor Cult Practices at Pampa Chica

local ceremonial centers described in the ethnohistoric record. As at these sites, Pampa Chica contains permanent, durable, and highly visible buildings that were designed, built, and used to store the human remains of dead individuals and to celebrate gatherings and collective displays. Thus, Pampa Chica also represents an explicit investment in a ritual program that involved the incorporation of human remains into public spaces, in which they could be preserved and easily accessed during collective celebrations. The following description of the architecture and activity areas of the site presents some of the particularities of the organization of activities at the site. Architecture Pampa Chica contains two independent structures. The two structures lie on an inclined terrain and are separated by approximately 140 meters horizontally and 10 meters vertically (Figure 6.2). Both structures are closed, orthogonal buildings that comprise several orthogonal

lui

rooms that are located on different terraced levels. Long walls made of field stones and adobe bricks were used in both of them to level the terrain and at the same time enclose and divide them. Structure 1 (Figure 6.3) is the largest of the two (45.19 x 37.87 meters) and is located in a higher and more secluded area of the site. It is divided into twenty orthogonal rooms that include patios, terraces, and closed rooms (R 1-20). Analysis of the location of these rooms on different terraced levels and their position within the internal access system of the building shows that they are organized into five distinct sectors (Sectors I \ ). Sectors 1 and III. located on lower terraced levels, are both large open spaces. Sector I has a large open patio in the center (R 20) and three long terraces to its sides (R 15 to the west, R 16 to the south, and a terrace that was not excavated to the north). Sector III also has a large open patio (R 9), but its patio has only one terrace that is located on its west side (R 7). On the north side of this terrace lies a small closed room (R 6) that is connected to the terrace by a narrow restricted access. Sec

S 77 - LURIN PAMPA CHICA

Figure 6.2. Structures 1 and 2 on the bottom of the Quebrada de Pampa Chica.

102

Jalh Dulanto

Pampa Chica Structural Phase 1 (associated contexts)

Figure 6.3. Structure I. Map include only walls located within excavated areas.

tors I and 111 are connected to each other by a narrow access connecting the terrace on the south side of the former with the open patio of the latter. They are also connected with the sector that lies immediately above them on the upper terraced levels (respectively. Sectors II and IV). In contrast to Sectors I and 111, Sectors II and IV are located on the upper terraced levels and are small closed spaces. I ach sector has a small closed patio (respectively, R 8 and R 19), flanked by two small rooms to the west (R 13 and R 14, and R 17 and R 18) and a narrow terrace or bench to the west (R 11 and R 1), and a small sunken area located in the center or to one side (R 8 and R 4).

Sector V, which was almost totally destroyed before the exca\ation of the site, is located on a terraced level immediately above Sectors II and IV and seems to have consisted of a single rectangular closed space (R 3). Sectors II and IV are not connected with each other by any type of access, and the wall separating them seems to have been tall enough to impede visibility between them (approximately 1.8 meters)."1 These two sectors, however, could have been connected with Sector V. One important characteristic of the organization of internal space in Structure 1 is the separation of open and closed spaces according to their location on lower or upper terraced levels. The characterization of upper

Ancestor Cult Practices at Pampa Chica

and lower sectors as either open or closed spaces is based on several lines of evidence. For example, lower sectors are two or three times larger than the upper ones, which suggests that they allowed accommodation of larger numbers of people than the upper sectors. Furthermore, lower sectors also had lower external walls than upper sectors (respectively, 0.6 and 1.8 meters) and were not roofed, which indicates that they were more easily visible from the exterior. Moreover, they are also easier to access than the upper sectors. Sector I is directly accessible from the exterior, and one has to trespass through Sectors I and III in order to go to the upper sectors. Last, and probably more important, the organization of their internal space favors the carrying out of public displays that comprise groups of spectators and performers: the terraces and benches are always clearly separated from the patios and spatially arranged looking to them. Upper sectors, in contrast, are internally organized in a manner that favors the carrying out of more proximate and intimate interactions between the participants of the same activity: the sunken areas are not only smaller but also closed and secluded. Another important characteristic of the organization of internal space in Structure 1 is the division of the structure into two symmetrical halves, each composed of a lower open sector and an upper closed one. Analysis of the internal access system of Structure 1 shows that once one is in Sector I, which is the only sector connected to the exterior, only two possible paths can be followed: one can continue to Sector II, the immediate smaller closed space in the north half, or one can move to Sector III and then Sector IV, respectively the lower open and upper closed sectors of the south half (Figure 6.4). Thus, Structure 1 shows a very particular pattern of internal organization. Movement within Structure 1 could be characterized as one marked by a transition from larger, more open spaces located on the lower terraced levels to smaller, more closed spaces, located on the upper ones, and the simultaneous separation of the participants into two symmetrical groups (as indicated by the two halves of the structure and especially the two separated upper sectors). As I show below, these are also two characteristics of the internal organization of Structure 2, and of the site as a whole. In contrast to Structure 1, Structure 2 (37.16 x 17.87 meters; Figure 6.5) is located on a lower and more accessible area of the site. It is divided into six orthogonal rooms that also include open and closed patios, terraces, and closed rooms (R 22-27). As in the case of Structure 1, these rooms are organized in distinct open and closed sectors that are located on different terraced

103

levels and show certain evidence of division of space into two symmetrical halves. Structure 2, however, also shows some important differences from Structure 1. It is divided into only two sectors, and overall it is less closed than Structure 1. Sector VII, for example, is an open space located on a lower terraced level. It is, in general, similar to Sectors I and III of Structure 1. It consists of a large open patio (R 24) and a long terrace located along its upper northeast side. The patio is connected to the exterior by a large though restricted L-shaped access. Sector VIII is a smaller and more closed space than Sector VII but also a larger and more open space than Sectors 11 and IV of Structure 1. It has a large though partially roofed patio (R 23) and two symmetrical pairs of rooms that are located on its upper side (R 22 and 27, and R 26 and 28). Although the patio seems to have been originally roofed, its external walls do not seem to have been tall enough (approximately 0.8 meters) to impede visibility from the exterior. The two symmetrical pairs of rooms, on the other hand, are closed spaces that contain two small rooms each. One is twice as large as the other, and it is the only one of the two that is connected to the large patio (R 3). Thus, as in the case of Structure 1, Structure 2 is divided into lower and upper, open and closed spaces. Movement between these spaces and the rooms that compose them is also marked by the transition from lower, more open spaces (Sector VI1) to upper, more closed ones (Sector VIII), and by the separation of the groups of participants into two equivalent groups (as reflected in the two symmetrical pairs of rooms). To summarize, analysis of architectural space in Pampa Chica suggests that the two structures were designed and built to be used complementarily as public spaces. Indeed, stratigraphy and radiocarbon dates indicate that the two were built and used and abandoned relatively at the same time (Dulanto et al. in press). The analysis also indicates that the communal gatherings that took place in the two structures involved two very different types of activities: some involved open displays; others more intimate interactions. The groups that participated in these communal gatherings interacted more closely during the open displays but were separated into two equivalent groups during the more secluded activities. 5 Activity Areas The analysis of activity areas presented in this section complements the analysis of architectural space. It gives us some idea about the specific characteristics of

104

Jalh Dulanto

SOUTH . NORTH HALF HALF SECTOR V

SECTOR II

SECTOR IV

UPPER

LEVELS

fit—j n n I 1E~) 1E31

1

• • • • • • • [ • • • • • • • • • • • • • • • •

LOWER || t LEVELS I

„ +

-ijTZl T7 I • • • • • •

K16

SECTOR III

R20

SECTOR I

Figure 6.4. Reconstruction of Structure I access patterns, based on the location and size of accesses connecting different rooms and also on the height of walls separating adjacent rooms and adjacent terraced levels.

Ancestor Cult Practices at Pampa Chica

N

A

103

Pampa Chica Structure 2 5

lOn

Figure 6.5. Structure 2.

both the more open and the more secluded types of practices defined thus far. Three different types of activity areas were identified in the architectural sectors defined at Pampa Chica: (1) at least one area in the upper terraced level of Structure 2 seems to have been used for the large-scale preparation and storage of food and probably also maize beer

(Sector VIII, R 22 and 23); (2) at least two areas in the lower terraced levels of Structure 1 and perhaps also Structure 2 seem to have been used, in turn, for some types of activities involving consumption of food and probably also beverages (Sectors I, III, and VII); and (3) two areas in the upper terraced levels of Structure 1 ,scem to have been used for some types of ritual acti\ ities in-

106

volving the repeated burial, removal, manipulation, and reinterment of human remains as well as the repeated burial of different types of goods (Sectors II and IV; see Human Remains, below). Identification of these activity areas is mainly based on the analysis of the distribution of different functional categories of features, artifacts, and ecofacts among different architectural areas of the site. Such analysis shows that different functional categories of archaeological remains appear associated with different types of architectural spaces. It also shows that the sets of functional categories that were found together within the same area seem to have been the result of types of activities that are consistent with the architectural characteristics of such areas. Features excavated in Pampa Chica, for example, were classified in four distinct functional categories: (1) hearths, (2) storage pits, (3) offering pits, and (4) burials. Hearths excavated in Pampa Chica were all clearly bounded concentrations of ash and charcoal that contained burned food remains (e.g., animal bones, shells, plant remains). Storage pits included several large pits (60-100 centimeters in diameter and 40-60 centimeters deep) that contained fragments of large neckless ollas as well as burned and unburned food remains. The fact that the diameter of these pits generally coincides with the maximum body diameter of the large ollas suggests that these pits were used to support complete vessels of this type. Offering pits, on the other hand, were all small (2454 centimeters in diameter and 16 centimeters deep) and contained only unburned food remains carefully arranged around their internal walls. Lastly, burials included pits of varied sizes whose main and sometimes only content was human remains. Neither pits classified as storage pits nor those classified as offering pits contained any type of human remains. These four functional categories of features tend to occur in different architectural sectors of the site (Table 6.1). Hearths and storage pits, for example, were found mainly on the upper, partially closed sector of Structure 2. Offering pits and burials, however, were found only on the uppermost closed sectors of Structure 1. No features were found in the open sectors of the lower terraced levels of either of the two structures. Something similar occurs with at least one class of artifacts: ceramic vessels. Fragments of ceramic vessels recovered at Pampa Chica were classified into six broad functional categories mainly on the basis of the analysis of their morphological attributes. These morphologicalfunctional categories include (1) bottles, (2) bowls, (3) jars, (4) small ollas, (5) large ollas, and (6) plates. These

Jalh Dulanto

categories correspond to a high degree to morphological-functional categories defined in cross-cultural ethnographic studies (see especially Henrickson and McDonald 1983). Small ollas from Pampa Chica, for example, closely resemble types of vessels cross-culturally defined as cooking pots; bowls and plates resemble those defined as serving and eating vessels; large ollas resemble those used for short- or long-term storage of dry or liquid foods; bottles resemble those used to transport and pour small amounts of liquids. As in the case of the categories of features presented above, these six broad functional categories of ceramic vessels tend to occur in different architectural sectors of the site. I did a correspondence analysis of the frequencies of pottery fragments per functional category and architectural sector (Table 6.2 and Figure 6.6). Plates and bowls (i.e., ceramic vessels used to serve and consume food) tend to occur in the open sectors of the lower terraced levels of Structure 1 (Sectors I, III. 1, and III.2, though also V).6 Small ollas, large ollas, and jars (i.e., ceramic vessels used to cook and store foods and beverages), in contrast, tend to occur in one specific room in one of the lower sectors of Structure 1 (R 6, Sector III.3) and in the partially closed upper sector of Structure 2 (Sector VIII); bottles (i.e., ceramic vessels used to transport and pour small amounts of liquids) tend to occur in at least one of the closed sectors of the upper terraced levels of Structure 1 (Sectors IV. 1 and IV.2). Finally, a similar pattern of differential distribution of remains among architectural sectors can be noted for at least two classes of non-artifactual organic remains: (1) animal bones and (2) shells. The upper sector of Structure 2 has significantly higher frequencies of these classes of organic remains, followed in order by the upper and then the lower sectors of Structure 1. These organic remains can be interpreted as the product of food preparation activities for several reasons. Animal bones, for example, often show marks of cutting and/or burning. Shells, likewise, were often recovered in concentrations of varied sizes that usually contained only shells of the same species and from which the food had always been extracted. Botanical remains also were often recovered in the form of concentrations of remains of the same species that usually included seeds and other parts of the plants (see below). To summarize, the analysis of the distribution of four distinct categories of features, six broad functional categories of pottery vessels, and at least two classes of nonartifactual organic remains among distinct architectural sectors of Pampa Chica shows a clear correlation between specific types of activity areas and types of architectural

Ancestor Cult Practices at Pampa Chica

107

ceramic vessels, they are the only ones that show Funerary Total Hearths Storage Pits Offering Pits high frequencies of small Contexts decorated bottles. Phase 1A Interestingly, the 1 0 0 0 1 Sector m types of activities that can be inferred from the differ0 3 0 Sector IV ences in the distribution of 0 0 17 9 8 Sector Vffl types of archaeological rePhase IB mains among different sectors are consistent w ith the 1 0 0 0 1 Sector I types of activities that can T 1 0 0 1 Sector II be inferred both from the 0 0 0 0 0 Sector DI architectural characteristics of each sector and from 2 0 0 0 SectoT IV the spatial arrangement of 1 1 0 0 0 Sector V these types of remains 14 14 0 0 0 Sector Vffl within each sector. For example, in the area exca3 43 27 8 rotal vated in the upper terraced le\els of Structure 2 (SecTable 6 2. Number of sherds per functional category of pottery vessel and uchitectiirai sector tor VIII), storage pits were found arranged in two parBig Ollas Small Ollas Plates Total Sector Bowls Bottles Jars allel rows u ith a large 84 0 0 21 0 1 63 0 hearth and several ash lenses in between, in a 132 39 6 68 7 10 II manner that closely re4 0 4 8 0 0 m.i 0 sembles the internal orga2 5 0 12 0 5 0 m.2 nization of maize-beer 0 0 14 0 0 14 m.3 0 production areas known 3 from archaeological and 0 X 0 1 1 3 IV1 ethnographic cases from 22 1 4 4 3 10 0 rv.2 the north coast (see 5 0 0 15 10 V 25 55 Bruning 1988; Camino 2 9 2 1 4 71 0 7 79 28 vn 19X7; Shimada 1994-221, 224. figs 8.40. 8.42). Also 549 39 137 48 271 18 36 Total in common with these better known cases, the area excavated in Pampa Chica spaces. The partially closed sector of the upper terraced levels of Structure 2 (Sector VIII), for example, shows a shows, comparatively to other areas of the same site a concentration of hearths, storage pits, and ceramic \es- concentration ot large ollas and jars of the t\pe normalU sels used for food preparation and storage, and large used in the preparation and storage of maize beer. Morequantities of non-artifactual organic remains that seem over, in comparison with other areas, it also shows the to be the by-product of food preparation activities. The highest variety of ceramic \ essels of different shapes and open sectors of the lower terraced levels of Structure 1 sizes that normally are used for storage and distribution (Sectors I and III), in contrast, show only a concentra- of maize beer (see Shimada 1994224). Furthermore, artion of ceramic vessels used for serving and consuming eas nearby yielded large concentrations of cobs, stems, food and liquids. They lack features and show the low- and leaves of maize that seem to have been left behind est concentration of organic remains in the site. The after the removal of the kernels needed for the preparaclosed sectors of the upper terraced levels of Structure 1 tion of maize beer. Also in common with ethnographic (Sectors II and IV) show a concentration of burials and cases is the fact that maize-beer production in Pampa offering pits. Although they contain different types of Chica seems to be associated with some form of largeTable 6.1. Number of features per functional category and architectural sector

]0H

Jalh Dulanlo

scale food production. The large number of hearths and concentrations of animal bones, shells, and botanical remains in the area suggests that the two types of activities probably took place at the same time. Interestingly, both types of activities occur within an area (R 22 and 23) that is closed enough to ensure controlled firing but open enough to provide the necessary ventilation. The situation is similar in the lower terraced levels of Structure 1 (Sectors I and III), and perhaps also of Structure 2 (Sector VII). In these areas the relative absence of features is consistent with their architectural characterization as open spaces designed for the gathering of people and public display of objects and persons. The presence of features like storage pits, burials, and perhaps e\en offering pits would have imposed limita-

tions on the movement of people within them. Furthermore, the relatively high concentration of ceramic vessels used for serving food and liquids and the lower concentration of organic remains (because these were all consumed) suggest that at least on some occasions the gatherings that took place in these areas involved some form of feasting. Finally, in the case of the upper sectors of Structure 1 it is interesting to note that both burials and offering pits occur only in the sunken areas in the center (R 8) or to the side (R 4) and that these are the only areas of the site that contain human remains. Interestingly, these areas where human remains occur are the most secluded and closed areas of the site, which suggests a certain intentional separation of the areas used for the storage of human remains and the series of activities that seem to

0.5 / • ni-i

0 \ Plates

\ Bowls /

I III-3

•

/IH-2 Large Ollas

\

/

• Jars

\

/

Small Ollas /

c

o '53 §-0.5 IV-2

-1.0

-1.5

IV-1 Bottles

open spaces closed spaces

-1.0

-0.5

0.5

1.0

Dimension I Figure 6.6. Functional categories of ceramic vessels in architectural spaces at Pampa Chica. Note the correspondence between cluster? of sectors and functional categories of ceramic vessels.

Ancestor Cult Practices at Pampa Chica

have involved the repeated manipulation of these remains (see below).

texts 8 and 9, respectively in R 17 and 13). Only one of the ten contexts contained a complete, articulated skeleton plus other types of remains (Context 10). The other Human Remains nine contexts contained only incomplete, disarticulated skeletons (Contexts 1 9), and in seven of the nine the Detailed analysis of the location and composition bones recovered were mostly small bones like epiphyses of different types of contexts that contained human re- and phalanges (Contexts 1 7). In all nine cases, regardmains shows that these contexts are the product of a com- less of whether they were pits or small rooms, the features plex sequence of activities involving the burial, removal, showed some mark of disturbance: the stones co\Lring manipulation, and reinterment of the remains of several the pit had been removed (Contexts 2 . 3 , 4 , 7) and or the dead individuals. bones found in them were mixed in no apparent order The sectors of the upper terraced levels of Structure (Contexts 3, 4. 5, 7). In at least four of these nine con1 yielded ten discrete contexts that contained human re- texts the features also contained bones from more than mains. Three of them were found in Sector II, in the north one individual (Contexts 2 4, X. 9). half of the structure (Contexts 5 and 6 [Figure 6.7] and A detailed comparison of these unusual contexts Context 9); six in Sector IV, in the south half (Contexts along nine different \ariables made it possible to iden2,4, 10 [Figure 6.8] and Contexts 1, 3, 8); and one (Con- tify a rather complex set of e\ents that seems to ha\e text 7) in Sector V. With the notable exception of some been involved in the treatment of human remains in small tufts of human hair found in Sector VI11, no hu- Pampa Chica. These nine \ariables are (1) the type of man remains were found in any of the other areas of the feature (e.g , pit, floor of closed patio, small room; (2) site. Of the ten contexts found in the upper terraced lev- its size; (3) its shape; (4) the presence absence of marks els of Structure 1, six are pits (Contexts 2. 3, 4, 5, 7, 10), of disturbance (e.g.. removed co\ennij stones, mixed two are concentrations of bones found on the floors of contents); (5) the minimum number of indi\ iduals within the sunken areas in the center or to the side of the closed the same feature, (6) the number of bones per skeletal patios (Contexts 1 and 4, respectively in R 4 and 8), and part; (7) the presence absence of articulated bones. (8) two are sets of bones deposited within small rooms (Con- the position and distribution of bones within the feature;

Figure 6.7. Sector II and associated contexts.

no

Jalh Dulanto

can calculate the minimum number of times the remains of different individuals were deposited and later removed. The idea is simple, if in a feature that can only contain one complete flexed individual we find part of the bones from at least three different individuals that were originally complete (see below), then these individuals were necessarily deposited in the feature at different times. The presence or absence of marks of disturbance provides further evidence of the reopening of a feature to remove the existing contents or deposit new ones.

Figure 6.8. Sector 11 and associated contexts.

and (9) the presence absence of bones of the same indi\idual in different contexts. These nine variables inform us about the sequences of events surrounding the treatment of human remains in Pampa Chica insofar as the} inform us about the formation and transformation of the contexts that contain human remains. For example, different types of features impose different limitations to the access and manipulation of the human remains they contain. Human remains stored in a room are easier to access than those buried in a pit. while human remains deposited on the floor of an open area are easier to access than those stored in a closed room. Thus, it is reasonable to suggest that the use of one or another type of feature to deposit human remains indicates, to some extent, whether it was desired to limit or facilitate access to the human remains (as required by different types of occasions). In the same manner, the size and shape of a feature impose limits to the number size, and position of individuals buried in them. They can be considered as an indication of the maximum number of complete individuals of a certain size who can be buried in a feature in a certain position. In turn, if we compare the maximum number of complete individuals a feature can contain with the minimum number of incomplete individuals actually recovered in it, then we

A basic assumption in this manner of calculating the minimum number of times a feature was reopened and its contents replaced is, of course, that there was certain intention to deposit in and extract from the features complete skeletons rather than only certain skeletal parts. The frequency of bones per skeletal part can be considered an indication of such intention. High frequencies of small bones like epiphyses and phalanges that are easily lost and left behind, which is the case in many of the contexts excavated in Pampa Chica (see below), tend to indicate that the intention was to remove complete skeletons rather than individual skeletal parts. The presence or absence of articulated bones, and their position within the pit, on the other hand, can be used to determine whether the remains of an individual were originally complete and, if so, in what position the corpse was disposed within the feature. If evidence indicates that the intention was to deposit and remove complete skeletons and not specific skeletal parts, then the presence absence of bones from the same individual in different contexts can be considered to be an indication of the movement of human remains from one context to another. The comparison of the ten contexts excavated in Pampa Chica along these nine variables permitted the identification of at least five distinct types of contexts that can be interpreted as the archaeological product of different stages in the sequence of events surrounding the treatment of human remains at the site. These five types of contexts or stages are the following: (1) undisturbed primary burials, (2) disturbed primary burials, (3) disturbed secondary burials, (4) concentrations of human remains on floors, and (5) small rooms containing human remains. There is only one example of an undisturbed primary burial. Context 10 consisted of a shallow circular pit (99 x 82 x 22 centimeters) that contained the complete and fully articulated skeleton of a male individual, 19 to 21 years old, who was flexed and wrapped in several layers of textiles, and buried resting on his

Ancestor Cult Practices at Pampa Chica

left side with a gourd vessel next to his head. The context was identified as a primary burial because the skeleton was complete and fully articulated and because the pit did not show any marks of disturbance. Thus, it was clear that the context was the product of the final disposal of an individual shortly after his death. Disturbed primary burials include several possible cases (Contexts 2, 3 [?], 4 [?], and 5 [?]) consisting of pits that contained incomplete skeletons of one or more individuals. These burial contexts show clear marks of reopening, removal of contents, and in some cases also replacement of contents. In all cases the pits were circular and rela tively small (respectively 170 x 60 x 42, 110 x 80 x 49, 70 x 66 x 117; and 104 x 98 x 140 centimeters). Although none of the pits was big enough to contain more than one flexed individual at a time, some of them contained the skeletal remains of more than one individual. Context 2. for example, contained the remains of at least two males (20-25 and 15-17 years old) and a third individual (>18 years old) whose sex could not be determined, while Context 4 contained the remains of at least two males (one 30—35 and the other 15-17 years old). Contexts 3 and 5 contained the remains of at least one individual each, both of them possible males (>l8 years old). In all cases, the skeletal remains found in the pits consisted mainly of small, disarticulated

III

0 1

Frequencies of skeletal elements recovered in the upper levels of Structure 1

2 Figure 6.9. Frequencies of different human bones found in the site \ott that the highest frequencies correspond to those bones located in the areas of the mummy bundle of a flexed individual that are most exposed to deterioration.

bones like phalanges and epiphyses, or bones that would have been located in the most exposed areas of the mummy bundle of a flexed individual (Figure 6.9), that is, bones that could be easily lost during the removal of a mummy bundle from its burial pit. This indicates that all contexts were opened at some point and their contents removed and in some cases replaced with the remains of another individual. Since in almost all cases (Contexts 3, 4, 5) the skeletal remains were found incomplete and disarticulated it is very difficult to know whether the contexts are disturbed primary or secondary burials. In at least one case (Context 2), however, the bones from the feet of one of the individuals were still

articulated and in their original position toward the side ot the pit, suggesting that they were left behind after the removal of a primary burial. There is at least one case of a disturbed secondarv burial (Context 4, though perhaps also Contexts 2 , 5 The remains of at least one of the indhiduals buried in Context 4 (the male 15 17 years old) could indicate that this context was the result of a complex sequence of events that included the burial of the complete or nearly complete skeleton of a male individual in a first pit (Context 2); the reopening of that first pit and the remo\al of mo t of the skeletal parts of the individual buried in it; the reburial or second burial in a second pit (Context 4)

112

of most of the skeletal parts that were extracted from the first pit; and the reopening of this second pit and the removal of most of the skeletal parts of that same individual from this second pit. At least two facts permitted the reconstruction of this four-step sequence. First, bones that seem to come from the same individual, a male 15 to 17 years old, were found in two contexts (Contexts 2 and 4). Not only are the skeletal remains of at least one of the individuals identified in Context 2 complementary with those of one of the individuals identified in Context 4. but also at least two bones, one from each context, articulate with each other (respectively, a distal epiphysis of a right femur and a proximal epiphysis of a right tibia). Second, the number and types of bones found in each context indicate that the skeletal remains of the individual in question were first deposited in Context 2 and then moved to Context 4. Not only were more bones recovered from Context 4 than from Context 2. but also larger bones, including part of the skull. The basic assumption behind this last observation is that the number and size of the bones left behind or lost during the removal of skeletal remains buried in a pit is a function of the deterioration and disarticulation of such remains. The deterioration and disarticulation of skeletal remains, in turn, can be considered a function of the time elapsed since the death of the individual, as well as of the number of times the remains have been removed and manipulated. Concentrations of human bones on floors comprise two cases (Contexts 1 and 6). Both are accumulations of bones that were found on the floors of the two closed patios of the upper terraced levels of Pampa Chica. Context 1 consists of a total of three small bones found scattered between Contexts 2 and 3 in Room 4, the closed patio of Sector IV. Context 6 is a clearly bounded concentration of fifteen bones that was found on the south side of Room 12, the small rhomboidal sunken patio of Sector 11. The location of these two concentrations in relation to some of the nearby pits and comparison of the sets of bones found in them with those of the nearby pits suggest that these two concentrations could be composed of bones lost during the removal, manipulation, and/or transportation to new locations of skeletal remains originally placed in some of the nearby pits (respectively, Contexts 3 and 5). Lastly, small rooms containing human remains also comprise two cases (Contexts 8 and 9). The small rooms are two located on the east side of the upper terraced levels of the site (respectively, Room 17 in Sector IV and Room 13 in Sector II) that seem to have been used for the short- or long-term storage of human remains. At some point, both of them contained the skeletal remains

Jalh Dulanto

of at least two individuals: one male (30-35 years old) and one individual of undetermined sex (12-17 years old) in Context 8, and one male (20-30 years old) and another individual of undetermined sex (>18 years old) in Context 9. Unfortunately, both small rooms were looted during a later occupation of the site (Phase 2) and most of their contents were spread around within a radius of approximately 2 meters, making it very difficult to determine whether the skeletal remains found in them were deposited there for the first time or had been previously buried in pits one or more times. However, in at least one case (Context 8), evidence suggests that part of the skeletal remains stored in the small rooms had been buried before in a nearby pit (Context 4). As in a case already discussed (see the discussion of disturbed secondary burials, above), the fact that at least two sets of bones, one from each context, are complementary with each other, and at least two bones, one from each set, articulate with each other (a right zygomatic and a frontal bone), suggests that the skeletal remains of one single individual were deposited in these two contexts at different times. Likewise, the fact that more and larger bones were found in Context 8 than in Context 4 suggests that the skeletal remains of the individual in question were first buried in Context 4 and later deposited in Context 8. Reconstruction of the Sequence of Events The detailed comparison of the location and contents often burial contexts excavated at Pampa Chica permits the identification of different types of processes involved in their formation and transformation (i.e., repeated burial, removal, transportation, storage, and reburial). Different types of contexts seem to be the result of different combinations of processes. Identification of these combinations of processes, in turn, permits the reconstruction of various possible sequences of events that seem to have surrounded the treatment of human remains at the site. Figure 6.10 summarizes these various possible sequences of events. The analysis suggests that skeletal remains recovered at Pampa Chica could have entered the archaeological record by being either stored in a small room or buried for the first time in a pit. If first stored in a small room, the skeletal remains either remained in the small room or were later extracted to be buried in a pit (?). If initially buried in a pit, however, the skeletal remains were later extracted and either moved to be stored in a small room or buried for a second time in another pit. In the process, and as a result of transportation and perhaps also manipulation of human remains, some bones

Ancestor Cult Practices at Pampa Chica

113

PRIMARY BURIALS

I

SMALL ROOMS

SECONDARY BURIALS

\7 FLOORS

Figure 6.10. Reconstruction of the movement of bones between contexts. The rectangles stand for different types of features that contained skeletal remains. The arrows represent movement of skeletal remains between different types of features.

ended up being deposited on the floors of the closed patios where the pits are located. Reburial of human remains in the same pit or movement between pits more than twice also could have contributed to the formation and transformation of contexts excavated at the site— but this is not evident in the sample analyzed. It is important to note that the last events in almost all possible sequences usually correspond to the extraction of the skeletal remains from a pit or small room. Of a minimum of eleven individuals found at the site, only three are represented by substantial skeletal parts. The other eight are only represented by small bones and bones that because of their location seem to have been left behind after the extraction of the most important and visible parts of the skeleton. Therefore, it is reasonable to conclude that, in most cases, human remains were housed only temporarily in Pampa Chica before being transported and disposed of somewhere else. It is also important to note that most of the human remains treated in Pampa Chica correspond to males. Of the eleven individuals recovered at the site, five were identified as males and two as probable males, while the remaining four were not sexed because of the lack of diagnostic traits. The age of these individuals extends from the 12 to 17 range up to 30 to 35 years old. No individuals were identified as either females or children (Salcedo 1996).

Conclusions In summary, analyses of the location of Pampa Chica relative to other sites and natural resources, the organization of architectural spaces and activity areas within the site, and the processes involved in the formation and transformation of the contexts containing human remains found

in it, permit not only the identification of important similarities between Pampa Chica and places of ancestor veneration described in the ethnohistoric record, but also discovery of some important differences between the two. The analysis of site location, for example, shows that Pampa Chica resembled the local ceremonial centers described in the ethnohistoric documents as communal sepulchers in at least two features. Like these sites, Pampa Chica was probably located overlooking one or more habitation sites and it also was probably located to mark the preferential access to certain critical resources of the communities that lived in those sites. Nonetheless, in contrast to them, Pampa Chica was not located close to these sites but at least 1 kilometer from them. In this manner it more closely resembles other types of places of ancestor veneration that, according to the documents, are outside the area of daily life activities—for example, places used for the enshrinement of individual mummified ancestors or places that were identified with more distant ancestral heroes. It is important to note, however, that distance alone is not enough to determine either the type of place of ancestor veneration or the genealogical position of the ancestors venerated in that place. The documents indicate that a general correlation between geographic and genealogical distance could be expected. The more distant the place of ancestor veneration from a habitation site, the more distant in the genealogical structure the ancestor venerated in that place was from the descendants who lived in that habitation site. Consequently, the more distant the place of ancestor veneration, the more inclusive the group of descendants that worshipped the ancestor associated with that place. But only the analysis of a representative sample of habitation sites and various types of places of ancestor venera-

114

tion will enable us to determine whether such correlation existed on the central coast during the late Initial Period and Early Horizon. If that is the case, it could be expected that more distant places will show more stylistic variability in associated material culture, and that the number of individuals buried in them will be reduced as human ancestors are replaced by superhuman ones. The analysis of the organization of architectural spaces and activity areas also shows important similarities and differences between Pampa Chica and some of the places of ancestor veneration described in the documents. Like the local ceremonial centers described in the documents as communal sepulchers, Pampa Chica is composed of permanent, durable, and highly visible buildings that were especially designed, built, and used to store human remains as well as to carry out rituals among large numbers of people congregated in that place. It also represents, in that sense, a significant investment on the part of one or more small groups in a program of ancestor veneration that demanded the incorporation of the human remains of a group of dead individuals into a public space. However, in contrast to the local ceremonial centers described in the documents, the analysis of architectural spaces also shows that Pampa Chica had a more marked separation between more closed and secluded areas used to store human remains and more open and accessible spaces used for the gatherings of people. Moreover, the analysis of activity areas indicates that the former were also used in more intimate types of ritual in which human remains were manipulated (consultation of ancestors and divination [?]), whereas the latter were used in more public types of rituals (public gatherings and displays) that involved feasting for which food and beverage were produced in large scale in other areas of the site. Interestingly, the analysis of architectural spaces also shows that the more intimate types of rituals involved manipulation of human remains and also that the production of food and beverage within the site involved the division of the group of participants congregated at the site into two equivalent groups (i.e., dual or moiety organization; but see note 5). Lastly, the analysis of the processes involved in the formation and transformation of contexts containing human remains also reveals important similarities and differences between Pampa Chica and the local ceremonial centers described in the ethnohistoric documents. The most important similarity between the two is perhaps that in both cases at least part of the activities carried out within the site frequently involved the repeated burial, removal, movement, storage, and reburial of human remains and probably even their use as ritual parapherna-

Jalh Dulanto

lia. However, Pampa Chica was not exactly a cemetery or communal sepulcher. Rather, the contextual analysis of human remains suggests that Pampa Chica was used mainly for the temporal dwelling of the remains of a very specific group of dead individuals: adolescent and adult males, Although the similarities and dissimilarities between the spatial and material dimensions of ancestor cult practices described in the documents (i.e., mainly for seventeenth century A.D. Caxatambo and Huarochiri) and those we can infer from the archaeological record (i.e., for seventh to third century B,C. Pampa Chica) are revealing, it is important to note that the sample analyzed is small (i.e., only one site and only ten contexts containing human remains). Only future extensive excavations of sites like Pampa Chica will allow us to determine (1) to what degree the patterns observed at Pampa Chica can be interpreted as the product of ancestor cult practices or other type of ritual practice and (2) to what degree the specific type of ancestor cult practices suggested here for Pampa Chica were typical of other late Initial Period and Early Horizon societies of the central coast and other Central Andean areas. Addressing these two questions is particularly important for future studies on the spatial and material dimensions of ancestor cult practices as well as Central Andean prehistory. The late Initial Period and Early Horizon have been characterized as a period of intense crisis for Central Andean coastal societies (Burger 1992:184—90). If proven, intensification of ancestor cult practices by small groups of villages in coastal areas during this period could be regarded as one more piece of evidence of such crisis (in terms of increasing competition between previously cooperating groups). A cknowledgments Archaeological investigations at the site of Pampa Chica were conducted as part of the Proyecto Arqueologico Tablada de Lurin directed by Dr. Krszyztof Makowski of the Pontificia Universidad Catolica del Peru. I would like to thank Dr. Makowski for the opportunity to work with him, I also thank Susan Gillespie, David Grove, Steve Leigh, one anonymous reviewer, and, especially, Helaine Silverman for their many constructive suggestions to improve this paper. Notes 1. In this study the late Initial Period is the period between the abandonment of the Initial Period coastal tradition of U-shaped temples and the beginning of the

Ancestor Cult Practices at Pampa Chica

Early Horizon (ca. 900/700-500 B.C.). Early Horizon is the period of Chavin influence in the central coast (ca. 500-200 B.C; see Burger 1992:58, 184), I situate Pampa Chica within these two periods on the basis of seven radiocarbon dates, all of them from hearths and other primary contexts found directly on occupation floors. These dates are the following: 2640 ± 70 B.P. (Gd-11197); 2540 ± 60 B.P. (Gd-11202); 2460 ± 60 B.P. (Gd-11200); 2440 ± 40 B.P. (Gd-7653); 2410 ± 70 B.P. (Gd-11197); 2330 ± 40 B.P. (Gd-7651); 2210 ± 40 B.P. (Gd-7648). 2. Mortuary ritual refers to the rites of passage that mark the separation of a deceased person from the living; ancestor enshrinement refers to rites that seem to have involved a change of ancestral status among deceased who have been already separated from the living; cyclical rituals of ancestor veneration refers to different types of rituals undertaken on the occasion of significant transitions in biological, environmental, agrarian, and even social cycles and that involve some act in which the ancestors are directly addressed. All three types of ritual occasions involved some form of ancestor veneration and, therefore, all three should be taken into account in the study of the organization of ancestor worship (see Salomon 1995; also see Morris 1991 on the distinction between mortuary rituals and ancestor cult that was made by Gluckman [1937]). 3. In coastal regions corpses were usually disposed in cemeteries rather than open sepulchers, preventing in this manner their frequent curation and direct manipulation. However, they still seem to have been regarded as important material referents of human ancestors. Indeed, the lack of frequent curation and direct manipulation of the corpses could be merely a response to the warmer and more humid coastal environment that prevents their preservation, rather than an indication of substantial differences in the organization of ancestor worship or in the forms of social organization behind these practices (cf Isbell 1997). 4. The original height of walls was calculated measuring the volume of the preserved walls and the volume of rubble associated with them. 5. The fact that the spatial organization of a building involves a separation of space into symmetrical halves indicates that at least the types of activities carried out in such spaces were probably organized in terms of a separation of one group of people into two equivalent groups. It does not necessarily imply that society as a whole was necessarily organized by a system of dual divisions, only that the types of activities carried out in that space were. To the extent that the activities in question are part of a program of ancestor veneration, it could

115

be argued that in these societies at least descent was probably organized by a system of dual divisions. 6. Sectors showing evident internal differences between open and closed spaces or in the amount and type of associated materials were subdivided for analysis into smaller spatial units. Sector III was subdivided into two open sectors (III.l, III.2) and one closed sector (III.3). Interestingly, it was readily evident during excavation that in contrast to the two open spaces, the closed one contained an unusually high number of fragments of large ollas used for storage. Sector IV, in turn, was subdivided into two distinct closed sectors (IV. 1, IV.2). 7. The number of individuals indicated for these contexts is always the MNI. Sex, age, and size as indicated by the individual bones were taken into consideration in determining the MNI. References Agurto, S., and L. Watanabe 1974 Inventario, Catastro, v Delimitacion del Patrimonio Arqueologico del Valle Bajo del Rio Lurin. Archivo del Centro de Investigacion y Restauracion de Bienes Monumentales. Lima: Instituto Nacional de Cultura. Briining, H. 1988 Estudios Monograficos del Departamento de Lambayeque. Tomo I: Lambayeque. Chiclayo: Dionisio Mendoza. Buikstra, J. 1995 Tombs for the Living...Or...for the Dead: The Osmore Ancestors. In Tombs for the Living: Andean Mortuary Practices, T. Dillehay, ed., pp. 229-80. Washington D.C.: Dumbarton Oaks. Burger, R. 1992 Chavin and the Origins of Andean Civilization. London: Thames and Hudson. Camino, L. 1987 Chicha de Maiz: Bebida y Vida del Pueblo de Catacaos. Piura: CIPCA. Doyle, M. 1988 The Ancestor Cult and Burial Ritual in Seventeenth and Eighteenth-Centwy Central Peru. Ph.D. dissertation, Department of History, University of California, Los Angeles. Ann Arbor, Mich.: University Microfilms.

116

Dulanto, J. 1992 El ciclo funerario en las comunidades rurales de Caxatambo en el Sigh XVII. Primer Encuentro de Estudiantes de Arqueologia de la Pontificia Universidad Catolica del Peru, Lima, 1994 Prospeccion arqueologica en el valle de Lurin, Willay 41:3-6. Dulanto, J,, L. Caceres, M. I. Velarde, and L. F. Villacorta In press Pampa Chica: un analisis funcional de un sitio de fines del Periodo Inicial y del Horizonte Temprano en la Costa Central del Peru. In El Formativo de la Costa Central, R, Burger and K. Makowski, eds. Lima; Fondo Editorial de la Pontificia Universidad Catolica del Peru. Duviols, P. 1979 Un symbolisme de l'occupation, de l'amenagement et de Texploitation de Tespace: Le monolithe Huanca et sa fonction dans les Andes prehispaniques. L Homme 19(2): 19—37. 1986 Cultura Andina y Represion: Procesos y Visitas de Extirpacion de Idolatrias, Caxatambo, Sigh XVII. Cuzco: Centro de Estudios Rurales Bartolome de Las Casas. Earle, T, 1972 Lurin Valley, Peru: Early Intermediate Period Settlement Development. American Antiquity 11A61-11. Engel, F. 1983 Prehistoric Andean Ecology. New York: Humanities Press and Department of Anthropology, Hunter College/CUNY. Feltham, J. 1983 The Lurin Valley, Peru: AD 1000-1532. Ph.D. dissertation, Institute of Archaeology, University of London. 1984 The Lurin Valley Project: Some Results for the Late Intermediate and Late Horizon Periods. In Current Archaeological Projects in the Andes: Some Approaches and Results, A. Kendall, ed., pp. 45-73. BAR International Series 210, Oxford. Gluckman, M. 1937 Mortuary Customs and the Belief in Survival after Death among the South-Eastern Bantu. Bantu Studies 11:117-36.

Jalh Dulanto

Henrickson, E., and M. McDonald 1983 Ceramic Form and Function: An Ethnographic Search and an Archaeological Application. American Anthropologist 85:630-43. Huertas, L. 1981 La Religion en una Sociedad Rural Andina (Siglo XVII). Ayacucho: Universidad Nacional San Cristobal de Huamanga. Isbell, W. 1997 Mummies and Mortuary Monuments: A Postprocessual Prehistory of Central Andean Organization. Austin: University of Texas Press. Lastres, J. 1953 El culto a los muertos entre los aborigenes peruanos. Peru Indigena 4(10/11): 63-74. Moodie, R. 1931 Roentgenologic Studies of Egyptian and Peruvian Mummies, vol. 3. Chicago; Field Museum of Natural History. Morris, I. 1991 The Archaeology of Ancestors; The Saxe/ Goldstein Hypothesis Revisited. Cambridge Archaeological Journal 1(2): 147-69. Palacios, J. 1988 La secuencia de la ceramica temprana del valle de Lima en Huachipa. Gaceta Arqueologica Andina 16:13-24. Patterson, T. 1966 Pattern and Process in the Early Intermediate Period Pottery of the Central Coast of Peru. University of California Publication in Anthropology, vol. 3, University of California Press, Berkeley. Rivera, M. 1995 The Preceramic Chinchorro Mummy Complex of Northern Chile: Context, Style, and Purpose. In Tombs for the Living: Andean Mortuary Practices, T. Dillehay, ed., pp. 43-78, Washington, D.C.: Dumbarton Oaks. Salcedo, L. 1996 Analisis e inventario de los restos oseos humanos excavados en el sitio Formativo Tardio Pampa

Ancestor Cult Practices at Pampa Chica

Chica S77 (Valle de Lurin). Proyecto Arqueologico Tablada de Lurin, Pontificia Universidad Catolica del Peru, Lima. Manuscript. Salomon, F. 1995 "The Beautiful Grandparents": Andean Ancestor Shrines and Mortuary Ritual as Seen Through Colonial Records. In Tombs for the Living: Andean Mortuary Practices, T. Dillehay, ed., pp. 315-53. Washington D.C.: Dumbarton Oaks.

117

Silva, J., R. Garcia Soto, K. Hirth, and J. Pinilla Blenke 1982 El valle del Rimac hace 2,500 afios: HuachipaJicamarca. Boletin de Lima 21.59—68. 1983 El Formativo en el valle del Rimac: HuachipaJicamarca. Arqueologia v Sociedad 9:2-83. Tello, J. C. 1929 Antiguo Peru: Primera Epoca. Lima: Comision Organizadora del Segundo Congreso Sudamericano de Turismo.

Shimada, I. 1994 Pampa Grande and the Mochica Culture. Austin: University of Texas Press.

Vreeland, J. M., and A. Cockburn 1980 Mummies of Peru. In Mummies, Disease and Ancient Cultures, A. Cockburn and E. Cockburn, eds., pp. 135—74. Cambridge: Cambridge University Press.

Silva, J. 1996 Prehistoric Settlement Patterns in the Chi lion River Valley, Peru. Ph.D. dissertation, Department of Anthropology, University of Michigan, Ann Arbor.

Zuidema, R. T. 1973 Kinship and Ancestor Cult in Three Peruvian Communities: Hernandez Principe's Account of 1622. Bulletin de I'lnstitut Francais d'Etudes Andines 2(1): 16-33.