La Arqueología de Cerro Baúl

1. Excavaciones Arqueológicas
Williams, P. R. and J. Isla 2002. Excavaciones Arqueológicas en el Centro
Administrativo Wari de Cerro Baúl. Gaceta Arqueológica Andina 26: 87-120.

2. Interacción Imperial in los Andes: Wari y Tiwanaku

Williams, P. R. and D. Nash. 2002 Imperial Interaction in the Andes: Wari and Tiwanaku
at Cerro Baúl. in W. Isbell & H. Silverman, eds. Andean Archaeology. New York: Plenum.
Pp. 243-265.

3. Arquitectura y Poder en la Frontera Wari-Tiwanaku

Nash, D. and P. R. Williams 2005. Architecture and Power on the Wari—Tiwanaku
Frontier . Archaeological Papers of the American Anthropological Association 14. Pp.
151-174.

4. Organización Política Wari

Nash, D. and P. R. Williams 2009. Wari Political Organization: The Southern Periphery.
Andean Civilization: A Tribute to Michael E. Moseley. J. Marcus and P. R. Williams, eds.
Los Angeles: Cotsen Institute of Archaeology Press, UCLA. Pp. 257-276.

5. Encuentros en el Reino Wari

Williams, P. R., D. J. Nash, M. Moseley, S. deFrance, M. Ruales, A. Miranda, D.
Goldstein 2008 Encuentros en el Reino Wari. Boletin de Arqueologia PUCP 9: 207-232.

6. Vistas al Apu
Williams, P. R. and D. J. Nash 2006 Sighting the Apu: A GIS Analysis of Wari Imperialism
and the Worship of Mountain Peaks World Archaeology 38(3):455-468.

7. Ritual Religiosa y Expansión Estatal Wari

Williams, P. R. and Donna Nash 2016. Religious Ritual and Wari State Expansion. In
Ritual and Archaic States, edited by Joanne M. A. Murphy. Gainesville: University Press
of Florida. 131-156.
 9
Architecture and Power on the Wari–Tiwanaku
Frontier
Donna J. Nash
The Field Museum, Chicago
and
Patrick Ryan Williams
The Field Museum, Chicago

ABSTRACT
The Wari Empire expanded and maintained control over many areas in the Andes for nearly four centuries (600–
1000 C.E.). This chapter documents changes in power relations and political institutions on the Wari–Tiwanaku
frontier. The settlements of both polities are well documented along their border in the Moquegua Valley of southern
Peru where Wari controlled their provincial settlements from the lofty heights of Cerro Baúl. We assess the changing
nature of the incorporation of different social groups within the Wari political structure of the frontier province over
the course of the Middle Horizon. As Tiwanaku social groups joined the Wari colony, new sets of public expressions
of power emerged, both from within these Tiwanaku groups and from the Wari administration itself. By examining
artifactual remains and the design of architectural spaces, we elucidate the changing power relations between Wari,
their subject populations, and their Tiwanaku neighbors.
Keywords: use of space, multiethnic interaction, state institutions, political organization, empire

W ari architecture is a material manifestation of power

and holds important clues to understanding how Wari
state officials managed resources and legitimized their con-
trol. Recent research at Cerro Baúl and its satellites on the
Wari southern frontier has elucidated the nature of power re-
lations, agency, and ideology between Wari representatives
and the various groups with which they interacted. Cerro
Baúl, located in the Moquegua Valley of Peru (Figure 9.1),
has been a prime focus of research seeking to understand
the relationship between the Wari and the Tiwanaku during
the Andean Middle Horizon, about 600–1000 C.E. (Williams
2001), because of the close proximity of large Tiwanaku
colonies in the middle valley to Wari settlements in the upper
drainage (Goldstein and Owen 2001; Moseley et al. 1991;
Owen and Goldstein 2001). In this chapter, we explore the
different architectural venues of power and the state institu-
tions they reflect in two phases of occupation in the upper
Moquegua drainage at the Wari provincial center of Cerro
Baúl and at surrounding subsidiary sites.
The Wari built an environment imbued with symbols
that were used to communicate their legitimacy in a num-
ber of ways (see Rapoport 1990). These symbols are most
explicit in iconographic representations of ceremony por-
trayed on portable objects that served as accessories to
state-sponsored activity. By studying the contexts of these
practices in their spatial venues, these activities can be un-
derstood in greater depth as well-developed institutions de-
signed to manage resources. The Wari repertoire of space
exhibits multiple mechanisms to maintain power and exert
influence. These spaces also represent activity sets that de-
fined Wari state institutions.
In their farthest southern province, the Wari exerted in-
fluence and managed resources through the monumental
center at Cerro Baúl. The site is dramatically placed on a
stunning mesa feature located between the Torata and Tumi-
laca tributaries of the upper Osmore drainage (also referred
to as the Moquegua Valley). Wari people did not originate
in the Moquegua study area but moved into it as foreign

Archaeological Papers of the American Anthropological Association, Vol. 14, pp. 151–174, ISBN 1-931303-20-7. ⃝ C 2005 by the American An-

thropological Association. All rights reserved. Permissions to photocopy or reproduce article content via www.ucpress.edu/journals/rights.htm.
152 Donna J. Nash and Patrick Ryan Williams

Figure 9.1. Middle Horizon sites in the Moquegua Valley.

Architecture and Power on the Wari–Tiwanaku Frontier
colonists. The colony was planned by the Wari state and
many structures were built with state-directed labor (Nash
1996). Seven sites in the valley exhibit Wari-related material
culture: Cerro Baúl, Cerro Mejia, El Paso, Cerro Petroglifo,
Cerro Trapiche (Moseley et al. 1991), Pampa del Arrastrado,
and El Tenedor (Owen 1994). There are also patches of do-
mestic terracing on the western slopes of Cerro Baúl (Figure
9.2). The settlements were accompanied by an extensive sys-
tem of agricultural terraces and cultivated slopes irrigated by
a canal over 10 kilometers in length (Williams 1997).
Architectural remains on Cerro Baúl are monumental
in nature and predominately constructed of double-faced
stone masonry, some of which was dressed. Cerro Mejia,
Cerro Petroglifo, El Paso, Cerro Trapiche, and Pampa del
Arrastrado have masonry remains resembling foundations,
and adobe superstructure is apparent in some buildings.
These sites have sectors of construction that exhibit simi-
lar spatial patterns to those of Cerro Baúl but are smaller
in scale and do not exhibit the same quality of architecture
(Nash 1996, 1997, 2002).
The Moquegua Wari colony is uniquely located within
sight of Tiwanaku settlements down valley. The Wari center
on Cerro Baúl may have been designed to manage relations
with the bordering Tiwanaku polity or to maintain the bound-
ary with a show of power and wealth (Isbell 1991). As the
Cerro Baúl colony was located on a border that Wari shared
with a politically strong and powerful polity, Wari institu-
tions may have been manifested or rendered with extra care.
It is notable that the Wari and Tiwanaku shared com-
mon religious iconography. Their sociopolitical interactions
may have had significant overlapping frames of reference.
Such interactions may have promoted orthodoxy (Washburn
1995) or hybridization (Prior and Carr 1995) in symbolism,
sumptuary custom, and ceremonial activity. The salient dif-
ferences in the spatial settings of state institutions as repre-
sented by the architectural ruins of the Wari and Tiwanaku,
however, demonstrate a significant difference in the orga-
nization of these states and the behaviors associated with
management and power legitimization.
Structure and the Use of Space
In the Wari colony, similar spatial venues or architec-
tural features were used at different levels of the political
hierarchy (Nash 2002). The state’s control of sumptuary cus-
tom regulated the parameters of variations in construction
quality, size, and complexity of the facilities. These subtle
differences can provide an accurate, though relative, guide
by which to gauge the status of spatial settings (or more
precisely their contained activity sets) and thus their impor-
153
tance or rank in the control hierarchy. The difference can be
seen as one of scale (see Kowalewski et al. 1983). Lower-
level administrators within a political system would exhibit
fewer symbols of legitimization or manifestations of effec-
tive power, whereas the pattern of management activity may
be similar but less complex.
Availability of raw material, skill of labor resources,
cosmological concepts, environmental setting, and the rank
of the officials using a spatial venue are all contributing fac-
tors that affect architectural size, quality of construction, and
the designed volume of the architectural forms (Zevi 1957).
Activity, however, is typically the most important factor de-
termining spatial design (Norberg-Schulz 1985). The space
required for a particular set of behaviors is of primary con-
cern in architectural design. The shape of the space is tightly
linked with the building’s intended use. The institutions man-
aging different facets of the state’s power (resource accu-
mulation, trade relations, religious legitimization, boundary
maintenance, and so on) involve different types of activities
and thus their built environments may exhibit variation in
design. Symbolism and the need to portray appropriate be-
havioral cues or ambiance are also significant in shaping the
volume of space (Rapoport 1990).
As Conklin (1990) has pointed out, changes in archi-
tectural form exhibit a particular inertia and thus may not
reflect changing activity patterns as quickly as they might
naturally be modified through interaction in such a dynamic
social arena as a semipermeable political frontier. Neverthe-
less, charting changes in the morphology of spatial venues
designed to contain political activity and power relations can
provide a trajectory along which institutional power struc-
tures were modified to address state needs. Institutions are
sets of practices and interactions that define the relationships
between people in a society. By attaching these behaviors and
their participants to architectural venues, we are seeking to
typologize activity through the material manifestation of its
context. Through documenting changes in the design of elite
and public spaces and activity sets in these settings, we ex-
amine how Wari state institutions developed. In this chapter
we apply this approach to the Middle Horizon architectural
remains in Moquegua.
The Wari Polity
The Wari state exhibits characteristics that are remi-
niscent of material remains from imperial entities such as
ancient Rome (Schreiber 1992). The Wari built centers of
power in many areas (Isbell and McEwan 1991; Kaulicke
and Isbell 2001). Through these installations they interacted
with local polities and populations, obtaining resources and
154
Figure 9.2. Wari colonial sites in the upper Moquegua drainage.
wielding influence. The Wari administrative elite controlled
vast resources by exerting authority over the labor of oth-
ers or influence over local leaders who possessed control of
labor and resources. The state’s need to interact with many
different groups had important implications for the struc-
Donna J. Nash and Patrick Ryan Williams
ture of state institutions. As an empire, state institutions re-
quired flexibility; one size would not effectively fit all re-
gions (see Doyle 1986; Hassig 1992; Hyslop 1993; Menzel
1959; Moseley and Mackey 1972; Sinopoli 1994; Van Buren
2000).
Architecture and Power on the Wari–Tiwanaku Frontier
Smaller political entities that controlled a more cosmo-
logically unified population through a smaller, less varied set
of elite administrators may exhibit less variety in the struc-
ture of edifices housing state institutions. The Wari, however,
were controlling disparate regions with different resources
and population sizes and in many instances coping with in
situ institutions to accomplish state goals through local per-
sonnel. Thus the variation documented between adminis-
trative centers reflects the need to be flexible and the Wari
state’s ability to adapt its institutions to regional conditions.
Wari architectural canons reflect a uniformity that is not
accidental but rather was purposefully designed. Neverthe-
less, the uniformity has been overstressed and subtle differ-
ences in the patterns of rooms and their orientation can have
significant correlates to the different activity sets accom-
modated in each space. To some degree similar but variant
strategies of control may have been executed at different lev-
els within the political hierarchy. Thus it is important that ex-
cavation and the examination of the organization of activities
be studied for many more structures. The findings presented
here may only pertain to the Cerro Baúl colony specifically,
because the region presented a unique set of situations that
were solved in relation to the groups that were brought to
the area or came under eventual control in this province. No
doubt the region’s proximity to Tiwanaku settlements and its
heartland on the shores of Lake Titicaca engaged elite offi-
cials from the Wari core in the political affairs of Moquegua,
perhaps more so than other provincial areas.
Architectural space exhibits standard characteristics but
may house different sets of activities (see Protzen 1993).
Space and architecture can take on symbolic representation
and reflect a cosmological idea or symbolize affiliation with
state, much as artifacts do when they portray a “corporate
style” (Moseley 1979). The use of a structure as a symbol is
but one example of the dialectic between the space people
design for themselves and the space that shapes the moods
and activities of people in its use (Bawden 1995; Geertz
1973; Kent 1990; Lawrence and Low 1990; Moore 1996).
Architectural features and the design of spaces proscribe
behavior and structure activity. Nevertheless, interpretation
of past behavior cannot be based on architectural form alone
but requires some reference to comparable spaces of known
function and, of course, the artifactual remains. Important
design features can be inferred from what was recorded of
Inka practice as well as from Wari iconographic imagery
depicting ceremony.
Inka Constructed Space
The Wari preceded the Inka by several centuries, and
we do not wish to imply that the Wari Empire operated in a
155
similar manner, but by examining the Inka Empire and how
it managed its resources and manipulated sociopolitical re-
lations we can gain insight into the complexities that should
be sought in the archaeological remains of Wari society. The
Inka Empire was an incredibly complex society whose ar-
chitecture incorporated a wide range of specialized activity
areas that accommodated the many institutions that grew as
the state developed.
The Sapa Inka, the absolute ruler of the empire, was
also considered a sacred individual, a living deity who was
served and worshipped even after his death (Betanzos 1996;
Moseley 1992; Rowe 1946). Therefore Inka statecraft was
seated in the pomp of political ceremony as well as in the
awe of religious observance. The ethnohistoric accounts of
the Inka polity describe public rituals in which the leader
himself took part, as well as other social mechanisms that
maintained the Sapa Inka’s access to labor resources from
every family within the empire’s vast territory (Murra 1980;
Rostworowski 1999; Rowe 1946). The following is a dis-
cussion of important Inka institutional spaces vital to the
activities of this great ancient Andean civilization. Ethno-
history provides a great deal of information about the work-
ings of the Inka polity; the following are merely highlights
that illuminate the importance of different venues in Andean
statecraft.
Public Space
Ethnohistoric accounts describe the Sapa Inka’s leading
role in large festivals such as Inti Raymi, an annual event as-
sociated with the June solstice and attended by the elites from
the province of Cuzco and by visiting leaders from through-
out the empire (Betanzos 1996; Cobo 1990; Garcilaso de
la Vega 1966). This festival and others like it would have
required a large open space. The large central plazas within
the capital were a setting of regular religious ceremony. In
hosting these gatherings and leading the rituals himself the
Inka emperor fulfilled his obligation to mitigate between the
gods and his subjects. He was responsible for the fertility
and abundance of the crops and animals. Therefore those in
attendance, the noble residents of Cuzco, and visiting lead-
ers were obligated to the Inka ruler because they accepted
his generosity, participated in the ceremony, and partook of
the feast. The public ceremonies tied the empire together.
Acting as host, the Sapa Inka legitimized his position and
cosmologically secured labor resources through establishing
the fealty of regional leaders for the coming year (Morris and
Thompson 1985).
Provincial capitals also had facilities for public events.
Huánuco Pampa and other large centers have large open
plazas with elevated platforms (Hyslop 1990; Morris and
156
Thompson 1985) from which the Inka or his representa-
tive could preside over ceremonies, accept tribute, or make
mandates over subordinate personnel (Cristóbal de Molina
1943:22). These simple architectural features created a space
in which a leader could interact with those being led. The
elite administrator’s elevated position was symbolized by the
elevated platform that provided a graphic display of the so-
ciety’s social hierarchy. The platform represented the power
of the Inka over the people in the province (Hyslop 1990;
see Moore 1996). Feasting in this venue created obligations
of the local people to serve the sponsor of provincial cere-
monies, the Inka state.
All public appearances of the Sapa Inka were formal
performances and spectacular processions. Ceremonies in
large open central plazas occurred regularly at the capital and
at the major provincial centers. The Sapa Inka was carried
upon a litter above everyone who gathered. He was accompa-
nied by a retinue of ranked individuals distinguished by their
dress and position relative to the ruler. Even military marches
were ordered ceremonial affairs (Cieza de León 1959; Cobo
1990). The order of the processions demonstrated the estab-
lished political order. They created in any space a sense of
appropriate behavior and in the formal organization of the
procession served as a visual cue to the audience of the ex-
pectations of an appropriate and acceptable set of behaviors.
These processions were dramatic, as exemplified by the de-
scription of the fateful royal parade in 1532 at Cajamarca
(Prescott 1942; Morris and Thompson 1985).
Sacred Space
The Qorikancha or enclosure of gold was a significant
structure in the Inka realm. It was the primary temple of
the sun and the center of the structured universe. Ceques,
imaginary lines that radiated out from this building, orga-
nized the sacred landscape and may have also regulated
rights to water and ritual responsibilities (Bauer 1998; Cobo
1956; Zuidema 1990). Conquistadors were informed they
had to fast for a year before entering this most sacred space
(Garcilaso de la Vega 1966). Ceremony reportedly associ-
ated with Inti Raymi and perhaps other events started with a
large group of elite personages in a plaza outside the struc-
ture but continued inside the temple. Ritual within this sacred
zone could only be witnessed by those of pure Inka blood
(Cobo 1990). Others experienced limited participation; all
but a few were excluded. Many types of ceremonial activities
were said to have taken place in the Qorikancha, yet entrance
was regulated. The architecture of the structure provides no
clue to this special quality. Unlike the sacred monumental
huacas of other Andean societies, which stand out as monu-
mental mountain effigies (Bawden 1996; Kolata 2003), the
Donna J. Nash and Patrick Ryan Williams
Qorikancha was not remarkable (Gasparini and Margolies
1980; Hyslop 1990). Clues to the Qorikancha’s exceptional
nature would only be discovered through comparative arti-
factual and ecofactual remains.
The rituals performed in the Qorikancha were another
act of legitimizing the role of the Inka, a materialization of
Inka ideology. Through acting on sacred knowledge within
the temple, the Inka were preventing others from knowing
how to maintain positive relations with the gods. Secret rites
that took place within the Qorikancha fulfilled the Sapa
Inka’s obligations to the populace by providing prosperity
to everyone in the Inka realm, a service only an initiated
Inka of pure blood could provide (see MacCormack 1991).
Other religious rites of this nature were conducted in sacred
fields (Bauer 1996); however, the archaeological remains of
these activities are not likely to be found.
Administrative Space
The Inka had a well-organized system of storage facil-
ities (Levine 1992). Qollqa were silo-like structures built
to keep a wealth of stores to support the state. It is unclear
how these were managed on a daily basis; however, Guaman
Poma (1980) and other chroniclers describe that officials
were in charge of tracking the inventory stored in the qol-
lqa and managing the distribution of these stored resources
to fulfill the needs of the state (Cieza de León 1959; Murra
1980). These goods were important for supporting the activi-
ties of dignitaries, bureaucrats, and other full-time specialists
in the service of the Inka and for people contributing their
labor on a part-time basis for construction projects or mili-
tary service. Stored goods were also needed to provide feasts
and gifts in the seasonal calendar in the ongoing process of
securing local participation in the Inka polity.
The qollqa are not settings of an institution but rather
are evidence that institutions of resource collection and man-
agement existed (Morris 1986). Several different institutions
coordinated to fill and use these structures. The qollqa are
associated with collection, storage, and eventual distribu-
tion. Assemblages inside a qollqa, however, cannot reveal
how relationships were established to collect these items or
how these goods were used to achieve political goals. The
regional distribution of qollqa has elucidated significant in-
formation about networks of resource movement (Morris
1992; Snead 1992). Structures used for the actual business
of resource management have not been described in detail
but have been reported from several Inka sites (Morris 1992;
Snead 1992). Wari examples will be described below.
Kallanka structures provided a venue for grand feasts
held during inclement weather (Garcilaso de la Vega
1966; Gasparini and Margolies 1980). Other occasions for
Architecture and Power on the Wari–Tiwanaku Frontier
gatherings in these structures were for more formal interac-
tions between officials of different ranks. Kallanka were rect-
angular structures with their longest axis built along one or
more sides of large plaza spaces. Typically they occur in pairs
and had many doors opening onto an important central plaza,
although the interior space was continuous and not broken
by dividing walls. The width of the space was so great that
these structures had interior columns to support the roof (see
Gasparini and Margolies 1980; Morris and Thompson 1985).
They were present in Cuzco along the large central plazas.
Kallanka are also found in the major provincial centers and
similar constructions are located at some of the minor cen-
ters and way stations, called tambo (Gasparini and Margolies
1980; Hyslop 1990).
Ethnohistoric accounts describe the kallanka as coun-
cil houses and places where justice was dispensed
(Gasparini and Margolies 1980). It has also been reported
that they were used to quarter traveling soldiers, dignitaries,
and guests (Hyslop 1990). The descriptions of kallanka sug-
gest that these gathering halls were the locations of negoti-
ation between the state and visiting leaders. They also seem
to have been a place where deals were made, gifts were ex-
changed, and compensations paid (see Betanzos 1996). Ne-
gotiations of this nature and many kinds of political affairs
require face-to-face interaction, an important facet of any
successful government. This kind of interaction may have
also taken place in more isolated, small-group discussions
within elite residential contexts, such as a specialized gath-
ering space within a palace.
The kancha, meaning “enclosure,” is an Inka-style con-
struction that is not unlike Wari elite residential structures.
These complexes are encompassed by a compound wall and
have individual roofed rectangular rooms that open onto
a central rectangular patio space. The Inka kancha differs
from the Wari elite residence in that the rectangular rooms
are typically not attached to each other and there may be
portions of the outer enclosure wall visible from the patio
(see Protzen 1993). Some kancha, however, closely resem-
ble their earlier counterparts in the Wari culture (Figure 9.3).
Gasparini and Margolies state that “the Qorikancha is basi-
cally analogous to other kancha-type structures. Its outside
appearance must not have been very different from that of
the Cuzco ‘palaces’ with well-finished walls and thatched
roofs” (1980:228–231). The Inka kancha was not adopted
by everyone; it seems to have been primarily a state style
of construction. It is found at most Inka constructed centers
and settlements. Hyslop suggests that kancha-style struc-
tures were components of palace complexes:
Identifying a kancha in Inka architecture does not neces-
sarily tell us about the activities carried out within it. It
157
is clear that kancha were the architectural basis for sim-
ple dwellings, just as they could compose the basis for
a palace or temple. It would seem that the rooms within
a kancha should have similar functions, but that is im-
probable, because sometimes kancha are composed of
rectangular rooms, some with doors and others that are
open-sided type. Much of the central sector of Cuzco was
composed of kancha, and these groupings around patios
served a variety of purposes: residences for kings and
persons of royal lineages (panaqa), special production
areas, and temples [1990:17].
Alone, the kancha were organized to provide space for
a variety of activities; however, together in a complex these
structures were associated with state-sponsored activities
and attached to Inka elite residence (Niles 1987).
Concepts of Space
The resemblance between Wari elite residences and Inka
kancha-style structures is undeniable. We suggest that there
are other important connections between Inka and Wari con-
cepts of space. These concepts and elements of the spatial
organization of Wari structures at Cerro Baúl and other sites
in the Moquegua colony will illustrate some of the institu-
tions operating in the Wari provincial government. We will
focus on three principal concepts of architectural manifes-
tations of power: subordination, exclusivity, and procession.
These concepts are artificial constructs, artifacts of our an-
thropological analysis that we will use in this discussion to
describe ways in which Wari elites manipulated activities
and their space. Thus the concepts are not mutually exclu-
sive and may operate in concert within some environments.
Elements such as size of the space, features, construction
quality, setting, and the spatial distribution of the associ-
ated artifact assemblages are significant considerations in
understanding the use of space. By analyzing manifestations
of these concepts and the description of these key elements,
we are able to categorize the nature of the institutions that
used the architectural space for their operation.
Subordination
Architecture reflects subordination when it structures
asymmetrical social interaction. This is expressed when a
feature elevates actors above others within a space, as with
a platform, or circumscribes the space that the dominant
individuals inhabit, as with a large recess or niche in which
a leader may not be elevated but is situated as the focus of a
room. This can also be expressed in a more portable medium
when an official is carried on a litter.
158 Donna J. Nash and Patrick Ryan Williams

Figure 9.3. Examples of kancha type structures.

Architecture and Power on the Wari–Tiwanaku Frontier
At a large scale the principle of subordination is demon-
strated in the settlement patterning of the Wari colony. Wari
elites inhabited the lofty heights of Cerro Baúl, while non-
Wari elites and commoners were restricted to smaller set-
tlements such as Cerro Mejia or Pampa del Arrastrado. The
walls circumscribing Cerro Baúl and the summit of Cerro
Mejia reinforce this subordination as well as express the
concept of exclusivity.
Exclusivity
Exclusivity refers to the restriction of entrance to or ob-
servation of certain settings where participation promotes
group cohesion. The excluded parties alternatively may be
allowed to watch a procession but not be a participant in this
symbolic behavior. Exclusivity may incorporate various lev-
els of inclusion as opposed to being a purely binary variable.
It can function between levels of a social hierarchy (class-
based discrimination) or between like groups participating at
the same level (faction-based discrimination). An important
result of exclusivity is the cohesion produced among those
in the selected exclusive group. Thus this concept has two
facets: it defines one group of people from another, the “in”
crowd and the “out,” at the same time as it binds members of
those groups together through participation and shared ex-
perience either in the special space or in activities segregated
from it.
Procession
Procession is an activity involving movement from one
location to another and can be conducted between meaning-
ful features in the landscape or take place in a built environ-
ment designed for activities of this nature. These activities
have great antiquity as reflected in early Andean monumen-
tal architecture (Conklin 1985; Moore 1996; Quilter 1999).
Platform mounds were built to emulate the mountainous
landscape, and ritual associated with these monuments may
have been focused toward controlling water and other natural
phenomena (see Kolata 2003).
Procession is implied by some Wari iconography and
related depictions from the Tiwanaku Gateway of the Sun.
These scenes show a central figure atop a tiered platform
holding staffs approached by two lines of beings coming
from different sides. These converging figures are depicted
as if in motion either running or flying (Cook 1994). Donnan
and McClelland (1999:167) have interpreted some Moche
iconography as depictions of ritual processions. In the pre-
sentation theme, a presiding individual at the top of a ramped
or stepped platform is approached and offered a cup.
159
The processional venues may be constructed to repre-
sent metaphysical characteristics of the natural landscape
and thus are designed to tie human ritual activities to the
supernatural. In so doing, processions serve as actions of
legitimization (Bauer 1996; Morris and Thompson 1985).
The organization of a procession, that is, the order in which
the figures stand, may represent the social order (see Cook
1994). Participation in these processions may reinforce and
reestablish the political hierarchies acting in a society.
Architectural Elements and Design
There are important elements, components, or qualities
of the constructed environment that can be changed, modi-
fied, or added to impact the nature of the activities possible
in the space. These elements include size, features, construc-
tion quality, setting, and the associated artifacts.
The size of a spatial venue is important and permits the
calculation of how many people participated in a particular
activity or how many people could witness a particular be-
havior. Size has an implication in understanding the degree
of exclusivity related to the context of the institution.
Features incorporated into the construction of a space
provide an understanding of the nature of the activity and
the flexibility of the intended use of a space. Benches, for
instance, demonstrate a proscribed setting for long-term in-
teraction and facilitate meals, feasting, and complex discus-
sions. The benches themselves encourage people to sit and
demonstrate where it is appropriate to sit. As such, bench
constructions serve as cues that direct people’s behavior
within a space. Thus platforms, dais structures, and benches
are features that can be accompanied by social custom to
invoke subordination and exclusivity.
Construction quality not only indicates relative sta-
tus differences to archaeologists but also was likely an
important element to the people interacting in the spaces.
Quality is recognized in several ways. The relative labor in-
puts into a structure are important. Factors such as the source
of the building materials, whether the stone was shaped, and
the height and thickness of the walls are examples of how la-
bor inputs could be quantified. More qualitative but equally
significant is the skill exhibited by masonry construction.
More time and skill is reflected when fieldstones are used but
walls are constructed so that they exhibit flat surfaces from
the careful selection and placement of stones. Construction
quality is subjective and its assessment by archaeologists
is based on the ruined state of the architectural remains.
Not all qualities that would have been notable are preserved.
Nevertheless, it is significant to consider these factors in un-
derstanding potential cues or nonverbal communication that
may have acted to structure behavior in the space.
160
Construction quality may not always be clearly ranked
but may be linked to other ideas described with words such
as sacred, profane, private, public, clean, dirty, and so on
(Rapoport 1990). Spaces can have more than one use, but
permanent features built into the spaces may help differen-
tiate how Wari designers would have categorized them. In
the case of Cerro Baúl, the preparation of the floor surface is
a good example that raises many questions. The floor types
vary from tamped clay to fine plaster and some floors are
paved with flagstones. As we are unaware of the sumptuary
laws for this cultural tradition it is impossible to judge which
was considered better, plaster or flagstones. It may also be
important to examine whether the nature of the space (res-
idential, religious, administrative, and others) corresponds
to one treatment or another. These differences may also be
temporal, with elite tastes changing over the course of a few
generations (Bourdieu 1996; Couture 2002).
The setting of an activity is significant to the nature of
an institution. The idea or label given to a space is culture
specific and linked to cultural norms and in some cases is
linked to a group’s cosmology (Rapoport 1990). At a regional
scale the idea of setting refers to the colony’s site hierarchy.
An elite residence at the provincial center may contain a dif-
ferent suite of activities from an elite residence at a tertiary
center even if the spatial arrangement is very similar (Nash
2002). Thus no single term can convey the nature of a par-
ticular setting and “thick description” is required (Bourdieu
1977; Geertz 1973; Ryle 1971). As archaeologists we can-
not grasp at these cosmological meanings; however, through
careful comparative study it may be possible to understand
spatial categories and perhaps use analogy to approximate
a meaningful label for the type of space and the behavior it
contained.
The associated artifact assemblage and the distribution
of these materials in the space provide the most important
information about the use of the space. But the nature of
the artifactual deposition must be treated with caution. It is
tempting to assume a “Pompeii scenario” and take the arti-
facts at face value (see Schiffer 1985). If examined carefully,
the distribution of the artifacts and the behaviors they imply
should provide some clues to the nature of the deposition.
Artifacts can be considered in two ways as they pertain
to activity in a space. Typically artifacts are thought of as
tools. Consumption vessels and serving wares, stone tools,
spindle whorls, and so on are used to infer a function or activ-
ity. Artifacts can also be used to establish ambiance. Fine ce-
ramic wares placed in niches can communicate the status of
the household; access to a variety of foodstuffs or rare items
likely represented the position of the bearer to others. Arti-
facts in a space and used in a space can promote or prohibit
activity as much as any permanent feature (Rapoport 1990).
Donna J. Nash and Patrick Ryan Williams
The elements of size, features, construction quality, set-
ting, and the artifactual assemblage will be discussed below
in association with the three concepts of behavior and their
embodiment in different social and temporal contexts. By
describing architectural venues with these elements and con-
cepts in mind we can begin to examine the different types of
power being exerted by Wari state officials in Moquegua and
the nature of the transformation or reorganization of power
relations over time.
Wari Institutional Space at Cerro Baúl
We divide the Middle Horizon occupation of Moquegua
into two time frames to compare changes in power relations:
the early Middle Horizon (600–800 C.E.) and the late Middle
Horizon (800–1000 C.E.). These time frames are still tenta-
tive in that we have not identified exactly when the reorgani-
zation took place (see Owen and Goldstein 2001; Williams
2001), but mounting evidence strongly supports at least two
phases of political organization during the Wari occupation
of the Moquegua region. This division allows us to assess
the changes in the materialization of power in monumen-
tal and elite residential architecture from one phase to the
next.
The Early Middle Horizon
In the early Middle Horizon, Wari established the prin-
cipal administrative center at Cerro Baúl and a secondary
center at Cerro Mejia (Figure 9.2). The first monument of
power is the Platform Complex aligned with Picchu Pic-
chu. It is located in Sector E on Cerro Baúl near the west-
ern edge of the summit (Figure 9.4). The complex includes
a 16- by-20-meter tiered platform with a staircase leading
down the south side into a walled rectangular sunken court
(Figure 9.5). Opposite the platform is a terraced structure, a
modification of natural features on the summit. It also has
a central staircase, which aligns with that of the platform,
leading down to the north. Standing on this terraced struc-
ture and looking across to the platform staircase beyond, one
sees Picchu Picchu, a sacred ice-capped mountain peak in
Arequipa, cradled by a dip in the intervalley ridge.
An in situ ritual burning event from within the con-
struction fill of the platform yielded a 14 C date of 1366±35
B.P. (600–770 C.E. calibrated at two sigma). This monument
had no peers in the Moquegua Valley at this time. While
modest in scale compared to the Akapana at Tiwanaku, it
is the first large-scale public monument in Moquegua. The
1,500-square-meter Cerro Baúl monument incorporates a
platform summit, a sunken court, and a terraced structure.
Architecture and Power on the Wari–Tiwanaku Frontier
Figure 9.4. The summit of Cerro Baúl.
The terraced structure consists of six terraces 20 meters in
length, each approximately 1 meter wide and approximately
50 centimeters high.
There are several scenarios that could explain the activ-
ities around which this space was designed. Currently, the
architectural elements and setting support the following in-
terpretation. Based on the quality of the modified southern
monument, the terraces or tiers may have provided a seating
area to view activity on the platform. Assuming seating ev-
ery meter on every terrace, 120 spectators could view events
from this structure.
The spectator’s focus is oriented toward the sunken court
and the platform beyond. The top surface of the platform is
only 1.2 meters above the lowest terrace. In the background,
the peaks of Picchu Picchu are unmistakably aligned with the
two staircases and the platform. The artificial space created
mimics the natural space beyond, and we hypothesize that
161
at least part of the public-scale rituals, incorporating over
100 individuals in the events, was focused on the distant
mountain peaks.
The alignment of descending and ascending staircases
oriented toward the Picchu Picchu peaks suggests that any in-
dividual in the space may have participated in the procession
during the course of the gathering. Ceremonial movements
may have linked the natural world with the cosmological
one, as mountains are seen as intermediaries between the
worlds in native Andean cosmology (Bolin 1998). Such a
performance is full of significance and may have served to
legitimate the roles of those in power. Nevertheless, even
though the platform is elevated, it does not place all the
spectators in a spatially inferior position to the ritual actors.
Ritual specialists may have performed the rituals on behalf
of higher ranked elites governing the province. We cannot
be sure who the actors were; however, the size of the space
162 Donna J. Nash and Patrick Ryan Williams

Figure 9.5. Early Middle Horizon monuments 600–800 C.E.

Architecture and Power on the Wari–Tiwanaku Frontier
suggests that just over 100 individuals could have been a part
of the event.
In a Wari colony with at least seven sites and a populace
of several thousand individuals (Williams and Sims 1998),
the defined setting of this activity reflects a certain degree of
exclusivity. The complex is uncircumscribed by walls and the
activities taking place on the platform could be noted, though
not in great detail, from settlements hundreds of meters be-
low in the Torata Valley. The platform complex exhibits the
concept of subordination, and the processional ceremonies
potentially enacted in this venue would have legitimized the
Wari governor’s right to power and perhaps established ranks
among the participants maintaining the political order.
In contrast to the events envisioned in this space, rituals
within the established walled compounds on the other side
of the mesa summit were much more intimate. Two classes
of architecture are notable in this regard. One class is com-
posed of the two D-shaped temple structures on the summit
of Cerro Baúl. The other class consists of patio-centered
compounds located on the summits of both Cerro Baúl and
the adjacent Wari settlement of Cerro Mejia.
The D-shaped structures are located on either side of the
later phase monumental core of Cerro Baúl. Both face north-
east, measure approximately 10 to 12 meters in diameter, and
have associated plazas larger than 10 meters square. The
western structure, Unit 10, has yielded the oldest 14 C date of
the two, 640–810 C.E. calibrated at two sigma. The eastern
structure, Unit 5, dates to 160 radiocarbon years later, based
on two dates processed from that space (Williams 2001).
These Unit 5 dates may represent use or remodeling rather
than original construction.
D-shaped structures have been described as ritual
venues of sacrifice and propitiation (Cook 2001). These sa-
cred spaces provided venues for political interaction through
participation in restricted religious rites. These actions may
have been limited to the highest levels of the Wari elite;
however, both temples open onto plaza spaces (Figure 9.5).
Perhaps similar to activities at the Inka Qorikancha, lower-
ranking individuals were allowed to witness or participate in
some acts in the plaza space but were not allowed to enter or
participate in the activities inside the D-shaped temple.
The space lacks a clear focus of subordination in the
architecture, with no single platform and 16 different wall
niches in the inner sanctum. The two D-shaped structures and
the complexes that envelope them, as in the first monument,
served to connect the natural and supernatural (Cook 2001).
They may have encompassed procession but are much more
exclusive in their access to the activities of the interior sacred
spaces. Furthermore, elaboration of the dominant position
of one space to another through features of subordination,
outside the exclusive access to the inner sanctum, is not
163
present in the D-shaped structures; instead, exclusive access
defined the relative status of the participants. This contrasts
with the use of features and concepts expressed in the next
class of architecture found in Sector A.
Unit 25 is a plaza compound approximately 10 meters
square, incorporating a 50-centimeter-wide bench around all
four sides. On the west side, the bench surface extends into
a two-by-four-meter niche or wall recess. The plaza has two
narrow entrances via corridors, one on the northeast side and
the other on the southwest side. A broad entrance leads to
an adjacent plaza to the east. Large quantities of serving and
storage vessels, currently under analysis, covered the floor
of this abandoned structure. Although no dates have yet been
processed from this context, six dates from this sector cluster
in the early period.
The space enclosed by this structure incorporates ap-
proximately 30 linear meters of bench seating, excluding
the elevated platform, and is approximately 100 square me-
ters in area. This kind of benched space is an appropriate
venue for gatherings. People sitting on the benches could
engage in conversations. If all of the available seating area
were used and each person occupied one meter of the bench,
a fair estimate for such a gathering would be limited to 30
people. Based on the assemblage of vessels associated with
this space, which included fragments hailing from as far
away as Cajamarca, 1,200 kilometers to the north, this was
an exclusive group of individuals.
Similar spatial configurations in the patio of Unit 9
suggest this activity may have been repeated, although at
a slightly smaller scale. Evidence of feasting is also present.
Unit 9 exhibits the organization of a Wari elite residence,
with rectangular abutting rooms opening onto a central patio
(Isbell 1989, 1991; McEwan 1987, 1991; Schreiber 1992).
As such, it represents concepts of space perpetuated in some
form by the Inka kancha of later times. Two radiocarbon
dates from Unit 9, 540–660 C.E. and 650–780 C.E. calibrated
at two sigma, show that this symbolic architectural form co-
incides with the earliest phase of Wari imperial expansion.
Both Unit 25 and Unit 9 seem to be part of a larger elite res-
idential complex and may represent different components
of a formal palatial structure. The enclosed plaza structure
was possibly a component of an elite residential compound,
designed to entertain a maximum of 30 people with no vis-
ibility to the outside world. A similar venue exists within a
more discrete elite residential structure and may have been
reserved for the highest ranking members of the society or
restricted to the use of the governor’s kin group.
On Cerro Mejia, Unit 145 resembles the construction
of Unit 9 to some degree. This structure has a slightly raised
platform and three stairs leading to an elevated rectilinear
space to the east (Figure 9.6). The highest stair extends across
164
Figure 9.6. Unit 145, summit of Cerro Mejia (from Nash 2002).
the front of this structure to form a bench. The room is open to
the patio but was roofed. The east wall has two niches placed
so that from the patio they could be viewed on either side of
the stairs approaching this sheltered dais. Similar structures
are modeled as Wari ceramic vessels (see Bonavia 1994:221,
fig. 167). Three other rectilinear rooms are arranged around
a central trapezoidal patio dating to about 700 C.E. (Nash
2002).
Unit 145 on Cerro Mejia is shaped like Unit 9 on Cerro
Baúl; however, the quality of construction is considerably
lower and an enclosure wall has not been located. Only the
eastern elevated room is paved with flagstones. A secondary
lower platform provides a place of honor for esteemed subor-
dinates. An intensive food preparation area was uncovered
in the northern room. This facility would have supported
feasting activities. Storage pits are located in the eastern and
southern rooms. The official living in this elite residence
may have distributed some resources or participated in gift
exchange, and the diversity of products in this residence sug-
gests that many goods were accepted as tribute or gifts (Nash
2002).
Unit 145 on Cerro Mejia exhibits qualities of subor-
dination, exclusivity, and procession all in one venue. It is
a multipurpose structure that reflects the modest access to
labor resources at this level in the political hierarchy. The of-
ficials at Cerro Mejia had one spatial venue to serve a variety
of activities. There are two elevated planes that allow two lev-
els of subordination. Exclusivity may not be expressed via
accessibility but rather through the explication of the politi-
cal hierarchy represented by the relative position in the patio
Donna J. Nash and Patrick Ryan Williams
space. Processions would have been limited, but as ranked
individuals ascended to the lower and upper platforms and
took positions similar to those depicted on textiles and ce-
ramic vessels, they were behaving in a symbolic way that
invoked an explicit cosmological meaning. The platforms in
the central patio of Unit 145 on Cerro Mejia exhibit the con-
cept of subordination, and processions enacted in this venue
would have legitimized Wari’s right to power and perhaps es-
tablished ranks among participants maintaining the political
order.
Cerro Mejia also exhibits platform construction in a
public venue. A set of four platforms is centrally located
among the summit architecture. Two platforms lie on the
eastern edge of an abnormally flat surface. Two smaller plat-
forms define the northern edge of the space (Figure 9.7).
This zone was modified to produce a plaza for the gather-
ing of a large group of people. These events were important
to establish continued subordination of the populace below
the provincial governor or his regional representatives. The
legitimizing performances may have included feasts that cre-
ated labor obligations or rituals that ensured productivity for
the coming year. Undoubtedly, this public space was an im-
portant venue of power for local administrators and state
officials.
Summary
During the early phase of Wari occupation, state insti-
tutions stressed activities that legitimized Wari power in a
public way. These power relations were set in a state-created
Architecture and Power on the Wari–Tiwanaku Frontier 165

Figure 9.7. Cerro Mejia summit structures (from Nash 2002).

166
venue evoking basic precepts of Wari belief and ritual struc-
tured to replicate the cosmological hierarchy within the hu-
man landscape. Additionally, spaces were designed where
an exclusive group of nobles could negotiate relationships
through feasting, gift exchange, and face-to-face interac-
tion. Some structure types contain architectural features that
elevate a small host of individuals relative to the rest. On
Cerro Baúl the Picchu Picchu platform emphasizes proces-
sion and various levels of inclusion; the other classes of
architecture emphasize exclusivity and sometimes marked
subordination.
Structures exhibiting similar components were present
on Cerro Mejia. These included elite residential structures
with patio spaces that permitted interaction in an atmosphere
of subordination. Platforms were set in a public space to
achieve a similar effect but with a larger group of people.
The structures on Cerro Mejia represent control at a different
scale and were no less important to achieving the state’s goals
in the region than those constructed on Cerro Baúl.
The Late Middle Horizon
During the second phase of occupation, new spatial
forms are introduced. The constructions are massive. The
sizes of the spaces are larger and thus the numbers of people
that could be brought together in these venues is consider-
ably greater. On the summit of Cerro Baúl, a massive recon-
struction event (Williams 2001) establishes at least six large
plaza complexes in Sector C (Figure 9.8), each greater than
30 meters across. These walled compounds exhibit different
forms, but all seem to be related to administration and lack
residential remains. For example, Unit 6 in Sector C is one of
three rooms in a series. It was elevated above the associated
plaza and entered via a small flight of stairs. This construc-
tion had two stories and a high bench along both long walls.
Inside this space there are no platforms or overt symbols of
subordination.
Another new kind of complex, typified by Unit 3 in Sec-
tor C, was designed to store large quantities of goods and
to manage these stores in a formal way (Figure 9.8). Unit
3 contains a cluster of four rooms with elevated wooden
floors, elevated thresholds, and food-related botanical mate-
rials (Williams 2001). The design of these storerooms would
have permitted increased air circulation and aided in the
preservation of perishable items. Similar elevated floors have
been found in small circular structures at the site of Cerro
Amaru in the vicinity of Viracochapampa, a Wari site in
the Huamanchuco area (Topic 1991; Topic and Topic 1984).
Later Inka qollqa in this region were built in a similar fash-
ion (Topic and Chiswell 1992). A rectangular room along
Donna J. Nash and Patrick Ryan Williams
the eastern wall of the Unit 3 compound was elevated above
the plaza and seems to have had a platform or bench built
along the exterior wall on either side of its entrance. The
organization of this compound suggests that these storage
facilities were directly managed by an individual accorded
elevated status by the architectural arrangement of Unit 3.
These new structures in Sector C are characterized by
a lack of domestic debris and have larger-scale architecture
than their earlier cousins, the small, sometimes residential,
patio compounds discussed earlier. The added monumental-
ity of these constructions may have amplified the feelings of
subordination people experienced in the vicinity of or inside
these structures. The examples that have been tested, Unit 3
and Unit 6, both use elevation to maintain a visual element
of status difference. These facilities are walled compounds
and thus access was through controlled passages that brought
people through large outer plazas and may have been strictly
regulated. Their size alone may have been awe inspiring.
Given that these large plaza compounds equaled or ex-
ceeded 1,000 square meters in area, they could accommo-
date up to ten times as many people as the small patio com-
pounds. The general paucity of material remains, in contrast
to the quantities associated with the smaller patio compounds
of earlier times, suggests a different function altogether for
these new monumental structures. These compounds did not
invite the interpersonal relations of the earlier structures but
stood as edifices to state administration.
We should note that the smaller patio compounds might
have continued in use into this period as areas for small gath-
erings. These structures may have become more formalized
over time. For example, Unit 1 on Cerro Baúl, which has
been described as a ritual libation hall where elite feasting
took place (Feldman 1998), was remodeled through its use
life and seems to have been of importance throughout the
entire Wari occupation (Williams 2001).
The central patio of Unit 1 is a 12-by-8-meter trape-
zoidal space (Figure 9.8), similar in form to that of Unit 145
on Cerro Mejia. Unit 1 has two rectangular rooms to the
northeast and northwest (Williams 2001). This scheme of
organizing space around a patio also occurs at the Wari site
of Pikillaqta in Cuzco (McEwan 1987). In the patio, Robert
Feldman (1998) describes a low terrace that cuts diagonally
from the northern corner to the middle of the southwest
wall. This low terrace may have been a triangular bench
or platform functioning to provide a place of superiority
during small-scale face-to-face interactions. Ceramic ves-
sels recovered in this context distinguish this structure from
others because of their high quality of production and the
presence of hybrid Wari–Tiwanaku design elements. It may
have been the prime venue for negotiations and exchange
between representatives of the two states.
Architecture and Power on the Wari–Tiwanaku Frontier
Figure 9.8. Late Middle Horizon monuments (800–1000 C.E.) on the summit of Cerro
Baúl.
Elsewhere at the site, at least one and possibly both of
the two D-shaped structures continued to be utilized in this
phase. Thus, we should not see the large plaza compounds
as replacing the small patio groups or D-shaped temples but
as complementing the nature of Wari power by adding a new
dimension of architecture. The potential reduction in the
number of these smaller units operating in the later phases
of the Wari occupation, however, may reflect a significant
shift in the way power was being expressed. An emphasis on
personal interaction in small exclusive venues was replaced
by more formal interaction in larger, more imposing spaces
that were nevertheless exclusive in their use of walls and
compartmentalization.
Concurrent with the shifts on the summit of Cerro Baúl,
there was an increase in Tiwanaku habitation both in the
Middle Valley around the sites of Omo and Chen Chen and
around Cerro Baúl. Public monuments were constructed in
167
association with these settlements, the most notable being
the Tiwanaku temple at Omo (Goldstein 1989) (Figure 9.9).
The Omo temple resembles features of the Akapana. It was
an imposing statement of Tiwanaku state power in the mid-
dle Moquegua Valley. Goldstein (1993) dates the structure
from a wood sample taken from the upper court entryway
lintel to 1160 ± 50 B.P., which is contemporaneous with the
reconstruction phenomenon of the large plaza complexes on
Cerro Baúl.
On the slopes of Cerro Baúl, a new type of monument
was introduced at approximately the same time. Three tri-
partite enclosures seem to mimic in some respect the mon-
ument at Omo with three courts but are much less elabo-
rate and rustic in nature. Owen (1998) investigated one of
these enclosures at La Cantera and dates the structure to ap-
proximately 900 C.E. based on radiocarbon dates of 1180 ±
80 B.P. and 1080 ± 70 B.P. We identified two further structures
168 Donna J. Nash and Patrick Ryan Williams

Figure 9.9. Late Middle Horizon monuments (800–1000 C.E.) in the Moquegua region.
Architecture and Power on the Wari–Tiwanaku Frontier
of the same style, El Paso and Sector J, both of which we
have sampled through excavation (Figure 9.9).
According to Moseley (personal communication, 2001),
all three structures appear to have been built over Wari agri-
cultural land used during the early Middle Horizon. Nash’s
excavations in the middle and lowest enclosures of the El
Paso complex, Unit 11A and 11B, revealed agricultural soil
beneath the floor and an earlier structure’s foundation. A
hearth associated with the earlier structure yielded a date of
1174±43 B.P. (780–940 C.E. calibrated at two sigma). Since
the hearth slightly predates the construction of the El Paso
structure, we infer the construction date for the structure to
be about 900 C.E. Ceramic assemblages from the surface
and excavations demonstrate a mix of Tiwanaku- and Wari-
influenced wares. Radiocarbon evidence from Sector J con-
firms the contemporaneity of all three structures with a date
of 1220 ± 52 B.P. (680–960 C.E. calibrated at two sigma).
Unlike the Omo temple or the platform complex in Sec-
tor E of Cerro Baúl, these tripartite enclosures contain no
artificial manipulation of elevational space. All are oriented
along hillsides to take advantage of the natural topography.
Thus, progression from the large lower court to the smaller
upper courts is always uphill. The remains of the enclosures
are low and built of fieldstone, not the ashlar blocks with
adobe superstructure described by Goldstein (1993) at Omo.
In contrast, La Cantera, Sector J, and the El Paso complex
have walls that do not exceed 70 centimeters in height and
have no subsurface foundation. There is no evidence for any
kind of superstructure. We estimate that the labor required
to construct the monuments was limited and the work was
perhaps accomplished by a small group over a matter of a
few weeks.
There is no midden associated with these complexes,
nor is there any architecture of subordination, as with the
platforms we saw in other contexts. The small rooms along
the sides of the plazas may represent areas where small
pagos, or offerings to the earth, were made, perhaps as part
of some progression from the large lower plazas to the more
restricted upper plazas. These complexes are compartmen-
talized as one ascends the slope, but the walls were not high
enough to shield external viewers from the activity that these
venues structured. As there are three of them in a relatively
small region and as each could accommodate a large group
of people (over 100), they do not seem to exhibit qualities of
exclusivity, although each may have represented a faction.
Exclusivity in this instance does not seem to have required
a visual cue or overt enforcement and may not have been a
significant factor.
In Moquegua, recent radiocarbon assays of contexts as-
sociated with Omo- and Chen Chen-style ceramics sug-
gest that the styles are contemporaneous (Goldstein and
169
Owen 2001). In the past, these two styles were associated
with wares from the Tiwanaku heartland (Goldstein 1989).
Omo-style wares were correlated to the Tiwanaku IV ves-
sels and Chen Chen-style ceramics were correlated with the
Tiwanaku V vessels. A more rustic Tiwanaku-related deriva-
tive style called Tumilaca was thought to postdate the
Tiwanaku occupation of the region. This style also seems
to be affiliated with the Middle Horizon but continues after
both states have withdrawn from the area. These styles may
be related to self-defined groups within Tiwanaku society
or may represent the distributions of wares sponsored by
competing groups of elites.
The scale of the sociopolitical units that constructed the
three complexes on the flanks of Cerro Baúl was probably
smaller than a valley-wide social group. We suggest each
of these structures may have been associated with a small
village on the slopes of Cerro Baúl inhabited in the later
half of the Middle Horizon and exhibiting locally produced
Tiwanaku- and Wari-influenced wares. The settlements of
Cancha de Yacango, Sector F, and Sector G may have built
and used these structures (Figure 9.2).
These uniquely local Tiwanaku-related enclosures, so
rustically constructed, may represent social groups’ or indi-
vidual community leaders’ attempts to establish their own
source of local monumental power. As such, they may even
represent a not-so-subtle resistance to the larger monument
of power, the Omo temple. Williams has argued elsewhere
that the late Middle Horizon may be a time of factionalization
in the Tiwanaku state (Williams 2002). These monuments
may represent manifestations of different social groups ex-
erting their own form of power. In the upper valley, these
monuments were built on the slopes of Cerro Baúl over for-
mer tracts of agricultural fields fed by the Wari irrigation
system. The associated Tiwanaku-related settlements would
have been dependent on the Wari economic resources pro-
vided by the Wari infrastructure of canal-fed agricultural
fields. The agricultural economy was apparently reorganized
in concert with the new construction of monumental spaces
on the summit of Cerro Baúl, perhaps in response to a shift
in the demographics of subordinate populations.
It seems that the venue of public ceremony shifted from
the summit of Cerro Mejia, with its platforms and state-
structured ritual, which was not occupied in the later Middle
Horizon, to more locally controlled phenomena. This shift
may reflect a conscious change in the Wari political system
to a more indirect strategy of control or the decline of state
prominence and power in the region. Equally, these monu-
ments may have been erected to compete with the Tiwanaku
state religious apparatus at Omo and perhaps served to resist
the Wari power structures created on the revered summit of
Cerro Baúl as well (see Brumfiel and Fox 1994).
170
Shifting Power, Changing Activity
In essence, examining the nature of power from the
Wari frontier in Moquegua documents a transformation at
two levels. At Cerro Baúl acts of legitimization are estab-
lished through ritual behavior at the Picchu Picchu-focused
platform complex and the D-shaped structures, both of
which exhibit exclusory mechanisms to reinforce and
maintain the political hierarchy. The Picchu Picchu complex
uses the visual cues of an elevated platform to symbolize
differences in power and control. These venues seem to
have continued in use throughout the Wari occupation and
were linked to a ritual calendar that allowed the Wari elite
to behave as mediators to the gods and thus displayed their
ability to ensure prosperity.
In the initial phase of Wari control in the Moquegua re-
gion, political relationships were built under the guise of
personalized hospitality and reciprocity. Face-to-face ne-
gotiations seem to have primarily taken place in elite res-
idential contexts and specialized gathering areas associ-
ated with a particular estate or palatial complex. Later,
the state built specialized nonresidential venues for face-
to-face negotiations, formalizing the behavior of establish-
ing political relationships. By taking this activity out of
the residential context, people coming to negotiate or es-
tablish relationships with Wari governors were doing so in
a setting that established their relationship with the state
rather than a particular governor and his estate. The hos-
pitality they received and the resulting obligation was not
loyalty attached to an individual leader but rather was
owed to the state and its representative. This may have
allowed the provincial governors to rotate these respon-
sibilities, substitute subordinate officials for some lev-
els of political interaction, or return more often to the
heartland leaving others in charge of administering the
province. These new monumental nonresidential construc-
tions may represent a material manifestation of a developing
bureaucracy.
On Cerro Mejia a central platform plaza complex pro-
vided a venue for lower-level elites, local or imported, to
establish obligations through feasting and festival. This re-
gional gathering was structured in association with Wari no-
tions of spatial organization and leaders occupying Wari-
style structures. The population living on domestic terraces
of Cerro Mejia does not exhibit local material culture and
may represent imported colonists from more than one re-
gion (Nash 2002). It is currently unclear how many small
settlements were participating in Wari society through con-
nections between local leaders and Wari officials at Cerro
Mejia. In the later phase of the Wari occupation, Cerro Mejia
is not occupied
Donna J. Nash and Patrick Ryan Williams
Local Tiwanaku-related forms of ritual architectural
compounds appear on the slopes of Cerro Baúl and at El
Paso. These structures allowed large groups to come to-
gether; however, no effort was made to unify the subordinate
population. Feasting and group ritual allowed lower-level
elites to establish obligations on their own behalf. These lo-
cal leaders may have acted as agents of the state but used
their own directed rituals and ceremonies to create domi-
nance within their own communities. The complexes likely
competed more with each other and Tiwanaku state religious
power than with the Wari administrative power, but these fac-
tional compounds cannot be ruled out as resistance to Wari
state structure. It is also evident that intimate interactions
may have remained important in some ongoing relationships.
Unit 1 on Cerro Baúl may have continued to be an impor-
tant context of diplomacy between the Wari and Tiwanaku
states. Nevertheless, in general it appears that intimate per-
sonalized interaction became less pronounced, and the state
and the formal state representative became the principles in
the operation of state institutions and resource management.
The power relations inferred from architectural design
elements and concepts we have defined as subordination, ex-
clusivity, and procession allow some interpretations of the
nature of changing Wari state institutions. The Wari state
managed a number of resources; this required establishing
their power through subordination, limiting access to their
power through exclusivity, and legitimizing their power and
the political hierarchy through acts of ritual procession. Hos-
pitality provided another mechanism of political interaction,
and spatial venues were carefully designed to promote face-
to-face interaction while exhibiting visual cues of subordi-
nation and exclusivity. The summits of Cerro Baúl and Cerro
Mejia exhibit other spatial venues that have unique charac-
teristics and remain to be tested. Continuing research of Wari
architecture, documenting the design of these spaces and the
activities conducted in them, will broaden our understanding
of the changing nature of the expansive Wari state and the
development of powerful Andean state institutions.
Acknowledgments
Funding to the authors for the research programs at
Cerro Baúl and Cerro Mejia has been provided by the Na-
tional Science Foundation (BCS-0226791 and 1602561-2),
the Heinz Family Foundation, the G. A. Bruno Foundation,
the Asociacion Contisuyo, the Field Museum, and the Uni-
versity of Florida. We also wish to thank Christi Conlee,
Dennis Ogburn, and Kevin Vaughn for the invitation to par-
ticipate in this stimulating volume and Elizabeth Brumfiel
and Jerry Moore for insightful comments on earlier versions
of this work. We are indebted to our colleagues and assistants
Architecture and Power on the Wari–Tiwanaku Frontier
who have worked with us over the past several years on the
Baúl complex. Finally, we thank the people of Moquegua for
participating with us in the process of discovering their great
heritage. Any flaws in interpretation are, of course, our own.
References
Bauer, Brian
1996 The Legitimization of the Inka State in Myth and
Ritual. American Anthropologist 98(2):327–337.
1998 The Sacred Landscape of the Inca: The Cuzco
Ceque System. Austin: University of Texas Press.
Bawden, Garth
1995 The Structural Paradox: Moche Culture as Political
Ideology. Latin American Antiquity 6(3):255–273.
1996 The Moche. Cambridge: Blackwell.
Betanzos, Juan de
1996 [1557] Narrative of the Incas. R. Hamilton and
D. Buchanan, trans. and eds. Austin: University of
Texas Press.
Bolin, Inge
1998 Rituals of Respect: The Secret of Survival in the
High Peruvian Andes. Austin: University of Texas
Press.
Bonavia, Duccio
1994 Arte e historia del Perú antiguo: colección Enrico
Poli Bianchi. Arequipa, Perú: Banco del Sur.
Bourdieu, Pierre
1977 Outline of a Theory of Practice. Richard Nice, trans.
Cambridge: Cambridge University Press.
1996 State Nobility: Elite Schools in the Field of Power.
Lauretta Clough, trans. Stanford: Stanford Univer-
sity Press.
Brumfiel, Elizabeth, and John Fox, eds.
1994 Factional Competition and Political Development
in the New World. Cambridge: Cambridge Univer-
sity Press.
Cieza de León, Pedro
1959 [1553] The Incas of Pedro de Cieza de León. Harriet
de Onis, trans. Victor Wolfgang von Hagen, ed.
Norman: University of Oklahoma Press.
Cobo, Bernabé
1956 [1653] Historia del nuevo mundo. Biblioteca de Au-
tores Espa noles. Madrid: Ediciones Atlas.
171
1990 [1653] Inca Religion and Customs. Roland Hamil-
ton, trans. Austin: University of Texas Press.
Conklin, William J.
1985 The Architecture of Huacas Los Reyes. In Early
Ceremonial Architecture in the Andes. Christopher
Donnan, ed. Pp. 139–164. Washington, DC: Dum-
barton Oaks.
1990 Architecture of the Chimor: Memory, Function,
and Image. In The Northern Dynasties: Kingship
and Statecraft in Chimor. Michael E. Moseley and
Alana Cordy-Collins, eds. Pp. 43–74. Washington,
DC: Dumbarton Oaks.
Cook, Anita
1994 Wari y Tiwanaku: entre el estilo y la imagen. Lima:
Pontificia Universidad Católica del Perú Fondo
Editorial.
2001 Huari D-Shaped Structures, Sacrificial Offerings,
and Divine Rulership. In Ritual Sacrifice in Ancient
Peru: New Discoveries and Interpretations. Eliza-
beth Benson and Anita Cook, eds. Pp. 137–164.
Austin: University of Texas Press.
Couture, Nicole
2002 The Construction of Power: Monumental Space and
Elite Residence at Tiwanaku, Bolivia. Ph.D. disser-
tation, Department of Anthropology, University of
Chicago.
Donnan, Christopher, and Donna McClelland
1999 Moche Fineline Painting: Its Evolution and Its
Artists. Los Angeles: UCLA Fowler Museum of
Cultural History.
Doyle, Michael W.
1986 Empires. Ithaca: Cornell University Press.
Feldman, Robert
1998 La ciudaduela Wari de Cerro Baúl en Moquegua. In
Moquegua: los primero doce mil años. Karen Wise,
ed. Pp. 59–66. Moquegua: Museo Contisuyo.
Garcilaso de la Vega
1966 [1609] Royal Commentaries of the Incas and Gen-
eral History of Peru. Harold V. Livermore, trans.
Austin: University of Texas Press.
Gasparini, Graziano, and Luise Margolies
1980 Inca Architecture. Patricia J. Lyon, trans. Bloom-
ington: Indiana University Press.
172
Geertz, Clifford
1973 The Interpretation of Cultures: Selected Essays.
New York: Basic Books.
Goldstein, Paul
1989 Omo, a Tiwanaku Provincial Center in Moquegua,
Peru. Ph.D. dissertation, Department of Anthropol-
ogy, University of Chicago.
1993 Tiwanaku Temples and State Expansion: A Ti-
wanaku Sunken-Court Temple in Moquegua, Peru.
Latin American Antiquity 4(1):22–47.
Goldstein, Paul, and Bruce Owen
2001 Tiwanaku en Moquegua: las colonias altiplánicas.
Boletı́n de Arqueologı́a PUCP (Lima) 5:139–168.
Guaman Poma de Ayala, Felipe
1980 [1615] Nueva crónica y buen gobierno. John Murra,
Rolena Adorno, and J. Urioste, trans. Madrid:
Historia 16.
Hassig, Ross
1992 War and Society in Ancient Mesoamerica. Berke-
ley: University of California Press.
Hyslop, John
1990 Inca Settlement Planning. Austin: University of
Texas Press.
1993 Factors Influencing the Transmission and Dis-
tribution of Inka Cultural Materials throughout
Tawantinsuyu. In Latin American Horizons. Don
Rice, ed. Pp. 337–356. Washington, DC: Dumbar-
ton Oaks.
Isbell, William H.
1989 Honcopampa: Was It a Huari Administrative Cen-
tre? In The Nature of Wari: A Reappraisal of the
Middle Horizon Period in Peru. R. M. Czwarno,
F. M. Meddens, and A. Morgan, eds. Pp. 98–114.
BAR International Series, 525. Oxford: British Ar-
chaeological Reports.
1991 Huari Administration and the Orthogonal Cellu-
lar Architectural Horizon. In Huari Administra-
tive Structure: Prehistoric Monumental Architec-
ture and State Government. William H. Isbell and
Gordon McEwan, eds. Pp. 293–315. Washington,
DC: Dumbarton Oaks.
Isbell, W. H., C. Brewster-Wray, and L. Spickard
1991 Architecture and Spatial Organization at Huari. In
Huari Administrative Structure: Prehistoric Monu-
Donna J. Nash and Patrick Ryan Williams
mental Architecture and State Government. W. H.
Isbell and G. McEwan, eds. Pp. 19–53. Washing-
ton, DC: Dumbarton Oaks.
Isbell, William H., and Gordon McEwan, eds.
1991 Huari Administrative Structure: Prehistoric Mon-
umental Architecture and State Government.
Washington, DC: Dumbarton Oaks.
Kaulicke, Peter, and William Isbell, eds.
2001 Huari y Tiwanaku: modelos vs. evidencias. Boletı́n
de Arqueologı́a PUCP (Lima) 5.
Kent, Susan, ed.
1990 Domestic Architecture and the Use of Space.
Cambridge: Cambridge University Press.
Kolata, Alan
2003 Tiwanaku Ceremonial Architecture and Urban Or-
ganization. In Tiwanaku and Its Hinterland: Ar-
chaeology and Paleoecology of an Andean Civi-
lization, vol. 2: Urban and Rural Archaeology Alan
Kolata, ed. Pp. 175–201. Washington, DC: Smith-
sonian Institution Press.
Kowalewski, Stephen, Richard Blanton, Laura Finsten, and
Gary Feinman
1983 Boundaries, Scale and Internal Organization. Jour-
nal of Anthropological Archaeology 2:32–56.
Lawrence, Denise, and Setha Low
1990 The Built Environment and Spatial Form. Annual
Review of Anthropology 19:453–505.
Levine, Terry
1992 Inka State Storage in Three Highland Regions: A
Comparative Study. In Inka Storage Systems. Terry
Levine, ed. Pp. 107–150. Norman: University of
Oklahoma Press.
MacCormick, Sabine
1991 Religion in the Andes: Vision and Imagination in
Early Colonial Peru. Princeton: Princeton Univer-
sity Press.
McEwan, Gordon
1987 The Middle Horizon in the Valley of Cuzco, Peru.
BAR International Series, 372. Oxford: British Ar-
chaeological Reports.
1991 Investigations at the Pikillacta Site: A Provincial
Huari Center in the Valley of Cuzco. In Huari
Administrative Structure: Prehistoric Monumental
Architecture and Power on the Wari–Tiwanaku Frontier
Architecture and State Government. William H.
Isbell and Gordon McEwan, eds. Pp. 93–120.
Washington, DC: Dumbarton Oaks.
Menzel, Dorothy
1959 The Inca Occupation of the South Coast of Peru.
Southwestern Journal of Anthropology 15(2):125–
142.
Molina, Cristóbol de
1943 [1553] Las crónicas de los Molinas, los peque nos
grandes libros de historia Americana, Ser. 1, Tomo
IV. Lima: D. Miranda.
Moore, Jerry
1996 Architecture and Power in the Ancient Andes:
The Archaeology of Public Buildings. Cambridge:
Cambridge University Press.
Morris, Craig
1986 Storage, Supply, and Redistribution in the Econ-
omy of the Inka State. In Anthropological History
of Andean Polities. John Murra, Nathan Wachtel,
and Jacques Revel, eds. Pp. 59–68. Cambridge:
Cambridge University Press.
1992 Huánuco Pampa and Tunsukancha: Major and Mi-
nor Nodes in the Inka Storage Network. In Inka
Storage Systems. Terry Levine, ed. Pp. 151–175.
Norman: University of Oklahoma Press.
Morris, Craig, and Donald E. Thompson
1985 Huánuco Pampa: An Inca City and Its Hinterland.
London: Thames and Hudson.
Moseley, Michael E.
1979 Peru’s Golden Treasures. Chicago: Field Museum
of Natural History.
1992 The Incas and Their Ancestors. New York: Thames
and Hudson.
Moseley, Michael, Robert Feldman, Paul Goldstein, and Luis
Watanabe
1991 Colonies and Conquest: Tiahuanaco and Huari in
Moquegua. In Huari Administrative Structure: Pre-
historic Monumental Architecture and State Gov-
ernment. William H. Isbell and Gordon McEwan,
eds. Pp. 121–140. Washington, DC: Dumbarton
Oaks.
Moseley, Michael, and Carol Mackey
1972 Peruvian Settlement Pattern Studies and Small Site
Methodology. American Antiquity 37:67–81.
173
Murra, John V.
1980 [1955] The Economic Organization of the Inca
State. Greenwich, CT: JAI.
Nash, Donna J.
1996 Cerro Petroglifo: Settlement Patterns and Social
Organization of a Residential Wari Community.
M.A. thesis, Department of Anthropology, Univer-
sity of Florida.
1997 Wari Architecture: A Case Study from the Mo-
quegua Valley, Peru. Poster presented at the 62nd
Annual Meeting of the Society for American Ar-
chaeology, Nashville.
2002 The Archaeology of Space: Places of Power in the
Wari Empire. Ph.D. dissertation, Department of
Anthropology, University of Florida.
Niles, Susan
1987 Callachaca: Style and Status in an Inca Community.
Iowa City: University of Iowa Press.
Norberg-Schulz, Christian
1985 The Concept of Dwelling: On the Way to Figurative
Architecture. New York: Electa.
Owen, Bruce
1994 Were Wari and Tiwanaku in Conflict, Competition
or Complementary Coexistence? Survey Evidence
from the Upper Moquegua Drainage, Peru. Paper
presented at the 59th Annual Meeting of the Society
for American Archaeology, Anaheim, CA.
1998 A Tiwanaku IV Temple at Cerro Baúl: Evidence
and Implications. Paper presented at the 38th An-
nual Meeting of the Institute for Andean Studies,
Berkeley.
Owen, Bruce, and Paul Goldstein
2001 Tiwanaku en Moquegua: interacciones regionales
y colapso. Boletı́n de Arqueologı́a PUCP (Lima)
5:169–188.
Prescott, William H.
1942 [1847] History of the Conquest of Peru. Everyman’s
Library. New York: E. P. Dutton.
Prior, John, and Christopher Carr
1995 Basketry of Northern California Indians: Interpret-
ing Style Hierarchies. In Style, Society and Per-
son: Archaeological and Ethnological Perspectives.
Christopher Carr and Jill Neitzal, eds. Pp. 259–296.
New York: Plenum.
174
Protzen, Jean-Pierre
1993 Inca Architecture and Construction at Ollantay-
tambo. New York: Oxford University Press.
Quilter, Jeffrey
1999 Moche Mimesis: Continuity and Change in Public
Art in Early Peru. In Moche Art and Archaeology
in Ancient Peru. Joanne Pillsbury, ed. Pp. 21–46.
Washington, DC: National Gallery of Art.
Rapoport, Amos
1990 The Meaning of the Built Environment. Beverly
Hills: Sage Publications.
Rostworowski de Diez Canseco, Marı́a
1999 History of the Inca Realm. Harry Iceland, trans.
Cambridge: Cambridge University Press.
Rowe, John
1946 Inca Culture at the Time of the Spanish Conquest.
In The Andean Civilizations. J. H. Steward, ed.
Pp. 183–330. Handbook of South American Indi-
ans, vol. 2. Washington, DC: Smithsonian Institu-
tion.
Ryle, Gilbert
1971 Collected Papers, Vol. 2: Collected Essays 1929–
1968. London: Hutchinson.
Schiffer, Michael B.
1985 Is There a “Pompeii Premise” in Archaeology?
Journal of Anthropological Research 41:18–41.
Schreiber, Katharina
1992 Wari Imperialism in Middle Horizon Peru. An-
thropological Papers 87, Museum of Anthropology.
Ann Arbor: University of Michigan.
Sinopoli, Carla
1994 The Archaeology of Empires. Annual Review of
Anthropology 23:159–180.
Snead, James
1992 Imperial Infrastructure and the Inka State Storage
System. In Inka Storage Systems. Terry Levine,
ed. Pp. 62–106. Norman: University of Oklahoma
Press.
Topic, John
1991 Huari and Huamachuco. In Huari Administrative
Structure: Prehistoric Monumental Architecture
Donna J. Nash and Patrick Ryan Williams
and State Government. William H. Isbell and Gor-
don McEwan, eds. Pp. 141–164. Washington, DC:
Dumbarton Oaks.
Topic, John, and Coreen Chiswell
1992 Inka Storage in Huamanchuco. In Inka Storage Sys-
tems. Terry Levine, ed. Pp. 206–236. Norman: Uni-
versity of Oklahoma Press.
Topic, Theresa, and John Topic
1984 Huamanchuco Archaeological Project: Prelimi-
nary Report on the Third Field Season, June–
August 1983. Trent University Occasional Papers
in Anthropology 1. Peterborough, Ontario.
Van Buren, Mary
2000 Political Fragmentation and Ideological Continuity
in the Andean Highlands. In Order, Legitimacy, and
Wealth in Ancient States. Janet Richards and Mary
Van Buren, eds. Pp. 77–87. Cambridge: Cambridge
University Press.
Washburn, Dorothy
1995 Style, Perception and Geometry. In Style, Society
and Person: Archaeological and Ethnological Per-
spectives. Christopher Carr and Jill Neitzal, eds.
Pp. 101–122. New York: Plenum.
Williams, Patrick Ryan
1997 The Role of Disaster in the Development of Agri-
culture and the Evolution of Social Complexity in
the South-Central Andes. Ph.D. dissertation, De-
partment of Anthropology, University of Florida.
2001 Cerro Baúl: A Wari Center on the Tiwanaku Fron-
tier. Latin American Antiquity 12(1):67–83.
2002 A Re-examination of Disaster-Induced Collapse in
the Case of the Andean Highland States: Wari and
Tiwanaku. World Archaeology 33(3):361–374.
Williams, Patrick, and Kenneth Sims
1998 Archaeological Population Estimates and Agrarian
Productivity. Poster presented at the 97th Annual
Meeting of the American Anthropological Associ-
ation, Philadelphia.
Zevi, Bruno
1957 Architecture as Space: How to Look at Architec-
ture. Milton Gendel, trans. New York: Horizon.
Zuidema, R. Tom
1990 Inca Civilization in Cuzco. Jean-Jacques Decoster,
trans. Austin: University of Texas Press.
BOLETÍN DE ARQUEOLOGÍA PUCP / N.° 9 / 2005, 207-232 / ISSN 1029-2004

LOS ENCUENTROS Y LAS BASES

PARA LA ADMINISTRACIÓN POLÍTICA WARI*

Patrick Ryan Williams,a Donna J. Nash,b Michael E. Moseley,c

Susan deFrance,d Mario Ruales,e Ana Mirandaf y David Goldstein g

Resumen

En el presente trabajo se analiza el rol de los encuentros y reuniones de diversa escala en la colonia wari de Moquegua, Perú (600-1000
d.C.). El papel de los festines cambió sustancialmente con la expansión de las entidades políticas Wari y Tiwanaku, e involucró múltiples
lugares y numerosos tipos de congregaciones. La evidencia resulta de excavaciones en dos sitios principales, Cerro Baúl y Cerro Mejía,
y se comparan los contextos de festines en ambos sitios con el argumento de que las reuniones o encuentros públicos fueron fundamentales
en múltiples escalas y en muchos lugares diferentes. Mediante el estudio tanto de los lugares de producción como los de consumo se busca
aclarar el mecanismo de la especialización de instalaciones en diversos contextos, así como los roles que los diversos miembros de la
sociedad wari desempeñaron en la producción de festines. Finalmente, por medio de la comparación del acceso diferenciado a productos
alimenticios y el empleo de alimentos únicos en lugares especiales, se aborda la diferenciación social en el acceso a los recursos en la colonia
wari de Moquegua. También se analiza la relación entre la cocina y la identidad en lo que constituía un encuentro colonial muy
cosmopolita en la sierra sur andina.

Palabras clave: Tiwanaku, diferenciación social, colonización, expansión política, cocina, identidad

Abstract

ENCOUNTERS AND THE FOUNDATION OF WARI STATECRAFT

In this work, we examine the role of public gatherings at various scales in the Wari colony in Moquegua, Perú (AD 600-1000). We
argue that the role of feasting changed significantly with the expansion of the Wari and Tiwanaku polities, and involved multiple venues
and numerous types of congregations. Our evidence stems from research at two main sites, Cerro Baúl and Cerro Mejía, and we
compare feasting contexts at both sites, arguing that public gathering was fundamental at multiple scales and in many different venues.
By examining both locales of production and locales of consumption, we elucidate the specialization in facilities in different contexts, as
well as the roles different members of society played in the production of feasts. Finally, by comparing the differential access to foodstuffs,
and the employment of unique foods in certain settings, we address social differentiation in resource access in the Wari colony in
Moquegua. We also examine the relationship between cuisine and identity in what was a very cosmopolitan colonial encounter in the
south Andean highlands.

Keywords: Tiwanaku, social differentiation, colonization, political expansion, cuisine and identity

* Traducción del inglés al castellano: Mónika Barrionuevo

a
The Field Museum, Chicago. Correo electrónico: rwilliams@fmnh.org
b
University of Illinois, Department of Anthropology. Correo electrónico: djnash@uic.edu
c
University of Florida, Department of Anthropology. Correo electrónico: moseley@anthro.ufl.edu
d
University of Florida, Department of Anthropology. Correo electrónico: sdef@anthro.ufl.edu
e
Instituto Nacional de Cultura. Correo electrónico: masega@terra.com.pe
f
Museo Contisuyo, Moquegua. Correo electrónico: amq_miranda@yahoo.com
g
Southern Illinois University, Department of Anthropology. Correo electrónico: djgoldst@siu.edu
208 WILLIAMS, NASH, MOSELEY, DE FRANCE, RUALES, MIRANDA Y GOLDSTEIN

1. Introducción

El rol de los encuentros o reuniones en el desarrollo de la administración política andina ha sido, por
años, el centro de gran parte de la discusión académica en los estudios arqueológicos. En los Andes
prehispánicos, los festines han sido estudiados como un componente importante de las relaciones
políticas y sociales de las sociedades complejas (Morris y Thompson 1985; Gero 2001; Cook y
Glowacki 2003). De hecho, muchas reuniones sociales incluyen la preparación y consumo de alimen-
tos de una naturaleza especial, de modo que pueden ser descritas como festines y están enmarcadas
en un ámbito supradoméstico. Los modelos arqueológicos que describen la economía política de los
estados andinos tempranos (Rowe 1946, 1982; Morris 1985; Rostworowski 1999; Moseley 2001)
incorporan ideas de reciprocidad e intercambios asimétricos de comida por trabajo o «festines de
trabajo» (work feasts, cf. Dietler y Herbich 2001). Sin embargo, las evidencias de estas actividades son
esquivas en el registro arqueológico.
Los líderes acumularon obligaciones laborales de sus seguidores por medio de la celebración de
«festines de patrocinio» (patron-role feasts, cf. Dietler 2001). Estos pueden haber sido organizados en
lugares públicos o en asociación con eventos rituales. Esta forma de manifestación de generosidad fue
una importante fuente de poder por medio de la cual las elites gobernantes acumulaban excedentes y
riquezas (Murra 1980, 1982; D’Altroy y Earle 1985; D’Altroy 2001). Es posible que eventos simila-
res, pero a una escala menor, fueran celebrados por parte de las elites locales para legitimar y mante-
ner el carácter de requerimiento y aprovechamiento de la reciprocidad entre los administradores y
sus subordinados (Isbell 1997; Kolata 2003a). De esta manera, los festines cumplen un papel impor-
tante en las descripciones de la interacción política del pasado andino, por lo que es de esperar que
estas prácticas estén bien representadas en el registro arqueológico.
Los modelos de administración política andinos que incluyen festines están ampliamente basados
en los registros etnográficos de los campesinos andinos y también en los registros históricos de las
prácticas estatales inka. Los registros etnohistóricos de los inka describen ceremonias a gran escala,
como el Inti Raymi, que restablecían la jerarquía de las obligaciones en intervalos regulares (Garcilaso
de la Vega 1966 [1609]). Esta festividad anual, mayormente realizada en el Cuzco o en cualquier lugar
en donde se encontrara el emperador (cf. Ramírez 2005), reunía a las elites de todo el ámbito inka.
Había intercambio de obsequios, se servían comidas especiales y los invitados, que eran entretenidos
y agasajados, presenciaban representaciones ceremoniales a gran escala, todo ello enmarcado en estas
reuniones, que reforzaban las relaciones de estatus entre las elites imperiales y los líderes locales. Por
otro lado, el Inka también alimentaba a grandes grupos de individuos comunes del pueblo durante los
proyectos de trabajo. Algunos cronistas describen que el Inka proporcionaba comida, alojamiento,
ropa y herramientas para aquellos que estaban al servicio del Estado (Guamán Poma 1980 [1615-
1616]; Betanzos 1996 [1551]).
Por lo general, gran parte de la evidencia arqueológica correspondiente a la política de celebración
de banquetes (commensal politics) en los Andes tempranos sugiere eventos compartidos por un grupo
relativamente pequeño de personas (cf. Brewster Wray 1989) o relacionados con depósitos rituales
(cf. Kolata 2003b). Estos hallazgos podrían representar lo que Dietler (2001) describe como un «festín
facultativo» (empowering feast), un evento en el que las autoridades de igual jerarquía operaban para
conseguir potestades específicas. Los acontecimientos en los cuales los líderes demostraban su rela-
ción con los ancestros u otros seres sobrenaturales fueron objetivos potencialmente significativos
para algunos banquetes y festines en los estados andinos tempranos. Sin embargo, como los motivos
para celebrar una reunión podían, muchas veces, encubrir algo, es difícil entender del todo las rela-
ciones sociopolíticas subyacentes que eran exteriorizadas en los diferentes eventos festivos. En con-
secuencia, para comprender el proceso político en general de la fundación del Estado wari es importante
que se tome en cuidadosa consideración los contextos en que se realizaron estos encuentros o reunio-
nes especiales. El presente trabajo examina su papel en las actividades sociopolíticas de la sociedad
wari en el centro provincial de Cerro Baúl y sus alrededores, pero primero se describirá el panorama
social del Horizonte Medio y la colonia wari en Moquegua con el fin de establecer un marco regional
para estas interacciones.

ISSN 1029-2004
 LOS ENCUENTROS Y LAS BASES PARA LA ADMINISTRACIÓN POLÍTICA WARI 209

2. El asentamiento

Cerro Baúl es una meseta geológica que se eleva 600 metros sobre el río Torata, en la árida sierra de la
cuenca del río Moquegua, en el sur del Perú (17,116° de latitud sur, 70,85° de longitud oeste, cf. Fig. 1).
Cuando las fuerzas imperiales procedentes de la ciudad de Wari, ubicada a 600 kilómetros al norte,
entraron en la región un poco antes de 600 d.C., tomaron la cima de la meseta y dos cerros adyacen-
tes, Cerro Mejía y Cerro Petroglifo, para implantar la colonia más sureña del imperio dentro del
territorio dominado por Tiwanaku, cuya capital se ubicaba cerca del lago Titicaca, en Bolivia. Duran-
te aproximadamente cuatro siglos, y con una población que llegaba a más de 1000 habitantes, este
puesto de avanzada fue único en su carácter debido que impulsó el contacto directo entre ambos
imperios (Williams 2001). En otras áreas, los dos regímenes estuvieron separados por zonas interme-
dias (buffer zones) de un ancho de casi 100 kilómetros (Lumbreras 1974).
Esta colonia califica plenamente como un «centro administrativo» establecido por un Estado,
definido por la presencia de cánones imperiales de arquitectura monumental y arte suntuario wari.
Todos los otros centros creados por el Estado fueron establecidos en las pampas cerca de la base de los
valles. Cerro Baúl rompe este patrón de una forma tan drástica que parece haber sido parte de una
decisión premeditada. La alejada cima de la meseta no fue habitada ni antes ni después de los
tiempos wari debido a que es un lugar completamente impracticable para vivir. Comenzando por el
agua, los recursos necesarios para la vida debían ser llevados a la cima de la meseta a costa de un
inmenso trabajo. No existe una razón económica viable para vivir por siglos en aquel lugar tan poco
atractivo y con un gasto tan grande. En todo caso, la decisión wari puede haber incluido una combi-
nación de tres factores que son difíciles de detectar arqueológicamente: la religión, la defensa y la
política.
Hoy en día, Cerro Baúl es un apu o montaña sagrada que los devotos escalan por una hora o más
para alcanzar la escarpada cima. En ese lugar pueden hacer «pagos» que incluyen la representación en
miniatura de granjas con casas, corrales, campos de cultivo y ganado por medio de piedras sueltas.
En el pasado, los estados andinos expansivos, como el inka, usurparon lugares sagrados para su
propio engrandecimiento, por lo que la ocupación wari de Cerro Baúl podría representar una usurpa-
ción similar o, por lo menos, una declaración de superioridad.
Se dice que la conquista inka de la región llevó a la población local a refugiarse en la impenetrable
cima de un gran bastión natural. La alta meseta, con sus cumbres escarpadas, es un formidable
baluarte geológico contra los asedios. Como acamparon alrededor de la base, los invasores recibie-
ron, eventualmente, la rendición de los locales gracias a la reducción de los abastecimientos de agua
y comida en la cima (Garcilaso de la Vega 1966 [1609]). En ese sentido, si el sitio no fue una táctica
militar utilizada durante la época wari, entonces Cerro Baúl podría haber sido, ciertamente, defendi-
ble. En todo caso, las decisiones políticas en la capital indujeron la presencia imperial wari en la
frontera de Moquegua. Esto resultó en un único caso de yuxtaposición cara a cara con su formidable
contemporáneo en el sur: Tiwanaku. Esta inusual situación surge como una explicación política para
Cerro Baúl, ya que la colonia habría podido hacer las veces de una especie de «embajada», un enclave
que representaba los intereses wari para la capital imperial, así como un lugar de encuentros entre las
entidades políticas respectivas. Una delegación gubernamental que residiera de manera opulenta en la
cima de este bastión sagrado podría haber generado una impresión política impactante. Si este
escenario es correcto, entonces el templo monumental en el complejo de Omo, en el valle medio,
habría sido la contraparte tiwanaku (Fig. 1). Este es el único santuario conocido de estilo imperial
construido fuera del núcleo altiplánico tiwanaku y podría representar, muy bien, una respuesta polí-
tica al enclave wari (Goldstein 2005).
Wari es caracterizado como un Estado relativamente secular y militar, mientras que Tiwanaku ha
sido entendido como eclesiástico y mercantil (Schreiber 1992; Kolata 1993). Ambos estuvieron so-
cialmente estratificados, con individuos comunes, sobre todo agricultores y pastores, que sostenían
artesanos, técnicos, especialistas religiosos y una clase jerárquica de nobles gobernantes. De manera
significativa, las iconografías wari y tiwanaku compartieron una deidad primordial con ojos con

ISSN 1029-2004
210 WILLIAMS, NASH, MOSELEY, DE FRANCE, RUALES, MIRANDA Y GOLDSTEIN

Fig. 1. Mapa de la sección superior del valle de Moquegua, Perú.

ISSN 1029-2004
 LOS ENCUENTROS Y LAS BASES PARA LA ADMINISTRACIÓN POLÍTICA WARI 211

Fig. 2. Vista de la colonia de Cerro Baúl.

bandas y rayos en la cabeza, llamada «Dios de los Báculos» (Front Facing Deity) y, en Cerro Baúl, los
artefactos con tales diseños provienen de contextos de bebida ritual.
Los dos asentamientos coloniales en Moquegua estaban a la vista uno del otro, pero se diferencia-
ban en su adaptación particular al medio como, por ejemplo, en el cultivo de las tierras desérticas que
requerían irrigación. Como en las prácticas locales tempranas, los pobladores tiwanaku construyeron
canales de irrigación relativamente pequeños para ganar áreas en las tierras planas del valle medio.
Aquí, la producción de cultivos sustentó no solo al complejo Omo, sino también a numerosos pue-
blos. La sierra más alta y accidentada fue un nicho económico no explotado que Wari transformó en
un paisaje agrícola por medio de la construcción de un sistema de canales a gran altura que se
alimentaba del río Torata, a más de 2600 metros sobre el nivel del mar. El hecho de compartir los
recursos de agua en épocas en que esta era escasa llevó necesariamente a estos dos grupos a establecer
relaciones de cooperación o conflicto a lo largo de toda su historia (Williams 2002).

3. La colonia wari

La ocupación foránea se concentró en los cerros adyacentes de Cerro Baúl, Cerro Mejía y Cerro
Petroglifo, hoy deshabitados por falta de agua (Fig. 2). Cada cerro tenía residencias de bajo estatus en
sus laderas y recintos de elite en la cima, pero las instalaciones preeminentes estaban sobre Cerro
Baúl. Las diferencias entre clases y rango estaban demarcadas por la ubicación y elaboración distin-
tiva de recintos y por el acceso diferenciado a los alimentos y bienes permanentes. Por ejemplo, los
prendedores, llamados tupu, fueron un distintivo de las mujeres de elite, y la cerámica fina, como los
keros, se decoraba según el rango. Ambos tipos de evidencia fueron recuperados solo de la cima de

ISSN 1029-2004
212 WILLIAMS, NASH, MOSELEY, DE FRANCE, RUALES, MIRANDA Y GOLDSTEIN

Cerro Baúl, mientras que los estudios realizados demostraron que otros bienes y alimentos también
fueron restringidos en su distribución. En este trabajo, los autores comparan las diferencias entre
sitios y la distribución de bienes y alimentos encontrados.
Los individuos que prestaban trabajo, servicios o rendían tributos esperaban reciprocidad en
forma de comida, bebida y presentes. Los festines y el intercambio de presentes fueron un puntal de
la política de celebración de banquetes, ya que creaban obligaciones de los subordinados respecto de
sus líderes. En este artículo se analizan algunas unidades domésticas de elite en la colonia wari que
presentan instalaciones especiales para la celebración de reuniones públicas. La bebida preferida fue
la chicha, una bebida alcohólica fermentada similar a la cerveza. En la colonia de Cerro Baúl, la
cantidad y calidad de la bebida servida, así como la comida, vajilla y presentes ofrecidos, variaban
respecto de la clase y rango. Mediante el estudio de los diferentes contextos en la colonia se busca
aclarar algo más acerca de cómo la gente participaba en estos encuentros o festines, y revelar la
infraestructura subyacente en la realización de estos.
Los asentamientos en la colonia wari estaban conectados por un sistema de irrigación que llevaba
agua a los cultivos y abastecía a los colonos de agua potable. El canal wari tenía su origen a mayor
altitud que el emplazamiento de la colonia y el agua era llevada ladera abajo por medio de un canal que
irrigaba los terrenos alrededor de Cerro Petroglifo y, posteriormente, alimentaba las terrazas de
Cerro Mejía, donde el canal tenía una capacidad máxima de carga de 400 litros por segundo (Williams
1997). Cerro Mejía y Cerro Baúl se encuentran muy cerca uno del otro en El Paso, un desfiladero
natural ubicado entre ambos promontorios. Aquí, el brazo sur del canal cruzaba un gran acueducto e
irrigaba las superficies sembradas, aterrazadas o no, en las laderas de la meseta. Los colonos sembra-
ban maíz, papas, tubérculos, leguminosas y otros cultivos; también, al parecer, introdujeron el árbol
de pimiento peruano (Schinus molle), cuyos pequeños frutos fueron usados en las bebidas. Casi todos
los elementos agrícolas para los festines fueron aportados por la infraestructura local. Con una capa-
cidad de irrigación de 324 hectáreas de tierra equivalente al sostenimiento de, aproximadamente,
2000 personas, los 20 kilómetros del sistema de canales fueron el proyecto más grande realizado en la
región incluso hasta tiempos actuales. A pesar de ello, los trabajos agrícolas fueron olvidados cuando
la colonia fue abandonada hacia 1000 d.C., pero la introducción de la irrigación de altura fue un
legado duradero recogido por las sociedades subsecuentes.
El asentamiento wari pudo haber continuado canal abajo según avanzaba la construcción del canal
y avanzaba la reclamación de territorios. Cerro Petroglifo, de planta oval, fue el primer sitio en
recibir el agua canalizada. Si bien es el más pequeño, también es el más cercano al río Torata y
presenta las instalaciones más compactas y formalmente planificadas, con sus alojamientos para el
pueblo en las laderas y viviendas para la elite en la cima, hacia el lado del río. Las viviendas de caña
se dispusieron sobre terrazas de piedra construidas en las laderas. A estas terrazas y la cima se accedía
mediante dos escaleras de piedra paralelas. Los recintos y patios de la cima incluyen instalaciones que
fueron dejadas sin terminar y la ausencia de basura doméstica indica que el asentamiento pudo no
haber sido habitado (Nash 1996).

4. Los encuentros en Cerro Mejía

Los fechados radiocarbónicos sugieren dos fases de ocupación colonial divididas hacia 800 d.C., con
Cerro Mejía ubicado exclusivamente en la fase más temprana y Cerro Baúl representado en ambas
(Tabla 1, Fig. 3). Cerro Mejía es un gran promontorio que alberga al asentamiento más grande de la
colonia, pero no el más denso ni el más suntuoso. Numerosas terrazas residenciales fueron erigidas
en la ladera alrededor de El Paso y las ocupaban individuos comunes. Gracias a la escasez de lluvia,
muchas actividades se realizaron en patios abiertos. Las viviendas unifamiliares y multifamiliares,
circundadas por paredes bajas de piedra, consisten de un patio abierto para actividades generales
junto a uno o más recintos techados para cocina, almacenamiento o descanso (Fig. 4). Grandes
paredes que descienden por las laderas dividen la ocupación en seis discretos barrios. Cada uno
contenía de ocho a 15 estructuras domésticas separadas de variada elaboración y tamaño, lo que
puede reflejar distinciones en parentesco o etnicidad (Nash 2002).

ISSN 1029-2004
 Tabla 1. Fechados radiocarbónicos para la colonia wari en Moquegua.

Desviación Rango de Rango de

Lab. N. o Fecha a.p. d13 C Material Contexto
estándar 1 sigma d.C. 2 sigmas d.C.

AA 46602 1454 35 560 645 540 660 -13,60% Madera CB-A 9A, Palacio
AA 46607 1308 37 660 770 650 780 -27,10% Molle CB-A 9B, Palacio
AA 58413 1252 33 680 810 680 890 -24,50% Carbón CB-A 25.1202, Palacio
AA 58414 1237 41 690 870 680 890 -21,17% Carbón CB-A 25.1208, Palacio
Beta 36969 1370 60 600 770 550 780 Carbón CB-A 2, fuera del Palacio
AA 46603 1294 35 675 775 650 780 -24,10% Frejol CB-A 7G, fuera del Palacio
Beta 36970 1270 60 660 860 650 900 Carbón CB-A 2, fuera del Palacio
GX 24708 1220 60 710 890 670 970 -26,00% Carbón CB-A 7B, fuera del Palacio
AA 58415 1201 35 770 890 710 960 -24,90% Carbón CB-A 24.2332, fuera del Palacio
Beta 36968 1400 60 580 690 530 780 Carbón CB-B 1.2, Chichería
Beta 36967 1090 70 880 1030 770 1160 Carbón CB-B 1.2, Chichería
TX 9280 1070 50 890 1020 860 1040 -27,10% Carbón CB-B 1.4, Chichería
TX 9281 900 40 1040 1210 1030 1220 -26,70% Carbón CB-B 1.4, Chichería
AA 46595 1310 44 660 770 640 810 -27,90% Madera CB-C 10, Templo en «D»
AA 53343 1195 44 770 900 690 980 -23,20% Carbón CB-C 26.1373, Templo en «D»
GX 24709 1140 55 780 990 770 1020 -27,80% Madera CB-B 5, Templo en «D»
TX 9279 1150 50 780 980 770 1000 -23,20% Carbón CB-B 5D, Templo en «D»
GX 24706 1400 45 600 670 540 720 -24,90% Carbón CB-C 3E, Sector administrativo
AA 58411 1240 33 690 860 680 890 -25,30% Madera CB-C 6.97-1226, Sector administrativo
GX 24707 1180 50 770 940 710 990 -23,90% Carbón CB-C 3E, Sector administrativo
TX 9278 1150 50 780 980 770 1000 -27,00% Carbón CB-C 3A, Sector administrativo
AA 46599 1366 35 644,4 685,2 600 770 -23,70% Carbón CB-E, Plataforma ritual
AA 46601 1438 35 600 655 540 670 -24,30% Madera CB-H 21, Sector residencial
AA 53345 1211 46 720 890 680 960 -24,10% Carbón CB-K 30.368, Sector residencial
AA 53346 1162 61 780 960 710 1000 -24,10% Carbón CB-K 30.372, Sector residencial
AA 41951 1313 58 650 780 620 880 -25,30% Carbón CM 3B, Sector residencial
AA 41956 1289 41 680 775 650 860 -21,60% Carbón CM 136, cima
AA 41952 1284 54 670 780 650 890 -24,00% Carbón CM 145, cima
LOS ENCUENTROS Y LAS BASES PARA LA ADMINISTRACIÓN POLÍTICA WARI

AA 41955 1269 42 680 780 660 880 -22,00% Carbón CM 118, cima
AA 41957 1268 40 685 780 660 880 -26,10% Carbón CM 5, Sector residencial
AA 46598 1236 35 690 870 680 890 -25,80% Carbón CM, Canal
AA 41954 1174 43 780 940 720 980 -22,50% Carbón CM 11, El Paso
AA 41953 886 40 1040 1220 1030 1250 -10,50% Carbón CM 8, Residencial

ISSN 1029-2004
213
214 WILLIAMS, NASH, MOSELEY, DE FRANCE, RUALES, MIRANDA Y GOLDSTEIN

calib. d.C. Cerro Mejía

Sectores B-E Sector A y faldas de
1500 Unidades 1, 3, 5, 10 Unidades 9, 25, 40 Cerro Baúl

1400

1300

1200

1100

1000

900

800

700

600

500
Beta-36968

Beta-36967

Beta-36969

Beta-36970
GX-24706

GX-24709

GX-24708
AA-46599
AA-46595
AA-53343
AA-24707

AA-53342
AA-46602
AA-46601

AA-46607
AA-46603

AA-53345
AA-53346
AA-53344
AA-41951
AA-41956
AA-41952
AA-41955
AA-41957
AA-46598
AA-41954
AA-41953
TX-9278
TX-9279

TX-9280
TX-9281

Fig. 3. Fechados radiocarbónicos de contextos wari en Moquegua.

La cima de Cerro Mejía estuvo demarcada por segmentos de un grueso muro de piedra de doble
cara, con un promedio de 1,5 metros de altura y más de 1 metro de ancho. Como la cima carece de
parapetos defensivos o montones de piedras para hondas, esta barrera más parece haber segregado
clases sociales. Tiene una abertura hacia Cerro Petroglifo, así como una entrada formal a la que se
accedía por medio de un largo tramo de monumentales escaleras de piedra que proceden de El Paso.
La ocupación en la cima estaba organizada alrededor de una plaza central demarcada por cuatro
plataformas bajas de piedra, dos hacia el este y dos hacia el noroeste. Dos grandes residencias
flanqueaban la plaza hacia el norte y hacia el sur. Las estructuras conformaban recintos de elite con un
patio rectilíneo abierto, rodeado en tres o más de sus lados por grandes recintos rectangulares que
fueron típicamente techados. Los muros de mampostería eran más altos y mejor construidos que los de
las viviendas de las laderas, pero, al mismo tiempo, eran inferiores en comparación con la arquitectura

ISSN 1029-2004
 LOS ENCUENTROS Y LAS BASES PARA LA ADMINISTRACIÓN POLÍTICA WARI 215

1m

Fig. 4. Complejo residencial Unidad 5 en las laderas de Cerro Mejía.

de la cima de Cerro Baúl. Las instalaciones de la cima de Cerro Mejía estaban dispersas sobre
grandes áreas abiertas y, de esta manera, se diferenciaban del patrón wari imperial de recintos amu-
rallados, contiguos y bastante aglutinados.
Los resultados de las excavaciones sugieren que la autoridad administrativa de Cerro Mejía ocu-
paba un complejo con cuatro patios al norte de las plataformas. Las actividades domésticas de carác-
ter privado se efectuaban en dos estructuras que flanqueaban el patio en sus lados oeste y sur (Fig. 5).
El foco visual del patio fue el estrado parcialmente cercado y ubicado hacia el este. Este recinto estaba
elevado alrededor de 1 metro sobre otros edificios y su alto estatus se señalaba porque estaba pavi-
mentado con losas de riolita y decorado con dos nichos ubicados, para su visibilidad, en el muro del
fondo, con lo que se les enmarcaba por una entrada bastante grande. Este amplio acceso se abría
hacia una larga banqueta exterior, a la cual se accedía desde una plataforma baja en el patio mediante
una gradería central de piedra (Nash 2002).
El edificio norte, equipado para la preparación de banquetes, tenía tres hornos abiertos para
preparar alimentos y una especie de cuatro «cajas de fuego» diseñadas para sostener grandes ollas de
hervido. Estas últimas fueron usadas, quizá, para reducir el grano o los frutos a una masa azucarada
y fermentada para producir chicha, la que podría acompañar la comida servida durante las ceremo-
nias realizadas en el estrado y plataforma adyacentes (Nash 2002). Sin embargo, las instalaciones
destinadas a los festines eran de carácter todavía más extenso y especializado en Cerro Baúl.

5. Los encuentros en Cerro Baúl

Una calle formal asciende 400 metros hacia la cima de la meseta desde El Paso. Bastante destruido
por deslizamientos, el pasaje escalonado estaba flanqueado por terrazas residenciales dispersas bajo el

ISSN 1029-2004
216 WILLIAMS, NASH, MOSELEY, DE FRANCE, RUALES, MIRANDA Y GOLDSTEIN

Fig. 5. El complejo patio de elite Unidad 145 en la cima de Cerro Mejía.

nivel de los acantilados de la cima. De manera similar a los muros de la cumbre de Cerro Mejía, los
acantilados dividían las clases sociales y proporcionaban defensa. Los restos semienterrados y mal
conservados de los recintos sobreviven en los acantilados a lo largo de casi 400 metros de las laderas
noreste y oeste de la cumbre de la meseta. Estas instalaciones, hechas con paredes de caña y sin
mayor elaboración, albergaron a la mayoría de residentes del cerro. Algunos pobladores del asenta-
miento produjeron artefactos y generaron desechos de piedras preciosas u otros materiales diagnós-
ticos del trabajo lapidario y artesanal, mientras que otros residentes estuvieron dedicados a prestar
servicios en las actividades de la cima.
Una impresionante arquitectura monumental fue erigida en la cima de la meseta nivelada de
manera artificial. Dispuestas en complejos amurallados, las estructuras aglutinadas de uno o más
pisos cubrieron alrededor de 3 hectáreas. En las construcciones de mampostería se empleó mortero
de tierra para afianzar los irregulares bloques de piedra cortados del casquete de la meseta (Fig. 6).
Canteras contiguas a manera de pozos de 10 metros de diámetro y 3 metros de profundidad ocupa-
ron la franja sureste de la cumbre de la meseta. Cortados en material conglomerado poroso, estos
pozos no sirvieron como cisternas, ya que toda el agua, al parecer, fue transportada desde El Paso,
sobre los 2590 metros. Otros materiales de construcción, como los grandes maderos de las vigas,
pastos (Stipa ichu) para los techos y toneladas de arcilla para el mortero y enlucido, hicieron de la
ciudadela una impresionante hazaña de trabajo corporativo. De la misma manera, para pavimentar
decorativamente patios, banquetas y los pisos de segundas plantas se usaron bloques de riolita proce-
dentes de Cerro Los Angeles, a 6 kilómetros de distancia.
El colapso de la mampostería no permitía apreciar su organización original, por lo que se dividió
la arquitectura de la cima en sectores por medio de letras, con el Sector A ubicado al noreste (Fig. 6).
Hacia 800 d.C. se produjo una extensa remodelación arquitectónica, con la nivelación parcial de

ISSN 1029-2004
LOS ENCUENTROS Y LAS BASES PARA LA ADMINISTRACIÓN POLÍTICA WARI 217

Fig. 6. Arquitectura monumental en la cima de Cerro Baúl.

ISSN 1029-2004
218 WILLIAMS, NASH, MOSELEY, DE FRANCE, RUALES, MIRANDA Y GOLDSTEIN

1m
N Vasija
Piedra/batán
Ceniza/fogón
D
Estructura 3

Acceso Estructura 2

Estructura 1
Acceso

B Estructura 4
A

Acceso
Fig. 7. Complejo de la chichería, o Unidad 1, en la cima de Cerro Baúl.

edificaciones tempranas y la construcción de nuevas estructuras que, en la actualidad, conforman

parte de las ruinas en superficie (Williams 2001). Las excavaciones arqueológicas muestran el estado
de las instalaciones más tardías al momento en que dejaron de ser usadas y mantenidas. El abandono
planificado y ordenado es lo que mejor explica por qué muchas estructuras carecen de objetos que
indiquen su uso original y por qué otros edificios presentaron evidencias de una clausura ritual que
implicó festines y consumo de bebidas. La clausura de los edificios con festines ceremoniales es
importante debido a que la gente dejó, de manera intencional, artefactos que son indicativos del
estatus y naturaleza de las estructuras.

6. Lugares de producción

La ceremonia de clausura más elaborada fue la correspondiente a una gran chichería que contenía
todo el equipamiento original necesario y las ofrendas finales (Fig. 7). Las excavaciones en estas
instalaciones produjeron numerosos tupus, lo que implica una importante participación de las muje-
res de elite en la producción de chicha (Fig. 8). El edificio, de planta trapezoidal, tenía compartimentos
separados para molienda, hervido y fermentación. El cuarto de molienda estaba techado y los análisis
químicos del suelo revelaron altos niveles de fosfato, lo que indica que en este lugar se realizaba el

ISSN 1029-2004
 LOS ENCUENTROS Y LAS BASES PARA LA ADMINISTRACIÓN POLÍTICA WARI 219

Fig. 8. Ejemplar de tupu

recuperado en las excava-
ciones de la Unidad 1.

procesamiento y/o derramamiento de materiales orgánicos. La chicha de jora preparada con maíz
requiere de la molienda de granos tiernos, la que fue efectuada en el recinto con batanes. El cuarto de
hervido norte contenía, por lo menos, siete fogones cada uno, con un par de piedras a manera de
pedestal. Estas soportaban originalmente las tinajas, ahora rotas y apoyadas sobre la ceniza acumula-
da y el muro posterior de la estructura. Gruesos depósitos de ceniza y desechos del fogón sobre el piso
contenían abundantes semillas de Schinus molle. En la actualidad, estas frutillas son hervidas o remo-
jadas por completo para liberar los azúcares en unos compartimentos de resina a través de sus hoyos
centrales. Después, la mezcla de almíbar hervida es fermentada para producir chicha de molle. En el
caso de Cerro Baúl, no está claro si se produjeron dos tipos de bebidas o si el molle fue adicionado
posteriormente a la chicha de jora para crear un preparado especial. Después de hervir la mezcla en
las tinajas, esta fue transportada al área de fermentación y colocada en 12 grandes tinajas alineadas en
la pared norte del patio central (cf. Feldman 1998). En ese lugar, el líquido se añejaba por espacio de
tres a cinco días, lo que dependía de la potencia deseada. Después de la fermentación, la bebida era
vertida en jarras para su consumo. Cada tinaja podía albergar alrededor de 150 litros de chicha, lo
que indica una capacidad de producción de, aproximadamente, 1800 litros por lote. Este hecho la
convierte en una de las más grandes chicherías preinka excavada en las Américas.
Si la chichería fue la instalación central que sustentó los ritos de libación de todo el complejo
monumental, entonces esta fue esencial para la economía política colonial. Debido a que los espacios
excavados han sido rellenados con restos de festines, libaciones y ofrendas ceremoniales, a lo más se
puede tener una visión imperfecta de cómo fueron usados estos espacios; sin embargo, el hecho de que
una instalación especial dedicada a la producción de chicha estuviera ubicada en el centro de la cumbre
de Cerro Baúl demuestra que la chicha era importante para las actividades estatales wari en el sitio. De
hecho, la naturaleza formal de la chichería y su capacidad de producción sugiere que los festines fueron
un elemento crucial e integral de la economía política empleada para el éxito de la entidad política wari
en Moquegua. No obstante, la interrogante permanece: ¿quién asistió y participó de estos encuentros?
La evidencia excavada comprende un escenario de clausura planificada que habría comenzado
tres o más días antes con la preparación de lo que podría ser el lote final de chicha. Cuando la bebida
estuvo lista para su consumo, era servida en jarras a los nobles, quienes habrían estado reunidos en el
patio frente al área de fermentación. En el complejo de la chichería los autores reconocieron por lo
menos siete distintos grupos de keros (Fig. 9). Estos grupos fueron clasificados por volumen, tamaño
y, quizás, por decoración. Los más simples y pequeños tenían una capacidad de 300 mililitros, pero
el volumen aumentaba hasta un penúltimo grupo de vasijas decoradas con paneles en negro y blanco,
idénticas a los vasos usados en la capital wari. Con una capacidad de 2 litros de líquido, los keros
decorados más grandes mostraban la cabeza del Dios de los Báculos, una imagen raramente representada

ISSN 1029-2004
220 WILLIAMS, NASH, MOSELEY, DE FRANCE, RUALES, MIRANDA Y GOLDSTEIN

Fig. 9. Cuatro grupos de vasijas recuperadas en las excavaciones de la Unidad 1.

ISSN 1029-2004
 LOS ENCUENTROS Y LAS BASES PARA LA ADMINISTRACIÓN POLÍTICA WARI 221

en la colonia. Estas vasijas podrían constituir una especie de híbridos, ya que las plumas estilizadas,
configuradas en bandas a lo largo del borde, recuerdan a las convenciones tiwanaku. Luego, la vajilla
de servicio y consumo fue tirada y rota, y las llamas consumieron el cuarto techado para molienda en
la chichería. Posteriormente, cuando los rescoldos se enfriaron, se colocaron seis collares y un braza-
lete de concha y piedra sobre las cenizas en un acto final de reverencia, lo que sugiere que el depósito
fue una ofrenda de algún tipo, más que el resultado de un saqueo en la cima del enclave.
La capacidad de producción de la chichería sugiere que el maíz y el molle fueron almacenados
antes de la producción de chicha. La instalación de almacenaje identificada más cercana fue una
hilera de tres recintos rectangulares continuos, ubicada a 15 metros al norte de la chichería. Cada
compartimiento, de 2 por 2 metros, era accesible por medio de una puerta externa con un umbral
elevado en 10 centímetros para mantener lejos a las plagas. La excavación produjo semillas de calaba-
za y molle, cáscaras de maní, huesos de camélido, cuyes y peces de mar, lo que representaba un
contenido muy diverso.
Un recinto del Sector C consiste de una instalación cuatripartita diseñada para el almacenamiento
de comida, con dos cuartos alineados abiertos hacia el este y dos hacia el oeste. Cada compartimien-
to, de 12 por 5 metros, tenía un pequeño acceso central elevado a 10 centímetros sobre la superficie.
En ambos lados internos de la entrada se colocaron maderos espaciados e introducidos en los muros
a lo ancho del cuarto y a 70 centímetros sobre el piso con el objeto de crear una superficie ventilada
por debajo de los alimentos vegetales almacenados allí. Si es que en algún momento existió un ducto
de ventilación en la parte superior del muro, como en las instalaciones inka (Morris 1992), la eviden-
cia de este se habría perdido al colapsar la arquitectura superior. De la misma forma que la chichería,
este depósito techado para alimentos también fue quemado. Así, la infraestructura que sustentó el
festín pudo haber compartido el mismo estatus de clausura ceremonial que la chichería. La chichería
en sí misma pudo haber sido un lugar para festejar y beber, si bien varios contextos en la cima de
Cerro Baúl muestran evidencias de esta importante actividad.

7. Lugares de consumo

Los recintos residenciales más lujosos de la colonia fueron construidos al noroeste de la chichería y
uno de los complejos excavados parcialmente podría corresponder al palacio del gobernador. A di-
chos recintos se accedía por medio de un estrecho corredor que llevaba hacia la entrada sur de un
espacioso patio sin techar. De acuerdo con su rango o actividades, las personas se podían acomodar
en banquetas de piedra labrada, de 20 centímetros de alto, dispuestas a lo largo de las paredes y sobre
el piso inferior (Fig. 10). Una estructura techada de planta en forma de «U», que se abre hacia el patio
a la altura de la banqueta, estaba dispuesta en el centro del muro oeste. A manera de comparación, los
pequeños y elevados edificios de planta en «U» funcionaron como «oficinas» de la elite para los
imperios chimú e inka, época en la cual se les denominó «masma». Las representaciones artísticas
chimú muestran una figura de pie en el centro del acceso a una estructura con planta en forma de «U»
que interactúa con individuos menores en el patio inferior (Moseley 2001). Si el patio de acceso o
masma en Cerro Baúl hubiera tenido una naturaleza similar, habría hecho las veces de «oficina del jefe
ejecutivo» que conducía la política colonial respecto de las elites subordinadas.
Un acceso en el muro opuesto de la estructura con planta en forma de «U» lleva hacia un patio
abierto de 8 por 8,2 metros, usado para la producción cerámica. Esto se hizo evidente debido a la
presencia de diferentes depósitos en el piso que contenían arcilla cruda, varios tipos de temperante,
así como herramientas para la molienda de materia prima e instrumentos para el acabado de las
vasijas. Al parecer, los productos acabados fueron usados durante los encuentros como bienes de
consumo o intercambiados como regalos. Los análisis químicos de cerámica producida para la gente
que residía fuera de la cima de Cerro Baúl como para los que vivieron en ella, procedente de los
puntos coloniales más amplios así como de diferentes centros de producción, señalan que los poblado-
res de este último usaban vasijas distintivas en materia prima y calidad que emulaban a la cerámica de
las elites de la capital imperial (Pérez 1995; Williams et al. 2003).

ISSN 1029-2004
 222

ISSN 1029-2004
WILLIAMS, NASH, MOSELEY, DE FRANCE, RUALES, MIRANDA Y GOLDSTEIN

Fig. 10. Áreas excavadas del Complejo Palacio (unidades 25, 40, 9) en la cima de Cerro Baúl.
 LOS ENCUENTROS Y LAS BASES PARA LA ADMINISTRACIÓN POLÍTICA WARI 223

A los lugares sagrados de la residencia palaciega se accedía por medio de dos pasajes en el lado
noroeste del patio de acceso, con el del norte abierto hacia un estrecho corredor que llevaba hacia un
gran grupo-patio con un patio abierto pavimentado rodeado de banquetas de 30 centímetros de alto,
las que, a su vez, están circundadas por cinco edificios contiguos techados y una plataforma ubicada
al este, también de 30 centímetros de alto (Fig. 10). La plataforma, pavimentada con losas y muy mal
conservada, hizo las veces, quizá, de una estructura tipo estrado para supervisar las ceremonias del
patio, como en las instalaciones de Cerro Mejía. Las excavaciones en cuatro de los recintos dispues-
tos al nivel del piso proporcionaron piedras de moler, lascas, husos y otros instrumentos textiles, así
como otros restos que indican actividades domésticas. Las instalaciones para la preparación de ban-
quetes, aún sin excavar, pueden encontrarse en otro lugar del complejo o, en todo caso, la comida y
la chicha para la realización de ceremonias especiales, como las de clausura, pudieron haber sido
llevadas desde otro lugar.
Las estructuras con planta en forma de «D» en la capital y otros sitios wari son interpretadas como
«templos de elite» (Bragayrac 1991; Cook 2001), y en Cerro Baúl se han encontrado dos de ellas (Fig. 6).
Estas se abren hacia el noreste, tienen paredes altas con nichos y un diámetro entre 10 y 12 metros.
Podían acomodar a pocos participantes y estaban orientadas hacia patios que pudieron albergar
audiencias modestas. El fosfato y nitrógeno en las muestras de piso de los edificios más pequeños
pueden indicar la presencia de comida y bebida, pero no pueden determinar si estos fueron un
resultado de uso normal o de ceremonias de clausura. Además, no hay evidencias de festines rituales
o rotura de vasijas. Las acciones posteriores al abandono asociadas con estas estructuras hacen difícil
las interpretaciones acerca de las actividades que se llevaron a cabo allí. A pesar de ello, fue quemado
un corredor techado que llevaba hacia el patio de la estructura más pequeña. Por otro lado, el aban-
dono ceremonial también estuvo presente en el recinto parcialmente investigado anexo al templo más
grande. La gran estructura con planta en forma de «D», denominada «Unidad 10», es parte de un gran
recinto amurallado. La Unidad 26 está dentro de este gran complejo y se ubica de manera adyacente
a la estructura con planta en forma de «D». Tiene un edificio central de planta rectangular, con una
plataforma interna hacia un extremo y un depósito en el otro, abierto hacia el Norte, así como una
banqueta frontal con un atrio. Dicha banqueta mira hacia una hilera de cuatro pequeños cuartos
contiguos construidos contra la esquina oeste del muro norte del recinto (Fig. 11). Tres de ellos eran
compartimentos tipo depósito dispuestos sobre la superficie, a los que se accedía por medio de un
muro frontal bajo.
El cuarto de la esquina era más elaborado y tenía una entrada que estaba enlucida y pintada. La
entrada era muy ancha y proporcionaba una vista del cuarto interno. Dos cistas circulares en el piso
del cuarto de la esquina cortaban el piso enlucido. La cista más grande contenía un entierro sin
ofrendas funerarias correspondiente a un adolescente prepúber en posición sentada y flexionada. La
otra contenía los restos de un infante dentro de una jarra de tamaño mediano; la jarra estaba cubierta
por un cuenco invertido. Si estas inhumaciones simples fueron ofrendas funerarias, entonces el
sacrificio habría conferido un carácter sagrado al singular contenido del recinto: un gran tambor de
cerámica decorado y con base redondeada (Fig. 12). A pesar de no haberse conservado, la piel del
tambor estuvo originalmente atada a través de 14 agujeros perforados alrededor del borde sin decorar
del instrumento antes de que este fuera quemado en el horno. El tambor pudo haber sido exhibido
boca abajo, pero no podría haberse tocado en esa posición ya que la iconografía está invertida en el
fondo del instrumento. Una banda roja con dos líneas negras rodea la concavidad central del instru-
mento sobre una hilera de cabezas estilizadas de aves pintadas de color negro. La base polícroma
contiene la representación de tres figuras andróginas serpentinas, cada una con un gorro en forma de
cono con rayas verticales y uno con una lanza (Fig. 12). La iconografía no es la correspondiente al
estilo Wari de Ayacucho y, más bien, estaría relacionada con Nazca Tardío, un estilo costeño contem-
poráneo en territorios 500 kilómetros al norte de la zona de Cerro Baúl. El recinto también contenía
un grupo de cuatro cuencos de estilo Loro, de filiación nazca tardío.
El tambor fue roto al momento del abandono, así como los cuatro cuencos, que podrían haber
estado guardados en el cuarto o fueron agregados a las ofrendas de clausura. La bebida durante los

ISSN 1029-2004
224 WILLIAMS, NASH, MOSELEY, DE FRANCE, RUALES, MIRANDA Y GOLDSTEIN

Fig. 11. Plano de planta del anexo del templo en la Unidad 26.
ritos finales está implícita en la presencia de un pequeño kero fragmentado y de finas tazas decoradas
y rotas en la banqueta, por debajo del atrio quemado del edificio central y en el área este del patio,
que también contenía restos de vajilla rota y huesos de camélido. Entre estos estaban los restos de un
animal adulto joven, posiblemente una llama. En aparente sacrificio, la bestia fue muerta y la carne
recortada, con marcas evidentes de cortes en el cráneo y partes postcraneales. Sus huesos fueron
colocados en un hoyo superficial con las partes anteriores del cráneo en dirección hacia el cuarto del
tambor. Esta forma difiere de otros rituales de abandono, si bien los rituales de libación, festines y
quema de techos son denominadores comunes.

ISSN 1029-2004
Fig. 12. Tambor de estilo Nazca. A la derecha, el desarrollo a color que muestra tres figuras danzantes (Dib.: A. Umire).
Fig. 13. Placa de cobre de estilo La Aguada.
 LOS ENCUENTROS Y LAS BASES PARA LA ADMINISTRACIÓN POLÍTICA WARI 225

8. Diferenciación social en la cocina y productos

Además de las distinciones en la arquitectura ritual y residencial, también es evidente una diferencia-
ción social en los bienes suntuarios, dieta y uso ceremonial de plantas y animales a lo largo del enclave
wari. Todos los habitantes consumieron maíz, quenopodiáceas, leguminosas, maní y ají, pero los
residentes de la cumbre de Cerro Baúl tuvieron un acceso exclusivo a la chicha de molle, coca y
tabaco (Tabla 2). Aunque se han identificado casi 50 especies diferentes de plantas procedentes de la
cima de Cerro Baúl, menos de la mitad de ellas están presentes en los contextos correspondientes a
los individuos comunes en las laderas. Además, a pesar de que se encuentran semillas de molle
aisladas en los contextos al exterior de la cima, solo en la meseta se han recuperado gran cantidad de
desechos de Schinus molle que indican la producción de chicha.
Los restos animales también indican rigurosas diferencias en la obtención y consumo de fauna y
uso ritual de los animales. El corpus de Cerro Mejía consiste, casi de manera exclusiva, de restos de
camélidos grandes y pequeños —presumiblemente llamas y alpacas domésticas—, con un uso fortui-
to de especies disponibles en el ámbito local como vizcacha, liebre andina, venado, si bien solo la
cornamenta y no la parte comestible, y un ave no identificada (Tabla 3). En consecuencia, se puede
inferir que los residentes de Cerro Mejía carecían de una fauna exótica.
En la cima de Cerro Baúl, los camélidos grandes y pequeños, así como el cuy, fueron la principal
fuente de proteínas. Los abundantes restos del último y su ausencia en Cerro Mejía sugieren que el
cuy fue destinado para la elite. Otros alimentos animales comprenden la vizcacha, dos especies de
venado —con la inclusión de elementos postcraneales—, perdices, pichones, palomas y, por lo me-
nos, 10 especies de peces del Pacífico, desde anchoveta hasta atún, todos encontrados en el palacio y
otros contextos monumentales (Tabla 4). También están presentes los restos de varios animales inter-
pretados como no comestibles y que podrían haber sido obtenidos localmente o de hábitats marinos
ubicados a 100 kilómetros de distancia. De baja importancia dietética, estas especies son difíciles de
adquirir y representarían ítems de valor simbólico o ritual. Estos incluyen por lo menos un león de la
montaña o puma (una falange distal o porción de la garra), un pequeño gato nativo, el zorro serrano,
dos perros domésticos juveniles, el cóndor andino (dos elementos de las alas: un cúbito distal traba-
jado y una falange distal de la especie Vultur gryphus), un búho enano, un atrapamosca (Colonia
colonus), por lo menos dos sapos y un único diente de tiburón mako (Isurus oxyrinchus). La mayoría
de estos restos de animales exóticos provienen de los contextos del palacio y la chichería.
Todos los animales encontrados en el enclave wari eran habitantes de la sierra y costa del Pacífico,
pero aún no se ha determinado cómo fueron obtenidos. Los camélidos fueron criados localmente y tres
especímenes tienen rasgos osteológicos isotópicos serranos (Kennedy 2003), pero algunos de estos
animales en el corpus de fauna provienen de la costa. El banquete final en el palacio incluyó los restos de
tres camélidos con rasgos isotópicos costeros, así como una gran variedad de peces marinos. No existe
evidencia de que los colonos wari pescaran o pastorearan a lo largo de la costa de Moquegua, por lo que
se deduce que los vecinos de las tierras bajas debieron haber producido productos costeros que llegaron
a Cerro Baúl por medio del intercambio con gente foránea o, en todo caso, los líderes de la costa
podrían haber intercambiado bienes en comida con los anfitriones wari del enclave.
En la colonia fueron recuperados otros tipos de bienes, quizás obtenidos o distribuidos por las
elites wari. Por ejemplo, la obsidiana, usada para cuchillos y puntas de proyectil, fue importada
principalmente de las minas cerca del núcleo wari, más que de las fuentes más cercanas usadas por los
tiwanaku (Burger et al. 2000). Esta aparece en todos los contextos, desde los espacios usados por los
individuos comunes hasta los de elite; sin embargo, la frecuencia de la obsidiana decrece mientras
mayor es la distancia a la cima de Cerro Baúl, como ocurre con la crisocola y el lapislázuli, usados
para las cuentas. Del palacio y la chichería se obtuvieron especímenes especiales de la concha ecua-
toriana Spondylus, de un origen más allá de las fronteras wari, y de la cerámica caolín del estilo
Cajamarca, procedente de la frontera norte del imperio. Originarios de casi 2000 kilómetros y 1300
kilómetros al norte, estos bienes exóticos se movieron presumiblemente hacia el sur a lo largo de las
rutas de intercambio hasta alcanzar las elites de la cima. En el palacio también se encontró una placa
de cobre del estilo La Aguada, de la región Catamarca en Argentina, aproximadamente 1300 kilómetros

ISSN 1029-2004
Tabla 2. Especímenes de flora recuperados de la cima de Cerro Baúl.

Flora de la cima de Cerro Baúl (48 especímenes)

Anacardiaceae Azioiaceae
Schinus molle (árbol de pimiento peruano) Trianthema sp.?
Chenopodiaceae Violiaceae
Chenopodium sp. (quenopodios) Viola sp.
Atriplex sp. Convulvucaceae
Poaceae Brassicaceae
Zea mays (maíz) Lepidium sp.
Stipa ichu (ichu) Plantaginaceae
Bromus sp. Plantago sp.
Andropogon sp. Nyctaginaceae
Eragrostis sp. Menispermaceae
Cactaceae Rubiaceae
Opuntia sp. (tunal) Gallium sp.
Haageocereus sp. (cactus) Papaveraceae
Echinopsis sp. (cactus) Papaver sp.
Cyperaceae Solanaceae
Rhynchospora sp. Capsicum sp. (ají)
Scirpus sp. Nicotiana sp. (tabaco)
Portulacaceae Fabaceae
Portulaca sp.
Amaranthaceae Arachis sp. (maní)
Amaranthus sp. Phaseolus sp. (haba)
Araeceae Haba sp. (haba)
Asteraceae Heliotropum sp.
Bidens sp. [mate medicinal] Acacia sp. [mate medicinal]
Tagetes sp. [mate medicinal] Cucurbitaceae
Verbesina sp. [mate medicinal] Lagenaria sp. (calabaza)
Malvaceae Cyclanthera sp.
Malva sp. Erythoxylaceae
Gossypium sp. [mate medicinal] Erythoxylon coca (coca)
Verbenaceae Papaver sp.
Verbena sp. [mate medicinal] Oxalidaceae
Geraniaceae Oxalis sp.
Erodium sp. Euphorbiaceae
Boraginaceae Euphorbium sp.
Heliotropum sp. [mate medicinal] Rosaceae
Scrophluariaceae Rubus sp.
Calceolaria sp. Azorella sp.? (llareta) [resina de incienso]

Flora de las laderas de Cerro Baúl (18 especímenes)

Anacardiaceae Fabaceae
Schinus molle Arachis sp. (maní)
(árbol de pimiento peruano) Phaseolus sp. (haba)
Cucurbitaceae
Chenopodiaceae Lagenaria sp. (calabaza)
Chenopodium sp. (quenopodios) Asteraceae
Poaceae Bidens sp. [mate medicinal]
Zea mays (maíz) Rudibeckia sp.
Stipa ichu (ichu) Stevia sp.
Bromus sp. Malvaceae
Verbenaceae
Cactaceae Verbena sp. [mate medicinal]
Haageocereus sp.(cactus) Brassicaceae
Echinopsis sp.(cactus) Lepidium sp.
Solanaceae Myrtaceae
Tabla 3. Restos faunísticos de contextos wari de la cima y laderas de Cerro Baúl.

Identificación Nombre común NME NMI Peso óseo

taxonómica (gr)

Cavia porcellus Cuy 616 17 95,9

Lagidium peruanum Vizcacha 13 2 10,6
Rodentia no identificado (pequeño) Roedor pequeño 74 12 4,4
Sigmodontinae Ratón 12 0 <,1
Canis familiaris Perro 59 2 16,5
Felis concolor Puma 1 1 0,9
Felis sp. Felino 1 1 0,4
Pseudalopex culpaeus Zorro serrano 1 1 1,5
Lama cf. glama Llama cf. llama 170 4 1946,7
Lama sp. Llama, alpaca 1390 9 6511,7
Lama sp. (grande) Camélido grande 114 3 1755,5
Lama sp. (pequeño) Camélido pequeño 11 1 118,1
Camelidae Camélidos 3 0 5,6
Hippocamelus antisensis Taruca 3 1 45,9
Odocoileus virginianus Venado 12 2 106,4
Camelidae/Cervidae Camélidos/cérvidos 19 0 7,6
Mamífero no identificado (grande) Mamífero grande no identificado 8477 0 9721
Mamífero no identificado (pequeño) Mamífero pequeño no identificado 30 0 71
Mamífero no identificado Mamífero no identificado 2913 0 353,8
Total mamíferos 13919 56 20773,5
Vultur gryphus Cóndor andino 2 1 3,6
Nothoprocta cf. ornata Perdiz 1 1 0,6
Nothoprocta sp. Perdiz 7 1 2,4
Columba maculosa Paloma 9 2 1,2
Zenaida auriculata Pichón 1 1 0,1
Glaucidium cf. peruanum Búho 1 1 0,1
cf. Muscisaxicola sp. Atrapamosca 1 1 <,1
Aves no identificadas Aves no identificadas 31 0 10,8
Total aves 53 8 18,8
Lacertilia Lagarto 29 2 0,5
Total reptiles 29 2 0,5
Bufo cf. arequipensis Sapo 3 2 0,5
Anfibios no identificados Anfibios no identificados 2 0 0,2
Total anfibios 5 2 0,7
Isurus oxyrinchus Tiburón mako 1 1 2,7
Total Chondrichthyes 1 1 2,7
Engraulidae Anchoveta 8 2 <,1
Clupeidae Sardina 82 4 1,2
cf. Atherinidae Pejerrey 1 1 <,1
Carangidae Cojinova 10 2 0,7
cf. Carangidae Cojinova 2 1 0,2
Cheilopogon sp. Pez volador 6 2 0,3
Labrisomus sp. Blenio
Scartichthys sp. Blenio 2 1 0,1
Sciaena deliciosa Lorna 1 1 0,1
Trachurus murphyi Jurel 192 7 47,8
Auxis sp. Atún 1 1 0,3
Osteichthyes no identificados Peces no identificados 1012 0 20,1
Total Osteichthyes 1317 22 70,8
Muestra Total 15325 92 20867,1
228 WILLIAMS, NASH, MOSELEY, DE FRANCE, RUALES, MIRANDA Y GOLDSTEIN

Tabla 4. Restos faunísticos de contextos wari de Cerro Mejía.

Identificación Peso óseo

Nombre común NME NMI (gr)
taxonómica

cf. Phyllotis sp. Roedor 5 3 <,1

Lagidium peruanum Vizcacha 1 1 2,1
Lama sp. Llama 264 3 1101,8
Lama cf. glama Llama 28 3 269,2
Cervidae Venado 5 1 51,9

Total mamíferos 303 11 1425

Aves no Aves no
identificadas identificadas 2 1 n/a

Total aves 2 1 n/a

Muestra total 305 12 1425

al sur (Fig. 13). Este inusual objeto podría haber llegado aquí por obra de intermediarios tiwanaku,
cuya nación ejerció una efectiva influencia en dirección sur, a unos 500 kilómetros de Catamarca.
Aunque en la cima de Cerro Baúl están presentes objetos de elite asociados con estilos distantes,
el intercambio con los vecinos cercanos fue muy diferente. La nobleza consumió alimentos exóticos
de la costa, pero no hay evidencia de artefactos o arte costeño. De manera significativa, el arte
tiwanaku está ausente en la colonia wari, aunque las dos poblaciones vivieron una junto a la otra por
siglos. Esta situación genera diversas expectativas. En este caso solo se puede especular que si el
enclave wari tuvo un carácter de «embajada», entonces las estrategias correspondientes apuntaron a
mantener una pureza en la identidad étnica y diferenciación política por medio de la prohibición de
la incorporación de los objetos simbólicos de sus vecinos imperiales.

9. Discusión

La complejidad social en la cuenca de Moquegua aumentó cuando las poblaciones wari y tiwanaku
colonizaron el área y residieron cercanas, entre sí, por casi 400 años. Los dos regímenes políticos fueron
étnicamente heterogéneos y el enclave en Cerro Baúl presentaba tres grupos distinguibles de residentes.
El grupo dominante, que, de manera presumible, procedía de la capital, habitó la cima de la meseta.
Dividida entre la elite de la cima y los individuos comunes de la ladera con un mínimo de parafernalia
wari, la población de Cerro Mejía procedía de algún lugar en la esfera wari y fungió de personal de
servicio hacia 800 d.C. (Nash 2002). Después de esto, pobladores relacionados con Tiwanaku comen-
zaron a habitar las aldeas en la región inmediata a Cerro Baúl (Williams y Nash 2002; Owen 2005). El
canal de Cerro Baúl abasteció tres de sus asentamientos y la presencia de obsidiana wari implica que
estos individuos tiwanaku constituyeron comunidades clientes del personal de servicio. Con la excep-
ción de la obsidiana, los elementos wari están ausentes en el corpus arqueológico tiwanaku local. Si los
objetos decorativos de estilo Wari consituyeron una insignia social utilizada solo por colonos de filiación
social wari dentro de la colonia, es probable que se aplicaran restricciones parecidas a sus vecinos
foráneos, los tiwanaku, en el valle superior fuera de la colonia, lo que resultó en una escasez de arte
suntuario tiwanaku en contextos de elite wari y viceversa en toda el área de Moquegua.
Junto con los festines, la libación ritual fue importante para ambas entidades. Valle abajo, en el
complejo arqueológico de Omo, se ubicó una instalación para la producción y el consumo de bebidas
de 4 por 4 metros (Goldstein 1993). Esta tuvo una sola tinaja para hervido, tres vasijas de fermenta-
ción y una capacidad de producción estimada en 360 litros. La molienda de grano se realizó en varias

ISSN 1029-2004
 LOS ENCUENTROS Y LAS BASES PARA LA ADMINISTRACIÓN POLÍTICA WARI 229

partes, pero la fermentación y el consumo en finos vasos keros se efectuaron en la estructura. Usada
tanto para la producción de chicha como de alimentos, la instalación más grande y de calidad más
fina en Cerro Mejía tuvo una capacidad de producción similar, pero el consumo se efectuaba en el
patio adyacente. La monumental chichería de Cerro Baúl, con compartimentos discretos para mo-
lienda, cocina y fermentación, es la planta de producción especializada de chicha más grande que se
conoce para esta época y los numerosos prendedores tupus encontrados implican la existencia de un
grupo laboral conformado por mujeres de elite.
Por razones desconocidas, Wari y Tiwanaku comenzaron a decaer hacia 1000 d.C. Si los tiwanaku
cayeron primero, no había necesidad para una embajada wari. No existen evidencias de una coac-
ción económica en el asentamiento o sus terrenos agrícolas, y las decisiones políticas pueden explicar
mejor la fundación y abandono del enclave. El contexto de la chichería proporciona la última fecha
para la ocupación de la cima de Cerro Baúl (Williams 2001) y sugiere que esta fue una instalación
importante hacia el final de ella. Al parecer, la chicha y la chichería habrían sido cruciales para
practicar ceremonias y rituales en el sitio. La chichería fue necesaria para realizar ritos de clausura y
podría haber sido la última estructura en ser destruida antes del abandono.

10. Conclusiones

La meseta de Cerro Baúl, un punto crucial en la sierra peruana, está coronada por ruinas monumen-
tales construidas por los colonos wari en la frontera más sureña de su nación. La cima es un lugar
completamente inadecuado para vivir en estado de opulencia, pero el enclave fue establecido como
una manifestación de poder y distinción. Tanto las provisiones para banquetes elaborados, así como
el agua y alimentos diarios, tuvieron que ser transportados a la cima de la meseta a costa de grandes
esfuerzos. Muchos de los elementos usados en estos encuentros viajaban decenas o cientos de kilóme-
tros para la realización de eventos especiales. La producción de chicha a gran escala tuvo una deman-
da similar, y la producción y consumo sobre la meseta fue una empresa costosa.
En ausencia de una razón económica evidente, la ocupación de la cima de Cerro Baúl está mejor
explicada por consideraciones políticas. Los autores proponen que los wari eligieron asentar una
delegación imperial sobre un bastión natural sagrado para establecer una representación política
impresionante en la única región donde se podía tener un contacto inmediato con el imperio tiwanaku.
Mientras los líderes locales continuaron con la celebración de «festines facultativos», los señores wari
establecían instalaciones especializadas para la realización de festines de producción y consumo sobre
la cima de la montaña. Estos eventos estaban imbuidos de nociones de estatus y jerarquía, tal como
se muestra en las plataformas especiales destinadas para las elites que presidían las celebraciones y la
presencia de vasijas de consumo diferenciado. Es probable que las instalaciones para realizar estas
fiestas a gran escala, en las que los señores regionales supremos recibían a sus subordinados, fueran
atendidas también por el personal de servicio de la elite. Las concentraciones de tupus en la cima de
Cerro Baúl, y especialmente en la chichería, sugieren que ciertas mujeres de elite fueron residentes
importantes del complejo de la cima y que, quizás, fueron los personajes principales en la producción
de chicha. Ellas trabajaban en instalaciones especializadas cuyo único propósito era la producción de
comida y bebida para los encuentros entre los líderes locales y los señores wari.
Tanto en los contextos wari como tiwanaku, los líderes coloniales presidieron los encuentros para
los individuos que participaban en ellos, ya fueran distinguidos o de menor rango. Estos eventos eran
importantes para constituir y mantener las relaciones sociales entre los miembros de las respectivas
comunidades coloniales. Dichos eventos difieren en clase y especialización de las fiestas efectuadas
en el centro provincial de Cerro Baúl, ya que este tuvo instalaciones únicas y calidad de alimentos de
carácter incomparable en toda la región para la realización de sus encuentros. En ese sentido, el
Estado wari bien pudo ser construido sobre la base de las fiestas comunales. Las elites provinciales
wari pudieron haber albergado a dignatarios foráneos, pero, además, celebraron eventos con sus
subordinados en los centros estatales. Estos festines tipo patróncliente constituyeron, quizás, el tipo
más importante de reunión existente y sirvieron de base para el Estado wari mediante el refuerzo de
la jerarquía social y la promulgación de sus ideales e ideología.

ISSN 1029-2004
230 WILLIAMS, NASH, MOSELEY, DE FRANCE, RUALES, MIRANDA Y GOLDSTEIN

REFERENCIAS

Betanzos, J. de
1996 Narrative of the Incas [traducción y edición de R. Hamilton y D. Buchanan], University of Texas Press, Austin.
[1557]

Bragayrac, D. E.
1991 Archaeological Excavations in the Vegachayoq Moqo Sector of Huari, en: W. H. Isbell y G. F. McEwan
(eds.), Huari Administrative Structure: Prehistoric Monumental Architecture and State Government, 71-80, Dumbarton
Oaks Research Library and Collection, Washington, D.C.

Brewster-Wray, C. C.
1989 Huari Administration: A View from the Capital, en: R. M. Czwarno, F. M. Meddens y A. Morgan (eds.), The Nature
of Huari: A Reappraisal of the Middle Horizon Period in Perú, 23-33, BAR International Series 525, London.

Burger, R. L., K. L. Mohr-Chávez y S. Chávez

2000 Through the Glass Darkly: Prehispanic Obsidian Procurement and Exchange in Southern Perú and Northern
Bolivia, Journal of World Prehistory 14 (3), 267-362, New York.

Cook, A. G.
2001 Huari D-Shaped Structures, Sacrificial Offerings, and Divine Rulership, en: E. P. Benson y A. Cook (eds.),
Ritual Sacrifice in Ancient Perú, 137-164, University of Texas Press, Austin.

Cook, A. G. y M. L. Glowacki
2003 Pots, Politics, and Power: Huari Ceramic Assemblages and Imperial Administration, en: T. L. Bray (ed.), The
Archaeology and Politics of Food and Feasting in Early States and Empires, 173-202, Kluwer Academic/Plenum
Publishers, New York.

D’Altroy, T. N.
2001 From Autonomous to Imperial Rule, en: T. N. D’Altroy y C. A. Hastorf (eds.), Empire and Domestic Economy,
325-339, Kluwer Academic Publishers, New York.

D’Altroy, T. N. y T. K. Earle
1985 Staple Finance, Wealth Finance, and Storage in the Inka Political Economy, Current Anthropology 26, 187-206,
Chicago.

Dietler, M.
2001 Theorizing the Feast: Rituals of Consumption, Commensal Politics, and Power in African Contexts, en: M.
Dietler y B. Hayden (eds.), Feasts: Archaeological and Ethnographic Perspectives on Food, Politics, and Power, 65-
114, Smithsonian Institution Press, Washington, D.C.

Dietler, M. y I. Herbich
2001 Feasts and Labor Mobilization: Dissecting a Fundamental Economic Practice, en: M. Dietler y B. Hayden (eds.),
Feasts: Archaeological and Ethnographic Perspectives on Food, Politics, and Power, 23-64, Smithsonian Institution
Press, Washington, D.C.

Feldman, R. A.
1998 La ciudadela wari de Cerro Baúl en Moquegua, en: K. Wise (ed.), Moquegua: los primeros doce mil años, 59-65,
Museo Contisuyo, Moquegua.

Garcilaso de la Vega, I.
1966 Royal Commentaries of the Incas and General History of Perú [traducción de H. V. Livermore], 2 vols., University of
[1609] Texas Press, Austin.

Gero, J. M.
2001 Field Knots and Ceramic Beaus: Interpreting Gender in the Peruvian Early Intermediate Period, en: C. F. Klein
(ed.), Gender in Pre-Hispanic America: A Symposium at Dumbarton Oaks, 12th and 13th October1996, 15-55,
Dumbarton Oaks, Washington, D.C.

Goldstein, P. S.
1993 House, Community, and State in the Earliest Tiwanaku Colony: Domestic Patterns and State Integration at
Omo M12, Moquegua, en: M. S. Aldenderfer (ed.), Domestic Architecture, Ethnicity, and Complementarity in the
South-Central Andes, 25-41, University of Iowa Press, Iowa City.

ISSN 1029-2004
 LOS ENCUENTROS Y LAS BASES PARA LA ADMINISTRACIÓN POLÍTICA WARI 231

2005 Andean Diaspora: The Tiwanaku Colonies and the Origins of South American Empire, University of Florida Press,
Gainesville.

Guamán Poma de Ayala, F.

1980 El primer nueva corónica y buen gobierno (edición de J. V. Murra, R. Adorno y J. Urioste), 3 vols., Siglo XXI,
[1615- México, D.F.
1616]

Isbell, W. H.
1997 Reconstructing Huari: A Cultural Chronology for the Capital City, en: L. Manzanilla (ed.), Emergence and
Change in Early Urban Societies, 181-227, Plenum Press, New York.

Kennedy, E.
2003 Isotopic Evidence for Camelid Herding Patterns in the Osmore Drainage, Perú, informe, University of Florida,
Gainesville.

Kolata, A. L.
1993 The Tiwanaku: Portrait of an Andean Civilization, Blackwell, Cambridge.

2003a The Social Production in Tiwanaku: Political Economy and Authority in a Native Andean State, en: A. L.
Kolata (ed.), Tiwanaku and its Hinterland: Archaeology and Paleoecology of an Andean Civilization, Vol. 2. Urban
and Rural Archaeology, 449-472, Smithsonian Institution Press, Washington, D.C.

2003b Tiwanaku Ceremonial Architecture and Urban Organization, en: A. L. Kolata (ed.), Tiwanaku and its Hinterland:
Archaeology and Paleoecology of an Andean Civilization. Vol. 2, Urban and Rural Archaeology, 175-201, Smithsonian
Institution Press, Washington, D.C.

Lumbreras, L. G.
1974 The Peoples and Cultures of Ancient Perú, Smithsonian Institution Press, Washington, D.C.

Morris, C.
1982 The Infrastructure of Inka Control in the Peruvian Central Highlands, en: G. A. Collier, R. I. Rosaldo y J. D.
Wirth (eds.), The Inca and Aztec States, 1400-1800: Anthropology and History, 153-171, Academic Press, New York.

1985 From Principles of Ecological Complementarity to the Organization and Administration of Tawantinsuyu, en: S.
Masuda, I. Shimada y C. Morris (eds.), Andean Ecology and Civilization, 477-490, University of Tokyo Press, Tokyo.

1992 The Technology of Highland Inka Food Storage, en: T. LeVine (ed.), Inka Storage Systems, 237-258, University
of Oklahoma Press, Norman.

Morris, C. y D. E. Thompson
1985 Huánuco Pampa: An Inca City and its Hinterland, Thames and Hudson, London/New York.

Moseley, M. E.
2001 The Incas and their Ancestors, The Archaeology of Perú, Thames and Hudson, London/New York.

Murra, J. V.
1980 The Economic Organization of the Inca State, Research in Economic Anthropology Supplement 1, JAI Press,
Greenwich.

1982 The Mit’a Obligations of Ethnic Groups to the Inka State, en: G. A. Collier, R. I. Rosaldo y J. D. Wirth (eds.),
The Inca and Aztec States, 1400-1800: Anthropology and History, 237-262, Academic Press, New York.

Nash, D. J.
1996 Cerro Petroglifo: Settlement Pattern and Social Organization of a Residential Wari Community, tesis de
maestría, Department of Anthropology, University of Florida, Gainesville.

2002 The Archaeology of Space: Places of Power in the Wari Empire, tesis de doctorado, Department of Anthropology,
University of Florida, Gainesville.

Owen, B. D.
2005 Distant Colonies and Explosive Collapse: The Two Stages of the Tiwanaku Diaspora in the Osmore Drainage,
Latin American Antiquity 16 (2), 45-80, Washington, D.C.

ISSN 1029-2004
232 WILLIAMS, NASH, MOSELEY, DE FRANCE, RUALES, MIRANDA Y GOLDSTEIN

Pérez, I.
1995 Informe «Excavaciones de salvataje y propuesta sobre conservación de estructuras descubiertas, área Museo de
Sitio en Wari», Universidad Nacional San Cristóbal de Huamanga, Ayacucho.

Ramírez, S. E.
2005 To Feed and Be Fed: The Cosmological Bases of Authority and Identity in the Andes, University of Texas Press,
Austin.

Rostworowski de Diez Canseco, M.

1999 History of the Inca Realm [traducción de H. B. Iceland], Cambridge University Press, Cambridge.

Rowe, J. H.
1946 Inca Culture at the Time of the Spanish Conquest, en: J. H. Steward (ed.), Handbook of South American
Indians. Vol. II, The Andean Civilizations, Bureau of American Ethnology, Bulletin 143, 183-330, Washington,
D.C.

1982 Inca Policies and Institutions Relating to Cultural Unification of the Empire, en: G. A. Collier, R. I. Rosaldo
y J. D. Wirth (eds.), The Inca and Aztec States, 1400-1800: Anthropology and History, 93-118, Academic Press,
New York.

Schreiber, K. J.
1992 Wari Imperialism in Middle Horizon, Perú, Anthropological Papers of the Museum of Anthropology 87, Ann
Arbor.

Williams, P. R.
1997 The Role of Disaster in the Development of Agriculture and the Evolution of Social Complexity in the South-
Central Andean Sierra, tesis de doctorado, Department of Anthropology, University of Florida, Gainesville.

2001 Cerro Baúl: A Wari Center on the Tiwanaku Frontier, Latin American Antiquity 12, 67-83, Washington, D.C.

2002 Rethinking Disaster-Induced Collapse in the Case of the Andean Highland States: Wari and Tiwanaku,
World Archaeology 33 (3), 361-374, London.

Williams, P. R. D. J. Nash, M. L. Glowacki, R. J. Speakman, H. Neff, M. D. Glascock, M. Ruales y J. Clark

2003 Using INAA to Assess Wari Ceramic Production and Exchange on the Tiwanaku Frontier, ponencia presentada
al 51.o Congreso Internacional de Americanistas, Santiago, Chile.

Williams, P. R. y D. J. Nash
2002 Imperial Interaction in the Andes: Huari and Tiwanaku at Cerro Baúl, Iwawi and Tiwanaku, en: H. I.
Silverman y W. H. Isbell (eds.), Andean Archaeology. Vol. I, Variations in Sociopolitical Organization, 199-242,
Kluwer Academic/Plenum Publishers, New York/Boston/Dordrecht/London/Moscow.

ISSN 1029-2004
Sighting the apu: a GIS analysis of Wari
imperialism and the worship of
mountain peaks

Patrick Ryan Williams and Donna J. Nash

Abstract

In the Andes, prominent mountains are revered as earthly spirits that protect, but may also punish,
their human constituents. These apu were often linked to distant ancestors and are considered the
most important local deities. During the phase of the earliest highland Andean expansive states (AD
600–1000), the Wari and Tiwanaku utilized mountain worship as a means of establishing hegemony
over local peoples who considered these mountains as places of ancestral origins. By usurping the
apu, or including them in the pantheon of imperial deities, the expansive state effectively held these
sacred places hostage and incorporated local belief systems into an imperial ideology. Recent
research has yielded new clues to the worship of mountain peaks, including the usurpation of a
unique geological mesa formation at Cerro Baúl as the basis for the Wari colonization of its southern
frontier. Furthermore, research on the mountain summit has revealed architectural complexes
oriented to, and presumably dedicated for, rites of veneration to the higher snowcapped volcanic
peaks visible from this mountain summit.

Keywords

Cognitive landscapes; architecture; state ideology; GIS viewshed analysis; Peru; Tiwanaku.

Introduction

In the ancient world, early expansive states utilized various means to exert their influence
over far-flung territories. Militarism and economic coercion were certainly important forces
in power strategies. Religion has also played a pivotal role in the expansion of early states. In
the Andes, a landscape of topographic diversity, mountain peaks were imbued with great
sacred power. Many of these peaks continue to be revered today. Usurping the power of
these sacred peaks was a prime concern of early Andean states, and the integration of
mountain deities into an imperial pantheon was a principal goal of early state administrators.

World Archaeology Vol. 38(3): 455–468 Archaeology at Altitude

ª 2006 Taylor & Francis ISSN 0043-8243 print/1470-1375 online
DOI: 10.1080/00438240600813491
456 Patrick Ryan Williams and Donna J. Nash

The early Wari state (AD 600–1000) was one of the most successful expansive polities in
South America. At its height, it held sway over 1300 kilometres of the Andean sierra from
Viracochapampa to Moquegua (Fig. 1). Its southern frontier intersected with its rival, the
Tiwanaku, in the Moquegua Valley. Here, the Wari established a grand bastion atop the
massif of Cerro Baúl (2590 masl). They built opulent palaces and temples on the flat

Figure 1 Viewshed map from Cerro Baúl. Grey shading indicates the area visible from the eastern
summit of the mesa. Triangles are peaks over 5000m calculated as visible from the summit. Bold lines
indicate the dominant features visible from Cerro Baúl.
 Sighting the apu 457

summit of the mesa that we have argued served as a sort of embassy with the cultures of
the southern realm (Moseley et al. 2005; Williams 2001).
A mountaintop, away from all natural sources of water and food production, seems an
improbable choice as the location of an imperial centre. Defensive concerns were certainly
met by such a location. The massive cliff faces could repel any frontal attack with just a
few defenders. Religious concerns are likely to have been as important, however. Locating
an imperial centre on the massif may have been a statement of religious power itself.
Perhaps as important is the link between the mesa summit and the distant snowcapped
peaks visible only from its summit. This visual link is reified in architectural complexes
inextricably oriented towards these distant apu, or sacred mountains.
In this paper we review the importance of mountains as ancestral spirits in Andean
societies. We then examine Wari expansion in relation to the incorporation of these
mountains into administrative ritual. Finally, we turn to the southern frontier of the Wari
realm and investigate the role that specific mountain deities may have played in the Wari
strategy of incorporation based on both modern ethnographic data of mountain worship
and GIS-based viewshed analyses of apu sighting. We also examine the construction and
orientation of ancient ritual architecture as evidence for a cult of the apu during Wari
times at Cerro Baúl.

Mountain worship in Andean societies

Snow-capped peaks are revered and venerated in many modern highland Andean
communities. Apu (Bolin 1998; Gose 1986), wamani (Isbell 1978) or jach’arana (Rasnake
1986, 1988), as they are variously called, serve as important members of indigenous
communities. At times they are considered parental figures or ancestors (Anders 1986;
Favre 1967) and by others as guardians or benefactors (Rasnake 1986). Regardless of the
relations described, the supernatural forces residing in these alpine environments are
focuses of ritual performance, whether the rite is conducted on the snow’s edge as a supra-
regional pilgrimage (Allen 1988), or the spirit is called to attend hearthside family services
(Bolin 1998), or ritual payments are made in specialized communal locations, where
surrogate altars or stone boxes pertain to specific familial groups and receive offerings sent
via burning or burial (Isbell 1978).
Even if these dramatic features cannot be seen from every location of a settlement and
its associated chacras (fields) and pastures, the members of different locales are ever aware
of the spirits’ oversight and the potential for good fortune or bad luck (Bolin 1998; Isbell
1978). These landscape animates are called by name as people traverse narrow passes or
engage in productive activities. These entreaties may be associated with symbolic acts such
as blowing on coca leaves or entail a series of ritualized performances accompanied by a
prescribed payment, or pagapu (Bolin 1998; Isbell 1978).
In its various forms, veneration of supernatural mountain personalities is deeply
embedded in the structural framework of Andean worldviews and plays an important role
in defining nested groups of kinship, community affiliation and larger regional identities
(Rasnake 1986; Eade and Sallnow 1991). Apu are the owners of plants and animals (Isbell
1978) and must be fed to maintain the fertility of crops, herds and people (Bolin 1998).
458 Patrick Ryan Williams and Donna J. Nash

Today, apu are appeased by annual meals in February and August, but their dates,
occasions and composition vary between highland groups, and may also differ between
different spirits that have complementary responsibilities and corresponding preferences.
Within a region, mountain peaks are ranked and form a hierarchy. The spirits hold
positions within the mytho-political order based on the relative height of their lofty
abodes. Highland lakes and mountain peaks communicate with one another between
regions. The higher-ranked entities give orders to their subordinates for the management
of resources throughout the natural landscape (Earls 1969; Gose 1986; Isbell 1978). These
features can be masculine or feminine in association and can be grouped as mates or have
sibling relations (Sallnow 1991).
The different apu have herds of their own that may parallel the holdings of their peasant
devotees. They may also hold great wealth in gold, silver and copper. In fact, Rasnake
(1988: 235) reports that the Yura people of Bolivia believe that the mountain spirits of the
region replenish the mines of Potosi by nightly sending caravans of vicuñas loaded with
silver. In relation to agricultural activities, the mountain spirits are the important
providers of water. Offerings often include chicha and or cane alcohol. Drinking in turn,
the participants of annual payments may start with Pachamama (earth mother) then toast
each regional apu according to their rank. In many instances, ethnographers describe these
libations being offered from seashells, objects that are related to the sea and the request for
water (Murra 1975). Humans thus replenish the liquid of the mountain spirit in an act of
reciprocity for the agricultural fertility that results from the mountain’s liquid (Gose 1986;
Harris 1982).
Mountain spirits can bestow gifts upon their adherents if proper restitution and
offerings are made. Families that fail to feed the hungry mountain spirits may lose
livestock or even family members. If offerings are wanting or ritual protocols are not
performed correctly, the apu may express its displeasure in a number of ways (Bolin 1998).
These spirits are considered ill-tempered and are easily vexed. Apu are often associated
with condors, a highland predatory bird that has been known to carry off younger animals
and feed on their hearts. Displeased apu may seek their revenge by sending such predators,
withholding water or sending stormy conditions. To a degree, these entities are venerated
out of fear.

Apu in the Prehispanic Andes

In the ancient Andes many of these patterns are apparent in the archaeological record. The
veneration of mountain peaks may date back as far as the Initial Period (2000–900 BC).
According to Carlos Williams, it seems that early coastal U-shaped centres turned their
backs on the sea and faced up river towards the important source of irrigation water
(Moseley 2001; Williams 1985). During later periods (c. AD 500), Moche ceramic vessels
model mountain peaks as scenes of human sacrifice in which the long hair from the
victim’s head flows downward in front of the body descending the peak’s slope (see Aimi
2003: 218–19, figures 187, 188). Alan Kolata (1993) has suggested that, at the site of
Tiwanaku, located near Lake Titicaca, the Akapana, one of the site’s principal pyramidal
monuments, was modelled as a sacred peak. Excavations conducted in the 1980s
 Sighting the apu 459

demonstrated that canals allowed for the ritual manipulation of water at the Akapana and
this activity was the primary focus of ritual performance during the early phase of use.
In Inka times (c. AD 1500), mountains were offered child sacrifices. Many of these
elaborate burials have recently been recovered from mountaintop sanctuary complexes in
southern Peru, Chile, and Argentina. Some of these interments and the associated ritual
performances, called capac hucha, may have been conducted to ask the mountain for
needed irrigation water or may have been associated with claiming legitimate rule over a
region and its resources (Anders 1986; Zuidema 1978, 1982). In many cases, the
individuals are well preserved, dressed in rich textiles and accompanied by a suite of
uniform grave goods including gold and silver figurines, objects of Spondylus and finely
crafted ceramic vessels, all of Inka Imperial style (McEwan and Van de Guchte 1992).
These Inka-sponsored offerings were a way for the empire to claim control of the
mountain’s contingent resources. Thus these rituals were a strategic way to claim resources
in conquered areas and to legitimize their control of water and landscape modification – as
intermediaries with these important chthonic entities (Zuidema 1982).

Wari expansion and mountain gods

Having reached pinnacle elaboration during Inka times, mountain veneration goes back to
the Early Formative Period (c. 1600–800 BC) in the Cuzco region (Zapata 1998) and was
also a key aspect of Wari state ideology during its occupation and control of water in this
important province of the empire (Glowacki and Malpass 2003). Furthermore, evidence
shows that mountain veneration may have been a legitimizing strategy throughout the
Wari Empire rather than just a characteristic of indigenous Cuzco cosmology.
Anders (1986), working at Azangaro, located in the core of the Wari Empire 15km from
the polity’s capital in Ayacucho, proposed that the site was specifically linked to mountain
veneration. She also suggested that other Wari sites and respective architectural complexes
may have been linked to propitiating mountain spirits. Most notably, she suggests that
the ‘wells’ found full of ritual offerings at Cerro Amaru may be akin to the stone boxes
used by modern populations in Chusci to make ritual payments to mountain deities (Isbell
1978). Significantly, the offerings at Cerro Amaru, located near Viracochapampa on the
northern periphery of the Wari Empire, consisted prominently of Spondylus priceps
(McCown 1945), a valued bivalve originating from warm Ecuadorian waters and an object
specifically linked to ritual requests for water during Inka times (Cobo 1956 [1653];
Glowacki and Malpass 2003).
Glowacki and Malpass (2003) describe the importance of landscape features to Wari
expansion and the placement of Wari sites adjacent to water features as well as
snowcapped peaks. They note that the Wari centre in the Sondondo Valley, Jincamocco, is
placed near the nevado Señal Carhuarazo; two Wari sites in the Huaro Valley are located
on the apu, Cerro Wiracochan; and they include in their listing Cerro Baúl, an apu in its
own right located in the Upper Moquegua Valley.
Anders’ (1986) research at Azangaro suggests links between snow-capped peaks in the
region and the boundaries of the Wari heartland. Anders lists the prominent wamani in the
area and associates four nevados with the quadripartition of the zone in the Middle
460 Patrick Ryan Williams and Donna J. Nash

Horizon and the Late Intermediate Period. Guaman Poma (1980 [1615]), a Spanish
chronicler particularly familiar with the Ayacucho region and its indigenous pre-Inka
kingdoms, described administrative units of the former autochthonous Chanka
Confederation polity as wamani. The size of these population segments is uncertain, but
it is significant that mountain spirits in the region are still called by this same name,
suggesting that territories were conceptually linked to the provinces of a mytho-political
montane segmentation.

Apu on Wari’s southern frontier

Don Juan Lopez Ventura has been performing the pago, or payment to the apu, for several
decades (Plate 1). He has also graciously performed the opening and closing pagos during
the 2002 and 2004 excavation seasons at Cerro Baúl. He learned his trade as a youth
through the teachings of his father, also a curandero. He apprentices two maestros, men
who know the art of pagos who work with him.
He was appointed President of the Comité de la Santı́sima Cruz de Cerro Baúl and
travels extensively in the South-Central Andes practicing his craft. A true shaman, he says,
is marked by a physical deformity. In his case it is his clouded and blind right eye. Don
Juan indicates that Cerro Baúl is an important apu in the chain of ancestral places in the
circum-Titicaca Basin. In interview with us, he employed a sketch to illustrate how the
cross on the mountain summit of Cerro Baúl is linked to other chains of apu. Baúl is
important not only for its unique form, but also because it visibly links the valley with the
snowcapped peaks that are also apu.

Plate 1 Don Juan Lopez offers the pago bundle to the apu on the summit of Cerro Baúl while the
senior author kneels beside him.
 Sighting the apu 461

To the north is Tata Picchu Picchu (Fig. 1), the most powerful of the apu visible from
Cerro Baúl. Picchu Picchu is of male gender, Don Juan says, and forms the first in a chain
of over a dozen apu peaks that stretches north through Arequipa to Cuzco (including
Machu Picchu, the famous Inka site) and on to Juliaca on the shores of Lake Titicaca.
To the east is Coline (or Arundane), which forms part of a chain of over a dozen apu
that stretch to the north east, ending at San Ignacio, Ichuño, and Sanchez Cerro. Don
Juan’s sketch also includes other apu strings that do not link with Cerro Baúl, but which
run from San Pedro, Bolivia, through the department of Puno, Peru, and end at La Paz
and Copacabana. Offerings are made to the apu to ensure success in business, marriage,
agriculture, etc. These offerings, or pagos, have many different forms, but almost always
include coca leaves, llama fat, sugar, alcohol and often cigarettes. Sometimes they include
the sacrifice of an animal.
The pago is a complex rite, but one aspect involves blowing on the coca leaves to bring
the prayers to the apu. As Don Juan blows the leaves he mentions each apu by name,
starting with those closest in the string and moving down. Pagos representing different
requests may invoke different apu, and a complex knowledge base surrounds each offering.
Don Juan indicates that charlatans abound in these rites, and that requests will not be
honoured if the apu are not supplicated properly.
While offerings can be made to apu even without being able to see them, visibility plays a
role in maintaining this complex of knowledge of the supernatural. The strings of apu that
tie this system together reference each other through inter-visibility and the strings are the
subject of ritual pilgrimages by shamans like Don Juan. For him, Moquegua’s apu
reference an extensive ritual landscape that covers the entire circum-Titicaca Basin,
hundreds of kilometres in diameter.
Each of the apu visible from Cerro Baúl forms part of a string of apu that link large
geographic areas together. They are integrative mechanisms that join distant regions to
one another. For many pilgrims, the peaks visible from Cerro Baúl are the apu of primary
reference. For Juan Lopez, they are part of a wider universe. These knowledge systems
connect ritual specialists who share this understanding of the regional apu. The apu thus
serve to join distant areas together; local apu may provide visual cues for specialists to a
wider network. Those same visual cues can help concretize distant connections to local
people and thus incorporate their cosmos into a larger system. Apu sighting is a way in
which unseen associations can be referenced in a visual landscape. To test this concept of
visualization further, we turn to a visual analysis of landscape and apu in the region
around Cerro Baúl.

Viewshed analyses of the southern Apu

The principal apu of the Moquegua Valley are generally not visible from the confines of
the verdant valleys. It is, however, nearly impossible for an observer to be at every point
on the landscape to observe visibility. Here, we turn to a visualization tool inherent in
topographic analysis – the viewshed. Utilizing the capabilities of geographic information
systems (GIS) technology, viewsheds map out the area within visual detection from a point
across a topographic surface. A viewshed calculates the area visible from a point on a
462 Patrick Ryan Williams and Donna J. Nash

landscape, much like a watershed calculates the topographic area that contributes to the
flow of water to a specific point.
In order to assess the inter-visibility of snow-capped mountain peaks with the valley
settlements and Cerro Baúl, visibility maps were generated for the principal mountain
peaks above the snow line at c. 5000m and Cerro Baúl itself. The data source is an SRTM
90m resolution digital terrain model of the southern Peruvian landscape created by the
Shuttle Radar Topography Missions (Fig. 2). Visibility analysis was restricted to an area
within 100km of the observer viewpoint, as it marks the limit of human sight on a clear
day in this environment. The results indicate that the promontory of Cerro Baúl has
greater inter-visibility with mountain peaks above 5000m in elevation than any other point
in the valley. It also illustrates several chains of hills that block the views of distant peaks
from almost all points except Cerro Baúl (Fig. 1).
In practice, some peaks are more clearly visible from Cerro Baúl than others identified
in the visibility analysis. Local topographic variations not entirely accounted for in the
relatively coarse 90m elevation data can affect visibility. For example, the Picchu Picchu
peaks, 30 degrees west of north from Cerro Baúl, sit in a local topographic dip relative to
the Baúl massif (Plate 2). This foreground dip makes these peaks much more visible than
their other northern counterparts. Likewise, the valley incised by the Torata River and the
Quebrada Cocotea to the north east make the peaks of Arundane and Hauilau much more
prominent than their neighbours (Plate 3).

Figure 2 Digital elevation model derived from the SRTM 90m resolution data set of the volcanic
chain of southern Peru. Vertical exaggeration is ten times horizontal resolution with view from south
to north.
 Sighting the apu 463

Plate 2 Off-centre photograph of the platform in veneration of the Picchu Picchu peak, indicated by
an arrow in the background.

Plate 3 Oblique aerial photograph from over the sacred rock showing the snowcapped peaks of
Hauilau and Arundane (indicated by arrows) visible on the eastern horizon. Architectural remains of
the Wari settlement are in the foreground.
464 Patrick Ryan Williams and Donna J. Nash

Thus, while over half a dozen peaks might be recognizable from Cerro Baúl, there are
but three peaks, Picchu Picchu, Hauilau and Arundane, which are predominant on the
landscape. Not surprisingly, two of these peaks (Picchu Picchu and Arundane/Coline) are
also the principal regional apu acknowledged by modern religious pilgrims and shamans
like Don Juan in addition to Cerro Baúl itself. As we now illustrate, it is most likely that
these were also the two peaks in addition to Cerro Baúl that the Wari also recognized as
the regional apu.

Ritual architecture and apu sighting at Cerro Baúl

Architectural orientations at Cerro Baúl are somewhat constrained by the topography of

the hill. There are two complexes, however, that have clear alignments to the most visible
mountain peaks of Picchu Picchu and Arundane. Both of these complexes are set apart
from the rest of the architectural core (Fig. 3), and have been denominated Sectors D and
E respectively. They both also utilize staircases, sunken plazas and elevated platforms to
create three dimensional synergies with the landscape beyond.

Figure 3 Map of the architecture on the summit of Cerro Baúl dating to the Wari occupation.
Orientation of the Sector E platform ritual pathway towards Picchu Picchu and the Sector D
complex encompassing the sacred rock towards Arundane are illustrated with arrows.
 Sighting the apu 465

Sector D is a complex constructed just west of the main Wari site around the largest
andesite boulder protruding out of the sandstone matrix of the mesa. The top of the
boulder sits 2–3m above the ground surface and is the highest natural point on the mesa
summit. It is a unique natural phenomenon and may have been viewed as a sacred rock
due to its central location, prominence and distinctive form. Around the great boulder, the
Wari constructed a 30 6 50m architectural complex composed of a western viewing
platform with a staircase leading down into a plaza that surrounded the great rock. The
walls of the plaza complex and the alignment of the viewing platform, staircase and the
summit of the boulder diverge from the more north-easterly orientation of the adjacent
ruins. They align perfectly with the central Arundane peak, however, and are the only
architecture at the site to conform to this orientation.
At the back of the viewing platform are two small rooms that frame the alignment of the
staircase with the boulder and the peak of Arundane on the distant horizon. The boulder
may have served as a surrogate for the mountain peak and for offerings made to it.
Unfortunately, looters have destroyed the area directly adjacent to the boulder and no
direct evidence of what these offerings may have been has yet been recovered.
Nevertheless, future research in the platform complex may yield new clues to the
relationship between the Wari and the Arundane apu. Other evidence for apu worship has
been recovered from the far west side of the mountain.
Sector E, denominated the Picchu Picchu Platform complex, comprises a 16620m
platform fronted by a sunken court (Plate 2), which is flanked by a terraced hillside. The
complex is oriented 30 degrees west of north and has two aligned staircases. One staircase
descends the terraced hillside. It channels its traffic through the sunken court toward the
platform, where a second staircase ascends the dais (Fig. 3 inset). Looking from the top of
the terraced hill across both staircases, one cannot deny the clear alignment of the
staircases with the peak of Picchu Picchu in the background.
Excavations within the platform yielded in situ burnt offerings dating to the
construction of the platform. A radiocarbon date processed on one of these platform
construction offerings yielded a calibrated age of 1366 + 35 BP (AA-46599; cal. AD 600–
770). The radiocarbon data clearly place the platform construction within the realm of the
first Wari occupations at the site (Nash and Williams 2005).
In fact, this complex is the only place on the site where Picchu Picchu falls perfectly
within the topographic dip in the foreground hills. A procession down the terraced hillside
staircase, across the sunken court and up the platform staircase would unmistakably move
directly towards the Picchu Picchu peaks. Furthermore, the cultural space created by the
complex, moving from a high terraced hillside into a sunken court and up the steps of an
elevated platform, replicates the natural topography beyond. The high point of Cerro Baúl
falls to the valley below and ascends through the dip in the distant hills to the glorious
peaks of Picchu Picchu in the distance.
Ritual progression through this space may have mimicked the movement of goods and
people from Moquegua to the peaks of Picchu Picchu and beyond to the Wari heartland.
The complex in Sector E was a ritual conduit symbolizing the connection of Baúl and the
Moquegua province to the peoples of the north. It linked the symbolic centre of
Moquegua with the symbolic points on the road to Ayacucho and thus tied the Moquegua
apu to a larger system of ideological significance.
466 Patrick Ryan Williams and Donna J. Nash

Discussion: the cult of the apu

The cults surrounding the apu of Arundane and Picchu Picchu thrived at Cerro Baúl during
the period of the expansive highland states, c. AD 500–1100. Baúl served as a nexus for the
integration of local ideologies of ancestry and supernatural power with regional and imperial
ideologies and as an apu itself. Arundane may have been viewed as the ancestral origin of the
local groups, as well as the apu most closely related to the altiplano and the Tiwanaku realm.
Picchu Picchu, on the road to Wari itself, was the apu associated with Wari regional
identity and, through its links to mountains closer to Wari, with the imperial identity.
Each apu had its own ritual complex, spatially segregated from the others and from the
rest of the architecture at the site. Each may have had its own ritual attendants dedicated
to that particular shrine. Future excavations in adjacent architectural complexes that
appear more residential may shed new light on this hypothesis. The clear orientation of
architectural features with the peaks on the horizon indicates the dedication of these
complexes to their respective apu. The use of three-dimensional architectural space to
highlight features within each complex and call out the features of the landscape beyond
made these linkages all the more impressive.
The settlement of Cerro Baúl itself placed those who resided on its summit in the same
league as the revered mountain deities. It drew resources from the surrounding populations
based not only on economic and political persuasion, but as a pivotal point in the religious
belief systems of the people themselves. Thus, political authority was reinforced by placing it
in a system of values that merged with local perspectives on sacredness and ancestry.

Conclusions

Mountains are important places in the Andean cognitive landscape. They serve as
protectors and punishers to local populations who view them as powerful deities. These
same peaks link people from different communities together as supplicants of the same
maximal social groups. As such, they are prime constituents for usurpation by state
ideologies of incorporation and control.
GIS-based viewshed analysis can provide a means of assessing which points on the
landscape are likely to be recognized by a large number of communities. In association with
ethnohistoric and ethnographic data, it can be a powerful tool for investigating the cognitive
landscapes of the past. The analysis presented here illustrates the importance of combining
ethnographic and archaeological data with modelled parameters. One of the most visible
peaks, Hauilau, did not form part a shaman’s account of apu, or of the architectural align-
ments visible from an ancient Wari site. Yet, in two other cases, architectural alignments with
prominent peaks highlighted by viewshed analysis are well documented in Wari structural
design. These same peaks are highlighted as important in the sacred landscape today.
Religious motivations can be a primary factor in extension of imperial control over a
region. As for the Inka, the inclusion of local religious traditions in an imperial ideology
was key to early state expansion in the Andes. On the southern Wari frontier, local deities
were incorporated into the state pantheon by establishing a state centred on the local apu,
by constructing architectural complexes oriented towards the distant snowcapped peaks
 Sighting the apu 467

and by performing rituals of veneration within these complexes that tied people and their
sacred places to the Wari elites.

Acknowledgements

Research on the architectural complexes discussed here was funded by the National
Science Foundation (BCS-0074410, BCS-0226791). Permits for excavations were
authorized by Peru’s National Institute of Culture via Resolución Directoral Nacional
numbers 613/INC (2001) and 646/INC (2002). Kenneth Sims supervised the excavations in
Sector E. Juan Lopez was kind enough to share his knowledge of the apu with us.

Patrick Ryan Williams,

The Field Museum, Chicago
Donna J. Nash,
University of Illinois-Chicago

References

Aimi, A. 2003. Perú: Tremila Anni di Capolavori. Florence: Firenze Mostre S.p.A.
Allen, C. J. 1988. The Hold Life Has: Coca and Cultural Identity in an Andean Community.
Washington, DC: Smithsonian Institution Press.
Anders, M. 1986. Dual organization and calendars inferred from the planned site of Azangaro: Wari
administrative strategies. PhD dissertation, Cornell University.
Bolin, I. 1998. Rituals of Respect. Austin, TX: University of Texas Press.
Cobo, B. 1956 [1653]. Historia del Nuevo Mundo. Biblioteca de autores Españoles, Tomos 91–2.
Madrid: Atlas.
Eade, J. and Sallnow, M. 1991. Contesting the Sacred: The Anthropology of Christian Pilgrimage.
New York: Routledge.
Earls, J. 1969. The organization of power in Quechua mythology. Steward Journal of Anthropology,
1: 63–82.
Favre, H. 1967. Tayta Wamani: la culte des montagnes dans la centre sur des Andes péruviennes.
In Colloque d’Etudes Pe´ruviennes. Publication des Annales de la Faculte´ des Lettres Aix-en-Provence
N.S. 61: 121–40.
Glowacki, M. and Malpass, M. 2003. Water, huacas, and ancestor worship: traces of a sacred Wari
landscape. Latin American Antiquity, 14(4): 431–48.
Gose, P. 1986. Sacrifice and the commodity form in the Andes. Man, 21(2): 296–310.
Guaman Poma de Ayala, F. 1980 [1615]. El primer Nueva corónica y buen gobierno [1615]
(eds J. V. Murra and R. Adorno, Quechua trans. J. L. Urioste), 3 vols. Mexico City: Siglo Veintiuno.
Harris, O. 1982. The dead and the devils among the Bolivian Laymi. In Death and the Regeneration
of Life (eds M. Block and J. Parry). Cambridge: Cambridge University Press.
Isbell, B. J. 1978. To Defend Ourselves: Ecology and Ritual in an Andean Village. Institute of Latin
American Studies. Austin, TX: University of Texas Press.
Kolata, A. 1993. The Tiwanaku: Portrait of an Andean Civilization. Oxford: Blackwell.
468 Patrick Ryan Williams and Donna J. Nash

McCown, T. D. 1945. Pre-incaic Huamachuco: survey and excavations in the region of Huamachuco
and Cajabamba. University of California Publication in American Archaeology and Ethnology, 39:
223–399.
McEwan, C. and Van de Guchte, M. 1992. Ancestral time and sacred space in Inca state ritual.
In The Ancient Americas: Art from Sacred Landscapes (ed. R. F. Townsend). Chicago, IL: Art
Institute of Chicago, pp. 359–71.
Moseley, M. 2001. The Inca and Their Ancestors. London: Thames & Hudson.
Moseley, M., Nash, D., Williams, P., DeFrance, S., Miranda, A. and Ruales, M. 2005. Burning
down the brewery: excavation and evacuation of an ancient imperial colony at Cerro Baúl, Perú.
Proceedings of the National Academy of Sciences, 102(48): 17264–71.
Murra, J. 1975. El Tráfico de mullu en la costa del Pacifı́co. In Formaciones Económicas y Polı´ticas
del Mundo Andino. Lima: Instituto de Estudios Peruanos, pp. 255–67.
Nash, D. and Williams, P. R. 2005. Architecture and power on the Wari–Tiwanaku frontier.
Archaeological Papers of the American Anthropological Association, 14.
Rasnake, R. 1986. Carnival in Yura: ritual reflections on ‘ayllu’ and state relations. American
Ethnologist, 13(4): 662–80.
Rasnake, R. 1988. Domination and Cultural Resistance: Authority and Power among an Andean
People. Durham, NC: Duke University Press.
Sallnow, M. 1991. Pilgrimage and cultural fracture in the Andes. In Contesting the Sacred: The
Anthropology of Christian Pilgrimage (eds J. Eade and M. Sallnow). New York: Routledge, pp. 137–153.
Williams, C. 1985. A scheme for the early monumental architecture of the central coast of Peru.
In Early Ceremonial Architecture in the Andes (ed. C. Donnan). Washington, DC: Dumbarton Oaks.
Williams, P. R. 2001. Cerro Baúl: a Wari centre on the Tiwanaku frontier. Latin American Antiquity,
12(1): 67–83.
Zapata, J. 1998. Los cerros sagrados: panorama del periodo formativo en la Cuenca del Vilcanota,
Cuzco. Boletin de Arqueologı´a PUCP, 2: 307–36.
Zuidema R. T. 1978. Shaft tombs and the Inca Empire. Journal of the Steward Anthropological
Society, 9: 133–78.
Zuidema, R. T. 1982. Bureaucracy and systematic knowledge in Andean civilization. In The Inca and
Aztec States 1400–1800: Anthropology and History (eds G. A. Collier, R. I. Rosaldo and
J. D. Wirth). New York: Academic Press, pp. 419–58.

Patrick Ryan Williams is Curator of Archaeological Science at the Field Museum in

Chicago. Educated at Northwestern and Florida, he has served on the faculty at Boston
University, and currently holds adjunct appointments at the Universities of Illinois-
Chicago, Florida and California Los Angeles, as well as at Northwestern University. He
has directed archaeological research in Peru since 1995, and is head of the ongoing
research programme at Cerro Baúl.

Donna J. Nash is Adjunct Professor of Anthropology at the University of Illinois at

Chicago. She has worked at the Florida Museum of Natural History and the Field
Museum in Chicago and continues to hold the post of Adjunct Curator at the latter. She
has conducted archaeological research in Peru since 1993 and is currently field director of
the research programme at Cerro Baúl.
 6
Religious Ritual and Wari State Expansion

Patrick Ryan Williams and Donna J. Nash

The complex relationship between religion and ritual has had a long history
of study in archaeology and the social sciences (Bell 1992; Insoll 2004; Ren-
frew and Zubrow 1994). We recognize that ritual is not an exclusive domain
of religion and that it is practiced in many different contexts. We also ac-
knowledge that domains of religion, politics, economics, and society inter-
sect in ritual practice and that a focus on religious ritual is only one aspect
of how ritual is used in society. However, we also maintain that actions and

proof
practices that reveal how people communicate with the supernatural forces
they perceive around them is fundamental to understanding the relation-
ships of power in society. We specifically examine this theme in relation to
understanding Wari state expansion, especially on its southern frontier. In
this chapter we focus on how ritual activities were used to create social ties
between representatives of expansive states and both subject populations
and competing peers. Specifically, we examine how ritual actions were used
within specific contexts to incorporate distinct populations into the social
framework of the Wari state on its southern frontier.
The focus on religion as a top-down force that subjugates the masses and
allows the elite to oppress the poor, as Karl Marx’s (1970 [1843]) approach
might advocate, oversimplifies the issue. Removing the concept of power
and inequality from the equation, as Clifford Geertz’s definition of religion
implies, fails to understand how religion and the state transform each other
(Asad 1983:237). As a refresher, Geertz (1973:90) argues that
religion is: (1) a system of symbols which acts to (2) establish power-
ful, pervasive, and long-lasting moods and motivations in [wo/]men
by (3) formulating conceptions of a general order of existence and (4)
clothing these conceptions with such an aura of factuality that (5) the
moods and motivations seem uniquely realistic.
132 · Patrick Ryan Williams and Donna J. Nash

It is, then, the dialectic between these two perspectives on religion: its
framework as a source of power, and its power as explanatory mechanism
to an individual’s lived reality that we seek to understand. It is the transfor-
mative power of religious ritual, both as a force of tradition and change, that
we focus on here. As Susan Kus (2012:12) notes, it is how Geertz’s “moods
and motivations” are “expressed and restructured in meaningful material
practice” that bridges this gap.
It is useful, in this context, to discuss what we mean by the expression
of power in a ritual context. We adopt an understanding of different ex-
pressions of power as advocated by Richard Blanton and colleagues (1996).
Corporate power refers to power expressed through structure, strength-
ened by enhancing the focus not on individual leaders but on different
groups and collectives, often represented by formal institutions. Exclusion-
ary power derives from networks of personal interactions and focuses on
the power of the individual leader to enact his or her will. More recently
elaborated in terms of autocracy and collective action, these sources of au-
thoritative power may structure representation of action (see Blanton, this
volume, and Feinman, this volume). Religious ritual can be a venue for the

proof
expression of either of these types of power. A priestly individual may use
ritual to enhance his or her own standing in a network of power, reflected
iconographically through his or her position as a mediator of supernatural
authority. Likewise, ritual can also emphasize an ideology or set of ideolo-
gies that promote a “cognitive code” (Blanton et al. 1996:2) that structures
the distribution of power among different corporate groups. Exclusionary
and corporate types are two axes of expression of power that can be forces
of maintaining convention or of spurring innovation.
Religious ritual can thus act as both a conservative force, enhancing the
role of traditional structures in society, and as a transformative influence,
adapting to changing circumstances to link the individual to new social
realities. This dual perspective on ritual both as a means of continuity and
conformity and as an agent of change is actually compatible with early
functionalist theories on the role of religion in society. Émile Durkheim
(1995 [1912]:257–58) asserts, for example,

That is why we can rest assured in advance that the practices of the
cult, whatever they may be, are something more than movements
without importance and gestures without efficacy. By the mere fact
that their apparent function is to strengthen the bonds attaching the
 Religious Ritual and Wari State Expansion · 133

believer to his god, they at the same time really strengthen the bonds
attaching the individual to the society of which he is a member, since
the god is only a figurative expression of the society.

Thus, as new sociopolitical actualities emerge, new ways of explaining the

relationship of the individual to society also come into play. Power dynam-
ics and inequalities are shifted as new political regimes take hold. Just as
Mark Aldenderfer (2010) argues that ritual was a key element in the devel-
opment of inequality in foraging societies, we argue that ritual plays a pre-
dominant role in the transformation of power structures in non-egalitarian
groups, that is, states. We argue this appears to be especially pronounced on
boundaries between social groups where states are expanding and negoti-
ating with local and regional power structures outside their areas of core
identity (see also Marcone, this volume).
We embrace a definition of ritual as the patterned ways of action and
belief that reinforce a social concept or construct. Religion is a set of beliefs
that structure the interaction between a community and the supernatural
forces it perceives in the world it inhabits. Thus, religious ritual defines

proof
those patterned actions that reinforce the relationship between a commu-
nity and those forces in its inhabited world it accepts on faith. We adopt a
practice-oriented approach to ritual and “focus on the ways that the experi-
ence of ritual and ritual symbolism promotes social orders and dominant
ideologies” (Fogelin 2007). Specifically, we are interested in understanding
how rituals construct and recontextualize social relationships. We investi-
gate in this instance less what ritual means than what it does for society.
Of course, the use of rituals in mediating social relations must be un-
derscored by belief systems and meaning. We draw on past ethnohistoric
and archaeological work that documents the importance of mountain apu
(deities) in Andean cultures as ancestral beings with power over life and
death, and specifically the role they played in the Middle Horizon states
of Wari and Tiwanaku (Williams and Nash 2006). The rituals of payment
to these apu and their veneration in special places on the landscape were
something that was common to both Wari and many of the peoples with
whom they interacted. The apu belief system thus formed a basis of com-
mon understanding of the supernatural world that provided a means for
these rituals to redefine social orders and hierarchies.
134 · Patrick Ryan Williams and Donna J. Nash

Wari Expansion and Rituals of Incorporation

The Wari expanded out of their capital in Ayacucho, Peru around 600 C.E.
and came to influence a 1,300-km stretch of the Andes Mountains (Isbell
1991; Schreiber 1992) (Figure 6.1). This period of expansion is known as the
Middle Horizon in Andean chronologies and refers to the period between
600 and 1000 C.E. when large-scale polities like Wari and their southern
Andean rivals the Tiwanaku held sway. While the Wari are often character-
ized as a secular and military empire, religious ideology played a prominent
role in their expansion. They worshipped ancestral deities and made offer-
ings to both ancestors and the supernatural in various ways. Some of these
rites took place in D-shaped temples, where rituals of ancestor worship
and decapitation propitiated their gods (Cook 2001). Their southernmost
outpost was located on the mountaintop of Cerro Baúl in southern Peru,
which contains two of the sacrosanct D-shaped temples (Williams 2001).
But Wari also used other rituals of incorporation in the different re-
gions they occupied. Both in the heartland and in several of the southern
provinces, one of the principal Wari rites of sacrifice was the intentional

proof
destruction of oversized ceramic vessels sculpted and painted in represen-
tation of elite human figures, known as face neck jars. They also smashed
other pottery finewares, including large oversized urns decorated with de-
signs of agricultural fecundity and the figures of earth fertility deities. These
pottery smashes have been recovered from localities both within the heart-
land (Cook 1984, 1985; Isbell and Cook 2002) as well as from provincial
areas primarily in the south coast (Anders 1990; Menzel 1968; Valdez 2011).
These ritual ceramic offerings may have served as a means to promote
a supernatural order in which Wari elites mediated the relationship be-
tween earthly bound humans and ancestral and animistic landscape dei-
ties. Mountain and creation deities, especially, had the power to visit abun-
dance or scarcity of irrigation waters on the desert coast of Peru. The ritual
vessels were forms for the storage and consumption of chicha, an alcoholic
drink often made from maize, whose fluid flows were closely linked to fer-
tility and the life-giving flow of water. The metaphor of chicha flowing from
vessels decorated with agricultural produce and ancestral deities likely re-
ferred to the flow of water from the mountains controlled by the same su-
pernatural forces depicted on the vessels.
Other scholars have argued for the direct manipulation of ancestral
remains in Wari provincial architectural complexes as a means of ensur-
ing loyalty and incorporating subordinate groups into Wari society. Gor-
 Religious Ritual and Wari State Expansion · 135

proof

Figure 6.1. Map of Peru, illustrating places discussed in the text.

don McEwan (1998) argues that a class of Wari architecture found at both
Pikillacta in Cusco and Viracochapampa in Huamachuco housed ceremo-
nies of ancestor worship designed to integrate conquered populations into
Wari social and political organization. The niched hall, a large (up to 15 × 42
m) rectangular structure with rounded corners and interior niches, is the
proposed locale of feasting ceremonies for both Wari and subject popula-
tions’ lineages (McEwan 1998). Presumably, the ancestral remains would
have been kept in these locales to ensure compliance. Offerings beneath
136 · Patrick Ryan Williams and Donna J. Nash

the floor reflect the links to Wari ancestor ideology (Cook 1992:360), and
secondary burials of human remains in the buildings suggest manipulation
of human bodies (McEwan 1998:76).
The niched hall as locale for the housing of ancestral remains of diverse
groups under Wari subjugation is not a given. William Isbell (2004) argues
we should not take the Inca model too literally. He discredits the argu-
ment that Middle Horizon mummies were held hostage in the provincial
centers. But the notion that ancestral rites took place in these structures is
not as easily dismissed. The question still remains, though, whether subject
groups were integrated through these niched hall ceremonies or if they
were entirely destined for the consumption of Wari residents of Pikillacta
itself. Thus, the nature of Wari strategies of incorporation through ritual is
still somewhat murky. In order to assess Wari strategies in an entirely dif-
ferent context, we turn to an analysis of ritual structures on Wari’s frontier
with its principal rival, Tiwanaku.
Tiwanaku was roughly contemporaneous with Wari in its expansive
phase between 600 and 1000 C.E. They built an enormous ceremonial cen-
ter and capital on the high Titicaca Plateau in modern-day Bolivia (Janusek

proof
2004; Kolata 2003). Tiwanaku held economic influence or established colo-
nies in both the eastern lowlands and the western coastal valleys (Goldstein
2005). Their religious structures were large platform mounds on which rit-
uals, including incense burning and the reinterment of human body parts
and sometimes human and animal sacrifice, were undertaken (Blom and
Janusek 2004; Moore 1996).
Our work in southern Peru has revealed the earliest direct evidence for
the contemporary worship of both religious traditions on the same sacred
mountain, tying the two faiths together as never documented before. The
mountain was the Wari controlled city of Cerro Baúl (Figure 6.2), a site that
linked local populations to regional entities as the only point within the
valley intervisible with all the surrounding mountain peaks of importance
in the regional apu network (Williams and Nash 2006). The contemporary
practice of both Wari and Tiwanaku ritual on the mountain summit illus-
trates the common ties that bind them and the ways in which they mediate
social relations. We turn first to a discussion of Wari D-shaped temples and
the practices of the Wari heartland as translated on the southern frontier.
proof Figure 6.2. Plan of the
summit of Cerro Baúl,
showing D-Shaped
temples and the Picchu
Picchu and Arundane
Temples.
138 · Patrick Ryan Williams and Donna J. Nash

D-Shaped Temples and Wari Ritual Practice

D-shaped temples have a floor plan in the shape of a capital letter D. Most
of these structures in both the provinces and the heartland measured ap-
proximately 10 m in diameter. They invariably looked out onto an open
plaza or courtyard. The entrance to the structure was by way of a door in
the center of the straight wall. A pair of niches flanked either side of the
door on the interior wall of the structure. Likewise, a set of four niches was
set in the curved interior wall in each of the other three directions. This
pattern of a total of 16 niches organized in 4 groups of 4 is repeated in every
known D-shaped structure.
D-shaped temples are relatively abundant at the capital and select sites in
the heartland, such as at Conchopata (Cook 2001). They are often found in
pairs in the provincial sites. D-shaped temples, earlier versions of which ap-
pear to have been round (Isbell and Cook 2002; Leoni 2006), are not found
at all Wari-affiliated sites, however. Pikillacta and Viracochapampa, for ex-
ample, have niched halls, which are rectilinear but are built with rounded
corners (McEwan 1998; Topic and Topic 2000). Also, some D-shaped struc-

proof
tures are located away from the core of settlements (Isbell 1989) or in sepa-
rate sites (e.g., at Sondondo). The latter appears to have been situated to
intervene in local mountain worship being placed between a local shrine
and its affiliated sacred peak in the Sondondo Valley (Schreiber 2004).
The use of relatively small, exclusive structures focused on mountain
peaks for ritual appears to date back to the Early Intermediate period in
the Wari heartland of Ayacucho. Leoni (2006), excavating the Huarpa (ca.
1–500 C.E.) occupation at Ñawinpukyo, describes a walled ceremonial
complex (the East Plaza) on the hill’s summit. In the center of the plaza a
rounded building with three to four concentric walls has a narrow door-
way oriented toward the snow-capped peak of Rasuwillka. The plaza, with
materials and radiocarbon dates indicating use during the Early Intermedi-
ate period, also contained remains of food preparation, camelid offerings
buried in the floor, and what appears to be a ritual pot smash of 63 vessels
composed of both decorated and plain wares.
Excavations at the heartland site of Conchopata demonstrate an affilia-
tion between circular and D-shaped structures (e.g., circular: EA100, EA-
143; D-shaped: EA-72, EA-33)1 and ritual practice (Isbell and Cook 2002).
Jose Ochatoma Paravicino and Martha Cabrera Romero (2002) excavated
EA-72 in 1997 and 1998. This area had been disturbed by settlement and
modern house construction. Thus, walls are not preserved to observe the
 Religious Ritual and Wari State Expansion · 139

presence or configuration of niches in these structures. Rescue excava-

tions revealed several intact layers. The description of the context suggests
layers of ritual deposition. Associated directly with the floor were several
pits containing camelid remains, smashed and relatively intact vessels of
several kinds, burnt human skulls, and a circular stone feature. Some of
the recovered vessels appear to have residue of contents interpreted as chi-
cha. The human skulls were burnt, found among burnt sediment, and each
were perforated in a manner consistent with trophy heads. The excavators
also found several clusters of smashed decorated pottery with elaborate
designs, somewhat later or overlying deposits appear larger with fragments
depicting warriors. Smashed coarse and undecorated pottery, stone tools,
and scattered camelid remains were also present. An immature camelid
burial appears to have been deposited above the human skulls and may
have formed part of that particular offering. Also just outside this structure
an adolescent female was found buried under the entrance (Tung 2012).
Similarly, William Isbell (2007) describes remains found in another
temple at Conchopata, EA-143. It also contained burnt human skulls and a
circular feature of standing stones. The human skulls were associated with

proof
bones from human hands and exhibit cut marks consistent with defleshing
and disarticulation (Tung and Cook 2006). Isbell also interprets the pres-
ence of articulated leg and toe bones from a deer, which are extremely rare
throughout the site, as the remains of a costume worn during ritual and
suggests that a series of fifteen pits left in the floor were probably associated
with vessels for the storage of chicha. This interpretation is likely based on
iconography on ceramic vessels that portray individuals wearing animal
headdresses with attached limbs (see Kroeber and Collier 1998:107). Recent
analysis of the skeletal remains from these two D-shaped structures shows
that the skulls represent the remains of men and children. Also many of the
individuals are foreigners, not residents of the Wari heartland (Knudson
and Tung 2011; Tung and Knudson 2008). The young female found just
outside the entrance to EA-72 was also a foreigner, and she may represent
a sacrificial offering (Tung 2012).
Wari D-shaped temples thus represent a locale of ritual practice that rei-
fies group belonging and exclusion (Nash and Williams 2004). These spaces
were likely locales where ritual consumption of chicha was accompanied by
the manipulation of other highly charged objects of religious importance,
including trophy heads, camelid offerings, and ceramics bearing iconogra-
phy with military themes. The human remains found in these spaces were
not those of the dominant Wari ethnic group but represented foreign men
140 · Patrick Ryan Williams and Donna J. Nash

and children from outside the group. Here ritual practice reinforced group
belonging, diacritical feasting, and ritual enactments that spoke to super-
natural power, self-importance, and the creation of an ethos of military
authority. Interestingly, the ritual practices documented in these heartland
spaces are not replicated everywhere D-shaped temples are found.
South of the Wari heartland, D-shaped temples are thus far best known
from Cerro Baúl (Williams 2001) (Figure 6.2). Here they were painted in
shades of white, blue-gray, and red. Small offerings were found buried in
pits around the exterior walls or in subfloor contexts in the center of the
room. These offerings include ceramic or gourd vessels or metal foil cut in
the shape of a llama (Williams and Isla 2002). The two D-shaped temples
at Cerro Baúl represent the site of sacrifice of these small offerings. In both
cases, their doorways are oriented to the mountain peak of Picchu Picchu
to the northwest. This orientation mimics the platform and sunken court
complex on the eastern side of Cerro Baúl, which we have previously iden-
tified as the Temple of Picchu Picchu, a Wari ritual complex built in the
seventh century C.E. in veneration of the mountain deity along the road to
Wari’s capital city (Williams and Nash 2006).

proof
Early Ritual Platforms in Wari’s Southern Province

Most discussions of Wari state ritual have centered on the D-shaped struc-
tures and the activities therein. Researchers have not documented platform
mound and sunken court complexes in Wari ritual contexts for the most
part. One exception is the semisubterranean temple Isbell describes from
the site of Wari in the Moraduchayoq sector. This cut-stone structure re-
sembles Formative period (2000 B.C.E.–400 C.E.) ritual structures of the
Titicaca Basin. Radiocarbon dates from the structure place its construction
in the seventh century (Finucane et al. 2007; Isbell et al. 1991).
In Wari’s southernmost province, D-shaped temples exist alongside
platform-sunken court complexes that are reminiscent of Tiwanaku and
earlier Titicaca Basin ritual practice. The evidence for the use of platform
mounds in the Moquegua province dates to the Formative–Middle Hori-
zon transition (ca. 600 C.E.), as documented by Kirk Costion (2009) at the
site of Yahuay Alta. This construction of a platform mound as a ceremonial
monument is coincident with the Wari colonization and occupation of the
mountaintop center of Cerro Baúl. Also dating to the first half of the sev-
enth century C.E., the colonization of Cerro Baúl was accompanied by the
 Religious Ritual and Wari State Expansion · 141

construction of the first platform-sunken court complex on its summit, the

Temple of Picchu Picchu.
The Temple of Picchu Picchu is a 16 × 20 m two-tiered platform fronted
by a sunken court that is overlooked by a hillside into which a series of
50-cm-high, stair-stepped benches were constructed. The complex is ori-
ented 30 degrees west of north and has two aligned staircases, one descend-
ing the bleacher-like hillside and the other ascending the platform. The two
staircases perfectly align as the central axis of the complex with the peak of
Picchu Picchu in the distance.
Several in situ burnt offerings were deposited when the platform was
constructed. A radiocarbon date processed on one of these platform con-
struction offerings yielded a calibrated age of 1366 ± 35 B.P. (AA-46599;
cal. 600–770 C.E.). The age places construction at the earliest stage of Wari
occupations at the site (Williams and Nash 2006).
This temple complex, dating to the first Wari occupations in the region,
may have preceded the D-shaped temples at the site. The earliest existing
and contemporary architecture on Cerro Baúl are the structures on the
eastern summit and the earliest constructions in the brewery area (Mose-

proof
ley et al. 2005). Thus, this early temple architecture at Cerro Baúl may not
have followed the canons of Wari D-shaped structures but that of the ritual
platforms of the south-central Andes. It should be noted, however, that
excavations in this early ritual platform had no identifiable Tiwanaku ce-
ramics. Indeed, only local Middle Horizon plainwares were located in the
construction fill of the platform.
The Picchu Picchu temple may have been Wari’s first attempt to indoc-
trinate local peoples into Wari religious hegemony. Instead of building the
first temple in the Wari style, Wari elites constructed a ritual space in the
familiar pattern of Titicaca Basin religious architecture that local people
would have been familiar with as pilgrims to altiplano sites. They oriented
the structure to the mountain peaks of the ancestors of some of these peo-
ples, the nevado of Picchu Picchu. The mountain of veneration may also
have been affiliated with some of the migrant populations that accompa-
nied Wari colonization of the region.
Wari ritual complexes were focused on the veneration of regional moun-
tain deities. This pattern has been noted for other Wari sites and regions
(Anders 1991; Glowacki and Malpass 2003). In Moquegua, we can identify
the Picchu Picchu temple as one of these sites. The Wari used Cerro Baúl,
the local sacred mountain, as a means of projecting and linking commu-
142 · Patrick Ryan Williams and Donna J. Nash

nities to more distant ancestral mountains related to Wari claims to le-

gitimacy. They enshrined Cerro Baúl itself as one of their apu. The use
of the apu of Cerro Baúl as a link to other important mountain deities is
also demonstrated in the Arundane Temple in the latter half of the Middle
Horizon.

Ritual Platforms in Late Middle Horizon

Wari and Tiwanaku ritual practices are distinct, and while Wari religious
practices had been previously documented on Cerro Baúl’s lofty summit,
no Tiwanaku rites had ever been recovered. In 2007 we excavated part of a
temple complex oriented toward the great peak of the east, the Arundane
volcano, on the road to Tiwanaku itself (Williams and Nash 2006) (Figure
6.3). The entire complex measures 50 m × 35 m and is constructed around
the largest volcanic boulder on the kilometer-long summit. The great rock
was surrounded by a central plaza 35 m square and a series of walls delimit-
ing intimate spaces around it. Excavations in 2010 demonstrated the Wari
constructed the plaza surrounding the great rock early in their provincial

proof
history. These constructions created a huaca, or sacred place, at the core of
the summit of the mountain peak, a place that likely represented the essen-
tial identity of the Cerro Baúl mountain deity for the Wari.
A subsequent addition to the complex articulated a western plaza via
a staircase and probably a monumental gateway built of coarse sandstone
and mortar (Figure 6.4). The walls of this gateway and surrounding rooms
were plastered and painted. A central patio off the western plaza formed an
alignment with the massive boulder and the peak of the Arundane volcano
to the east. Radiocarbon dates from activities in the central patio indicate
the use of this annex space dated to sometime after 750 C.E.
CB07—43—0470 1229 ± 34 B.P. 43B-CQ50 S. profile
CB07—43—0471 1212 ± 34 B.P. 43B-CQ50 E. profile
Excavations in the western plaza, where two rooms flanked an elevated
patio, revealed the evidence for Tiwanaku ritual activity on the last-use
surface (Figure 6.5). The western patio of the temple complex contained an
elevated altar 2 m square, made of stone and mortar, around which several
burnt offerings had been made. Buried in the center of the altar were the
skeletal remains of a human foot. Human body parts consecrated sacred
spaces in both Tiwanaku and Wari contexts (Blom and Janusek 2004).
 proof
Figure 6.3. Plan of the Arundane Temple complex.

Figure 6.4. Artist reconstruction of the Arundane Temple.

proof Figure 6.5. Plan
of the Arundane
platform, illustrat-
ing locations of
chemical enrich-
ment of soils and
Tiwanaku ritual
ceramic finds.
 Religious Ritual and Wari State Expansion · 145

An in situ analysis of the soils in the plaza surrounding the altar using an
Innov-X Alpha portable X-ray fluorescence instrument revealed patterns of
elemental enrichment reflecting two types of offerings. In the forward part
of the plaza, facing the mountains to the east, the soil was enriched in alkali
and alkaline metals, such as potassium, calcium, and strontium. These were
likely part of compounds, perhaps burned as incense, whose remains were
strewn on the floor. The area immediately around the altar was enriched
in transition metals, such as manganese, titanium, and copper, relative to
the rest of the structure. Micro fragments of chrysocolla and chalcopyrite
were also recovered from these areas in excavation. The offering of ground
semiprecious stones was part of the rites that took place around the altar,
known as challa in the southern Andean religious tradition.
The telltale sign that this was a Tiwanaku ritual area came from the ce-
ramics recovered from the altar and surrounding patio. The only ceramics
recovered were not Wari but Tiwanaku ritual incense burners, used only in
Tiwanaku ritual contexts. Fragments of at least six incense burners with a
high variety of forms characteristic of the diversity of burner forms found
in the ritual center of Tiwanaku were recovered from excavations in the

proof
temple (Figure 6.6). These vessels were used to burn offerings to the gods
and were created in the form of supernatural llamas and pumas, two of the
principal animals worshipped by the Tiwanaku.
A preliminary analysis of the ceramic pastes of the incense burners from
the six vessels recovered from around the altar revealed that only one was
likely to have been made locally. We also analyzed the ceramic matrix of
the six incense burners with X-ray fluorescence using the same portable
Innov-X Alpha instrument employed in the soils analysis. The instrument
was used in soil mode, calibrated with fundamental parameters, to analyze
40 additional samples of Tiwanaku ceramics collected from residential sites
in the same valley.
The local samples cluster in a broad series of groups that we have pre-
viously identified using both neutron activation analysis and inductively
coupled plasma mass spectrometry as local clay sources. One of the incense
burner fragments matched this local paste, but the other 5 were chemically
distinct from the local ceramics and from samples of 50 local clay sources
we collected (Sharratt et al. 2009). Four have elevated levels of strontium,
indicating a different geological source than the local clays and the local
Tiwanaku pots from the surrounding valley. One has elevated levels of bar-
ium and vanadium that differentiate it from the local ceramic group. We
hypothesize that the source locality for these ceramics are the Katari and
a b

c
proof
d

Figure 6.6a–g. Illustrations

of Tiwanaku incense burners
from the Arundane Temple.

e
 Religious Ritual and Wari State Expansion · 147

f g

Tiwanaku Valleys of highland Bolivia but await confirmation via higher

resolution analyses and comparison with materials from those areas.
Radiocarbon dates from the complex suggest the Tiwanaku use of the
structure began around 800 C.E. or slightly thereafter. Wari occupation

proof
on the summit commenced at 600 C.E. and endured until just after 1000
C.E. Thus, Tiwanaku ritual practice is coincident with Wari occupation
of adjacent and surrounding structures. We hypothesize that as Tiwanaku
populations in the valley grew dramatically in the years between 800 and
1000 C.E., the Wari sought ways to incorporate Tiwanaku ritual practice
into their state institutions. This strategy of ideological incorporation was
met by transforming a component of the local Baúl huaca shrine into a
Tiwanaku temple that linked the Cerro Baúl huaca to the snow-capped
mountain peak along the road to the Tiwanaku heartland. Access to the
ritual space and the practices that were reinforced in it were monitored and
controlled by the Wari lords, and religious belief became a powerful means
of mediating the relationships between distinct demographics. This is even
more evident in the programs of ritual constructions surrounding Cerro
Baúl at the end of the Wari reign.

A Landscape of Ritual Interactions

Surrounding the sacred mountain peak of Cerro Baúl are no less than three
subsidiary tripartite temples encircling the northern side of the great mesa
(Figure 6.7), some of which have been previously excavated (Owen and
Goldstein 2001). These temples, associated with local villages of Tiwanaku
 148 · Patrick Ryan Williams and Donna J. Nash

proof
Figure 6.7. Plan of the Cerro Baúl colony and the rustic tripartite temples on the slopes.

or other south-central Andean groups, are all oriented toward the moun-
tain summit of Cerro Baúl. They link ritual actions of mountain worship
undertaken by local communities with their local protector. Meanwhile,
temples on the summit of Baúl interface with regional mountain deities of
import to both Wari, Tiwanaku, and local peoples.
Each of these temple complexes has a lower elevation open court en-
closed by a low fieldstone wall less than 1 m in height. The lower courts
measure from 30 to 40 m on a side with an entrance at the midpoint of
 Religious Ritual and Wari State Expansion · 149

the downslope wall. This lower court provides access to a more restricted
central court that is slightly narrower than the length of the lower court
but generally less than 10 m in width. This middle court represents a por-
tal from the more open an unrestricted lower court and the more highly
restricted upper court. The upper court is a much more intimate space. It
is typically much smaller than the lower or middle courts, often less than
10 m on a side. It is divided in small rooms of 1–3 m length and width. The
upper courts represent the locales of small-scale ceremony presided over
by a specialist and attended by only a privileged few.
Excavations in the upper courts of these rustic temple complexes pro-
vided some clues to the offerings made within their walls. At the Yacango
temple, Bruce Owen discovered a cache of Tiwanaku blackware keros, or
ritual drinking vessels, in the uppermost court of the complex (Owen and
Goldstein 2001). They were apparently intentionally buried or sacrificed in
a pit within a small room in the intimate upper court. At Coplay, Patrick
Ryan Williams (2002) recovered a silver ring located within a pit in a small
room in the uppermost court of the complex. The pit also contained the
remains of heavily eroded animal bones that were likely deposited at the

proof
same time. And at El Paso, Donna Nash (2002) found a disarticulated hu-
man skull in the center of the middle court of the complex.
Each offering speaks to distinct ritual activities that resulted in these
deposits. At Yacango, the cache of keros presumably speaks to a small com-
munal event in which it is likely that toasts and imbibement preceded the
interment of several drinking vessels. This context parallels Marcone’s (this
volume) feasting areas that lack Wari material culture at Lote B in the Lurín
Valley. The item of personal adornment interred in a small pit in the con-
fines of the upper court of the second temple may reflect a personal or
individual rite of sacrifice. The disarticulated human skull may reference
the interment of body parts in sacred spaces seen in the Arundane Temple
and in both Tiwanaku monuments and within Wari D-shaped structures
discussed previously. It is interesting that the El Paso complex is the one
rustic temple on the edge of the town of Cerro Mejia, a town Nash (2012)
argues may have been populated by communities from different regions of
Peru.

Discussion

On its southern frontier, Wari elites created and encouraged different forms
of ritual practice than were permitted in its heartland. Here ritual practice
150 · Patrick Ryan Williams and Donna J. Nash

was embedded in architectural complexes that reflected local and regional

traditions that appealed to the large demographic that the Wari state was
attempting to attract. The platform–sunken court complex was built early
in the Wari settlement’s history as a strategy for incorporating a newly ac-
quired territory into the Wari realm. The Wari intent was not only to orient
local worship on the mountain they occupied but also to tie that worship to
distant mountain peaks over which the Wari also claimed hegemony. The
platform–sunken court complexes became so successful, they were con-
structed elsewhere: in the Tiwanaku’s main provincial town, and especially
by the several small Tiwanaku villages that came to surround Cerro Baúl
in the late 800s C.E. This demonstrates that the practices Wari encouraged
in the Early Middle Horizon within the sacred precinct of Cerro Baúl was
adopted as an effective power strategy by local community leaders in the
latter years of Wari hegemony.
These religious venues were thus recreated by local groups within the
Wari sphere of influence. The replicates not only served to enhance the
prestige of local elites but also played into the Wari initiative to turn their
own provincial center into a regional sanctuary of power and importance.

proof
In so doing, the Wari elites on Cerro Baúl enshrined themselves as the
intermediaries between different local communities and the deities they
worshipped. Local elites also capitalized on the ritual transformations
propagated by the new system. In their local shrines of worship oriented
toward Baúl, they exercised new rites that reaffirmed their relationship to
the land and toward their communities. This parallels Marcone’s argument
(this volume) that local-scale ritual that did not involve Wari material cul-
ture were transformed by the processes that took place in the Early Middle
Horizon. We would argue, however, that the ritual contexts within the Wari
settlement at Cerro Baúl does indeed represent a direct Wari interface with
high-level local elites.
The mountain of Cerro Baúl served to integrate the local population of
Tiwanaku settlers into a regional system of ritual and economic exchange
promoted by the Wari. It highlights the association in archaic states be-
tween religious belief and politico-economic relationships. The main
networks of connection between the key province of Moquegua and the
respective capitals of Wari and Tiwanaku were codified, understood, and
justified via ritual practice.
The mountain remains sacred to Andean peoples today, and shaman
continue to worship on its summit. Some of those specialists link Cerro
Baúl to regional apu as far away as Machu Picchu and the Calawaya area on
 Religious Ritual and Wari State Expansion · 151

the eastern slopes of the Titicaca Basin (Williams and Nash 2006). These vi-
sions of Cerro Baúl’s place in the cosmos cast it as the nexus of four ceques,
or ritual lines, linking regional peaks over several hundred kilometers.
Wari lords co-opted Tiwanaku ritual to assert their own agenda. They
recognized the strategic importance of scripting ritual in order to legit-
imate their own positions of power. The shared practice of ritual inter-
twined Tiwanaku and Wari people, as a fundamentally Tiwanaku practice
was hosted within Wari walls. In doing so, the Wari inserted Tiwanaku
beliefs into its own cosmology and established hegemony over the sacred
peak that served to link local populations to the wider pantheon of moun-
tain deities.
This practice of incorporation, the act of usurping Tiwanaku ritual prac-
tice within the walls of Wari administrative architecture, was fundamental
to Wari strategies of expansion on its southern frontier. Faced with a bur-
geoning population of Tiwanaku settlers, incorporating Tiwanaku practice
into Wari cosmology was a fundamentally pragmatic means of creating an
imperial strategy of expansion.

proof
Conclusions

Gary Feinman (1995) points out that the corporate mode of chiefly power
focuses on collective ritual, construction of public monuments, and the
integration of social segments. While certainly both network and corporate
strategies were used in Wari expansion, the treatment of alternative reli-
gious ideologies in Wari’s southern province reinforces the employment of
corporate modes of action in this realm of Wari hegemony. Architectural
monuments were the structures that housed religious rituals. The construc-
tion of these public monuments was an overt overture to accepting the re-
ligious rituals they traditionally housed. And acceptance of religious ritual
performance was a means of social integration of a diverse set of constitu-
ents in Wari imperial practice.
One way this means of social integration may have taken place is by
forming a ritual place of cross-cultural appeal to a broad spectrum of elites,
one that formed an elite ideology of ritual practice. Here we document
the incorporation of diverse forms of corporate architecture and diverse
collective and individual religious rituals into the Wari colonial enterprise
on the southern frontier at the site of Cerro Baúl. We argue that creat-
ing cosmopolitan society was part of the Wari strategy of expansion and
that incorporation and tolerance of belief systems was key to Wari success.
152 · Patrick Ryan Williams and Donna J. Nash

Dogmatic fundamentalism was likely part of certain Andean religious sys-

tems of the time, but in the Wari case, pragmatic incorporation of distinct
religious rites was key to creating a successful expansive state, especially on
the frontier.
This poses certain challenges to the archaeologist, for we often mistak-
enly associate distinctive material culture as evidence for rejection of state
culture or simplify it as resistance to power. In fact, diversity in ritual mate-
rial expression may be a sign of pluralism in society rather than a rejection
of state power. This work shows that a larger regional context is neces-
sary for understanding how power and ritual intersect in social relations,
and that material expression can manifest many different forms of ritual
interaction.

Notes
1.Isbell 2007 renamed the structures as follows: EA-100 = BC-D; EA-143 = BC-C; EA-
72 = BC-G, which suggests that EA-33, which does not appear on the earlier map, may
correspond to BC-F.

References Cited
Aldenderfer, Mark
2010
proof
Gimme That Old Time Religion: Rethinking the Role of Religion in the Emer-
gence of Social Inequality. In Pathways to Power, edited by T. Douglas Price
and Gary Feinman, pp. 77–95. Springer, New York.
Anders, Martha
1990 Maymi: Un sitio del horizonte medio en el valle de Pisco. Gaceta Arqueológica
Andina 17:27–39.
1991 Structure and Function at the Planned Site of Azangaro: Cautionary Notes
for the Model of Huari as a Centralized Secular State. In Huari Administrative
Structure: Prehistoric Monumental Architecture and State Government, edited
by William H. Isbell and Gordon F. McEwan, pp. 165–197. Dumbarton Oaks
Research Library and Collection, Washington, D.C.
Asad, Talal
1983 Anthropological Conceptions of Religion: Reflections on Geertz. Man, n.s.
18(2):237–259.
Bell, Catherine M.
1992 Ritual Theory, Ritual Practice. Oxford University Press, Oxford.
Blanton, R. E., G. M. Feinman, S. A. Kowalewski, and P. N. Peregrine
1996 A Dual-Processual Theory for the Evolution of Mesoamerican Civilization.
Current Anthropology 37(1):1–14.
 Religious Ritual and Wari State Expansion · 153

Blom, Deborah E., and John W. Janusek.

2004 Making Place: Humans as Dedications in Tiwanaku. World Archaeology
36(1):123–141.
Cook, Anita G.
1984 The Middle Horizon Ceramic Offerings from Conchopata. Nawpa Pacha
22/23:49–90.
1985 The Politico-Religious Implications of the Huari Offering Tradition. Dialogo
Andino 4:203–222.
1992 The Stone Ancestors: Idioms of Imperial Attire and Rank among Huari Figu-
rines. Latin American Antiquity 3(4):341–364.
2001 Huari D-Shaped Structures, Sacrificial Offerings, and Divine Rulership. In
Ritual Sacrifice in Ancient Peru, edited by Elizabeth P. Benson and Anita G.
Cook, pp. 137–163. University of Texas Press, Austin.
Costion, Kirk E.
2009 Huaracane Social Organization: Change over Time at the Prehispanic Com-
munity of Yahuay Alta, Perú. Ph.D. dissertation, Department of Anthropol-
ogy, University of Pittsburgh.
Durkheim, Émile
1995 [1912] The Elementary Forms of Religious Life. Translated by Karen E. Fields. Free
Press, New York.
Feinman, Gary M.

proof
1995 The Emergence of Inequality. In Foundations of Social Inequality, Vol. 1, edited
by T. Douglas Price and Gary M. Feinman, pp. 255–279. Plenum, New York.
Finucane, Brian Clifton, J. Ernesto Valdez, Ismael Pérez Calderon, Cirilo Vivanco Poma-
canchari, Lidio M Valdez, and Tamsin O’Connell.
2007 The End Of Empire: New Radiocarbon Dates from the Ayacucho Valley,
Peru, and Their Implications for the Collapse of the Wari State. Radiocarbon
49(2):579–592.
Fogelin, Lars.
2007 The Archaeology of Religious Ritual. Annual Review of Anthropology 36:55–
71.
Geertz, Clifford
1973 The Interpretation of Cultures: Selected Essays. Basic Books, New York.
Glowacki, Mary, and Michael Malpass.
2003 Water, Huacas, and Ancestor Worship: Traces of a Sacred Wari Landscape.
Latin American Antiquity 14(4):431–448.
Goldstein, Paul.
2005 Andean Diaspora: The Tiwanaku Colonies and the Origins of South American
Empire. University Press of Florida, Gainesville.
Insoll, Timothy
2004 Archaeology, Ritual, Religion. Routledge, London.
Isbell, William H.
1989 Honcopampa: Was It a Huari Administrative Centre. In The Nature of Wari:
A Reappraisal of the Middle Horizon Period in Peru, edited by R. Michael
154 · Patrick Ryan Williams and Donna J. Nash

Czwarno, Frank M. Meddens, and Alexandra Morgan, pp. 98–114. BAR In-
ternational Series. Archaeopress, Oxford.
1991 Wari Administration and the Orthogonal Cellular Architecture Horizon. In
Wari Administrative Structure: Prehistoric Monumental Architecture and State
Government, edited by W. Isbell and G. McEwan. Dumbarton Oaks Research
Library and Collection, Washington, D.C.
2000 Repensando el horizonte medio: El caso de Conchopata, Ayacucho, Perú. Bo-
letín de Arqueología PUCP 4:9–68.
2004 Mortuary Preferences: A Wari Culture Case Study from Middle Horizon
Peru. Latin American Antiquity 15(1):3–32.
2007 A Community of Potters or Multi-Crafting Wives of Polygynous Lords. In
Craft Production in Complex Societies: Multicraft and Producer Perspectives,
edited by Izumi Shimada, pp. 68–96. University of Utah Press, Salt Lake City.
Isbell, William, Christine Brewster-Wray, and Linda Spickard.
1991 Architecture and Spatial Organization at Huari. In Huari Administrative
Structure: Prehistoric Monumental Architecture and State Government, edited
by W. Isbell and G. McEwan, pp. 19–53. Dumbarton Oaks Research Library
and Collection, Washington, D.C.
Isbell, W. H., and A. G. Cook
2002 A New Perspective on Conchopata and the Andean Middle Horizon. In An-
dean Archaeology II: Art, Landscape, and Society, edited by Helaine Silverman

proof
and William H. Isbell, pp. 249–305. Springer, New York.
Janusek, John Wayne.
2004 Identity and Power in the Ancient Andes: Tiwanaku Cities through Time. Rout-
ledge, New York.
Knudson, Kelly J., and Tiffiny A. Tung.
2011 Investigating Regional Mobility in the Southern Hinterland of the Wari Em-
pire: Biogeochemistry at the Site of Beringa, Peru. American Journal of Physi-
cal Anthropology 145(2):299–310.
Kolata, Alan.
2003 Tiwanaku and Its Hinterland: Archaeology and Paleoecology of an Andean
Civilization. Vol. 2, Urban and Rural Archaeology. Smithsonian Institution
Press, Washington, D.C.
Kroeber, A. L., and D. Collier
1998 The Archaeology and Pottery of Nazca, Peru: Alfred L. Kroeber’s 1926 Expedi-
tion. AltaMira, Lanham, Maryland.
Kus, Susan
2012 Matters of Belief: Middle-Range Theory, Religion, and the “State.” Archaeo-
logical Papers of the American Anthropological Association 21(1):11–22.
Leoni, J. B.
2006 Ritual and Society in Early Intermediate Period Ayacucho: A View from the
Site of Ñawinpukyo. In Andean Archaeology III: North and South, edited by
William H. Isbell and Helaine Silverman, pp. 279–306. Springer, New York.
Marx, Karl
1970 [1843] Critique of Hegel’s “Philosophy of Right.” Translated by A. Jolin and J. O’Malley.
Cambridge University Press, Cambridge.
 Religious Ritual and Wari State Expansion · 155

McEwan, G. F.
1998 The Function of Niched Halls in Wari Architecture. Latin American Antiquity
9(1):68–86.
Menzel, D.
1968 New Data on the Huari Empire in Middle Horizon Epoch 2A. Nawpa Pacha
6:47–114.
Moore, Jerry D.
1996 The Archaeology of Plazas and the Proxemics of Ritual: Three Andean Tradi-
tions. American Anthropologist 98(4):789–802.
Moseley, M., D. Nash, P. Williams, S. DeFrance, A. Miranda, and M. Ruales.
2005 Burning Down the Brewery: Excavation and Evacuation of an Ancient Impe-
rial Colony at Cerro Baúl, Perú. Proceedings of the National Academy of Sci-
ences 102(48):17264-17271.
Nash, Donna
2002 The Archaeology of Space: Places of Power in the Wari Empire. Ph.D. Dis-
sertation, Department of Anthropology, University of Florida, Gainesville.
2012 El establecimiento de relaciones de poder a través del uso del espacio residen-
cial en la provincia Wari de Moquegua. Bulletin de l’Institut Français d’Études
Andines 41(1):1–34.
Nash, Donna, and P. R. Williams
2004 Architecture and Power on the Wari–Tiwanaku Frontier. Archeological Papers

proof
of the American Anthropological Association 14(1):151–174.
Ochatoma Paravicino, J., and M. Cabrera Romero
2002 Religious Ideology and Military Organization in the Iconography of a D-
shaped Ceremonial Precinct at Conchopata. In Andean Archaeology II: Art,
Landscape, and Society, edited by Helaine Silverman and William H. Isbell,
pp. 225–247. Springer, New York.
Owen, Bruce, and Paul Goldstein.
2001 Tiwanaku en Moquegua: Interacciones Regionales y Colapso. Boletín de Ar-
queología PUCP 5:169–188.
Renfrew, Colin, and Ezra Zubrow (editors)
1994 The Ancient Mind: Elements of Cognitive Archaeology. Cambridge University
Press, Cambridge.
Schreiber, Katharina.
1992 Wari Imperialism in Middle Horizon Peru. Museum of Anthropology, Univer-
sity of Michigan, Ann Arbor.
2004 Sacred Landscapes and Imperial Ideologies: The Wari Empire in Sondon-
do, Peru. Archeological Papers of the American Anthropological Association
14(1):131–150.
Sharratt, N., M. Golitko, P. R. Williams, and L. Dussubieux
2009 Ceramic Production during the Middle Horizon: Wari and Tiwanaku Clay
Procurement in the Moquegua Valley, Peru. Geoarchaeology 24(6):792–820.
Topic, John, and Theresa Topic
2000 Hacia la comprensión del fenómeno Huari: Una perspectiva norteña. Boletín
de Arqueología PUCP 4:181–217.
156 · Patrick Ryan Williams and Donna J. Nash

Tung, Tiffiny A.
2012 Violence, Ritual, and the Wari Empire. University Press of Florida, Gainesville.
Tung, Tiffiny A., and Anita G. Cook.
2006 Intermediate-Elite Agency in the Wari Empire: The Bioarchaeological and
Mortuary Evidence. In Intermediate Elites in Pre-Columbian States and Em-
pires, edited by C. Elson and A. Covey, pp. 68–93. University of Arizona Press:
Tuscon.
Tung, Tiffiny A., and Kelly J. Knudson.
2008 Social Identities and Geographical Origins of Wari Trophy Heads from Con-
chopata, Peru. Current Anthropology 49(5):915–925.
Valdez, Lidio M.
2011 Una ofrenda de cerámica ceremonial Wari en La Oroya, Valle de Acarí, Perú.
Revista Chilena de Antropología 20:189–204.
Williams, Patrick Ryan.
2001 Cerro Baúl: A Wari Center on the Tiwanaku Frontier. Latin American Antiq-
uity 12(1):67–83.
2002 Proyecto arqueologico Cerro Baul: Informe final. Instituto Andino de Estu-
dios Arqueológicos Instituto Nacional de Cultura, Lima.
Williams, Patrick Ryan, and Johny Isla
2002 Excavaciones arqueológicas en el Centro Administrativo Wari de Cerro Baúl.
Gaceta Arqueológica Andina 26:87–120.

proof
Williams, Patrick Ryan, and Donna J. Nash
2006 Sighting the Apu: A GIS Analysis of Wari Imperialism and the Worship of
Mountain Peaks. World Archaeology 38(3):455–468.