Advances in Titicaca Basin Archaeology-1 (Obra Completa) (Stanish, Cohen y Aldenderfer) PDF
Advances in Titicaca Basin Archaeology-1 (Obra Completa) (Stanish, Cohen y Aldenderfer) PDF
Advances in Titicaca Basin Archaeology-1 (Obra Completa) (Stanish, Cohen y Aldenderfer) PDF
Advances in
Titicaca Basin
Archaeology-1
Edited by
Charles Stanish,
Amanda B. Cohen, and
Mark S. Aldenderfer
Contents
Authors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
List of Figures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii
List of Tables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xii
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii
CHAPTER 1
CHAPTER 2
CHAPTER 3
CHAPTER 4
CHAPTER 5
CHAPTER 6
CHAPTER 7
CHAPTER 8
CHAPTER 9
iii
iv
Authors Affiliations
Mark S. Aldenderfer
University of California, Santa Barbara
Christine A. Hastorf
University of California, Berkeley
Elizabeth Arkush
University of California, Los Angeles
John W. Janusek
Vanderbilt University, Nashville
Matthew S. Bandy
Institute of Andean Studies, Berkeley
Cynthia J. Klink
University of California, Santa Barbara
Cecilia Chvez
Programa Collasuyu, Puno, Peru
Joel Myres
University of California, Irvine
Lisa M. Cipolla
University of California, Santa Barbara
Rolando Paredes
Instituto Nacional de Cultura, Puno, Peru
Lawrence S. Coben
University of Pennsylvania, Philadelphia
Aime Plourde,
University of California, Los Angeles
Amanda B. Cohen
University of California, Los Angeles
Matthew T. Seddon
SWCA Environmental Consultants, Salt Lake
City
Edmundo de la Vega
Programa Collasuyu, Puno, Peru
Charles Stanish
University of California, Los Angeles
Kirk L. Frye
University of California, Santa Barbara
Tiffiny Tung
Vanderbilt University, Nashville
List of Figures
1.1
1.2
1.3
1.4
1.5
2.1
The Ro Huenque Valley, with the location of the 1997 survey blocks . . . . . . . . . . . . . . . . . . . . . . . . . 14
3.1
3.2
3.3
3.4
3.5
3.6
4.1
4.2
4.3
4.4
4.5
4.6
vii
viii
5.1
5.2
5.3
5.4
5.5
5.6
66
67
67
69
70
5.7
5.8
6.1
6.2
6.3
6.4
96
97
98
99
7.1
7.2
8.1
8.2
8.3
8.4
8.5
8.6
8.7
8.8
8.9
8.10
8.11
115
116
118
120
121
123
124
126
127
128
129
9.1
9.2
9.3
9.4
136
137
138
142
144
147
10.1
10.2
10.3
77
79
87
149
151
152
154
156
157
158
159
LIST OF FIGURES
10.11
10.12
10.13
10.14
10.15
11.1
11.2
11.3
11.4
11.5
11.6
ix
Map and plans of excavation blocks in Chiji Jawira, including: (1) northeast block with
burnt surface and pit kiln (Area A), and (2) north block with remnant adobe foundations.
After Rivera Casanovas 1994:Figures 8.1, 8.12, and 8.27 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
View of excavation block in Chiji Jawira (facing the urban core), showing adobe building
foundations. Photo by Wolfgang Schuler, courtesy of Alan Kolata . . . . . . . . . . . . . . . . . . . . . . . . . . 162
Elaborate escudillas from residential and mortuary contexts in Late Tiwanaku IV
occupations in the Putuni area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
Typical serving wares from Akapana East 1M (ac) and Akapana East 2 (de). Sherds from
Akapana East 2 include tazones decorated with continuous volutes (a, b) and nonlocal
Omereque-style bowls (c, d) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
Serving vessels characteristic of Chiji Jawira, including ceremonial keros with llama motifs
(a, b) and Derived-style vessel sherds (c). After Alconini Mjica 1995:Figure 74 and Rivera
Casanovas 1994:Figures 12.112.2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
The Chucuito-Cutimbo and Juli-Pomata survey areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
Late Intermediate, or Altiplano period, ceramics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
Altiplano period settlement patterns in the Chucuito-Cutimbo and Juli-Pomata
survey areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
Typical Altiplano period settlement pattern. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
Site of Cutimbo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
Distribution of Altiplano period above-ground tombs in the Chucuito-Cutimbo and
Juli-Pomata survey areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
12.1
12.2
12.3
12.4
12.5
12.6
13.1
13.2
13.3
13.4
14.1
14.2
14.3
14.4
14.5
14.6
14.7
14.8
14.9
14.10
14.11
14.12
14.13
14.14
14.15
The Lake Titicaca Basin, with Late Horizon towns, ceremonial sites, and the road. . . . . . . . . . . . . 215
The Copacabana Peninsula, with Late Horizon ceremonial sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216
Intinkala . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
Orcohawira . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
The Horca del Inka . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
Traces of a wall at the Horca del Inka . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
Carved rocks at Copacati. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
Carved rocks on the northwest flank of Copacati . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
Playa Chatuma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
Close-up of Playa Chatuma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
Carved rock on the beach at Playa Chatuma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
Playa Chatuma and surrounding area. After Stanish et al. 1997:86 . . . . . . . . . . . . . . . . . . . . . . . . . . 225
Carved rocks at Pukara Chatuma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
Carved rocks at Playa Chinchin Jalave . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
Carved stone near Sillumocco-Huaquina . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
14.16
14.17
14.18
14.19
14.20
14.21
14.22
14.23
14.24
14.25
14.26
14.27
14.28
228
229
230
231
231
232
233
233
234
235
235
236
237
15.1
15.2
15.3
15.4
15.5
15.6
15.7
15.8
15.9
15.10
15.11
15.12
15.13
15.14
15.15
15.16
15.17
15.18
15.19
15.20
15.21
15.22
15.23
15.24
15.25
15.26
15.27
15.28
15.29
15.30
15.31
15.32
15.33
15.34
244
245
245
246
247
248
248
249
249
250
251
251
252
252
253
253
254
254
255
256
257
258
258
259
259
260
260
261
262
262
263
264
264
265
LIST OF FIGURES
xi
16.1
Late Horizon ceramics from the Taraco Peninsula: (ac) Taraco Polychrome, (dn) Chucuito
Black on Red, (o) Chucuito Polychrome, (pae) Saxamar (Pacajes-Inka), (afam) Inka
Imperial . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
16.2 Early Colonial period ceramics from the Taraco Peninsula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272
16.3 Schematic representation of a contemporary Aymara residential compound . . . . . . . . . . . . . . . . . 276
16.4 Habitation area and splash zone: (a) dispersed, (b) nucleated residential patterns . . . . . . . . . . . 277
16.5 Pacajes-Inka phase settlement pattern. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 280
16.6 Distribution of population by site size: (a) Pacajes-Inka, (b) Late Pacajes . . . . . . . . . . . . . . . . . . . . 281
16.7 Rank-size diagrams: (a) Pacajes-Inka, (b) Late Pacajes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
16.8 Late Pacajes phase settlement pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283
16.9 Change in population density (index/250 ha) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
16.10 Profile of population index density change: west to east . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
17.1
17.2
17.3
17.4
17.5
17.6
17.7
17.8
17.9
17.10
17.11
17.12
17.13
17.14
List of Tables
2.1
2.2
2.3
2.4
2.5
2.6
3.1
3.2
3.3
3.4
3.5
3.6
3.7
3.8
3.9
3.10
3.11
3.12
3.13
3.14
3.15
3.16
3.17
3.18
3.19
10.1
10.2
11.1
12.1
12.2
13.1
Settlement pattern data from the Chucuito-Cutimbo and Juli-Pomata surveys . . . . . . . . . . . . . . . 203
16.1
16.2
16.3
16.4
Preface
ters of orthography, we permitted each author to
use their preferred spellings and conventions.
Many people have provided assistance in
the long process of preparing this volume for
publication. We would like to thank Karen Doehner for her assistance in the earlier stages of
manuscript preparation. She also translated
some of the text. Matthew Bandy provided
much technical support throughout the project
as well as contributing a paper. We thank our
superb copyeditor, Dr. Janet Dunn, who proved
to be exacting and diligent. Ms. Carol Leyba designed this volume, and we once again thank
her for her outstanding work. The Director of
Publications of the Cotsen Institute of Archaeology at UCLA, Dr. Julia Sanchez, has provided
immense professional help in this work. Finally,
we thank Mr. Lloyd Cotsen for his generous
support of the Institutes mission, including the
publication of archaeological monographs and
books.
xiii
1.
Introduction to
Advances in Titicaca Basin Archaeology-1
Charles Stanish and Amanda B. Cohen
Neira (1967), Hyslop (1976), and Faldn Arancibia (1985, 1991). This work, combined with ethnohistorical research on the later periods of
prehistory, provided a broad outline of the culture history of the region (e.g., Bouysse-Cassagne
1991; Lumbreras 1974b; Lumbreras and Amat
1968). Work outside of the far northern and far
southern areas of the basin was very rare until
the late 1980s. The systematic analysis of raised
fields in the Huata area by Erickson, as mentioned above, and the rescue excavations by
Nez and Paredes at Isla Esteves (1978) represented the first intensive research away from either Pukara or Tiwanaku core areas.
In spite of this work, research in the Titicaca
Basin has been surprisingly sparse compared to
other areas of the world where first-generation
states developed. It is important to remember
[FIGURE 1.4]
1.2]
1.3
FIGURE 1.4. Map of the Lake Titicaca area with selected sites mentioned in text
CHRONOLOGY
Two kinds of chronologies are employed in the
Andes. One is historical, characterized by the famous horizon marker methodology. The most
common one, of course, is that of John Rowe, developed in collaboration with Dorothy Menzel
from the analysis of pottery from Ica on the
south coast of Peru. The second chronology is a
stage-based one, first developed in detail by
Luis Lumbreras (1974b, 1981). There has been a
large literature on the strengths and weaknesses
of these kinds of chronologies and this debate
will not be repeated here.
In general, most contemporary archaeologists working in the Titicaca region utilize the
stage chronology for a number of reasons as
[FIGURE 1.5]
1500
1000
North
West
South
Stage
Ica
Inca
Inca
Inca
Inca
Expansive
Inca
Late Horizon
Colla
Lupaqa
Pacajes
Altiplano
States
Regional
Period
Late
Intermediate
Tiwanaku V
Tiwanaku
Tiwanaku
Expansive
Middle
Horizon
Upper
Formative
Early
Intermediate
Period
Late
Huaa
Tiwanaku
Tiwanaku
500
AD/BC
500
1000
Tiwanaku IV
Early
Huaa
Pukara
Cusipata
Qeya
Late
Sillumocco
Early
Sillumocco
Kalasasaya
Late
Chiripa
Middle
Chiripa
Early
Chiripa
Qaluyu
Late
Titinhuayani
Early
Horizon
Early
Titinhuayani
Middle
Formative
Pasiri
Pasiri
1500
Early
Formative
2000
Late
Archaic
ADVANCES IN
TITICACA BASIN RESEARCH
This volume follows a traditional chronological
ordering of articles. In Chapters 2, 3, and 4, Mark
Aldenderfer, Cindy Klink, and Lisa Cipolla discuss recently discovered Archaic sites in the region. The work by Aldenderfer and his
colleagues from the University of California,
Santa Barbara, has substantially changed our
understanding of the Titicaca Basin Archaic. Aldenderfer and Klinks contributions represent
the first point typology in the region, indicating
that there were Early, Middle, Late, and Terminal Archaic occupations. (To date, no Paleo-Indian sites have been found in the Lake Titicaca
Basin.) Klinks and Cipollas work represent
some of the first analyses of systematically collected reconnaissance data on Archaic period
settlement distribution published for the region.
Klink discovered over 150 Archaic period sites
in the western Titicaca Basin of the Huenque
drainage, while Cipolla discusses and analyzes
almost 100 sites in the Huancan-Putina region
located in the north.
The work of Klink and Cipolla indicates that
there were heavy concentrations of Archaic period peoples up the rivers away from the lake.
Their work, both here and in other versions yet
to be published, suggest that the immediate lake
area was not as conducive to human occupation
as were the river valleys and dry puna in the uplands. Around the Terminal Archaic at 4400
3600 BP, Klink sees a shift toward the present
lake edge marking, perhaps, the initial shift
toward greater use of more lakeshore-oriented
land-use systems that were characteristic of all
later periods (Klink, Chapter 2, this volume).
The data from the Huancan-Putina region reported on by Cipolla in Chapter 4 support this
model as well. Likewise, excavations at an openair site on the Island of the Sun indicate initial
occupations during the Terminal Archaic that
also support this model (Stanish et al. 2002).
Cipollas chapter represents the first systematic presentation of Archaic data from the northern Titicaca region. She notes the similarities and
differences between the northern and the western basin, comparing her data to that of Aldenderfer and Klink. Like the Ilave area in the
The first ranked societies constitute the Middle Formative in our stage chronology. The most
prominent Middle Formative political centers
are Qaluyu in the north and Chiripa in the
south. The earliest Chiripa and Qaluyu pottery
dates to at least 1400 BC. These well-made vessels appear to be associated with a new emphasis on feasting and possibly the development of
prestige-good economies in simple chiefly societies. Middle Formative societies rose and fell
with some regularity throughout the region
(Stanish 2003). By 800 BC or so, the sites of
Pukara and Tiwanaku developed as two of a
number of important chiefly centers in the region. By 200 BC, if not earlier, these two sites rose
in prominence among their peers and constituted two of several major centers. The period in
which complex regional polities (complex chiefdoms in some typologies) developed is defined
here as the Upper Formative. Stanish (1999,
2003) defines these polities as complex chiefdoms. They are characterized by the development of two groups, elites and commoners, plus
a distinctive corporate architecture, a regional
settlement system, heightened economic exchange, and increased production of preciosities.
Christine Hastorf (Chapter 5, this volume)
provides an exhaustive review of Upper Formative period settlements for the Titicaca region.
She characterizes the period as one defined by
the emergence and consolidation of multiple
community ceremonial centers around the basin. She goes on to note that the time period is
also dominated by regional ethnic polities with
a cohesive iconography (Hastorf, Chapter 5,
this volume). She marks the end of the Upper
Formative with the shift to Tiwanaku dominance around the middle of the first millennium
AD. Based upon a superb data set from her work
at Chiripa (Hastorf 1999a, 1999b), she is able to
divide the Upper Formative into Early, Late, and
Terminal phases. Her chapter nicely summarizes
the iconography, architecture, economy, and settlement distributions of Upper Formative sites.
Joel Myres and Rolando Paredes in Chapter
6 provide a report about one of the major islands
on the northern side of the lake. Their chapter
describes a fascinating pattern not described in
any other area of the basinpaired sunken
10
in the region. This area is famous in the alternative literature as the home of Paititi and Akakor,
lost cities built in the forests perched above the
mysterious Amazon. In Chapter 15, Lawrence
Coben and Charles Stanish discuss the results of
a brief reconnaissance into the area. This reconnaissance was facilitated by the construction of a
new road to the hydroelectric plant at San
Gabn. The results of just a few days observations were striking. Vast numbers of late prehistoric sites were discovered along an ancient
road. The pre-Hispanic road generally followed
the course of the modern road, with some obvious differences where foot bridges could pass
over streams and other topographic features.
The Carabaya region is outside of the Titicaca Basin hydrological drainage but is clearly
part of the circum-Titicaca Basin in cultural
terms. Sites with Colla- and Lupaqa-related pottery abound along the roads. One site in particular, first noticed by Coben and Stanish in a
photograph in the office of the alcalde of
Ollachea, is a fairly large and well-preserved
Inca site that is located in a similar topographical location as Machu Picchu. The site of Illincaya (also spelled Illingaya) is perched above a
protected hill at the edge of a road. Below the
site are hot springs. The high-forest/low-forest
ecotone begins just below the town of Ollachea,
as it does at Machu Picchu. In fact, the existence
of a major road with Late Intermediate and Inca
period sites that cuts directly into the forest suggests a pattern that is similar to the settlement
system from Cusco to Machu Picchu. These data
raise the compelling possibility of more settlement systems created by late prehistoric polities
up and down the eastern slopes of the Andes.
Certainly, the Carabaya region was known to be
a major source of gold for the Inca state. It is significant that the first major use of the road is preInca, associated with northern Titicaca Basin
polities. Our understanding of the eastern slopes
is essentially nonexistent, and this chapter raises
exciting new possibilities for future work.
Matthew Bandy and John Janusek provide a
fascinating look at early Colonial period settlement shifts and demographic trends in Chapter
16. They first offer a refinement of the Late horizon and Early Colonial ceramic chronology of
the Pacajes area. They then provide a very so-
phisticated analysis of the archaeological indicators of household size and material correlates of
household artifact dispersals as they relate to assessing population estimates. Their analysis of
the demographic and settlement shifts provided
some very surprising results. While they did in
fact discover substantial drops in the population
of the Tiwanaku Valley and Katari Basin, they
found that the population of the Taraco Peninsula increased during the Early Colonial period.
They attribute this to the differences in administration between these areas. While the areas that
experienced population decline were held by individual Spaniards as encomiendas, the Taraco
area was a direct Crown holding. Their data
therefore indicate that indigenous peoples preferred to live outside of the encomienda system in
favor of living in Crown territories. The political
tensions between encomenderos and the Spanish
Crown are legion in the historical literature.
Such archaeological data serve to complement
and enhance our understanding of the complexities of early Colonial life, again providing a
wholesome check on the often biased documentary data.
Chapter 17 of this volume catalogs and describes a number important sites found throughout the western and northern basin areas.
Reconnaissance has been conducted by Programa Collasuyu members over many years.
Most of these sites are outside of individual
members research areas but are too important
to not be published.
Work continues in the Titicaca Basin at an
ever-increasing rate. This volume reflects a small
portion of the exciting results that continue to
come out. At the present time, very important
work is focused on the Formative period occupations in the northern basin. Excavations by
Elizabeth Klarich (2004) at Pukara have uncovered domestic elite areas. Amanda Cohen
(2003a, 2003b, 2004) has excavated two Formative period sites in the Pukara Valley including
both ceremonial and domestic contexts. Rolando
Paredes and his colleagues continue to work in
the northern Basin. In the southern basin, survey
projects are ongoing, including investigations by
Matthew Bandy and Carlos Lmuz. Excavations
at sites in the south are currently being conducted by John Janusek, Javier Escalante, Debo-
11
2.
Archaic Period Research in the
Ro Huenque Valley, Peru
Cynthia J. Klink
INTRODUCTION
OUR UNDERSTANDING of the Archaic1 period in
the Lake Titicaca Basin2 remains limited, hampered by a lack of research aimed specifically at
illuminating this early prehistoric period in the
region. Nearly all archaeological investigations
in the region have focused exclusively on Formative period and later complex societies (see this
volume; also Bermann 1994; Kolata 1996b;
Mujica 1987; Stanish and Steadman 1994; Stanish
et al. 1997). Preceramic period research on the Altiplano, while still in an incipient stage, began
with Aldenderfers Proyecto Chmak Pacha, a
long-term archaeological project centered in the
Ro Ilave Valley, Peru (Aldenderfer 1996). I
present in this paper the initial data from another
recently initiated Archaic period research project,
situated in the Ro Huenque Valley in the southwestern lake basin, Peru (Figure 2.1).
The Ro Huenque research project has two
goals: to develop a firmer understanding of Preceramic settlement and adaptations on the Peruvian Altiplano, and to examine the process of
initial human movement into and settlement of
the region. Field research took place in two
phases. The first phase (199798) involved intensive, systematic survey of four survey blocks in
the river valley, and was aimed at obtaining a
representative sample of sites and documenting
their general characteristics and distribution.
The second phase (1998) entailed intensive surface collection at a series of Early through Late
Archaic period sites, and enables more detailed
examination of changes in the use of specific lo-
[FIGURE 2.1
13
14
CYNTHIA J. KLINK
FIGURE 2.1. The Ro Huenque Valley, with the location of the 1997 survey blocks
15
The Paleo-Environment
The 12,000 to 8,000 BP period, although for the
most part cooler and moister than modern conditions, marked the onset of a long-term drying
trend that began in fits and starts and accelerated
more rapidly after 9000 BP. The transition from
Late Pleistocene to Holocene conditions was
more pronounced in the Bolivian than in the Peruvian Cordillera, due to the greater extent of
glacial cover in the former (Baied and Wheeler
1993; Seltzer 1990, 1993). Glacial retreat in the Bolivian Cordillera began around 12,000 BP, and
was marked by a series of still stands and readvances until about 10,000 BP (Abbott, Seltzer,
Kelts, and Southon 1997; Seltzer 1993). After this
time, glaciers rapidly retreated to near-modern
limits by 9000 BP, and most valleys were completely deglaciated before 8000 BP (Abbott, Seltzer, Kelts, and Southon 1997; Seltzer 1990, 1993).
Bofedales and small lakes formed in deglaciated
terrain in both the Bolivian and Peruvian Cordilleras (Abbott, Seltzer, Kelts, and Southon 1997;
Baied and Wheeler 1993; Graf 1981). Paleo-lake
Tauca (Titicaca) began to drop from its high
stand at 510 m above modern levels sometime
after 11,000 BP (Clapperton 1993), and likely approached modern levels around 9000 BP (Abbott,
Seltzer, Kelts, and Southon 1997). Recently deglaciated areas underwent pioneering vegetation
stages (Graf 1981), and nonglaciated areas developed more extensive vegetation cover, which
consisted of mixed grass-shrub lands likely with
16
CYNTHIA J. KLINK
METHODOLOGY
The 199798 survey phase utilized a probabilistic, systematic, and intensive survey design. The
locations of the four survey blocks were selected such that two blocks were located in each
environmental zone. The blocks were evenly
spaced along the valley and were of approximately equal size; each also contained roughly
comparable features (mountains, terraces, bofedales or lagoons, etc.). All terrain within each
block was systematically examined by pedestrian survey, with typically 10-m and no greater
than 15-m intervals between crew members.
Sites were operationally defined as locationally discrete artifact clusters, and their borders
typically delimited based on where artifact densities fell below two to three artifacts within 10
m. Isolated projectile point finds were also recorded. Components refer to the chronological
distinctions perceptible within the archaeological
materials recovered at a site. Three artifact collection strategies were used, depending on site
preservation, size, and content: grab samples,
transect collections, and point-provenience collecting. Information on site size, artifact density,
geomorphic and environmental setting, the
kinds and distribution of artifacts or structures
present, and state of preservation was recorded
for all sites. Locations were fixed using a Global
Positioning System (GPS).
The data analysis and interpretation generally follow the Preceramic periodization established by Aldenderfer (1998) at the site of Asana
in the adjacent Peruvian high sierra, as modified
for the Peruvian Altiplano by Aldenderfer and
Klink (Aldenderfer and Klink 1996; Klink and
Aldenderfer 1996). The Ro Huenque projects
chronological scheme has four Archaic periods:
the Early Archaic, circa 10,0008,000 BP; the Middle Archaic, circa 80006000 BP; the Late Archaic,
circa 60004400 BP; and the Terminal Archaic,
circa 44003600 BP. Projectile point style crossdating provides the means for identifying chronologically distinct Archaic components at individual sites, and utilizes the projectile point
typology defined by Klink and Aldenderfer
(Chapter 3, this volume). This typology is based
on excavation data from the sites of Asana (Al-
17
PROJECT
RESULTS
18
CYNTHIA J. KLINK
Early
Archaic
Middle
Archaic
Late
Archaic
Terminal
Archaic
151
29
70
69
22
4.59
0.88
2.13
2.10
0.67
Time-calibrated count*
7.25
17.50
21.56
13.75
Time-calibrated density
(per km2)*
0.22
0.53
0.67
0.42
Count
2)
Density (per km
# Because of sites with multiple components, the total count is greater than the total number of sites.
* The time-calibrated count and calibrated density are calculated as the component count or the component density divided
by the number of 500-year blocks comprising each Archaic period. They account for the variable time lengths of different
periods and facilitate direct inter-period comparisons.
adjacent regions, such as the far southern Peruvian coast (Sandweiss et al. 1998) and the northern Chilean highlands (Santoro 1989; Santoro and
Nuez 1987), is marked by a notable decline or
general absence of sites. Middle Archaic period
migration into the Lake Titicaca Basin may have
been one response to environmental changes associated with the onset of notable early to midHolocene climatic warming. The high degree of
similarity in Early and Middle Archaic projectile
point styles between the lake basin, the Chilean
highlands, and the Peruvian coast minimally argue for cultural contact between these areas at
this time (Klink and Aldenderfer 1996). The second possibility is the reorganization of settlement
and mobility patterns within the basin, with
greater mobility potentially contributing to
higher site frequencies during this period. The
shift to drier conditions on the Altiplano likely
would have reduced environmental productivity
in the lake basin (but to a lesser extent than in
other, already relatively drier areas such as the
coast, or the more southerly Chilean highlands),
and would have increased resource patchiness.
These environmental responses could have fostered either more frequent residential moves
(lower productivity) or greater logistical mobility
(increased patchiness).
Other than a modest increase in component
count and density, there is relatively little difference between the Middle and Late Archaic components. The Terminal Archaic marks a period
of more pronounced change, with component
count and density decreasing substantially. Like
the earlier increase, this pattern could have re-
19
Middle Archaic
(n = 70)
Late Archaic
(n = 69)
Terminal Archaic
(n = 22)
Dry puna
55%
54%
58%
50%
Wet puna
45%
46%
42%
50%
Dry/wet puna
density ratio*
1.17
1.13
1.31
0.94
* This ratio is calculated as the site density in the dry puna zone divided by the site density in the wet puna zone. It accounts
for the slight size difference in total survey area coverage between the two zones. Values greater than 1.00 indicate greater
use of the dry puna; values less than 1.00 indicate greater use of the wet puna.
Middle Archaic
(n = 70)
Late Archaic
(n = 69)
Terminal Archaic
(n = 22)
Block 1
48%
39%
39%
18%
Block 2
7%
16%
19%
32%
Block 3
21%
27%
30%
45%
Block 4
24%
19%
12%
5%
Middle Archaic
(n = 70)
Late Archaic
(n = 69)
Terminal Archaic
(n = 22)
Mt./upland*
34%
15%
13%
10%
Permanent water**
45%
52%
55%
64%
Other/ephemeral
resources***
20%
33%
32%
27%
cates an erroneous assessment of the early prehistoric settlement potential of these zones, which
was based on relative differences in modern climatic conditions and population distributions.
Second, the persistence of this pattern, despite
other major changes in Preceramic settlement,
implies that such large-scale environmental variability played a relatively minor role in Archaic
period developments in the Titicaca Basin, until
perhaps Terminal Archaic times. Instead, it seems
that geographic position within the valley and local environmental variability exerted greater influence (see Tables 2.3 and 2.4).
[TABLE 2.3 , 2.4]
20
CYNTHIA J. KLINK
Terminal Archaic represents a significant reorientation in settlement strategies and the initial
shift toward relatively greater use of the zone
closer to Lake Titicaca (Aldenderfer 1996; Aldenderfer and Klink 1996). Comparison of Terminal Archaic and Post-Archaic site densities
between the Ro Ilave (Aldenderfer 1996) and
Ro Huenque survey zones solidly buttresses
this argument. The Terminal Archaic component densities are slightly higher in the Ilave
Valley (0.81 sites/km2) than in the Huenque
Valley (0.67 sites/km2), and this density differential increases even more substantially in the
Post-Archaic period. The Post-Archaic component density in the Ro Ilave Valley (14.05 sites/
km2) is double that of the Huenque Valley (6.94
sites/km2).
The distribution of components by resource-access zone (Table 2.4) also reveals notable changes over the course of the Archaic.
Resource-access zones were defined based on
geomorphic site settings. Mountain/upland
settingswhich include mountaintops, mountain slopes, and fans at mountain basesshould
provide access to dispersed resources such as
taruca (deer) and vicua in queual forest and
tola shrub land (Aldenderfer 1998), and to
smaller, rocky-habitat animals such as viscacha
and some ground-nesting birds. Permanent water settings include river terrace margins and the
immediate edges of lakes and bofedales, and
would have provided access not only to water,
but to all animals found habitually using these
habitats, particularly the vicua (Aldenderfer
1998), and likely a variety of birds. All other geomorphic settings, such as the lower-elevation
hills and ridges interspersed across the pampas
and the pampas themselves, likely would have
provided access to limited and more unpredictable resources. The bunch-grass-dominated
plant communities found in these areas have
very low productivity and generally provide
only low-grade forage for herbivores (Aldenderfer 1998), although rodents such as chinchilla
can be found here. During the wet season, shallow pans on the pampas surface fill with water
and provide temporary camelid grazing areas
for modern-day pastoralists.
Throughout the Preceramic period, the use
of mountain/upland areas decreased while the
21
22
CYNTHIA J. KLINK
Middle Archaic
Late Archaic
Terminal Archaic
All
Mean
Standard deviation (m2)2
n
6963
6625
24
5092
4795
66
5855
6287
63
5341
4485
18
Single component3
Mean
Standard deviation (m2)
n
5084
0
1
2672
2046
7
3325
1237
2
1383
915
5
sites1
1. All single- and multicomponent sites with diagnostic material for a specific component, not including isolated point finds,
when quantitative data were recorded. One exceptionally large Ceramic period site overlying a much smaller Preceramic
occupation was not included because it significantly skewed average size computations.
2 Standard deviations nearly equal or exceed the mean in cases when the size distribution does not conform perfectly to a
normal distribution.
3. Sites with material diagnostic of only a single Archaic period, not including isolated point finds, when quantitative data
were recorded.
Single-component sites
Low-density sites
Moderate-density sites
High-density sites
Early Archaic
(n = 25)
Middle Archaic
(n = 68)
Late Archaic
(n = 65)
Terminal Archaic
(n = 20)
8%
18%
18%
20%
40%
35%
34%
25%
52%
47%
48%
55%
Early Archaic
(n = 1)
Middle Archaic
(n = 7)
Late Archaic
(n = 2)
Terminal Archaic
(n = 5)
0%
29%
100%
40%
100%
57%
0%
60%
0%
14%
0%
0%
1. All single- and multicomponent sites with diagnostic material for a specific component, not including isolated point finds,
when artifact density data were recorded. One exceptionally large Ceramic period site overlying a much smaller Preceramic occupation was not included because it significantly skewed average size computations.
2. Sites with material diagnostic of only a single Archaic period, not including isolated point finds, when artifact density data
were recorded.
the two data sets and the extremely small sample of single-component Early Archaic sites.
The single-component data suggests that mobility may have become more complexly organized during the Middle Archaic, with a wider
array of longer- and shorter-term sites used,
and/or higher reoccupation frequencies at
some sites. This greater complexity suggests a
shift toward a more logistically organized
(sensu Binford 1980) mobility pattern relative
to the Early Archaic. Subsequent trends are difficult to discern because of conflicting trends
between the two data sets. There was either no
change from the Middle to Late Archaic (all
sites), or a shift back to a less complex, residentially organized (sensu Binford 1980) mobility
pattern more like the Early Archaic, but characterized by higher mobility (single-component
sites). The Late to Terminal Archaic trend is
clearer, as both data sets indicate a shift toward
more longer-term occupations during Terminal
Archaic times. However, the magnitude of this
change is uncertain, as is whether it was accompanied by a shift from residential to more
logistically organized settlement (single-component sites) or not (all sites).
DISCUSSION
The surprisingly high Archaic site densities in
the Huenque Valley greatly strengthen the hypothesis that the majority of Archaic period developments played out in the more interior
sections of the Lake Titicaca Basin rather than
along the lakeshore (Aldenderfer and Klink
1996; Klink and Aldenderfer 1996). This is further supported by comparisons of site densities
between the Ilave and Huenque Valleys that indicate that substantially greater use of areas
closer to Lake Titicaca first occurred during
Post-Archaic times. This shift in land-use focus
makes sense within the context of the clearly
documented early through mid-Holocene dry
phase in the broader region, which first began to
ameliorate during the Terminal Archaic and
most closely resembled modern conditions after
circa 3600 BP. As noted earlier, mid-Holocene
warming caused a significant drop in the level of
Lake Titicaca and dramatically increased salinity
levels and reduced lacustrine productivity (Selt-
23
24
CYNTHIA J. KLINK
CONCLUSION
Our knowledge about Preceramic developments
in the Titicaca Basin is in a fairly rudimentary
stage. This analysis of the site survey data from
the Ro Huenque Valley project allows us to propose a firmer outline of Archaic period land-use
and settlement trends in the region, and highlights important issues and new directions for future research. Inherent limitations in this data set
point out the need for more detailed examinations of site structure for aiding interpretations of
changing land-use and settlement strategies over
time. Ongoing analyses of data from the 1998 intensive collection research phase should bring
these subjects into much sharper focus.
NOTES
1.
2.
3.
A Projectile Point Chronology for the
South-Central Andean Highlands
Cynthia Klink and Mark Aldenderfer
INTRODUCTION
OUR GOAL WITH this chapter is to define a temporally sensitive projectile point typology for the
south-central Andean highlands. The past ten to
fifteen years have seen an increasing amount of
archaeological research in the altiplano and the
adjacent puna and high sierra of Peru, Bolivia,
and Chile (Albarracin-Jordan 1996a, 1996b; Aldenderfer 1998; Bermann 1994; Hastorf 1999b; Janusek 1999; Kolata 1996b; Mujica 1987; Prssinen
and Siiriinen 1997; Santoro and Nuez 1987;
Stanish 1997; Stanish and Steadman 1994; Stanish
et al. 1997). Most of these investigations have focused on the more recent portion of prehistory
the Formative through Inca periodsand these
efforts have produced both better-defined ceramic chronologies and a more refined understanding of cultural developments during these
periods. In comparison, the Preceramic period remains relatively unknown, and projectile points
in general, regardless of age, have received scant
attention. It is becoming increasingly clear that
this knowledge gap results from a lack of research focused on these topics. Recent survey
projects by the authors document a substantial
Preceramic human presence within the region
(Aldenderfer 1996; Klink 1998). Other researchers, examining later time periods, not only are
encountering Preceramic sites during surveys,
but also are finding that projectile points often
are common items of post-Preceramic period material culture (Stanish and Steadman 1994; Stanish et al. 1997; Lisa Cipolla, Aime Plourde,
Amanda Cohen, and Matt Bandy, all personal
Background
This typology is based on projectile point data
from excavated and radiocarbon-dated contexts
within the south-central Andean highlands (Figure 3.1), and it should be viewed as directly applicable only within this region. The northern
limit of this region is defined by the Ro Camana/Majes drainage in the Department of
Arequipa, and in the south between the Ros Camarones and Camia in northern Chile. The
western flanks of the Andes define the western
limits of the region, while to the east, it includes
the high puna stretching through Peru, Chile,
and Bolivia eastward to the continental divide.
As more sites in the highlands are added to the
database, it will be possible to expand the geographic scope of the typology.
The projectile point assemblages from the
sites of Asana (Aldenderfer 1998) and Quelcatani (Aldenderfer, in press), located in southern
Peru (see Figure 3.1), are the foundation of this
chronology. These sites are the only deeply stratified, well-dated, extensively excavated sites
within the region that contain Preceramic
through Formative period and later deposits.
Asana is a deeply stratified, open-air site located
at 3435 m elevation, in the high sierra ecozone of
the western Cordillera. Excavations covered
[FIGURE 3.1]
25
26
Sites: (1) Toquepala, (2) Asana, (3) Caru, (4) Quelcatani, (5) Patapatane, (6)
Hakenasa, (7) Las Cuevas, (8) Tojo-Tojones, (9) Sumbay, (10) Ichua, (11) Omo,
(12) Tumatumani, (13) Jiskairumoko, (14), El Panteon, (15) Tumuku, Quelkata,
(16) Titijones, (17) Tiwanaku, (18) Yara, (19) Acha-2, Camarones-14.
FIGURE 3.1. Select sites and locations in the south-central Andean highlands
28
FIGURE 3.3. Types 1A1B, 2A2C. Type 1A: (a) Asana, level PXXVII, (b) Patapatane, after Santoro and Nuez 1987:Figure 5-13, (c) Caru, after Ravines 1967:Figure XXVI-#14. Type 1B: (d)
Asana, level PXXIV, (e) Patapatane, after Santoro and Nuez 1987:Figure 5-14. Type 2A: (f)
Asana, level PXXV, (g) Quelcatani, level WXXXV. Type 2B: (h) Quelcatani, level WXXXI(a), (i)
Quelcatani, level WXXXI. Type 2C: (j) Quelcatani, level WXXXI, (k) Quelcatani, level WXXXV,
(l) Caru, after Ravines 1967:Figure XXVII-17.
assessment of similarities and differences between types. The Asana projectile points were
directly examined for this analysis. All measurements for the Quelcatani specimens came from
digital photographs. Measurements for tools
from other sites were generated from scaled
drawings published in the original texts.
1A. Diamond/Foliate
This type (Figure 3.3ac, Table 3.1) is defined
based on three examples from Asana, two
30
Asana,
L. PXIX
Patapatane,
Fig. 3b
Patapatane,
Santoro and
Nuez 1987:
Fig. 5-#17
Caru, Fig. 3c
Caru,
Ravines 1967:
Fig. XXVI-#15
8700
9400
87009400
8600
8200
8200
8200
8200
Length
38
38
und
38
53 (est)
47 (frag)
36 (est)
33 (frag)
36
Shoulder angle
35
20
55
45
30
40
45
Tool width
16
17
21
24
25
19
21
Haft length
15
19
17
16
24
15
16
Blade length
23
19
und
22
29 (est)
21 (est)
20
1.54
1.00
und
0.92
1.16 (est)
1.11
0.95
Site
Age (BP)
Asana,
L. PXXVII,
Fig. 3a
Asana,
L. PXXV
0.65
1.00
und
0.73
0.83 (est)
0.71
0.80
Haft angle
75
70
70
65
75
75
65
Basal width)
2.38
2.24
und
1.58
2.12 (est)
1.89
1.71
1B. Pentagonal
This type (Figure 3.3de, Table 3.2) is defined
based on two examples from Asana and two
from Patapatane (Santoro and Nuez 1987). Its
diagnostic characteristic is its pentagonal, or 5sided, form. The fifth side refers to the base,
which is quite broad (basal widths of 48 mm),
and either straight or very gently convex. These
specimens also differ from the type 1A examples
in having less steeply contracting haft angles
(7585), relatively higher-waisted hafts (haft/
blade length ratios of 0.901.45), and squatter
blade elements (blade length/tool width ratios
of 0.690.90). No examples have blade edge
modification other than the shoulder spines.
At Asana, one point comes from a level
dated to ca. 94008700 BP, and the other from a
level dated to ca. 8700 BP. The two Patapatane
specimens are from a stratum dated to ca. 8200
BP. None were recovered from Quelcatani,
which dates to ca. 7300 BP and younger. This
form is present, but uncommon, in surface assemblages from the Ro Huenque and Ro Ilave
Valleys. Type 1B is considered diagnostic of the
Early Preceramic, or the period dating to ca.
10,0008,000 BP.
[TABLE 3.2
REGIONAL COMPARISONS
The undated basal stratum at SU-2 in the Arequipa highlands also contains a point very similar to our type 1B (Neira 1990). As discussed
above, this stratum underlies a stratum crossdated to ca. 6200 BP. Lavalle et al. (1985) also recovered one roughly comparable point from the
basal level at Telarmachay, which dates to ca.
90007200 BP. None of Ricks (1980) point types
at Pachamachay closely match our 1B style.
However, some examples of his types 2E (ca.
32
Site
Asana,
L. PXXVII
Asana,
L. PXXIV, Fig. 3d
Patapatane,
Fig. 3e
Age (BP)
Patapatane,
Santoro and
Nuez 1987:
Fig. 5-#15
87009400
8700
8200
8200
Length
35
27
35
34 (est)
31 (frag)
Shoulder angle
40
50
45
45
23
18 (est)
20
19 (est)
Haft length
19
16
17
18
Blade length
16
11
18
16 (est)
0.84
0.69 (est)
0.90
0.84 (est)
1.19
1.45
0.94
1.13 (est)
Haft angle
85
85
80
75
Basal width
1.52
1.50 (est)
1.94
1.79 (est)
Tool length/width
90008000 BP) and 2B (80006500 BP) are reminiscent of our type 1B.
2A. Foliate
This type (Figure 3.3fg, Table 3.3) is defined using three examples from Asana and one from
REGIONAL COMPARISONS
Ravines type P9 at Toquepala closely resembles
our type 2A: una pieza larga y delgada de hoja
ms o menos ovalla espiga constituye aproximadamente la mitad de la longitud total de la
pieza. La base es de forma convexa (1972:
147). Type P9 occurs nearly exclusively in strata
4 and 5, the two basal strata at the site. Ravines
2B
Asana, L.
PXXV, Fig. 3f
Asana,
L. PXIX
Asana,
L. PXII
Quelcatani,
L. WXXXV,
Fig. 3g
Quelcatani,
L. WXXXI(a),
Fig. 3h
Quelcatani,
L. WXXXI,
Fig. 3i
8700
8600
80008100
72007300
60007000
60007000
Tool length
46
32
41 (est)
35
26
17
Shoulder angle
30
25
15
20
25
Tool width
18
15
18
15
14
11
Haft length
20
15
26
20
17
10
Blade length
26
17
15 (est)
15
1.44
1.13
0.83 (est)
1.00
0.64
0.64
.77
0.88
1.73 (est)
1.33
1.89
1.43
Haft angle
75
70
75
80
65
65
Basal width
2.56
2.13
2.28 (est)
2.33
1.86
1.55
Site
Age (BP)
Tool length/width
rower bases (12 mm) and more noticeably contracting stems (65 haft angles) than other Series
2 examples. Both specimens have rounded
shoulders, straight bases, and lack blade edge
modification.
Both tools were recovered from levels dating
to ca. 70006000 BP. Type 2B is absent from
Asana and rare in the Ro Huenque and Ro Ilave Valleys. This limited distribution and the
lack of clear regional correlates suggests that
type 2B is a localized style diagnostic of the later
Middle Archaic, dating to approximately 7000
6000 BP.
REGIONAL COMPARISONS
There are no good correlates for type 2B. Ravines (1972) type P6 at Toquepala has a diamond-like form similar to our type 2B, but his
illustrated examples are noticeably larger, have
angular shoulders and pointed bases. Type P6
occurs exclusively in strata 2 and 3 at the site,
which Ravines estimates to date to ca. 60005000
BP. At Pachamachay (Rick 1980), one illustrated
example of Ricks type 3B and one of his type
3A, both dated to ca. 90006500 BP, are only
vaguely similar to our type 2B. At Telarmachay,
some of Lavalle et al.s (1985) type 1.4b specimens very roughly resemble our type 2B; these
34
Quelcatani,
L. XXXI(a)
Quelcatani,
Quelcatani,
L. WXXXI, Fig. 3j L. WXXXV, Fig. 3k
Asana,
L. PX
60007000
60007000
72007300
8000
Tool length
30
32
28
29
Shoulder angle
25
15
15
Tool width
17 (est)
17
15
16
Haft length
17
16
19
19
13
16
10
0.76 (est)
0.94
0.60
0.63
1.31
1.00
2.11
1.90
70
75
75
80
Blade length
Haft angle
Basal width
Tool length/width
1.76 (est)
1.88
1.87
1.81
2C. Pentagonal
This type (Figure 3.3jl, Table 3.4) is defined using three examples from Quelcatani and one
from Asana. Its diagnostic characteristic is its
pentagonal form, with the fifth side of the
pentagon being the base. The base is quite broad
(basal widths of 45 mm), and either straight or
very gently convex. Other than overall shape
and basal width, the 2C specimens do not stand
out in any particular way and have average
metric characteristics compared to other Series 2
examples.
Two Quelcatani examples were found in ca.
70006000 BP levels, and the third in a ca. 7300
7200 BP level. The Asana specimen was recovered in level PX, which is dated to ca. 8000 BP.
Type 2C forms are relatively common in the Ro
Huenque and Ro Ilave Valleys. We consider
type 2C diagnostic of the Middle Archaic, or the
period dating to ca. 80006000 BP.
[TABLE 3.4
REGIONAL COMPARISONS
In his excavations at Caru, Ravines (1967) recovered a similar point, also shown here as Figure
3.3l, from a stratum radiocarbon dated to ca.
8200 BP. This point differs slightly in having a
small, shallow basal concavity. Ravines (1972)
also illustrates a couple of points from Toque-
SERIES 3. UNSTEMMED,
UNSHOULDERED FOLIATE FORMS
Diagnostic characteristics of Series 3 are (1) absence of a stem, (2) absence of a shoulder, (3)
large size, and (4) elongated shape. These are the
generic foliate forms ubiquitous at Preceramic
sites across the broader Andean region. We define five major subtypes (3A, 3B, 3C, 3D, 3F) apparently local to the region, based on variation
in haft form and blade edge modification. In our
presentation of subtype 3D, we also discuss one
other potential variant (3E) that occurs rarely in
the south-central highlands and seems to be confined primarily to the western Cordilleran flank.
Type 3A spans the entire Early and Middle
Archaic, dating to ca. 10,0006,000 BP. Type 3B is
restricted to the Middle Archaic, or ca. 8000
6000 BP. Types 3C and 3D are extremely longlived forms, enduring throughout the entire Preceramic. The tentative type 3E likely is restricted
FIGURE 3.4. Types 3A3F. Type 3A: (a) Asana, level PVI, (b) Asana, level PXXII.
Type 3B: (c) Asana, level XVIIc, (d) Asana, level XIX, (e) Asana, level PI. Type 3C:
(f) Hakenasa, after Santoro and Nuez 1987:Figure 8-57, (g) Las Cuevas, after
Santoro and Nuez 1987:Figure 4-6. Type 3D/3E: (h) Asana, level XVIIb (3D), (i)
Asana, level XVIIc (3E variant), (j) Asana, level XIX (3D). Type 3F: (k) Hakenasa,
after Santoro and Nuez 1987:Figure 8-59, (l) Sumbay (SU-3), after Neira
1990:36, example B.
36
TABLE 3.5. Type 3A: Wide contracting haft with straight base
Site
Age (BP)
Width: 15 mm
from Base
Width: 10 mm
from Base
Width: 5 mm
from Base
Basal Width
Quelcatani, L. WXXXI(a)
60007000
17
15
13
10
Quelcatani, L. WXXXI(a)
60007000
17
15
13
7100
16
15
12
60007800
16
15
13
10
Asana, L. PVI
60007800
18
16
13
Asana, L. PXX
8700
16
14
12
8700
20
18
16
10
87009400
und
und
12
Asana, L. XXI
Asana, L. PXXVI
REGIONAL COMPARISONS
Points with wide, contracting hafts and straight
bases (whether from unshouldered, unstemmed
forms or not) occur at Early and/or Middle Archaic contexts at several sites. They date to ca.
8200 BP at Patapatane (Santoro and Nuez 1987)
and at Caru (Ravines 1967), and to before 6200
4D/4F
(n = 5)
Basal width
510
35
1016
1013
1218
1314
(n = 2)
1620
blade
element*
BP
Site
Age (BP)
Length
Max. width
Tool length/width
Basal width
Asana,
L. XVI
Asana,
L. XVIIa
Asana,
L. XVIIc,
Fig. 4c
Asana,
L. XIX,
Fig. 4d
Asana,
L. XX
Asana,
L. PI,
Fig. 4e
60006500
6500
6800
7100
7100
60007800
45 (est)
34 (frag)
61
38 (est)
28 (frag)
38 (est)
37 (frag)
52 (est)
32 (frag)
49 (est)
37 (frag)
18
22
18
20
20
19
2.50 (est)
2.77
2.11 (est)
1.90
2.60 (est)
2.58 (est)
Regional Comparisons
Santoro and Nuez (1987) found two similar
points at Patapatane that appear transitional between our types 1A, 2A, and 3B. One came from
a ca. 8200 BP stratum, the other from an undated,
overlying level. Dauelsberg (1983) describes
finding serrated-edge, convex-based lanceolates
at Tojo-Tojone in a small excavation unit that
also produced a radiocarbon date of 9580
+1950/1540 BP. This date must be treated cautiously given (1) the small size and depth of the
unit (0.50 x 1.0 m, 40 cm depth), (2) the large
sigma range for the date, (3) the uncertain con-
38
Site
Hakenasa, Hakenasa,
Santoro
Santoro
and Nuez and Nuez
1987: Fig.
1987: Fig.
8-#55
8-#56
Las
Cuevas,
Fig. 4g
Age (BP)
4400
4400
4400
4400
83009500
83009500
Length
44 (est)
24 (frag)
43 (est)
30 (frag)
40
39
30
34
19
21
19
25
18
21
Max width
2.32 (est)
2.05 (est)
2.11
1.56
1.67
1.62
15
15
16
19
11
11
16
18
18
22
14
14
18
20
19
24
17
17
15
21
16
25
16
20
Basal width
17
Tool length/width
REGIONAL COMPARISONS
To find correlates for this point type we must
look either north to the central Peruvian highlands or south to the Chilean-Argentine salt
TABLE 3.9. Type 3D: Convex base foliates without blade-edge modification
Site
Asana,
L. XVIIb,
Fig. 4f
Asana,
L. XVIIc,
Fig. 4i (3E
variant)
Asana,
L. XVIII
Asana,
L. XIX,
Fig. 4j
Asana,
L. XX
Asana,
L. XXI
Asana,
L. PXXI
Age (BP)
65006800
6800
68007100
7100
7100
7100
8700
Length
39
46 (est)
44 (frag)
No est
27 (frag)
40 (est)
36 (frag)
47 (est)
39 (frag)
53 (est)
29 (frag)
50 (est)
36 (frag)
Max. width
14
17
22
15
18
19
18 (est)
2.79
2.71 (est)
no est
2.67 (est)
2.61 (est)
2.79 (est)
2.78 (est)
Tool length/width
Basal width
REGIONAL COMPARISONS
Type 3D correlates. Essentially identical points occur at many sites across the broader Andean region throughout the entire Preceramic era.
Ravines (1972) type P3 (large) and P4 (small) foliate points at Toquepala occur in all excavated
strata at the site, although they tend to decrease
in frequency over time. Type 3D-like foliate
points are radiocarbon dated to Early to Middle
Archaic contexts at Tojo-Tojone (Dauelsberg
1983), a Late Archaic stratum at Hakenasa (Santoro and Nuez 1987), and Early through Late
Archaic contexts at several sites in the Arequipa
highlands (Neira 1990).
In central Peru, points essentially identical
to our type 3D are found in all three Preceramic
horizons at Lauricocha (Cardich 1958). At Telarmachay, Lavalle et al.s (1985) 3D-like foliate
styles (their types 1.1a, 1.1b, 1.1c, 1.5, 1.6) are
40
3F Regional Correlates
Hakenasa, Fig. 4k
Toquepala,
Ravines 1972: p. 142
SU-3, Fig. 4l
Age (BP)
4400
5000?5500
54006200
54006200
Length
60 (est)
39 (frag)
no est
23 (frag)
72 (est)
60 (frag)
78 (est)
57 (frag)
Tool width
25
28
25
27
Basal width
24
19
19
18
Concavity width
18
14
11
12
Concavity depth
2.40 (est)
no est
2.88 (est)
2.78
Haft width*
n/a
n/a
20
22
Haft width/tool
width*
n/a
n/a
0.80
0.80
Site
Tool length/width
lustration lacks a stratum affiliation. Ravines estimates the combined age of these two strata as
ca. 55005000 BP. This is substantiated in part by
the fact that these strata contain points similar to
our Late Archaic type 4D, and overlie strata with
points similar to our Middle Archaic type 3B. We
consider type 3F diagnostic of the Late Archaic,
or the period ca. 60004400 BP.
REGIONAL COMPARISONS
Schiappacasse and Niemeyer (1984) recovered
two similar points from Camarones-14 that date
to ca. 6600 BP. Comparable forms also are known
from dated Late Archaic contexts at the site of
SU-3 in the Arequipa highlands (Neira 1990).
SU-3 contains three prehistoric cultural strata
that span the Late Archaic period, the lower two
of which produced radiocarbon dates of ca. 5400
BP and 6200 BP, respectively. Neira defines three
styles of elongated, lanceolate to triangular
points with basal concavities at SU-3, which occur throughout all of the prehistoric strata. Of
these, Neiras type Sumbay II-B most closely resembles our type 3F, except for the presence of a
very broad stem (see Figure 3.4l). The Sumbay
II-C style is also a close match, but has a more
triangular form and contracting haft margins.
We are not aware of other excavated and
dated correlates for our type 3F. However, sur-
face finds of 3F points are abundant in the southwestern Lake Titicaca region. Neira (1990)
reports numerous surface finds at the sites of Tumuku and Quelkata, as well as other localities
around the community of Chichillape, located in
the upper Ro Huenque drainage. Aldenderfer
(1985) encountered examples at Titijones, in the
high puna ecozone of the Ro Osmore drainage.
Given this, and given its ubiquity in the Ro
Ilave and Ro Huenque project areas, type 3F may
be a more localized style occurring primarily in
and around the southwestern Titicaca Basin.
REGIONAL COMPARISONS
During excavations at the site of Yara, on the
southern Peruvian coast, Rasmussen (1998) recovered a fairly similar point near a hearth radiocarbon dated to ca. 7800 BP. This specimen
differs in being much longer and wider than the
highland examples, but has the same broadstemmed, triangular-bladed, contracting haft
form. Comparable points apparently also have
been found at Tojo-Tojone. Santoro and Nuez
state that:
Recent excavations in Tojo-Tojone by
Santoro and Dauelsberg (ms) recovered
diagnostic stemmed, triangular points
previously unknown from this site
(Dauelsberg 1983). These points were
found only in the lowest level of the site,
which lacked the thick, lanceolate points
diagnostic of the overlying Early Archaic
strata. The Tojo-Tojone stemmed triangular points are similar to those from
Las Cuevas I (Figure 3.4, nos. 1 and 2).
[Santoro and Nuez 1987:68]
The previously recovered lanceolate points
(Dauelsberg 1983) were serrated-edge, convexbased forms similar to our Middle Archaic type
3B and were generally associated with a radiocarbon date of 9580 (+1950, 1540) BP. This suggests that Tojo-Tojones lower stratum, and the
stemmed points it contains, date to some time in
the Early Archaic (pre-8000 BP). The results of this
more recent work have not yet been published
42
FIGURE 3.5. Types 4A4F. Type 4A: (a) Las Cuevas, after Santoro and Nuez 1987:Figure 4-2,
(b) Hakenasa, after Santoro and Nuez 1987:Figure 6-22. Type 4B: (c) Asana, level PXXIX, (d)
Las Cuevas, after Santoro and Nuez 1987:Figure 4-3. Type 4C: (e) El Panteon, level XXII, (f)
El Panteon, level XXII. Type 4D: (g) Asana, level VIII, (h) Hakenasa, after Santoro and Nuez
1987:Figure 7-41, (i) Hakenasa, after Santoro and Nuez 1987:Figure 7-46. Type 4E: (j) Omo,
after Goldstein 1993a:Figure 3.7a (left), (k) Omo, after Goldstein 1993a:Figure 3.7a (center), (l)
Tumatumani, after Stanish and Steadman 1994:Figur e 174-#0309, (m) Tumatumani, after
Stanish and Steadman 1994:Figure 174-#0373. Type 4F: (n) Asana, level VIII, (o) Asana, level
VIII, (p) Hakenasa, after Santoro and Nuez 1987:Figure 7-39 .
TABLE 3.11. Type 4A: Triangular-bladed broad-stemmed forms with contracting hafts
Site
Hakenasa, Fig. 5b
Asana, L. PXXVII
83009500
8300
87009400
Length
28
35 (est)
28 (frag)
und
Age (BP)
Shoulder angle
80
60
60
Tool width
22
18
23
Haft width
11
12
17 (est)
0.50
0.67
0.74 (est)
11
no est
Blade length
20
24 (est)
no est
0.91
1.33 (est)
no est
0.40
0.46 (est)
no est
Haft angle
70
65
no est
Basal width
<1
no est
no est
1.27
1.94 (est)
no est
Tool length/width
4C
Asana, L. PXXIX,
Fig. 5c
9400
83009500
30003600
30003600
Length
29
29 (est)
21 (frag)
26
27
Shoulder angle
50
45
90
90
Tool width
11
13
24
18
Haft width
10
0.82
0.69
0.42
0.39
Site
Age (BP)
Haft length
10
11
Blade length
19
20 (est)
15
19
1.73
1.54 (est)
0.63
1.06
0.53
0.45 (est)
0.73
0.42
Haft angle
80
70
70
75
Basal width
2.63
2.23
1.08
1.50
Tool length/width
44
REGIONAL COMPARISONS
Ravines (1972) type P7 at Toquepala strongly resembles our type 4D. The Toquepala P7 style
was recovered exclusively in strata 1 and 2,
which Ravines estimates to date to ca. 55005000
BP. This is substantiated in part by the fact that
these strata overlie those containing points similar to our Middle Archaic type 3B. Menghin and
Schroeder (1957) recovered several points very
similar to our 4D form at the site of Ichua.
While their stratigraphic control was fairly poor
and the site lacks radiocarbon dates, they noted
that these points came from the lowest excavated deposits. The points recovered from the
upper strata resemble our Terminal Archaic to
Formative period types 5C and 5D, suggesting
that the broad stemmed types in the lower level
date to the Late Archaic period.
Site
Quelcatani, Quelcatani,
L. WXXIX
L. WXXIX
Asana,
L. VIII,
Fig. 5g
Hakenasa,
Santoro and
Hakenasa, Nuez 1987:
Fig. 5h
Fig. 7-42
Hakenasa,
Fig. 5i
Age (BP)
5600
5600
4600
4400
4400
4400
Length
35 (est)
28 (frag)
no est
33 (frag)
37 (est)
30 (frag)
no est
31 (frag)
no est
34 (frag)
37 (est)
35 (frag)
Shoulder angle
70
80
50
75
70
75
Tool width
16
18
17
14
16
19
13
15
13
10
11
13
0.81
0.83
0.76
0.71
0.69
0.68
Haft length
no est
12
no est
no est
Blade length
27 (est)
30
25 (est)
25
26
29 (est)
Haft width
Haft width/tool width
1.69 (est)
1.67
1.47 (est)
1.79
1.63
1.53 (est)
0.30 (est)
no est
0.48 (est)
no est
no est
0.28 (est)
Haft angle
80
70
80
80
75
85
Basal width
no est
no est
no est
no est
11
no est
10
no est
no est
12
2.19 (est)
no est
1.83 (frag)
2.18 (est)
no est
2.21 (frag)
(no est)
2.13 (frag)
1.95 (est)
Roughly comparable points have been recovered at SU-3 throughout the ca. 62005400 BP
levels, and at other sites in the Sumbay region
(Neira 1990). However, the Sumbay forms integrate aspects of both our types 4D and 3F. They
have the large size, broad stem, and parallelsided haft of the 4D form, but have basal concavities like type 3F. The Late Archaic SU-3
strata also contained an expanding-haft point
form, Sumbay II-E (Neira 1990). Similar broadstemmed, expanding haft points have been
found in undated contexts around Huancan in
the northern Lake Titicaca Basin (Lisa Cipolla
and Aime Plourde, personal communication
2000). Only a few expanding haft forms were recovered during the Ro Ilave survey, and none
are known from the Ro Huenque Valley.
gua Valleys low sierra. Its diagnostic characteristics are its (1) narrow-stemmed form, (2)
barbed shoulders, (3) elongated shape, and (4)
narrow width. These points have either contracting or parallel-sided hafts. Two have convex
bases, the other a straight base. None have blade
edge modification.
All three examples were found in Structure 7
at M12, in levels dating to the Omo phase AD
500650, which was contemporaneous with the
later portion of Tiwanaku Phase IV (Classic Tiwanaku). M12 is one component of the larger
Omo site complex, a Tiwanaku colony directly
occupied by settlers from Tiwanaku (Goldstein
1993a, 1993b). Goldstein suggests that these
points were likely imported to M12 from the Tiwanaku heartland because there is no evidence
of lithic manufacturing debris, and further notes
that they were made of only two material types,
including exotic obsidian. Neither Asana nor
Quelcatani contained type 4E points, and this
style is rare in the Ro Huenque and Ro Ilave
Valleys. Based on these data, and the regional
[TABLE 3.14]
46
4E Regional Comparisons
Tumatumani,
Stanish and
Steadman
1994: Fig.
Tumatumani,
Tumatumani,
174-302
Fig. 5m
Fig. 5l
Omo-M12,
Fig. 5j
Omo-M12,
Fig. 5k
Omo-M12,
Goldstein
1993: Fig.
3.7a (right)
500650
500650
500650
Uncertain
Uncertain
Uncertain
Length
35
20
29
27
14
16
Site
Age (AD)
Shoulder angle
115
100
125
130
115
100
Tool width
13
13
13
12
Haft width
0.31
0.38
0.38
0.42
0.38
0.44
Blade length
32
17
24
23
11
13
2.46
1.31
1.85
1.92
1.38
1.44
0.09
0.18
0.21
0.17
0.27
0.23
Haft angle
75
65
90
70
60
70
Basal width
2.69
1.54
2.23
2.25
1.75
1.78
at the site of Tiwanaku itself, Omo (a major Tiwanaku colony), and Tumatumani (an important Tiwanaku satellite center), clearly associates
this point type with the growth and spread of
the Tiwanaku state. That access to or use of this
style may have been reserved for the socio-political elite is implied by both the rarity of 4E
points in the Ilave and Huenque Valleys, areas
without substantial Tiwanaku centers that likely
contained primarily more rural agro-pastoral
settlements, and the use of high-quality, exotic
materials. The lack of evidence for lithic manufacture at Omo, and the limited range of raw
materials used to produce the Omo and Tumatumani points, further suggest controlled production and/or trade.
Asana,
L. VIII,
Fig. 5n
Asana,
L. VIII,
Fig. 5o
Asana,
L. VIII
Age (BP)
4600
4600
4600
4400
4400
Length
25
26
29
32
Shoulder angle
35
n/a*
50
60
Tool width
17
14
11
Haft width
12
12
Site
4400
4400
30
29
30
60
115
90
15
16
14
17
12
14
10
12
0.71
0.85
0.73
0.80
0.88
0.71
0.71
Haft length
10
10
10
Blade length
16
16
20
22
20
22
22
0.94
1.14
1.82
1.47
1.25
1.57
1.29
0.56
0.63
0.45
0.45
0.50
0.32
0.36
Haft angle
80
85
90
85
80
75
65
Basal width
1.47
1.86
2.64
2.13
1.88
2.07
1.76
Tool length/width
REGIONAL COMPARISONS
Aldenderfer has recovered Type 4F points during his 19971999 excavations at the site of Jis-
SERIES 5. UNSTEMMED,
UNSHOULDERED TRIANGULAR FORMS
The diagnostic characteristics of Series 5 are (1)
absence of a stem, (2) absence of a shoulder, and
(3) squat, triangular shape. Over 400 Series 5
points were recovered from securely dated contexts at Quelcatani, providing an unparalleled
data set for temporal analysis. Initial analyses
attempted to define temporally sensitive types
48
REGIONAL COMPARISONS
At Pachamachay in central Peru, Ricks (1980)
Types 5B and 8Bdated to ca. 45003500 BP and
ca. 45001500 BP, respectivelyare similar in
size to our 5A, but have a much more elongated,
foliate shape. Further to the south in Chile, 5Alike points apparently date to the Early Preceramic. Nuez Atencio (1992) recovered one
point at Tuina (dated to ca. 108009000 BP),
which resembles our Type 5A. Another similar
point at Tambillo dates to ca. 8600 BP (Nuez
Atencio 1992). Both of these points have the
more rounded, squat shape characteristic of our
5A type, but are notably larger than any of our
5A examples.
Hakenasa, Fig. 6a
Hakenasa, Fig. 6b
Hakenasa, Santoro
and Nuez 1987:
Fig. 9-70
Age (BP)
4400
4400
4400
4400
Length
23
20
28
25
Width
20
13
22 (est)
20 (frag)
22
Basal width
1.15
1.54
1.27 (est)
1.14
Site
Tool length/width
Hakenasa, Fig. 6c
FIGURE 3.6. Types 5A5D. Type 5A: (a) Hakenasa, after Santoro and Nuez 1987:Figure 9-68,
(b) Hakenasa, after Santoro and Nuez 1987:Figure 9-69, (c) Hakenasa, after Santoro and
Nuez 1987:Figure 9-67. Type 5B: (d) Hakenasa, after Santoro and Nuez 1987:Figure 9-80,
(e) Quelcatani, level WXXIV. Type 5C: (f) Hakenasa, after Santoro and Nuez 1987:Figure 983, (g) Quelcatani, level WXXIV, (h) Quelcatani, level WXXV, (i) Quelcatani, level WXXV.
Type 5D: (j) Asana, level IIb, (k) Asana, level IIa, (l) Quelcatani, level WXVIII, (m) Quelcatani,
level WXX, (n) Quelcatani, level WXVIII .
Site
Hakenasa,
Santoro and
Nuez 1987:
Fig. 9-79
Hakenasa,
Fig. 6d
Quelcatani,
L. WXXIV
Quelcatani,
L. WXXIV,
Fig. 6e
Quelcatani,
L. WXVIII
Quelcatani,
L. WXVIII
Age
4400 BP
4400 BP
3660 BP
3660 BP
2800 BP
2800 BP
Length
31
24
44
20
11
15
Width
24 (est)
22 (frag)
22
34
14
10
14
Basal width
12 (est)
15
Tool length/
width
1.29 (est)
1.09
1.29
1.43
1.10
1.07
50
REGIONAL COMPARISONS
Within the south-central Andean highlands, 5B
style points have been recovered from only one
other site, Las Cuevas, where one was found in
the ca. 95008300 BP stratum (Santoro and
Nuez 1987). In the Chilean-Argentine salt
puna, 5B-like points appear to be an Early Preceramic style. Aschero (1984) recovered two comparable points from a ca. 9200 BP stratum at Inca
Cueva-4. Nuez Atencio (1992) also shows two
analogous points, one from Tambillo dated to ca.
8600 BP and one from Tuina dated to ca. 10800
9000 BP. In central Peru, two of Ricks (1980) Pachamachay styles resemble our type 5B: his type
TABLE 3.18. Type 5C: Large triangular forms with concave bases
Site
Hakenasa,
Fig. 6f
Quelcatani,
L. WXXIV
Quelcatani,
L. WXXIV,
Fig. 6g
Quelcatani,
L. WXXI
Quelcatani,
L. WXXV,
Fig. 6h
Quelcatani,
L. WXXIV,
Fig. 6i
Age
4400 BP
3700 BP
3700 BP
3000 BP
3800 BP
3660 BP
Length
31
34
35
30
30
34
Width
21
25
22
26
29
25
Basal width
19
18 (est)
15 (est)
23
23
21
Concavity
width
14
13
11
13
18
16
1.48
1.36
1.59
1.15
1.03
1.36
Concavity
depth
Tool length/
width
52
TABLE 3.19. Type 5D: Small triangular forms with concave bases
Tool
Concavity Concavity Length/
Width
Depth
Width
Age
Length
Width
Basal
Width
4400 BP
21
15
15
10
1.40
4400 BP
27
15
13
11
1.80
3600-4300
17 (est)
14 (frag)
16
15
11
1.06
BP
3600 BP
15
13
13
1.15
Asana, L. IIA
3600 BP
17 (est)
9 (frag)
10
1.70
3660 BP
23
16
14
11
1.44
2400 BP
26
16
14
1.63
2700 BP
21
11
10
1.91
Quelcatani, L. EXIV
3000 BP
16
13
11
1.78
Site
REGIONAL COMPARISONS
Aldenderfers 19971999 excavations at Jiskairumoko in the Ro Ilave Valley recovered several 5D points from levels dating to ca. 3600
3400 BP. Type 5D points were also found at the
Formative- to Tiwanaku-aged site of Tumatumani, along the Lake Titicaca shore, but cannot
be firmly dated because they were recovered
from either excavation units containing mixed
construction fill or surface contexts (Stanish and
Steadman 1994). Many of Ravines (1972) illustrated Type P1 points at Toquepala are very similar to our Type 5D. Most P1 points were found
in the uppermost stratum (stratum 1), with the
remainder from stratum 2.
lands to central highland Peru. This is most obvious in the distribution of unstemmed,
unshouldered foliate points (Series 3 types),
which represent long-standing traditions in both
regions throughout the Preceramic, as well as in
the south-central low sierra and coast.
Second, geographic homogeneity in point
styles diminishes substantially through time.
Over the course of the Archaic, the number of
identifiable styles appears to increase, while the
geographic extent of each style decreases. More
localized styles first begin to develop in the Middle Archaic (Types 2B and 3E), and this trend accelerates in the Late and Terminal Archaic.
During the Late Archaic, most ties to the central
Peruvian highlands disappear, and while links to
the adjacent western valleys continue, more local
highland styles (Type 3F) and variants develop.
Local variation is best seen in Type 4D, where
different attribute combinations appear in different portions of the south-central highlands. Type
4D-like points have concave bases in the Arequipa highlands, but straight or convex bases in
the southwestern Titicaca Basin. The expanding
stem 4D variant known from the Arequipa region also appears fairly common in surface contexts in the northern Lake Titicaca Basin, but is
rare to nonexistent in the southwestern portion
of the basin. This development of more regional
styles implies that interaction spheres began to
focus within, rather than outside, the southcentral highlands and that social-cultural developments increasingly played out over this more
localized stage. The proliferation of even more
localized styles in the Late Archaic hints that
identity markers became more important, or that
social or political differentiation intensified at
this time.
Third, ties between the south-central highlands and the more southerly salt puna apparently were always very weak, as the entire
character of the Preceramic projectile point traditions in these two regions differs sharply. The
ubiquitous Archaic period tradition of unstemmed, unshouldered foliate points (Types
3A, 3B, 3D, 3E and 3F) seen in the south-central
Andes is rare further south, where unstemmed,
unshouldered, triangular forms (Types 3C, 5B
and 5C) are common. Most other point types
found in the south-central highlands (Series 1
54
and 2, and Types 4A4E) are essentially unknown in the salt puna. In addition, the blade
and micro-lithic tradition documented in the
Atacama region beginning in the Late Archaic
or ca. 5000 BP (Nuez Atencio 1992)does not
occur in the south-central highlands. This clear
divergence suggests that these two regions experienced very different cultural-historical trajectories, likely with only limited contact or
interaction, at best, between the two regions
throughout most of prehistory.
Finally, it is apparent that projectile point
studies also have significant potential to contribute to our understanding of post-Archaic societies. Typically, research emphases move away
from lithics toward ceramics, architecture, and
settlement patterns when more complex societies
are studied, yet clearly projectile points provide
additional avenues for investigating these societies and can even help form new questions about
them. For example, while it appears that bow
and arrow technology developed during the Terminal Archaic, it first becomes heavily integrated
into technological systems during the Formative
period. It is during this time frame that the first
sedentary agro-pastoral villages and complex social and political institutions developed in the
Lake Titicaca Basin (Aldenderfer 1997; Hastorf
1999b; Mujica 1987; Stanish and Steadman 1994;
Stanish et al. 1997). What role did the bow and
arrow play in these developments? At least one
important function may have been territorial defense and warfare, as suggested by rock art in the
Ro Huenque Valley (Klink 1999). However, the
bow and arrow never completely replaced other
weapon systems. Why were multiple technologies maintained, and what might this reveal
about hunting practices and subsistence economies? Projectile point frequencies also appear to
drop precipitously in the Titicaca Basin beginning with the expansion of the Tiwanaku state,
and this trend continues through the Late Horizon. How might the reduced use of projectile
points relate to state expansion and control? The
identification of a probable elite Tiwanaku
point style (Type 4E) allows an additional means
of investigating topics such as social identity and
status, and economic issues such as controlled
production and trade.
CONCLUSION
Projectile points constitute one of the most abundant material culture indicators of prehistoric
occupation and use of the south-central Andean
highlands, particularly for the Preceramic period. This is demonstrated by the results of several recent survey projects in the region, which
document not only their ubiquity but also their
wide range of morphological variability. However, the inability to confidently assess their cultural and temporal affiliations has severely
limited the potential to gain insight from these
finds. The projectile point chronology presented
here moves us considerably closer to not only
our immediate goal of being able to use the surface archaeological record as a direct source of
information about the prehistory of the southcentral highlands, but also our long-term goal of
better understanding the history and trajectory
of cultural developments in the region. While
this typology marks a significant step forward, it
should be viewed as a work in progress rather
than a final statement. Although its foundation
is solid, it is clear that the currently available
data set of excavated and radiocarbon-dated
projectile point assemblages does not fully reflect the diversity of point forms known from
surface contexts. We expect, and hope, that future research will generate additional data and
enable refinement and expansion of this typology.
4.
Preceramic Period Settlement Patterns
in the Huancan-Putina River Valley,
Northern Titicaca Basin, Peru
Lisa M. Cipolla
THE SURVEY
INTRODUCTION
A SYSTEMATIC SURFACE survey by Stanish and
Plourde in 1999 and 2000 recovered almost one
hundred Preceramic sites in the HuancanPutina River Valley of the Titicaca Basin (Stanish and Plourde 2000). The earliest evidence for
occupation in the northern basin is indicated by
the presence of Early Archaic sites dating to
80006000 BC. These dates are consistent with
data from the southern and western portions of
the basin (Aldenderfer 1998; Aldenderfer and
Klink 1996; Klink, Chapter 2, this volume; Stanish et al. 1997), and indicate that the region was
not occupied earlier, as no Paleo-Indian period
points to date have been found in the Titicaca
Basin. Furthermore, the changes in the Preceramic settlement pattern in the northern basin
are also consistent with changes seen elsewhere
in the basin. While these similarities may not
seem surprising, it is useful to note that the Titicaca Basin covers an area of 50,000 km2, and
thus the homogeneity of the data has profound
regional implications for understanding the
early prehistory of the area. It is not until the
Late Archaic period that notable differences between the north and elsewhere in the region appear. Most notable, it is at this time that
differences between the northern basin and the
southern basin began to form, a pattern that
presages the differences seen in the earliest Ceramic periods in the region, such as Qaluyu and
Chiripa.
AREA
55
56
LISA M. CIPOLLA
METHODOLOGY
The survey methodology used by Stanish and
Plourde was an intensive, systematic coverage
of the entire valley from the ridgetops to the valley bottom. All sites were recorded, and representative samples of all artifact types present on
the surface were collected. Relative artifact densities and site sizes were recorded as well. An
analysis of all lithic material collected was conducted, with particular attention given to identifying projectile point styles and material types
recovered. Artifact densities per site for the Preceramic periods were not calculated for this
study, nor were differences in site sizes over
time. Most of the Preceramic sites are multicomponent sites, and a detailed study aimed at isolating site size per time period has not been
conducted at this time.
The identification of each Preceramic time
period for this survey is based on the projectile
point typology developed by Klink and Al-
[FIGURE 4.1]
57
[FIGURE 4.2 ]
RESULTS
The Early Archaic Settlement Pattern
(10,0008,000 BP)
Only six sites containing diagnostic Early Archaic points were found in the survey (Figure
4.3). Interestingly, five of the sites are all clustered in Zone 3 along the Ro Cala Cala, a tributary of the Ro Huancan. The other site is
58
LISA M. CIPOLLA
59
60
LISA M. CIPOLLA
61
DISCUSSION
62
LISA M. CIPOLLA
CONCLUSION
The Preceramic settlement patterns for the
Huancan-Putina Valley are similar to those
from elsewhere in the region during the Early
and Middle Archaic periods. The initial occupation of the northern basin began 10,0008,000
years ago with a small number of sites and a
gradual increase in population over time. Population levels throughout the Preceramic do not
appear to have ever been particularly high given
the small number of sites in comparison to the
size of the valley itself. The subsistence patterns
indicate a focus on the upland puna areas of
fresh-water sources throughout the entire Preceramic, with increasing diversification of resources not occurring until the Late Archaic
period. It is also during the Late Archaic period
that significant differences in projectile point
styles from elsewhere in the basin occur. The
question of why these differences occurred,
however, remains to be answered. It is clear that
63
NOTES
1.
2.
In fact, while our crew was working at site Hu113 near Putina in November of 2001, we witnessed a portion of a river terrace collapse due
to active downcutting of the river terrace on its
western edge. A stretch of river terrace, approximately 20 m in length, fell into the Ro Putina.
Fortunately, no previously recorded sites were
at that location. However, we noticed significant erosion along the river terrace at sites HU101 and 102 in the last two years. Almost 75%
of each of these sites had eroded into the Ro
Putina at the time of this fieldwork.
Other points of this same style have been
recorded elsewhere in the northern basin, and
will be published in the future (Eduardo
Arizaca Medina, INC-Puno, personal communication 2002).
5.
The Upper (Middle and Late) Formative
in the Titicaca Region
Christine A. Hastorf
INTRODUCTION
THE UPPER FORMATIVE phase (circa 800/600 BC
to AD 400) in the south central highlands surrounding Lake Titicaca spans over 1000 years in
some locations.1 This period witnessed the creation of a series of ritually specialized and intensive agricultural-herding polities. Some
consider this a time of social stratification with
regional yet autonomous long-lived centers. The
name for this phase in the Bolivian and Peruvian
sequence uses the Formative terminology to describe the first settlements on the landscape as
well as evidence of the onset of territoriality
(Lumbreras 1974a; Ponce Sangins 1970, 1971,
1981:13). The Upper Formative Bolivian time
span falls within the phasing of the later part of
the Early Horizon and the Early Intermediate
period of the Rowe Peruvian sequence (Pozorski
and Pozorski 1997:7; Rowe 1962). This phase culminates in the marked cultural development of
the first pre-Hispanic large-scale polity, Tiwanaku, that expanded throughout the Lake Titicaca
region of modern Peru and Bolivia.
There are at least three documented regional
centers with civic-ceremonial architecture dated
to the beginning of this time phase, and surely
more will be uncovered in future systematic surveys. Throughout the first 500 years (contemporaneous with the Early Horizon), on the south
coast (particularly the Paracas Peninsula and the
Ica Valley) and the northern highlands (e.g.,
Chavn de Huantar), centers with special architecture were built. These early Titicaca centers
are most often located close to the lake shore,
[FIGURE 5.1
65
66
CHRISTINE A. HASTORF
Chvez, Gregorio Cordero M., Juan Faldn, Alfred Kidder II, Carlos Lmuz, Elas Mujica, Jose
Luis Paz Soria, Maks Portugal Zamora, Max Portugal Ortz, Lee Steadman, Aime Plourde,
Amanda Cohen, Elizabeth Klarich, and William
Whitehead, but also the scholars of later phases,
including Wendell Bennett, Carlos Ponce Sangins, Marion Tschopik, Juan Albarracin-Jordan,
Sonia Alconini, Marc Bermann, Deborah Blom,
Edmundo de la Vega M., Kirk Frye, John Janusek, Alan Kolata, James Mathews, Claudia Rivera Casanovas, Oswaldo Rivera Sundt, Mathew
Seddon, and Charles Stanish. We now have a series of ceramic sequences and settlement patterns
that give a more realistic sense of the events that
occurred during this dynamic time period.
CHRONOLOGY
The dates of the Upper Formative phase in this
chapter are 800/600 BC to AD 400 and have been
separated into three subphases for this volume.
This time span does not equate to the occupation
THE ARCHAEOLOGY
The Formative trajectory was an expanding dynamic that began as early as the Early Lower
Formative (ca. 2000 to 1500/1100 BC), when the
climate became progressively wetter (and perhaps warmer) and the lake level increased,
making the Titicaca Basin more arable and hos-
CHAPTER 5: THE UPPER (MIDDLE AND LATE) FORMATIVE IN THE TITICACA REGION
67
68
CHRISTINE A. HASTORF
well as lineage and ancestral ties. With the development of Pukara in the Late Upper Formative (the Cusipata phase), there is a shift in
image emphasis, with the incorporation of north
coastal (Cupisnique) and western Paracas images. These images suggest an active communication with residents of the coast and beyonda
communication that supports status validation,
ritual elaboration, and identity building (Helms
1979, 1998).
This Pukara style incorporated the power
images of a shamanistic religious cult that is
generally associated with the acquisitional
power gained from dead enemies and is most
notably seen in trophy head images on the coast
(Proulx 1999; Williams, Forgey, and Klarich
2001). While heads had been included in the
PaAjano images, in the Late Upper Formative
they took on a more significant dimension as
they were held by kneeling felines or humans,
which suggests a stronger sense of unequal
power. By the Qalasasaya times, around 200 BC,
we begin to see a more overt political authority
with controlling civic cults built upon the earlier
religious foundations of incorporation and kingroup worship; this authority is reflected in the
architecture at the centers that are dotted around
the lake.
There is also strong evidence for the intensification of several economic bases throughout
the Terminal Upper Formative phase, including
intensive raised-field agriculture, a strong lacustrine economy, and herding (Erickson 1988a,
1996; Graffam 1990). Once the lake shore stabilized, sometime around AD 200, and the hydrology was steady (Binford and Kolata 1996), a
series of intensive economic strategies were developed, including raised fields, qochas, or artificially enlarged seasonal ponds used to feed and
water camelids (Flores Ochoa 1987), and low
aqueducts that channeled hillside water for erosion control and water use.
Both the landscape and iconographic evidence suggest an expansion of decision-making
control by certain kin or ayllu groups in order to
unify and channel labor from a web of nested
kin groups scattered across the landscape. The
political evidence for these changes is seen in religious manifestations and ancestor ideologies.
Toward the end of this phase, the political struc[FIGURE 5.3 ]
tures seem to be expanded, as seen in the civicceremonial elaboration that bred a sense of centralization. There is evidence for group rituals
and feasting, limited access to exotic goods,
larger and more active local spheres of influence,
and distinctions between centers.
CHAPTER 5: THE UPPER (MIDDLE AND LATE) FORMATIVE IN THE TITICACA REGION
69
The artifact evidence suggests that local alliances were establishing a complex of ritual and
economic strategies that led to a more codified
life cycle of ayllu boundaries with increasingly
stronger ethnic markers. Political leadership is
only inferred, but some form of kin-based consolidation and economic accumulation was occurring during these 500 years, as seen in the
evolution of increasingly elaborated family burials centered around an adult female at the site of
Chiripa (Hastorf 1999a, 1999b, 2003; Steadman
and Hastorf 2001).
As these ethnic identities developed and
consolidated, it appears that interest in items of
exotic status also increased, along with the use
of fertility and surreal imagery, entering both
from the coast and the eastern forests. These
items included sodalite beads, metal jewelry,
and warm valley plants.
FIGURE 5.4. Example of the classic Upper Formative
stone carved surreal face
west of Chiripa, allows us to see how a small settlement with only two civic buildings and several hectares of occupation was brought into
Chiripas sphere by around 500 BC (Beck 2004).
More excavation work in sites of this time period will allow us to place these enclosures and
the associated ritual material into more of a social context through associated residence patterns.
This phase has also been also called the
Sunken Court Tradition by Stanish (1994b) because each of the larger sites in this time period
has a sunken court, often with an associated
raised platform mound. While they are more
common in this phase, sunken courts began in
the Middle Formative (around 1000 BC), developing from walled ritual burial enclosures. We
know that this sunken enclosure tradition was a
powerful and important form of ritually marking sacred space and territory, allowing for reunions associated with the ancestors. As such,
sunken enclosures continue today, as seen on the
island of Amantan (Niles 1987b; Spahni 1971).
Such continuities in the Andes resonate with the
long dure of ancestral importance of land
claims, fertility, and local power.
SITE DISTRIBUTIONS
When Bennett wrote his article surveying the archaeology of the Lake Titicaca Basin in 1950, he
mentioned six sites, five of which were dated to
the Formative by diagnostic pottery sherds.
While today we know of at least a dozen central
sites with sunken enclosures potentially dating
to this phase, there should be many more as yet
unreported sites (Figure 5.5).
A center is defined as a site with a sunken
enclosure that also may have a raised platform
mound. Sunken enclosures are not found at every site, as is the case at the site of Alto Pukara
(Beck n.d., 2004). Clusters of smaller sites, probably farmsteads without special architecture, are
found across the landscape, as in the site of San
Bartolom-Viscachani (Herhahn 2001; Stanish
and Steadman 1994).2
The centers hug the lakeshore and often are
in view of the magnificent snow-peaked range
of mountains to the east, still important deities
today. At least by the end of this phase, the
larger sites were regularly spaced around the
lake shore in more densely populated regions
like the Taraco, Copacabana, and the Huata Peninsulas, as well as along rivers.3 Charles Stanish
(1997) and Clark Erickson (1996) on the western
and northern shore, Eduardo Casanova (1942)
and Karen Chvez and Sergio Chvez (1997) on
the Copacabana Peninsula, Carlos Lmuz (2001)
[FIGURE 5.5 ]
70
CHRISTINE A. HASTORF
CHAPTER 5: THE UPPER (MIDDLE AND LATE) FORMATIVE IN THE TITICACA REGION
71
72
CHRISTINE A. HASTORF
CHAPTER 5: THE UPPER (MIDDLE AND LATE) FORMATIVE IN THE TITICACA REGION
73
74
CHRISTINE A. HASTORF
CHAPTER 5: THE UPPER (MIDDLE AND LATE) FORMATIVE IN THE TITICACA REGION
Zamora 1967). At Corralpata there are two platforms, one of which is within terrace walls (Paz
Sora 2000). The Tambo Kusi, Santiago de Huata,
and Titimani centers probably expanded
throughout this subphase. Their stone anthropomorphic, camelid, arms, and serpent designs
seem most similar to the Chiripa (PAjano) style.
Unfortunately, these sites do not have absolute
dates. Portugals ceramic, site, and stone descriptions place them as Early Upper Formative
sites. Based on a theme of sacred deity fertility
rituals, their pattern also substantiated a model
of the independent but communicating alliances
around the lake, much like the Sillumocco, Copacabana, and Qaluyu clusters.
ECONOMY
Subsistence at these sites would have been focused on lake resources as well as on the beginning of raised-field shoreline farming and
herding (de France 1997; Erickson 1996; Kent
1982; Moore et al. 1999; D. Steadman 1997). The
centers in this time demonstrate evidence for
long-distance trade in permanent goods, which
some centers had begun to include in their
forms of inter-group competition (Bandy 2001a).
There were also some inland centers on rivers during this phase, Qaluyu being one example. While these inland centers might have
focused on herding, they also might have been
involved in trade or the production of special
goods (such as at Tambo Kusi [Corralpata]),
trading into the eastern slopes (Faldn Arancibia
1991; Paz Sora 2000).
Often considered evidence for increasing political consolidation, exotic, imported goods,
most notably copper (from an as yet unknown
southern source) and obsidian (from the Colca
Valley to the north), are found at some of these
sites (Browman 1980; Burger et al. 1998a; Burger
et al. 1998b; Osgood Brooks and Glascock 1996).
Although currently lacking artifactual evidence,
we could hypothesize that exotic plants used in
hallucinogenic rituals were brought in as well at
this time, as is illustrated in Cupisnique ceramics in the Early Horizon times and at San Pedro
de Atacama in the Middle Horizon. Plants such
as coca (Erythroxylum sp.), Anadenanthera (A.
colubrina, A. peregrina), and tobacco (Nicotiana
rustica) surely would have been present in the
75
76
CHRISTINE A. HASTORF
At Chiripa, both Browmans and TAPs excavations provide evidence for farming and herding (Browman 1978a, 1980, 1986; Hastorf et al.
1997; Moore 1999; Whitehead 1999a). The Chiripa middens have yielded plant and animal food
remains. The domestic plant material includes
Chenopodium seeds and tubers, and both Solanum
and Scirpus from the lake. To date, only one instance of maize has been found at any of these
sites. Lee (1997) reports one kernel and one
glume uncovered at Chisi, which dates to
around 200 BC. The many large animal bones are
primarily from camelids (Kent 1982; D. Steadman 1997). There is a tremendous amount of fish
bone at Chiripa as well as in the Copacabana
sites, supporting the interpretation of a strong
lacustrine emphasis in the diet (de France 1997;
D. Steadman 1997). Both areas have many birds,
but these are represented more frequently in the
Chiripa collections than on the peninsula overall
(de France 1997). At this stage in the research,
the same plants and animal taxa turn up in both
ritual and domestic settings.
On-site cooking and eating at Chiripa is suggested by the recovery of serving bowls and locally made cooking pots that are burned on the
outside (L. Steadman 1999, 2002). Chipped
stone, mainly of local origin, is also common,
with debitage occurring in the middens and on
the sites surface (Matthew Bandy, personal
communication). There is regular evidence for
stone adzes at the site, which would have been
used in woodworking, among other things
(Mathews 1992b). This range of material suggests the completion of daily tasks at Chiripa as
well as at the other Upper Formative sites, sites
that include both ritual centers and farmsteads.
Evidence for the beginning of intensive
farming at Chiripa is seen in the nearby raised
fields (Graffam 1990) as well as in artificial terreplanes that reach out into the water. Bandy
(2001b) reports that there was a large influx of
stone hoes from the north onto the Taraco Peninsula at this time. The northern lakeshore region
of Pancha also has early evidence of raised
fields, as shown by the detailed excavations of
Clark Erickson (1996). Intensive herding is indicated at Chiripa by the absence of evidence for
large wild animals in the faunal assemblage
(Moore et al. 1999). Ecological evidence suggests
CHAPTER 5: THE UPPER (MIDDLE AND LATE) FORMATIVE IN THE TITICACA REGION
FIGURE 5.6. The Chisi sunken enclosure that has been completely excavated and reconstructed by the Yayamama
Archaeological Project
77
78
CHRISTINE A. HASTORF
CHAPTER 5: THE UPPER (MIDDLE AND LATE) FORMATIVE IN THE TITICACA REGION
FIGURE 5.7. Plan of Upper Houses at Chiripa with central enclosure and two entrances (Hastorf 1999b)
79
80
CHRISTINE A. HASTORF
Huaricoto structures at a number of Casma Valley Initial and Early Horizon sites, dating up to
200 BC (Pozorski and Pozorski 1996). A structure
with a central hearth and benches has also been
reported in the south coast Acar Valley at Hachas (Riddell and Valdez 1987:8; Silverman 1996:
116). Continuing on in time, there are a series of
later buildings that suggest similar activities as
well, most notably Enclosure 2 at Pukara and the
Putuni at Tiwanaku (K. Chvez 1988).
One last important point about the ceremonial precinct at Chiripa during the Upper Formative time is that the mound could have had at
least two and possibly four contemporaneous
sunken enclosures during the lifespan of these
Upper Houses. The Amantan Island in the
northern lake (Figure 5.5) continues to have a
similar phenomenon today, with a male and a
female enclosure simultaneously being used for
community rituals (see Arkush, Chapter 14, this
volume; for the Pachamama and Pachatata on
Amantan see Spahni 1971; Niles 1987b). Such
dualism in ritual architectural designation at
Chiripa is illustrated by the two structure types
as well as by the two groups of Upper Houses.
While this dualism may have originated during
the Early Formative and continued throughout
the Upper Formative, only further excavations
lower in the mound will clarify this.
The continuing evidence of the type of sacred storage discussed above is illustrated in the
Upper Houses at Pukara. Nakandarari excavated a truncated step pyramid in sector BG
(work reported in Wheeler and Mujica 1981:29).
This is quite a small structure and is most like
the Upper Houses of Chiripa. Most important,
however, is that there were two niches in this
Upper Formative (Epoch III) structure, each
holding stone sculptures (one a human body
and the other a head) (Wheeler and Mujica
1981:29). Both of the plastered and painted statues had been placed in the niches that were then
filled in. The stone styles are of a type associated
with the transition from the PaAjano/
Yayamama tradition into the Pukara-Tiwanaku
stone style. There are probably other structures
and even simple sunken enclosures waiting to
be uncovered at Pukara that will demonstrate
the redundancy we see at Chiripa. This northern
Formative tradition has been defined primarily
by two ceramic surface treatments (incised redon-cream), and one common form (the flat-bottom bowl) (Franquemont 1986; Rowe and Brandel 1969). While these ceramic types share a lip
shape with later vessel forms, they reflect an
early stage of the ceramic evolution at the site.
These ceramics are less similar to the contemporaneous late Qaluyu ceramics (Mujica 1987).
ICONOGRAPHY
Ceramic trumpets with modeled felines found at
the Upper Formative centers also suggest elaborate rituals within the courtyards (Bennett 1936;
K. Chvez 1988; S. Chvez 1992). Perhaps most
significant, although difficult to date, is the evidence for the surreal, carved stone stelae that
stood within these enclosures (Browman 1972;
K. Chvez 1988, 1997; S. Chvez and K. Chvez
1975; Lmuz 2001; Portugal Ortz 1981, 1989).
The literature has drawings and photographs of
the wonderful stone carvings found around the
basins shores.
Although somewhat speculative, there is
some logic in the seriation that is discussed by
both Browman and K. and S. Chvez. It seems
that the first stone figures were made on irregular stones or slabs with what Browman calls
Asiruni figures (Figure 5.2; Browman 1972).
These have been found in both the north and
south of the basin. The images include the frog,
lizard, snake, and suche (catfish), all of which involve water-earth and amphibian fertility images. These are ethnographically associated with
sympathetic agricultural fertility magic (Denise
Arnold, personal communication 1999; Browman 1972:Figure 1; S. Chvez and K. Chvez
1975:Figure 11). The images suggest mythic
power and mystery, but are not in any way surreal or fearful. They elicit moisture and fertility
and are most likely associated with the growing
cycle, the lacustrine world, and Pachamama.
Chvez and Chvez conservatively date these to
pre-Pukara. However, they seem to be earlier,
beginning in the Chiripa phases some time after
800 BC.
Increasingly surreal, anthropomorphic faces,
in addition to animals (especially camelids),
were added on to the imagery of these sculpted
stones through time in the Upper Formative.
The most notable classic images are from the
CHAPTER 5: THE UPPER (MIDDLE AND LATE) FORMATIVE IN THE TITICACA REGION
81
82
CHRISTINE A. HASTORF
POLITICS
What do these bits of data tell us about the political nature of these times? While there is a clear
amplification of more formal ritual ceremony
around the lake, we have less direct evidence for
the associated politics. There is a kin-based and
perhaps even family ritual focus, as seen in Chiripas fourteen individual structures. These larger
groupings were formed into two ayllus or moieties, most likely representing some sort of an extended lineage. Such structures, all at one
location, could mean that the Chiripeos developed a stronger nested social order than elsewhere in the basin at this time. In this scenario,
certain family lines maintained a structure in the
sacred ritual precinct along with its associated ritual cycle and social interaction. With affiliation
came communal obligation. How goods were
produced remains unclear, but it is possible that
since agricultural production was more intensive,
labor was extracted, at least periodically, and used
in part on the local raised fields, especially as the
lake began to retreat around 450 BC. These Upper
House mound structures were clearly well maintained, rebuilt often, and replastered even more
often (Hastorf et al. 2001; Hastorf 2003). These
participatory rituals and feasts would function to
cement social units, intensify social authority, provide links to the regional ancestors, and also demonstrate generosity and piety in propitiating the
ancestors. The in situ upright stones (as with the
Asiruni figures at Arapa, Chiripa, Chisi, Santiago
de Huata, and elsewhere), the stone female or
plant conopas (as at Titimani), and the trumpets
that were used in ceremonies all suggest that such
rituals took place inside the enclosures. Such an
escalation of group consolidation is further seen
in the new ritual imagery and paraphernalia that
entered the region, the more elaborate ceramics,
and the increased presence of exotic goods like
obsidian and copper. Due to lack of systematically collected data on economic and political systems, the extent of regional economic production
centralization is not yet clear. The subsistence and
craft production data we have thus far suggests
that people were quite decentralized and community-sufficient. There is little to no evidence for
specialization. The notable exception involves
hoes that were mined, formed, and traded from
Incatunuhuiri (Frye and Steadman 2001).
CHAPTER 5: THE UPPER (MIDDLE AND LATE) FORMATIVE IN THE TITICACA REGION
83
84
CHRISTINE A. HASTORF
CHAPTER 5: THE UPPER (MIDDLE AND LATE) FORMATIVE IN THE TITICACA REGION
(Portugal Ortz 1981), dating by the style of image is not accurate enough. We await further research along the eastern side of the lake to fill in
information about this region.
In the southwest, we now have evidence of a
settlement with civic structures at Tiwanaku.
Ponce Sangins (1971) calls this phase the village Tiwanaku or Tiwanaku I phase, suggesting that the site was still comparatively small.
The recent Formative evidence from Kolatas
and Vranichs projects place the size of the settlement between 1.0 and 2.0 km2, spread across the
center as well as toward the river in the Kkaraa
sector of the site (Janusek 2003). Ponce Sangins
first identified this phase, called Qalasasaya,
based on thirty-five whole ceramic vessels recovered from burials, which he uncovered in his
1957 excavations while he was refining Bennetts
test trench results. There has been some controversy about the stratigraphy and dating of this
early Tiwanaku sequence, but the ceramic types
that are associated with the phase have been
found in other parts of the site, as well as at
other sites in the region, and thus both Tiwanaku I and III are still considered real phases.
From this Qalasasaya phase we have PaAjano
figures in a sunken enclosure that suggest that
this site also joined in the same basin-wide ceremonialism. The same range of artifacts spans the
adjoining valleys to the north and south of the
Tiwanaku Valley. From the Tiwanaku middleand lower-valley surveys we learn that there
were some fairly major settlement pattern
changes, consisting primarily of people moving
downslope and into fewer settlements, many
close to Tiwanaku proper (Albarracin-Jordan
and Mathews 1990; Mathews 1995). The secondary site of Tilata was sampled by excavation and
found to be occupied during these later Formative times (Mathews 1992b, 1995). Kantapa and
Quesani are Tiwanaku Valley sites with continuing occupation (Albarracin-Jordan et al. 1993).
The most detailed domestic Formative excavations at the shoreline site of Lukurmata have
been by Marc Bermann (1994). His earliest excavated houses dated to this Tiwanaku I phase.
Lukurmata is considered to have developed into
an important Tiwanaku center, but its economic
focus at this point was probably based on its own
85
86
CHRISTINE A. HASTORF
ECONOMY
The evidence we have suggests that the economic production continued to intensify over
this time period around the lake shore. The
raised fields near the lake nearest to Pukara
were expanded (Erickson 1988a, 1996). Recent
Koani Pampa evidence to the north of Tiwanaku
shows a notable expansion in field building during this phase of the Upper Formative (John Janusek, personal communication 1994; Kolata
1986, 1996). This increased intensive farming is
also suggested in the expanding local economic
base seen at Lukurmata (Bermann 1994; Graffam
1990). The botanical evidence at Lukurmata includes the important local domestic food crops,
tubers, and Chenopodium (Lennstrom et al. n.d.).
Where we have evidence, herding also seems to
remain important. This is seen indirectly in the
Pukara ceramic iconography where tethered llamas are being held by a front-faced female fecundity deity (S. Chvez 1992). The overall
expansion of settlements into the valleys suggests a gradual increasing focus on pampa farming during this time. At Pukara, there might
have been craft specialists, especially for ceramic
production of decorated liquid holding vessels
and trumpets (S. Chvez 1992).
The only excavated evidence for domestic
houses comes from Lukurmata. The ovoid,
somewhat irregular houses that Bermann (1994)
uncovered seem to have packed clay floors and
were approximately 3.0 x 5.0 m in size. These
houses have indoor hearths, providing us with a
sense that people lived and processed food in
small family units. Experimental production and
harvesting of the agricultural fields suggests
there were permanent habitation near the highyielding fields (Erickson 1988a, 1996; Janusek
2001). From the layout of the individual fields in
evidence across the pampas, we do not know if
there was lineage land ownership, if herds were
owned and inherited by individuals, or if all
productive resources were overseen by ayllus
(larger groups).
RITUAL ARCHITECTURE
A new architectural feature constructed in this
phase was the stepped platform, seen most
clearly at Pukara (Figure 5.8), with smaller versions at Sillumocco-Huaquina and Titimani.
[FIGURE 5.8]
CHAPTER 5: THE UPPER (MIDDLE AND LATE) FORMATIVE IN THE TITICACA REGION
87
88
CHRISTINE A. HASTORF
Chiripas small buildings, holding the most sacred items, including the ancestors themselves
for a series of different groups brought together
in one place. Much like the Inca, the Tiwanaku
might have brought in the most sacred ancestors
of conquered or allied peoples in order to unify
the spirits and therefore their ideologies. The
larger scale of this enclosure at Tiwanaku, like
the six enclosures at Pukara, is larger than the
one mound plus the three or so sunken enclosures at Chiripa, providing us with a solid trajectory of increasing ritual-political materiality.
Although the rituals could have still been kinship-based, they became much more elaborate,
with more fancy (and more labor intensive)
buildings and a larger space in which to congregate.
The Qalasasaya structure suggests political
ritual influence well beyond a single Chiripa
polity, as it would have incorporated a population that spanned many villages. The change in
ceremonial scale parallels the change seen on the
coast at about 1500 BC (Moore 1996b), although
in the Titicaca Basin this shift occurs around 200
BC. Exclusion in ritual space is another feature
shared with coastal residents. It is as if the ceremonies were not accessible to all as only a select
portion of the populace could see the event.
What was transpiring in the smaller communities as they were incorporated into this new
level of commitment and interaction?
It is likely that the rural residents continued
to have smaller versions of these feasting,
drinking, musical, ancestor-worshiping, fertilityrequesting rituals in ceremonial spaces and on
hilltops. With time, such locations eventually
were built up into platforms or built down into
the earth. Certain kin groups and settlements
gained local prestige and therefore access to labor
for such construction. While each community
contributed in its own way, they also probably
made periodic visits to the extant centers to gain
spiritual inspiration and cultural identity, returning home to reproduce at least portions of the increasingly elaborate ritual complex.
ICONOGRAPHY
At Pukara, mica and sand-tempered cream-onred ceramics persist, while accompanying new
polychrome, incised wares (S. Chvez 1992). The
CHAPTER 5: THE UPPER (MIDDLE AND LATE) FORMATIVE IN THE TITICACA REGION
89
90
CHRISTINE A. HASTORF
CHAPTER 5: THE UPPER (MIDDLE AND LATE) FORMATIVE IN THE TITICACA REGION
91
92
CHRISTINE A. HASTORF
(Albarracin-Jordan et al. 1993; Girault 1977; Portugal Ortz and Portugal Zamora 1975, 1977). So,
while the overall settlement pattern for the Tiwanaku Valley shows aggregation at the center,
there were still people moving around the valley.
Along the western shore, these were the final
days of the substantial polity centers of Sillumocco-Huaquina, Palermo, Tumatumani, Incatunuhuiri, and Ckackachipata (Stanish et al.
1997). After 900 years of peer-polity interaction,
competition, and nested identities in the north,
Pukara was abandoned, and the northern basins political web faded. The causes of this demise are almost unstudied at this point, if we
assume that it was not Tiwanakus hegemony
that initiated the fading of the multi-group ritual
center. Is this demise linked to the AD 100
drought? Is it linked to political shifts on the
west coast and to the north in the greater Cusco
region, or to new pressures from the east? Were
trade routes for important ritual paraphernalia
cut off, influencing both the economics and the
rituals? This seems to be an important period of
study for the future.
ECONOMICS
On the brink of expansion, with all ingredients
present, Tiwanakus religious cult, like Pukaras
before, was integrating the nested supra-ayllus
expandable local economic base (evident in
raised-field farming by the lake, herding in the
dried southern pampas, and the agricultural terraces along the hillsides). Fishing in the lake,
along with caravan trading (marine shells were
carried into the region), also took place during
this subphase (Kolata 1986). This diverse subsistence base was not only productive, but could
provide surpluses and therefore specialization.
Long-distance trade probably became more regular during this phase, with trade networks extending to the coast, the north, the east, and the
south, and involving such items as iconography,
ritual esoterica, copper, gold, silver, obsidian,
semiprecious stones (malachite, turquoise, and
lapis lazuli), tropical wood, marine shells (Browman 1980), coca, hallucinogenic drugs, and
probably spices (such as chile peppers, although
we have no material evidence for this yet).
The ceramic traits associated with the Qeya
phase include a lack of complete burnishing and
CHAPTER 5: THE UPPER (MIDDLE AND LATE) FORMATIVE IN THE TITICACA REGION
93
CONCLUSIONS
This 1000-year phase saw the evolution of a mix
of indigenous political-religious structures that
were interwoven within a religious world infused
with external exotic beliefs and images among local ancestral rituals. These beliefs united people
with their landscape while increasingly making
them feel similar. These dynamics worked to
bring people together, yet fueled them with potential competing power nexus emanating from
the local families as well as from the ritual enclosures, the sites of periodic group gatherings, and
renewal of group community. This pattern
around the lake created an ebb and flow of alliances that allowed loose (acephalous) polities to
develop throughout the basin. While redundant
at many levels, the symbolic and political geography allowed some communities to expand their
influence through ritual events at certain critical
times. In many parts of the basin, the communal
feasts in small chambers (feasts meant to communicate with and placate the ancestors) changed to
larger feasts, intensifying the potential for stratification through group cohesion while developing
visual cues to power in larger enclosures. As
Pauketat (2000) notes for southeastern North
America, some people join in these building endeavors because of their belief in and desire for
these social bonds between the living and the
94
CHRISTINE A. HASTORF
dead. These activities can lead to rules that go beyond the ritual manifestations that initiated the
process.
By the end of this phase, a shift could be
seen from communality and consumption in the
creation and use of public works and collective
rituals to individuals and specialists guiding the
rituals of surplus from an inner core. Part of this
escalatory cycle was developed from increased
food yields and labor to build and maintain
more ritual centers. The place of warfare and
threats associated with this must have been ever
present, only becoming evident in the later
Pukara ceramic and stone carving imagery.
These on-the-ground activities probably expanded throughout this Upper Formative period, along with the acquisition of long-distance
goods and practices that aided the ascendance of
certain centers around the lakeshore, perhaps as
Stanish (2003) and Bandy (2001b) propose, and
as seen in the settlement locations along these
trade routes. Was it the agricultural base, the
many herds and grasslands, the fish and birding
resources of the lake, the plant resources, the access to trade or traders, or the links to the most
sacred mountains that allowed some communities to elaborate their ritual activities? Ritual authority grew at some centers, building on certain
family lineages and moieties, and shifting the
role of the shaman and the power of dead ancestors over life. This became manifest in the architecture and its expansion and elaboration at key
centers around the lake. We know that the larger
Formative centers were associated with agricultural intensification, pockets of raised fields, canals, and agricultural terraces. Control over
production was not strong until the scale was
large, as at Pukara and Tiwanaku, yet there are
patches of raised fields associated with some of
these enclosure sites that suggest a cohesion of
not only architectural construction and ritual but
also of production and trade in these earlier Formative times. Perhaps the organizational leaders
were the same as the local ritual leaders and the
shamans. Centralization was not the way, however. Rather, an overlapping culture of different
family and activity groups developed. A regularization of site architecture developed and
eventually site hierarchy also developed. The
Upper Formative was a time of growth and elab-
oration throughout the basin, not just of population but also of ritual practices and political
connectedness. It exemplifies the onset of an Andean state.
ACKNOWLEDGMENTS
Many scholars have contributed to this compilation, most prominently all of the members of the
Taraco Archaeological Project, Matt Bandy, Jos
Luis Paz, Amanda Cohen, Lee Steadman, Bill
Whitehead, and Charles Stanish, who provided
much of the western and northern data. Much of
the specific archaeological knowledge is still in
the process of being collected and thus receiving
information from all my colleagues who work
around the Titicaca Basin was crucial. The excavations at Chiripa were in part supported by
The National Science Foundation, the National
Geographic Society, and the Stahl Foundation. I
also want to thank the Archaeology Institute in
La Paz and the community of Chiripa, which enabled and enlivened our work there. I would
like to dedicate this chapter to Karen Mohr
Chvez, who completed so much fundamental
research on this time period to the benefit of us
all. We hope her memory will live on in these
sites and works, like those whom we study who
are gone.
NOTES
1.
Some label the first 400 years the Middle Formative, focusing this phase only on the rise
and decline of Pukara.
2. We are unable to locate each individual place
name on a map given the very large number of
site and place names included in this chapter.
We have instead identified on the maps the
area where the sites are found (e.g., Huata Peninsula, lower Tiwanaku Valley, etc.) and recommend that readers consult the appropriate
cited works for more detailed locations [the
editors].
3. There are two Huata (or Huatta) Peninsulas:
one in the north in Peru and one in the southeast in Bolivia.
4. The Cero Cupe referred to here is different
from the one reported on in Stanish et al.
(Chapter 17, this volume).
6.
Pukara Influence on Isla Soto,
Lake Titicaca, Peru
Joel Myres and Rolando Paredes
ally abandoned, the site was presumably recognized as an ancestral huaca during the Late
Horizon. The nature and location of these ruins
contribute an added spatial and temporal dimension to a model of an Early Intermediate period cultural expansion originating from Pukara
and extending southward. Subsequent work on
the Titicaca islands should provide additional
data to establish the extent and relationship of
these unique island complexes.
INTRODUCTION
ARCHAEOLOGICAL RUINS significant to the prehistoric cultural development of the altiplano are
located on Isla Soto in the northern part of Lake
Titicaca, Peru. We refer to these ruins as the Isla
Soto Archaeological Complex. Characteristics
that define the Soto complex include paired
sunken courts oriented along an east-west axis
and surrounding terraces that architecturally enhance the courts. In association with the sunken
courts was a partial monolithic sculpture in the
Pukara style. These features, coupled with its
prominent location overlooking the island and
lake, make this an ideal location for a ritual site.
Assigning a date to the Isla Soto archaeological complex is possible through a comparison of
the architectural features and other artifacts on
the surface. Similar sunken courts used as temples or shrines for ceremonial purposes are
found throughout the Andean zone, and particularly the Titicaca Basin. We propose that the archaeological structures on the island comprise a
limited use location built primarily for ceremonial purposes. In addition to the sunken court
complex, the northern edge of the island retains
cleared plazas situated on a bluff overlooking
the lake. In this chapter, the existence and placement of other ruins located on other Titicaca islands are noted, and the relationship with Isla
Soto of these other island sites is considered.
Based on these observations, we propose
that the Isla Soto complex as a ceremonial center
was established during the Early Intermediate
period, during Pukara times. Although eventu-
ISLA SOTO
Isla Soto is located in the northeastern margin of
the lake near Conima on the mainland. The current population of the island numbers about 100
inhabitants. It is divided into two communities
Cabrauyo and Cijataneand together these
comprise the residences of the entire population
of the island. Both Quechua and Aymara speakers inhabit the island. The island is relatively
bare, and due to the scarcity of resources some of
the inhabitants practice seasonal migrations to
other areas, particularly to San Juan del Oro to
the north. The dynamics of such migrations have
been previously studied (Collins 1988) and are
important to understanding the population demographics of the current residents.
The island entered the scientific literature in
the nineteenth century. While writing from the
sacred Inka Island of the Sun, Ephraim Squier
(1877a) commented on Isla Soto, the island visible in the distance to the north. Squier notes that
Isla Soto contained the remains of ancient architecture (Squier 1877a:333), but curiously does
95
96
ARCHAEOLOGY
The principal site, like the others on the island, is
attributed by the locals to a giant, puka calson,
who came across the waters from the south. This
site is strategically located on a leveled mound
atop a small plateau adjacent to Cabrauyo (Figure 6.2). Offset and surrounded by a series of terraces modified from the natural terrain, this site
provides a vantage to view the entire island and
the encompassing lake and mainland shoreline
in all directions. These three levels of terraces
surround an elevated mound and create a type
97
98
99
agricultural terraces in the area. They do, however, resemble terraced platforms reminiscent of
other altiplano sites, including Pukara and Tiwanaku. A structure was located between two
plazas on the principle level. This structure consisted of two compartments and was enclosed
by rocks. The contents of these units had been
removed. It is not possible to determine whether
this structure was built at the same time as the
plazas or represents a later addition. However, it
does appear that these units served as small
burial towers, or chullpas (Hyslop 1977a).
At the top of the hill overlooking the cleared
plaza was a circle of stone placed in the ground
with a break or opening to the south. The circle
measured 2.0 m in diameter. It was not possible
to determine the function of this circle. It is
likely that these remains are recent and not associated with the plazas that the site overlooks.
We found virtually no datable ceramic fragments on the surface. Because of this, our dating
of the Isla Soto Pre-Columbian presence on the
island is limited and based on the isolated sculpture fragment and the described architectural
features.
DISCUSSION
Sunken courts are not unique to the Titicaca Basin. They are found in the sierra and coastal areas of the central Andes. Although the earliest
examples from central and northern Peru are circular, rectangular versions appear in the sierra
during ceramic times at Chavn de Huantar
(Moseley 1992). In the altiplano, the long tradition
100
of rectangular sunken plazas has been well established (K. Chvez 1988; Hastorf, Chapter 5,
this volume). The importance of the sunken
court throughout the region and through time
can be seen at the site of Chiripa in Bolivia (Hastorf et al. 1999b) and later at the regional centers
of Pukara, in the northwestern Lake Titicaca region, and at Tiwanaku to the south. While the
design and features of the sunken court were
modified through time, the importance and
prominence of the sunken court among altiplano
sites remained intact. The architecture of the
sunken court at Chiripa follows a similar pattern
of niches inset into the walls, a single centralized
location situated among surrounding structures,
and a deliberate distinction between levels
above- and below-ground. Such utilization of
opposing space fits the architectural canon of
sunken courts of the altiplano by restricting the
size of the ceremonial area. The example from
Isla Soto does vary somewhat from the elements
of the Chiripa model by incorporating features
such as dual courts, terraces, and stone constructions. These features are more reminiscent of
later Pukara architecture. Despite the absence of
associated enclosing structures, the platform is
circumscribed by a small wall defining the topographic limits of the level in the form of a U. The
significance of the U hilltop orientation to
present-day Aymara has been considered. Moseley states that sanctuaries of the type seen at Soto
serve a powerful mountain spirit that influences
meteorological phenomena and that the shrine is
employed for rain-bringing rites to benefit crops
(Moseley 1992:137). Other U-shaped structures
in the Titicaca Basin include one found at Tumatumani near Juli (Stanish and Steadman 1994)
and the site of Huajje near Isla Esteves (Stanish
et al., Chapter 17, this volume).
The visual impact of the placement of this
site on the central and most prominent flat area
of the island is enhanced by a series of terraces.
The plateau has been modified and is sectioned
along a roughly north-south axis creating east
and west halves. The symmetry of this configuration implies an underlying ideology consistent
with Andean concepts of duality, as seen in the
dual courts, the division of platform ground
space into halves, and the partitioning of space,
both above and below ground.
(Paredes 1985) and its appearance here may support its utility as a satellite ritual installation
founded during an expansion southeastward
from this regional center. Moreover, similarly
constructed compounds have been recorded
across the lake on Isla Amantan (Niles 1987a).
The combined results of the site layout and the
architectural feature found on the island in
terms of dual courts, orientation, and a circumscribed platform enclosure support a Pukara influence from the northwest as opposed to strong
influences from the south. It is very likely that
investigations in the Moho region may identify
the existence of comparable structures. Based on
the resemblance between the Isla Soto sunken
court complexes and those of similar arrangements at Pukara, we would date this complex to
late Early Intermediate times circa AD 200400.
The archaeological remains of Isla Soto provide a glimpse into the place of this island
within the framework of the cultural development in the Titicaca Basin. Reports of ancient ruins on the major islands of the lake provide an
intriguing picture of their importance and placement within the cultural history of the region as
well as their continued significance to the modern population. This is especially true in an unproductive island such as Soto (Bandelier 1910).
Of interest are the similarities between ruins reported on Amantan (Niles 1987b) and those described on Soto. Their importance is emphasized
by the effort invested in these locations and reflected in the significance of their structural features and configurations. While little work has
been done to examine the extent of these finds,
Isla Soto is just one of the islands on the lake that
contains significant archaeological remains.
ADDITIONAL CONSIDERATIONS
The Titicaca islands are of considerable importance in Andean history. The most renowned of
the Lake Titicaca islands are Titicaca and Coati,
or the Islands of the Sun and Moon, respectively.
In one of the earliest recorded explorations of
the Titicaca Basin, Ephraim Squier (1877a) devoted considerable attention to these most venerated islands of the lake, which together
constituted sacred centers and held particular
prominence in Inka mythology (and see more re-
101
102
7.
Tiwanaku Expansion into the
Western Titicaca Basin, Peru
Charles Stanish, Kirk Lawrence Frye,
Edmundo de la Vega, and Matthew T. Seddon
THE TIWANAKU STATE was the largest autochthonous polity to develop in the Andes south of
Cuzco before the Spanish conquest (Figure 1.3).
Sometime in the first millennium AD, the Tiwanaku state extended its influence over a vast territory covering parts of four modern nations and
a variety of ecological zones. Tiwanaku artifacts
have been found over an area of approximately
400,000 km2, an area the size of modern California.
Defining the processes of state formation
and territorial expansion of the Tiwanaku polity
outside of its core territory stands as one of the
most important and intriguing problems in Andean prehistory. Tiwanaku expansion involves a
number of related questions. What was the nature of the Tiwanaku political economy? Was Tiwanaku an expansionist state at all, or was its
influence in areas outside of the Tiwanaku Valley a result of social, religious, or other nonpolitical factors? Was Tiwanaku a conquest state
structurally similar to the later Inca empire, or
did it expand through more subtle political and
economic strategies by establishing exchange relationships and colonies in targeted provincial
areas? If it was indeed an expansionist state,
when did Tiwanaku expand? Was its expansion
early, during, or before its Tiwanaku III or Qeya
period, or did its major expansion occur later?
These questions are central in understanding the
processes of the Tiwanaku expansion throughout the vast south central Andean region. In this
chapter, we present settlement data from the
103
104
105
THE
Tiwanaku sites are easily identified by the presence of a high density of decorated Tiwanaku
pottery on the settlements and, in some cases,
architectural modifications using Tiwanaku canons. Tiwanaku pottery in the region is recognized by very distinctive shapes and design
motifs. Certain Tiwanaku fineware pastes are
also distinctive and represent imports into the
region. The best Tiwanaku marker, however, is
pottery in the form of kero, tazn (flaring sided
bowl), and incense burner shapes. In most cases,
the pottery was locally made, but a number of
106
107
108
109
110
111
and substantially modified previous Upper Formative period sites such as Tumatumani. The
earlier terraces were rebuilt into larger ones with
the same architectural pattern (Stanish and
Steadman 1994). Testing at the site of Palermo
also indicated that Tiwanaku substantially modified this site as well. Clear Upper Formative period levels were separated from Tiwanaku levels
by a burning event, which may have been associated with the arrival of Tiwanaku in the area
(Stanish et al. 1992). The site was constructed into
an Akapana/Kalasasaya/semisubterranean temple complex typical of core Tiwanaku sites as
well as external ones (Goldstein 1993b), indicating a strong Tiwanaku interest in establishing
site layouts similar to those of the homeland.
Testing at the site of Sillumocco-Huaquina indicates a similar pattern (see de la Vega, Chapter 8,
this volume). Here, again, Tiwanaku modified a
previous Upper Formative period occupation at
the site. In this case, the entire top terrace of a terraced hill was added during the Tiwanaku period.
It is significant that there is no evidence of
Tiwanaku settlement in its earlier or pre-Tiwanaku IV phases. The lack of Qeya and earlier sites
in the region is significant. The occupation of the
region occurred only after Tiwanaku had coalesced into a powerful polity in its core territory
in the Pacajes region. The Tiwanaku state expanded into the Lupaqa region at a time when
the area was controlled by complex Upper Formative polities.
112
INTERPRETATIONS
Our understanding of the nature of the Tiwanaku presence in the Titicaca Basin outside of the
Tiwanaku heartland continues to change as we
acquire new data. Our research demonstrates
that the Tiwanaku presence along the western
lake edge was substantial. Wherever we have
done reconnaissance and survey south of the Ilave River, we find large Tiwanaku sites. In areas
where we carry out systematic, 100% full-regional coverage survey south of the Ilave River,
we find very dense clusters of Tiwanaku sites.
Settlements are heavily concentrated along the
lake edge and rivers, and dramatically decrease
in density into the puna. We have isolated more
than 100 sites in the western Titicaca region,
113
114
Between these two polar extremes of Tiwanaku polity economy exist a series of alternative
models that vary according to the degree of economic and political hierarchy in the Tiwanaku
core, and the degree to which coercive means
were or were not used to bring together different
ethnic groups and polities into a Tiwanaku
sphere of influence. Most importantly, the model
proposed by Albarracin-Jordan (1996b) argues
for a segmentary organization of the Tiwanaku
state. In this model, the Tiwanaku polity was
built on Andean sociological organizations of
ayllu and marka.
Our research reported on here helps to redefine the nature of Tiwanaku expansion and geopolitical control. Tiwanaku was not a small
version of the Inca state. It was a first-generation
state with very different, and less complex political and economic institutions. Tiwanaku control
was highly selective in nature, expanding along
roads and creating enclaves or colonies outside
of its core territory. Our data do support the
models of a centralized political organization,
suggesting that Tiwanaku coerced populations
along the western edge of lake Titicaca south of
the Ilave into their polity. In this sense, Tiwanaku is best conceived of as a expansive polity
with a restricted core south of the Ilave and Escoma areas, and a selective control of roads and
enclaves outside of this zone. In short, our current data are quite compelling that Tiwanaku
was indeed an expansive polity of some complexity, similar to Wari and Moche, but one that
did not create the same kind of administrative
organizations as the later Andean empires.
ACKNOWLEDGMENTS
The authors thank members of the Programa
Collasuyu, our friends in the town of Juli, Fresia
Sardn, Percy Calisaya, and Rolando Paredes.
We gratefully acknowledge the support of the
National Science Foundation, the Cotsen Institute of Archaeology at UCLA, the Wenner-Gren
Foundation for Anthropological Research, the
John Heinz III Foundation, the Field Museum of
Natural History, and the Dean of Social Sciences,
UCLA.
8.
Excavations at Sillumocco-Huaquina
Edmundo de la Vega
FIGURE 8.1. Tiwanaku sites in the Juli-Pomata area and the Moyopampa Complex
115
116
EDMUNDO DE LA VEGA
117
The presence of the diverse site types associated with the extensive waru warus and hydraulic system suggests that these sites were
articulated within a local agricultural system. The
differences between sites reveal the distinct role
each site played within the settlement system.
SITE DESCRIPTION
Sillumocco-Huaquina is found on the western
end of the Pampa de Moyo (Moyopampa), at the
foot of the western slope of the Zapacollo Hill.
The site is on a small natural knoll that has been
extensively terraced along its slopes and top.
The surface bears an abundant scatter of sherds
and lithics spanning several occupational periods from the Formative period to the present.
Several post-Tiwanaku funerary structures are
also found on the south slopes (chullpas and collared tombs). The site covers an area of approximately 3.5 ha, and rises up to 16 m above the
pampa. The general orientation of the site is
southeast to northwest, with its main access
probably at the southeast. Raised fields and remains of canals are found in the pampa surrounding the site, as well as terraces along the
western slopes of Zapacollo. Although the terraces have no architectural features, there are at
least four small sites associated with them, identified by surface finds as Tiwanaku.
Terraces
The site consists of five levels of irregular terraces, which correspond to the irregularity of the
natural terrain (Figure 8.3). At the southeastern
end, where the height of the knoll is only 5 m
from base to top, only three terraces can be identified, T1 being the thickest, at 2 m. On the other
hand, at the northeastern end, where the knoll
height reaches 16 to 17 m, all five terraces can be
recognized, varying in thickness between 2 m
(T1, T3b, T4c, and T4d) and 4 m (T2 and T4b).
A striking feature that is observed is the way
the terrace retaining wall follows a broken line
with angles up to 90. This can best be observed
in the T2 terrace where, at the northern and
southern extremes, they form a segment of the
Andean cross, or squared cross, with three angles pointing out and two pointing in. A similar
118
EDMUNDO DE LA VEGA
pattern with an irregular squared cross is revealed in a floor plan of T1. Both the site dimensions and the terrace configurations are very
similar to those of the pyramid of Akapana, in
the urban ceremonial center of Tiwanaku (Manzanilla et al. 1990:84). Another significant aspect
is the presence of inclined earthworks on three
of the T1 corners that seem to have functioned as
entries to the uppermost terrace. The average
size of the earthworks is 6 x 3 m, spanning an
average depth of 2 m.
Funerary Structures
A variety of funerary structures are concentrated on top of the knoll and on the south and
southeastern sectors. These include chullpas,
collared tombs, and subterranean tombs. A total
of four chullpas were identified (Tombs 2, 3, 4,
and 8) on the T3a terrace. The largest of these
was Tomb 8, with an external dia7 of approximately 4.5 m (although the tomb wall is almost
completely destroyed) and with large stone
blocks (up to 1.1 m long) strewn around it. Tomb
3 shows a double-ring rock base. The exterior
ring (3.1 m in diameter) consists of fragments of
worked stone, while the interior ring consists of
field stones (2.3 m in diameter). Some cranial
fragments and human long bones were seen inside this structure.
Tombs 1 and 2, atop the knoll, are seemingly
the most important, not only because of their
prominent location on the site, but because they
are the only ones to have been built on a large
platform (8.6 x 6.5 x 1.7 m). Tomb 1 is a collared
tomb with an external diameter of 2.4 m, delimited by seven large stone blocks embedded vertically. Tomb 2, on the other hand, is a chullpa,
recognized only by the remains of its foundation,
which measures 3.8 m in its external diameter.
Cists of subterranean tombs have been found on
T2 and T3c. Cist openings range between 1 and
1.5 m in width, and the average depth is 0.7 m.
The remaining funerary structures are collared tombs, formed by vertically arranged stone
blocks of various sizes and enclosing a circular
area with internal diameter between 1.5 and 2 m.
These funerary components of the site all seem to
correspond to the post-Tiwanaku period.
SITE EXCAVATION
The excavation strategy theoretically was to entail two stages of excavation. Initially, the excavation of test units was planned with the
119
Excavation Area 2
This area is located on the northeastern arm of
T1, and lies 5 m from the northern corner of the
terrace (Figure 8.3). Three test units were excavated in this area: one 2 x 2 m square and two 1m extensions of the north and east wall profiles.
A sequence of three levels of occupation was represented in a depth of 2.2 m: two Formative levels and one Tiwanaku level. Above these were
post-Tiwanaku fill and a cap of recent deposit.
The Formative occupation was partially exposed in the northern profile and was deposited
directly over bedrock. A wall made of large
stone blocks was seen to run east-west, parallel
to the T1 wall, which is thought to be the base of
the first retaining wall of the terrace. Formative
ceramics consist of fiber-tempered wares. Superimposed on this, the second Formative occupation was also revealed in the north profile.
Several disturbed burials were associated with
this level.
The remains of a circular structure, associated with the Tiwanaku occupation and approximately 3.0 m in diameter, was evidenced by a
fieldstone block foundation that was arranged in
two rows. The structures interior showed a red
compacted clay floor. To the west of the structure, a thick layer (5060 cm deep) of guano was
found, leading to the interpretation that the area
had been a corral relating to the habitations. This
hypothesis is reinforced by the analysis of faunal
remains, which demonstrates large quantities of
both adult and young camelids and a full range
120
EDMUNDO DE LA VEGA
121
122
EDMUNDO DE LA VEGA
(Figure 8.6) in all levels, although in the uppermost levels (2 and 3) there are also fragments of
what appears to be Altiplano style ceramics. This
is a significant find, as it helps to place the terrace
construction into a chronological framework.
The presence of Tiwanaku and Altiplano ceramics in the fill eliminate the possibility that the
construction date from the Formative period.
The absence of Altiplano ceramics in the lower
level (where only Formative and Tiwanaku
sherds are found) leads us to conclude that the
terrace was built during the Tiwanaku period.
The fill content includes an abundance of lithic
materials (primarily hoe fragments and lithic debris, consisting primarily of andesite and, less
frequently, basalt, as well as quartzite, red jasper,
and chalcedony) and sandstone and rhyolite
pebbles. The faunal remains reflect a heavy reliance on camelids (99% of the bone recovered is
camelid bone), and include vertebrae, ribs, scapulae, and phalanges. These faunal remains represent individuals ranging in age from neonate to
juvenile and adults. The bones had been cleaned
and burned; some exhibit butchery marks. The
remaining 1% of the faunal assemblage is unidentifiable, but could possibly be rodent or bird.
Stratum B is a thin, compacted lens of
brown (7.5YR5/2) sandy clay, mixed with gravel
and slivers of charcoal. It is found beneath the
rock wall and functions as a foundation for the
wall. It was not excavated. Stratum C is a culturally sterile level of natural soil that corresponds
to bedrock. It consists of a red clay (2.5YR5/6)
that is compacted and homogeneous.
[FIGURE 8.6
In Stratum B, Formative, Tiwanaku, and Altiplano ceramic fragments are mixed in the deposit, the most notable of which is a kero
(drinking cup) base retrieved from the wall base.
Additionally, hoe fragments, andesite, quartzite,
and basalt debris, and fragmentary camelid remains were found in this stratum.
Stratum B is defined as terrace fill. It consists of a brown (7.5YR5/2), compact, sandy
clay, with 1- to 5-cm pebbles, and contains charcoal fragments and ceramics. Its average depth
is 0.1 m. The T1 retaining wall rests directly
upon this level. Although infrequent, Formative
and Tiwanaku sherds are present. Andesite
flakes continue to be present. However, other
raw material types (quartz, jasper, and rhyolite)
are also represented.
Stratum C is again the natural sterile deposit. Compact and homogeneous red clay
(2.5YR5/6) predominates, although there are a
few whitish inclusions. The excavation of this
level was abandoned at a depth of 0.7 m.
Retaining Wall East of the T1 Terrace. The orientation of this wall runs southeast to northwest,
with a length of 22.5 m and a width between 0.4
and 0.5 m, depending on the dimensions of the
stone blocks. Based on the current terrace height
(0.8 to 1 m), it is calculated that the original wall
height was 2 m. The terrace construction seems to
have been done over the red clay natural deposit
by first digging a trench to use as a foundation. A
clay mixture was placed at the bottom of the
trench before arranging the stone blocks over it.
The wall consists of a single row of masonry
comprised of irregular blocks of various sizes:
large blocks (1 x 1 m, 90 x 50 cm, and 60 x 40 cm),
medium-sized blocks (25 x 12 cm, 40 x 20 cm,
and 20 x 12 cm), and small blocks (15 x 8 cm and
10 x 5 cm). The small blocks seem to have been
used as shunts to plug the empty spaces left between larger blocks. The stone blocks seem to
have been paired randomly and are cemented
with clay. Irregular blocks were primarily used,
although some appear to have been lightly
shaped and others have worked or cut surfaces.
Preliminary interpretations of these data enable us to define two important things: (1) construction of the uppermost terrace took place
during the Tiwanaku period, and (2) fill material
FIGURE 8.6. Test Unit 1, Formative and Tiwanaku period pottery fragments
123
124
EDMUNDO DE LA VEGA
Excavation Area 3
PROFILE 2
Excavation Area 3 is situated by the wall of the
T3 terrace at the corner where heavy machinery
may have truncated the retaining wall and
dragged the fill some 2 to 3 m toward the interior of the terrace (Figure 8.3). A highly impacted area was selected to be cleaned,
excavating a 1.6 x 1.1 m (at surface), 0.3 m (at
base) trench. Several levels were distinguished
within five strata (Figure 8.7).
Stratum A (Level 1) consists of an agricultural surface. The soil is a homogeneous, semicompact, brown (10YR4/3) sandy clay with
gravel (13 cm), pebble (610 cm), and rock (> 10
cm) inclusions and many roots. The average
depth is 0.25 m, and ceramic fragments reflect
the Formative, Tiwanaku, and Altiplano periods.
[FIGURE 8.7
Resting directly atop a natural deposit, its average thickness is 0.2 m. Because of its stratigraphic position, no evidence of walls or floors
was found, although this stratum seems to correspond to the earliest occupation of this sector
of the site. It is interesting to note that the surface of this level is so flat and even that it appears to have been leveled. Fragments of
Formative ceramics were recovered, but the majority of ceramics recovered are plain and nonclassifiable. Nonetheless, the fiber and mica
tempering in the ceramics recovered is suggestive of the Formative period. It seems that the
content of this level is randomly mixed.
For Stratum E (Level 6), the natural deposit
yields two well-defined and distinct superimposed levels. The uppermost, E1, is a coarse
sand (15 mm grain size), generally compact
and homogeneous, reddish gray in color (5YR5/
2), and including some small (13 cm) cobbles.
The lower level, E2, consists of compact gravel
and rolled cobbles that could possibly have been
the source for the cobbles in Stratum C. Neither
of these layers yielded any evidence of human
occupation.
The data from Profile 2 confirm two observations made from the excavation of Profile 1: (1)
that overall, the terrace construction occurred
during the Tiwanaku period, and (2) that construction materials consisted of recycled earlier
deposits. At the same time, two new observations can be made: (1) it is possible that the terrace was constructed directly over previous
occupations (Stratum D), which were leveled
prior to construction, and (2) natural rocky deposits were utilized for construction materials
(Stratum C).
Excavation Area 4
This excavation area is situated along the western slope of the knoll (Figure 8.3). The excavation units were placed on the T4b terrace with
the objective of finding evidence of an occupational component on the lower terraces. Although two test units were excavated, no
significant discoveries were made. On the basis
of the relative frequency of Formative ceramics
in the lower levels, it can be postulated that a
significant Formative occupation would likely
be found in this area, either because of the great
125
[FIGURE 8.8 ]
126
EDMUNDO DE LA VEGA
wall fragments. The thickness of this varies between 0.1 and 0.15 m. For the most part, the ceramic assemblage reflects the Formative period,
although various Tiwanaku elements are also
present. Faunal remains and lithic materials decrease sharply in this stratum compared to the
concentration of these found in Stratum A. Stratum C is a sterile, natural rocky deposit of a light
brown (2.5YR6/2) color.
TEST UNIT 2
In Area 4, the 1 x 2 m extension of the eastern
wall of Test Unit 1 was designated as Test Unit 2
and excavated to follow a rock concentration
that had appeared in Test Unit 1 and was
thought to be a collapsed wall. The content of
Excavation Area 5
Excavation Area 5 is situated on the northern sector of T2 below Area 1 (Figure 8.3). This location
was selected for comparative purposes, in order
to contrast T1 with T2 and to define the functional
or occupational differences that might exist be-
127
FIGURE 8.9. Stratigraphic profiles of Units 1 and 2 in Area 5, east and west profiles
128
EDMUNDO DE LA VEGA
3 and part of Level 4, however, fragments of Tiwanaku-like wares (in terms of paste and finish)
or Late Intermediate wares were also found (Figure 8.11). The lithic collection includes andesite,
basalt, and quartzite debris, as well as stone
hoes and adzes. An abundance of camelid and
fish remains was also recovered.
Stratum B is a gravel stratum mixed in with
Stratum B that blends into the underlying bedrock. With a depth between 15 and 30 cm, it is
dark grayish-brown (10YR4/4) color. A collapsed human cranium was found on the rock
surface, covered by this gravel, and encrusted in
the northern profile. It is possible that the skull
is associated with a disturbed burial from Stratum C.
Stratum C is an occupational level wedged
between Strata B and C. A semicircular stone
configuration contained the partial remains of a
looted burial; only three vertebrae, a pelvis, the
left leg and the articulated bones of the left foot,
and fragments of the right leg were present. The
cranium and other disarticulated bones were
strewn outside the tomb. Formative ceramics
were found in association with the burial, including a clay mask. A compact gray (7.5YR6/4) clay
lens, thought to be a floor, was found in this
level. Clearly seen in the south and west profiles,
it showed a variable thickness between 5 and 20
cm. Slightly below this, but still in association
with the clay lens, was a 10-cm (average) lens of
ash (Stratum C) containing Formative ceramics.
This level lies directly over the Stratum C fill and
is directly associated with it. The first fill that was
used for terrace construction for this occupation
seems to have been severely disturbed and totally covered by the second fill (Stratum B).
Stratum C (Levels 6 or 711) shows that the
first terrace construction phase consists of Fill 1,
a light brown (7.5YR5/2) compact, sandy clay,
with some air bubbles. The matrix has intrusive
pockets of clayey soil, mixed with ash (C-1, C-2,
C-3, and C-4) and it is covered by large stone
blocks. This fill lies over the natural soil of Stratum D. The material associated with this consists
of Formative ceramics and a large quantity of
andesite, basalt, and quartzite lithic debris, as
well as fragments of stone hoes. Abundant camelid bones, as well as fish scales, constitute the
faunal assemblage.
[FIGURE 8.11
129
FIGURE 8.11. Formative and Tiwanaku pottery from Units 1 and 2, Stratum B, Area 5
The excavation results of this area were especially significant, enabling the identification of
the Formative period for the terrace construction
(Strata C and C). It is still unclear whether the
later Fill B, which raised the level of the terrace,
pertains to the Late Formative or the Tiwanaku
occupation of the site. In identifying Formative
130
EDMUNDO DE LA VEGA
Distribution of Tiwanaku artifacts and features suggests that the northern slopes of the upper terraces were primarily occupied; the
Tiwanaku occupation was not very extensive on
the southern slopes, perhaps reflecting the elite
nature of the occupation. Another important aspect of the Tiwanaku presence is the fact that
there were at least two phases of occupation. No
in situ remains of the first have been identified,
although it is clear that the remains were used
for extending the T1 and T3 terraces. The second
phase is recognized by a reconstruction of the
terrace complex. It has not yet been possible to
establish the chronology of these events.
Additionally, the association of terrace complexes for cultivation with the raised fields and
canal irrigation systems further confirms that
the ceremonial function of Sillumocco-Huaquina
would have necessitated administration and
control of the agricultural production. This helps
to better delineate and explain the Tiwanaku
presence at the site and Tiwanaku expansion in
the region.
Two models have been proposed. One sees
Tiwanaku as a unified state, while the other perceives it as a fragmented state. In order to address these two models we must look to the
similarities between the architecture at Sillumocco-Huaquina and the Akapana pyramid at
Tiwanaku.
MODEL OF TIWANAKU AS A
UNIFIED STATE
This model defines Tiwanaku as a state that expanded its domination and control over a vast
territory with minor political entities, and then
governed its subordinates with strict control under a central bureaucracy (Kolata 1983, 1985;
Mathews 1992a; Ponce Sangins 1981, 1991, 1992;
and see Stanish 1992, 2002, 2003; Stanish and
Steadman 1994; and this volume, Chapter 7).
Taking this perspective, the replication of sacred symbols, as in the case of the Akapana pyramid, must have been done under the direct
orders of the state bureaucracy. In the study of
states such as Tiwanaku, the public architecture
built for civic, military, or religious functions is
an indicator of the degree to which the social or-
ganization has evolved. The large-scale ceremonial architecture is interpreted as the manifestation of the economic and political power of
the state, inasmuch as it reflects the ability to organize and control an immense work force, and
it also reflects the hierarchical class structure as
well as the ideological framework of the state.
When these kinds of structures are found outside the nuclear centers, it demonstrates the extent of dominance and expansion, the degree of
political integration, and the mechanisms of social control (Goldstein 1993a).
In the study of Andean state expansion,
typical architectural features of public structures (such as ushnu, or sunken courtyards) seen
outside the nuclear area are an expression of
control over that political entity. In the case of
the Inkas, sites like Hatun Xauxa (DAltroy
1981), Hunuco Pampa (Morris 1982; Morris and
Thompson 1985), Huancay Alto (Dillehay 1977),
and Hatuncolla (Julien 1983) functioned as provincial centers or capitals, administering for the
region. For the Wari time period, Pikillacta,
Wiracochapampa, and Jinkamocco had a similar
function (Isbell 1985, 1987; Isbell and Schreiber
1978; McEwan 1990; Schreiber 1992).
Two main monumental construction styles
have been recognized for the Tiwanaku period:
(1) pyramids (Akapana, Pumapuncu, Wila
Pukara), and (2) walled enclosures (Kalasasaya,
Putuni, semisubterranean temple) (Manzanilla
et al. 1990:83). To these we must also add the important architectural features of doorways and
ramps that give access to the important spaces
(Goldstein 1993b:24).
Two expansion spheres with provincial centers have been attributed to Tiwanaku on the basis of ceremonial architecture featuring sunken
courtyards similar to the semisubterranean temple found in Tiwanaku proper. The first is in one
of the Pacific valleys, represented by the central
site, Omo, in Moquegua (Goldstein 1989, 1993b).
The second sphere is in the circum-lacustrine region, represented by Pajchiri (Bennett 1936),
Pokotia (Lumbreras 1974b), Chiripa (Browman
1978a, b), Pachatata (Niles 1988), Lukurmata
and mounds PK-5 and PK-6 (Kolata 1985), and
Tumuku (Stanish et al. 1997).
The only known replication of the Akapana
pyramid is at Sillumocco-Huaquina. The only
131
132
EDMUNDO DE LA VEGA
SUMMARY
In conclusion, the following assertions can be
made about Sillumocco-Huaquina:
The architectural similarities to the Akapana pyramid suggest that SillumoccoHuaquina had primarily a ceremonial function and the residential component of the
site was reserved for the elite. This, however, does not rule out the possibility that it
may also have served an administrative
function for control of the agricultural production on the raised fields associated with
the site.
133
NOTES
1. Politically, the site is located in the Department of Puno, Province of Chucuito, District
of Juli.
2. Translated by the editors: una jerarqua especfica; algunos realizados al nivel de los segmentos ms simples, otros al nivel de los
ndulos territoriales, y otros a nivel regional o
multiregional (Albarracin-Jordan 1996a:218).
9.
The Tiwanaku Period Occupation on the
Island of the Sun
Matthew T. Seddon
THE ISLAND OF THE SUN is located at the southern end of Lake Titicaca (Figures 1.3 and 9.1). The
island is approximately 2 km wide and 12 km
long. The topography of the island is very steep,
with five peaks that rise 200 m above the lake.
The island also sits in a unique ecological zone
within the Titicaca Basin. Due to the lake and
other local climatic factors, ambient temperatures are higher for the island than the average
for the region, permitting the cultivation of
maize (Zea mays) in quantities not possible elsewhere in the Titicaca Basin. These ecological
characteristics, along with other factors, may account for the long pre-Hispanic occupation of the
island, which spans from the Late Archaic (ca.
30002000 BC) through the Inka period (AD 1400
1532) (Figure 1.4).
During the Inka period, the island served as
an important shrine/pilgrimage area, one of the
three most important in the Inka empire (Bauer
et al. 1996; Bauer and Stanish 2001; Betanzos 1987
[1551]; Calancha 1981 [1638]; Cieza de Leon 1959
[1532]; Cobo 1990 [1653]; Dearborn, et al. 1998;
Ramos Gaviln 1988 [1621]; Stanish and Bauer
2004). The focus of Inka religious activity was a
large sandstone outcrop at the northern end of
the island known as the Titikala (Figure 9.2). An
elaborate sacred ritual complex was erected by
the Inka in this area (Dearborn et al. 1998).
A number of chronicles suggest that the island was an important ritual area for polities in
the Titicaca Basin prior to the arrival of the Inka
(Cobo 1990 [1653]:92-93; Ramos Gaviln 1988
[1621]:39). This chapter focuses on the Tiwanaku
occupation of the island. Research indicates that
[FIGURE 9.1 ]
HISTORY OF ARCHAEOLOGICAL
RESEARCH ON THE
ISLAND OF THE SUN
After the initial visits of the Spanish priests and
the chroniclers Alonso Ramos Gaviln (in the
late 1500s; Ramos Gaviln 1988 [1621]) and
Bernab Cobo in 1616 (Cobo 1990 [1653]), the island escaped the historic record until visits in
the nineteenth century by naturalists such as
Ephraim Squier (1877b). At the end of the nineteenth century, Adolph Bandelier (1910) conducted the first survey of the island. This survey
was not systematic, consisting primarily of asking the local occupants the locations of archaeological sites. Bandelier assigned most of the
occupations he recorded to the Inka period, although he did acknowledge the presence of a
pre-Inka component, which he called Chullpa
(Bandelier 1910:165). He also unknowingly provided the first published account of one of the
major Tiwanaku sites on the island, Chucaripupata (Bandelier 1910:225).
The archaeologist Alberto Perrin Pando (1954)
carried out excavations at the site of Wacuyo
(Wakuyu) in the early 1950s. Wacuyo is located at
the south end of the island in the community of
135
136
MATTHEW T. SEDDON
Yumani. Perrin Pando excavated eight subterranean tombs, all with Tiwanaku artifacts. The results will be discussed in greater detail below.
In the late 1980s and early 1990s, a project
led by Johan Reinhard of the Woodlands Mountain Institute and Carlos Ponce Sangins of the
Bolivian Archaeology Institute conducted underwater explorations on a shallow reef located
off the northern tip of the island (near the island
of Khoa) (Ponce Sangins et al. 1992). In their ex-
137
FIGURE 9.2. The Island of the Sun and major Tiwanaku period sites
In 1994, Proyecto Tiksi Kjarka began intensive archaeological work on the island (Bauer
and Stanish 2001). The project began with an intensive, total coverage, pedestrian survey of the
entire island, which recorded approximately 180
sites, including a substantial Tiwanaku occupation. To date, two of the Tiwanaku sites recorded
in this survey (Chucaripupata and Titinhuayani)
have been excavated. The combined survey and
excavation data provide us with our best body
of information thus far on the relationship of the
SURVEY RESULTS
The survey recorded thirty-five sites with surface evidence of Tiwanaku period occupation
(Figure 9.3). Virtually all of these sites were identified through the presence of Tiwanaku IV/V
kero (drinking cup) and incensario fragments on
the surface. Direct evidence of Tiwanaku III or
138
MATTHEW T. SEDDON
Qeya-style ceramics is virtually absent, these ceramics being present only at the site of Qeya
Kollu (upon which the Tiwanaku III style is
based). Thus, for the moment, the first clear and
extensive Tiwanaku period occupations can only
be defined for the Tiwanaku IV and V periods.
Tiwanaku sites on the island range from
0.01 ha to over 4.0 ha in size. Fifteen are 0.11.0
ha, seven are 1.02.0 ha, two are 2.03.0 ha, and
four are 3.0 ha and over in total occupation size.
Smaller sites (0.11.0 ha) generally consist of 50
m2 to 1 ha of light ceramic scatter, with occasional Tiwanaku diagnostics, over a series of
terraces. The larger sites (over 3 ha) generally
consist of denser scatters of ceramics over large,
usually modified, hilltops or ridges with many
descending terraces (type III sites in the survey
of the Juli-Pomata region of Peru [Stanish et al.
1997:108]). The majority of these sites cluster in
the areas of high agricultural production on the
[FIGURE 9.3]
A second major aspect of the Tiwanaku settlement pattern is the presence of a line of sites
along the southwestern ridge of the island (Figure 9.3). This line runs from the major site of
Apachinacapata to the northern, and possibly
ritual, area. This line certainly indicates the presence of a Tiwanaku period road connecting the
southern to the northern ends of the island. This
southwestern side of the island is not occupied
today, and shows little occupation in other periods. Given this, occupation along the road during the Tiwanaku period indicates that travel
along the road was heavy during the Tiwanaku
period. This suggests the possibility of processes
akin to a pilgrimage route at this time.
In summary, these settlement data indicate
an intense occupation of the island during the Tiwanaku period. Sites range from small (less than
1 ha) hamlets to a number of large settlements. A
heavily traveled road led from the southern end
of the island to the area of the later Inka sacred
precinct. Occupation was also high in the area of
the later Inka precinct, suggesting that the area
may have had a similar function in the Tiwanaku
period as well. The major Tiwanaku period occupations bear further description.
139
140
MATTHEW T. SEDDON
141
142
MATTHEW T. SEDDON
CONCLUSIONS
the ceramics were locally manufactured. The exact nature of the relationship of the site to the Tiwanaku center is discussed in Seddon (1998). It is
clear that very strong affiliations with Tiwanaku
are reflected in the adoption of the Tiwanaku
canon for both the forms and decorative motifs of
ceramics manufactured on the island.
10.
Residential Diversity and the Rise of
Complexity in Tiwanaku
John W. Janusek
ALTHOUGH TIWANAKU WAS for over six hundred years (AD 5001150) one of the most influential centers in the pre-Hispanic Andes, the site
has remained a great mystery. Early travelers
and archaeologists noted Tiwanakus extant
monumental edifices and exquisite stone sculptures (Posnansky 1945; Squier 1877a; Stbel and
Uhle 1892), but the apparent isolation of these
ruins fueled the notion that Tiwanaku was an
uninhabited ceremonial center. Nowhere were
there clear surface indications for associated
habitations, and what is more, the ruins appeared incongruously in what was considered a
barren, windswept environment. Ephraim
Squier remarked over a century ago, after traveling through the region:
We can hardly conceive of remains so extensive as those of Tiahuanaco, except as
indications of a large population, and as
evidences of the previous existence on or
near the spot of a considerable city. But
we find nowhere in the city any decided
traces of ancient habitations, such as
abound elsewhere in Peru, in connection
with most public edifices. Again, the region around is cold, and for the most
part arid and barren.... This is not, prima
facie, a region for nurturing or sustaining a large population, and certainly not
one wherein we should expect to find a
capital. Tiahuanaco may have been a sacred spot or shrine, the position of which
was determined by an accident, an augury, or a dream, but I can hardly believe
that it was a seat of dominion [Squier
1877a:300].
143
144
JOHN W. JANUSEK
CHAPTER 10: RESIDENTIAL DIVERSITY AND THE RISE OF COMPLEXITY IN TIWANAKU 145
form of society that is rooted in urbanism and incorporates an institutionalized political structure
(Cohen 1981; Wright and Johnson 1975). Complexity also refers to a high degree of social differentiation, or heterogeneity (Blau 1977;
McGuire 1983; Smith 1994:144). Social differences
may take numerous forms, but archaeologists
tend to focus on hierarchy and role. Hierarchy,
an expression of inequality (Fried 1967; compare
to McGuire 1983), is grounded in institutionalized differences of wealth, status, or power
(Brumfiel and Earle 1987; Weber 1947:152153).
Role, an expression of specialized livelihoods, is
grounded in complementary differences that ultimately foster social integration. Hierarchy itself
is considered a form of integration to the extent
that leadership and bureaucracies provide specialized services for society (see, for example,
Service 1975; Wright and Johnson 1975). Following this traditional view, complex societies are
considered highly ordered systems in which
other potential forms of heterogeneity, such as
faction, lineage, ethnicity, or gender either are
not considered or are considered primitive and
potentially disruptive survivals.
Urbanism is the concrete expression of state
organization and social complexity. Archaeologists have not yet articulated a comparative definition of the preindustrial city that is much more
systematic than Childes (1936, 1950). Central to
Childes vision is the idea that the city, though an
extensive and densely populated focus of diversity, is a social community sui generis. The reasons behind urban development may be
manifold, but urban societies are held together
by two processes: first, by organic solidarity,
through which Peasants, craftsmen, priests and
rulers form a community... largely ...because
each performs mutually complementary functions, needed for the well-being of the whole
(Childe 1950:16); and second, by ideological devices that mask or render natural the appropriation of social surplus by a tiny ruling class
(Childe 1950:16). As a diagnostic expression of
complex society, the city incorporates occupational specialization and class differentiation, but
remains an integrated community with a shared
sense of well-being. According to Childe (1950:
16), in ancient cities there is simply no room for
skeptics and sectaries.
The idea that Tiwanaku was a complex society characterized by urbanism is an intriguing
hypothesis, but one that has not yet been fully
explored at the principal site itself. Until recently, no household contexts had been systematically excavated. In this chapter, I seek to
examine the nature of Andean social complexity
and urbanism at Tiwanaku. As mentioned, complexity refers in part to the development of social relations and institutions such as status
differences, which manifest relations of inequality, and craft specialization, which manifest
functional differences. Here I compare excavated residential areas in Tiwanaku in order to
determine the extent to which social inequality
and specialized production characterized the urban settlement, and to clarify the spatial configuration of such forms of heterogeneity.
Nevertheless, an examination of Tiwanaku
urbanism may demand a deeper examination of
the nature of social complexity. Archaeological
research around the world indicates that hierarchy and specialization are not the only constituents of preindustrial social complexity (see, for
example, Brumfiel 1994a; Crumley 1995;
McGuire 1983; Smith 1994; Stein 1994). Ethnic
identity, lineage divisions, factional differences,
and other forms of heterogeneity may be endemic features of an entire social order and its
centralizing institutions. For example, the ruling
classes of the Aztec and Inka empires maintained and even celebrated kinship and lineage
principles traditionally thought to be characteristic of primitive societies (Brumfiel 1994a; Conrad and Demarest 1984; Zuidema 1990). In many
polities of Asia, Africa, and Mesoamerica (Demarest 1992; Fox 1987; Geertz 1980; Southall
1988; Smith 1994; Tambiah 1977), lineage segments and rival factions formed the basis of both
socioeconomic differentiation and political centralization. The very centralizing institutions of
such polities were rooted in kinship ties and factional competition.
Responding to the implications of such interpretations, some archaeologists have attempted
to develop alternative theoretical approaches to
social complexity. One of the most intriguing
ideas is the concept of heterarchy. Crumley (1987,
1995) and others (Brumfiel 1994a; King and Potter 1994; White 1995) have developed the concept
146
JOHN W. JANUSEK
to criticize the usual equation of order and complexity with hierarchy. Archaeologists often construct models of social complexity based on
differences in the degree of hierarchy and political centralization. But complex societies always
incorporate organizations that are unranked, or
that possess the potential for being ranked in a
number of ways (Crumley and Marquardt 1987:
163), and ranking itself is often one among other
elements of social differentiation. Heterarchy
reminds us that forms of order exist that are not
exclusively hierarchical, and that interactive elements in complex systems need not be permanently ranked relative to one another (Crumley
1995:3). As King and Potter (1994:84) remark, the
concept of heterarchy does not negate hierarchy,
it subsumes it.
I examine evidence for characteristics of
complexity in Tiwanaku that, not being overtly
functional, may have differentiated urban populations as discrete social groups. I explore group
affiliation and identity as a potential heterarchical dimension of complexity in Tiwanaku. Local
identity may be based in kin-based groups (see,
for example, King and Potter 1994) or in interest
groups and factions cross-cutting kinship or
class ties (Brumfiel 1994a). In either case, social
identity may be an important dimension of such
differences. Through time, groups may shift in
relative status or power, or an entire social order
may become hierarchical as an elite class crystallizes, but rifts of status or class may well follow
or coexist alongside ethnic, clan, kin, and other
totemic boundaries (see, for example, Brumfiel 1994a; Conrad and Demarest 1984; Fox 1987;
Zuidema 1990). Thus, a hierarchical order may
come to subsume a heterarchical social order,
but even as ranked relations shift through time,
and as states rise and fall, perduring social
boundaries may continue to form part of a societys basic organizing principles.
of radiocarbon measurements, has been established for Tiwanaku and its surroundings (Janusek 2001). This chronology builds on the
chronology originally proposed by Carlos Ponce
Sangins (1981), with certain significant alterations based on a series of critiques that have
been raised over the past ten years (AlbarracinJordan and Mathews 1990; Janusek and Alconini
1994; Mathews 1995). Ponce Sangins argued
that Tiwanaku developed over three evolutionary stagesvillage, urban, and imperialthat
he divided into five major periods (IV). In this
paper, the time span (AD 5001150) coeval with
Tiwanaku urban expansion and regional influence is termed the Tiwanaku period, which includes two principal phases: Tiwanaku IV and V.
This chapter focuses on the Tiwanaku IV phase
(AD 500800), when Tiwanaku emerged as a major urban and ceremonial center at the center of a
pan-regional polity. In many contexts we can
distinguish distinct early (AD 500600) and late
Tiwanaku IV (AD 600800) subphases. In general, though, domestic occupations and artifact
types demonstrate great continuity over these
300 years.
The beginning of Tiwanaku IV was marked
by the abrupt appearance of an entire range of
red-slipped (and to a lesser degree blackslipped) serving and ceremonial forms (Figure
10.2). Although diverse, these vessels demonstrated a relatively high degree of standardization in certain canons of form, treatment, and
iconography. Most common were keros, tazones,
and vasijas, but a number of special serving
types, including escudillas, cuencos, and fuentes,
were more selectively distributed across contexts and throughout society. Tiwanaku-style
iconography depicted themes that developed
out of earlier (Kalasasaya and Qeya) decorative
styles, but it was strikingly different in execution
and meaning. In Early Tiwanaku IV, Tiwanaku
remained a relatively small settlement (most
likely < 2 km2), and the range of specific ceramic
types and variants remained relatively low. After AD 600, in the Late IV period, Tiwanaku expanded into a major urban settlement, and
ceramic forms and iconography demonstrated
greater stylistic diversity. Certain forms and motifs became more standardized, marking the
crystallization of a Tiwanaku corporate style.
[FIGURE 10.2]
CHAPTER 10: RESIDENTIAL DIVERSITY AND THE RISE OF COMPLEXITY IN TIWANAKU 147
FIGURE 10.2. Major classes, types, and variants of Tiwanaku-style ceramic vessels
For example, most escudillas had wide rims decorated with stylized mythical imagery (Figure
10.13), while sahumadores tended to display stylized condor wings and feathers (Figure 10.8).
Nevertheless, almost every major vessel type
multiplied into a variety of forms, treatments,
and iconographic repertoire. Meanwhile, nonlo-
cal vessels formed a small but significant component of many Late Tiwanaku IV residential
assemblages. Foreign and foreign-inspired serving vessels at Tiwanaku included the Omereque
style, most common in the Cochabamba Valley
region to the southeast (Ibarra and Querejazu
1986); the Yampara style, characteristic of the
148
JOHN W. JANUSEK
Chuquisaca Valley south of Cochabamba (Janusek et al. 1997); and the Yura style, most characteristic of the highlands and valleys of the Potos
region south of Tiwanaku (Ibarra and Querejazu
1986). In most cases, nonlocal characteristics of
form, treatment, and iconography corresponded
with distinctive pastes and tempers (Janusek
1994:126128; Rivera Casanovas 1994).
RESIDENTIAL ARCHAEOLOGY
AT TIWANAKU
This project was originally conceived as an exploration in household archaeology, and one
fundamental objective was to define the material
constitution of the household unit, what I define
as a minimal coresidential social group with corporate roles (Janusek 1994:81). To this end, I
sought to encounter the the smallest artifactual
and architectural assemblage repeated over a
settlement (Stanish 1989a:11). A great deal of
research has demonstrated the promise that
household archaeology holds for investigating
economic, social, and even religious organizations in the past (Ashmore and Wilk 1988; Bermann 1994, 1997; Manzanilla 1996; Palka 1997;
Smith 1987; Stanish 1989a; Wilk 1983; Wilk and
Netting 1984; Wilk and Rathje 1982).
However, this research highlighted several
problems in doing household archaeology at a
major pre-Hispanic settlement such as Tiwanaku. First, contrary to traditional wisdom in the
Andes (for example, Bolton and Mayer 1977;
Custred 1977; Orlove and Custred 1980), the
household was not necessarily the primary or
primordial corporate unit in the past (Collins
1986; Harris 1981). Todays Andean households
are indigenous responses to five hundred years
of Western administration, markets, and cultural
values (Stanish 1992:1920), during which time
more inclusive corporate organizations, such as
ayllus (Isbell 1977:91) and lineages (Abercrombie
1986; Collins 1986), have eroded. Second, and
related to this point, although household archaeology is often considered to be primarily
concerned with reconstructing past domestic
life, in Tiwanaku (just as in major pre-Columbian centers such as Teotihuacan) the conventional divisions between domestic, ritual, and
specialized spaces and activities break down
(Ames 1995; Feinman and Nicholas 1995; Janusek 1999). Much of Tiwanaku was devoted to
activities not explicitly domestic, and residences
often incorporated ritual practices and specialized economic activities into domestic life.
Places where people slept, ate, and raised children often were the same places they conducted
periodic ceremonies or plied a particular trade.
Finally, depositional complexity at Tiwanaku further complicated the prospect of conducting an effective project of household
archaeology. Repeated cycles of building, disposal, abandonment, and reconstruction had
continuously disturbed previous archaeological
contexts. Tiwanaku was an extraordinarily dynamic site. At least 40% of excavated occupations consisted of middens and refuse pits,
many several meters in cross-section and 0.5 to
1.0 m deep. On the positive side, the sheer volume of refuse and the extent of space devoted to
depositing refuse point to Tiwanakus significance as a major center. The patterns indicate
that through time, household life was increasingly linked to the role of the settlement as a center of social and ceremonial convergence.
The theoretical and practical complications
of doing household archaeology in Tiwanaku invite a more encompassing research focus. Beyond uncovering repetitive household units, we
revealed significant diversity across the settlement, and exposed in each residential area the
remains of a wide spectrum of life spanning domestic, specialized, and ritual activities. Thus,
rather than attempting to elucidate an arbitrarily
defined group of domestic activities, I seek to
illuminate certain concrete rhythms of daily
life (see Ensor 2000). This perspective promises
to illuminate some of Tiwanakus internal dynamics and the complexities inherent in its social organizations and interactions.
CHAPTER 10: RESIDENTIAL DIVERSITY AND THE RISE OF COMPLEXITY IN TIWANAKU 149
[FIGURE 10.3 ]
FIGURE 10.3. General plan of the Akapana East sector in Tiwanaku: (1) Akapana East 1M,
(2) Akapana East 1, and (3) Akapana East 2
150
JOHN W. JANUSEK
similar to small adobe seats (patillas) in traditional Aymara houses. On the earliest surface,
two basalt cobbles with pounding scars were
found directly in front of the platform, suggesting that generalized lithic production was one
activity carried out here. Although the room
contained no obvious sleeping platforms, its
north quarter was compact and free of ash, suggesting that simple beds of textile or ichu grass
were laid along the warm wall next to the
hearth.
An outdoor patio with ten superimposed
surfaces occupied the west side of the structure.
Charcoal from the second surface provided a
calibrated radiocarbon date of AD 540 100
(Beta-5549, Stuiver and Pearson 1993), which
corresponds to the Early Tiwanaku IV period.
Changes in ceramic assemblages on the subsequent eight surfaces indicated that occupation
continued through Late Tiwanaku IV. The surfaces revealed artifacts representing a wide
range of activities, indicating that this was a key
locus of domestic activity. The first three floors
were associated with the opening for a deep pit,
the restricted globular shape of which suggests
that it served as a storage chamber. Like contemporary Aymara subterranean bins, this chamber
would have served to store dry consumable
goods, such as tubers and quinoa. During the
packing of the fourth occupation surface, the
chamber was converted into a refuse pit.
An outdoor midden and refuse zone occupied the area north of Structure 1. Layers of
greenish-gray ash with taquia and domestic
refuse alternated with more compact layers of
laminated silty clay (5YR4/3), the latter most
likely from seasonally eroding adobe walls. The
area along the north wall of the kitchen contained chunks of burnt adobe and broken ollas,
where inhabitants discarded hearth cleanings. A
great deal of the artifactual material was burned,
as in many secondary contexts at Tiwanaku, indicating that trash was burned periodically to
discourage scavengers.
The north excavation block in Akapana East
1M revealed a second domestic structure. The
foundation wall of the compound continued
through this area, broken by a staggered section
that appears to have been an entrance. Structure
2 was not as well preserved as Structure 1, and
CHAPTER 10: RESIDENTIAL DIVERSITY AND THE RISE OF COMPLEXITY IN TIWANAKU 151
152
JOHN W. JANUSEK
FIGURE 10.5. Views of areas associated with Structure 1, Akapana East 1M: (a) small kitchen area with
corner hearth, and (b) early surface of outdoor patio showing artifact distributions
CHAPTER 10: RESIDENTIAL DIVERSITY AND THE RISE OF COMPLEXITY IN TIWANAKU 153
ture 1, N7868 E5362). Several of the bones (including many vertebrae, hand bones, and feet
bones) were missing, suggesting either secondary interment or that the burial had been disturbed by later activity.
AKAPANA EAST 1
By the Late Tiwanaku IV phase (AD 600800), the
area along the west bank of the moat was the
backyard for a nearby residential compound
or group of compounds (Figure 10.3). Excavations in Akapana East 1 revealed a deep well,
several ephemeral hearths, and over twenty
deep amorphous pits filled with immense quantities of ash, taquia, and refuse. Such pits were
common throughout Tiwanaku, and they typically tore into huge sections of prior occupations.
The refuse in the pits reflected an incredibly diverse range of domestic activities. The energy
that went into the excavation of these pits, and
the amount of materials deposited in them, suggest that they served two purposes. The pits
functioned as adobe quarries that provided the
clay loam necessary to form adobe and tapia
bricks for wall construction. Once excavated, the
pits served as refuse deposits for ash and refuse
generated at the growing settlement.
The pits yielded large quantities of olla,
tinaja, and serving ware sherds, fragmented and
butchered camelid bone, broken ground stone
tools, food remains such as quinoa seeds and
maize kernels, ash, and camelid dung (taquia)
fuel. The growing center generated immense
amounts of material, including elaborate crafted
objects, domestic implements, and waste. Still,
the pits represented more than simply domestic
garbage pits. Adobe quarries in contemporary
Tiwanaku often take years to fill with ash and
refuse. Most of the pits in Akapana East 1 presented 13 strata of ash, indicating that they
were filled relatively quickly. Also, remains of
partially reconstructible tinajas and elaborate
serving wares were often strewn about a single
pit. In these massive pits the type and sheer
abundance of refuse, as well as its manner of its
disposal, indicate that they held more than the
waste generated during everyday domestic activities. They also contained the abundant refuse
generated during periodic ceremonies and
feasts.
154
JOHN W. JANUSEK
AKAPANA EAST 2
Excavations outside of the moat revealed substantial evidence for occupation during the Late
Tiwanaku IV and Early Tiwanaku V periods
(Figure 10.6). One excavation unit in Akapana
East 2 reached precultural red clay directly under Late Tiwanaku IV occupations, indicating
that this area was first occupied later than the
first occupations in AkE 1M. Much of the occupation had been disturbed by post-abandonment quarrying, deflation, and plowing. What
remained consisted of a single domestic structure surrounded by several features, occupation
surfaces, and wall foundations. All of the visible
foundations were oriented6 to 8 degrees east
of north, following the orientation of the monumental complex and the AkE 1 residential complex.
Two large foundations, supporting compound walls similar to those in AkE 1M, surrounded a single structure. Only the northeast
corner of a structure with three superimposed
occupation surfaces remained intact. The foundations consisted of adobe blocks on cobble
[FIGURE 10.6 ]
CHAPTER 10: RESIDENTIAL DIVERSITY AND THE RISE OF COMPLEXITY IN TIWANAKU 155
rounded pestle-like crushers for processing peppers, dried meat (charqui), and seeds. Butchering
also was conducted in residential compounds,
as represented by the substantial quantities of ad
hoc, unretouched, or quickly retouched lithic
scrapers and knives, and by butchering marks
on the bones of camelids, guinea pig (cuy), vizcacha, and birds. Dry foods were stored in bins
and in underground pits with small openings,
and liquids were fermented and stored in jars
and large tinajas. Cooking was represented by
hearths, deep ovens, and enclosed kitchens of
varied sizes, and also by great quantities of
cooking vessel sherds, taquia fuel, and ash.
Other common domestic activities included the
production of bone and stone tools, which was
represented by significant quantities of lithic
debitage and cut and splintered bone. Surprisingly, while spindle whorls were prevalent, evidence for weaving was uncommon in Akapana
East.
Two common implements served specific activities that remain unclear (Figure 10.7). Mandible tools consisted of the ascending ramus and
posterior body of a camelid mandible, which
was broken off just behind the last molar and intentionally polished (Bermann 1994:Figure
12.10; Janusek 1994:Figure 7.7). Several examples from the Moquegua Valley of Peru were
hafted to a wooden handle (Goldstein 1989:Figure 50). The production of these bone implements was a general domestic activity in
Tiwanaku. They apparently were used as cleaning or burnishing tools, perhaps for camelid
hides in the production of leather goods. Conical
stone or fired clay objects, known locally as
trompos, were also common in Tiwanaku domestic contexts, and they often appear in small
caches. They may have served as plugs for small
container openings, or as weights in the exchange of foodstuffs or other goods.
Domestic life in Tiwanaku also included a
wide range of activities not restricted to the material reproduction of the household. Ceramic
sahumadores were used to burn some substance
with high lipid content (Michael Marchbanks,
personal communication 1991), which may have
been resin or camelid fat (Tschopik 1950:208)
(Figure 10.8). Most likely, sahumadores served
both quotidian and ritual ends. Like similar
[FIGURE 10.7 ]
156
JOHN W. JANUSEK
1992:51). Fetal burials in Tiwanaku, like those today, were undoubtedly offerings dedicated to
the well-being of a houses inhabitants. Placing a
human infant under a compound wall (Akapana
East 1M) and a llama under the corner of an individual house (Akapana East 2) accords well
with the different scales of social grouping.
While the fetal llama was likely dedicated to a
house and its inhabitants, the human baby, under the compound wall, may have been dedicated by and for the entire social group living in
the compound.
CHAPTER 10: RESIDENTIAL DIVERSITY AND THE RISE OF COMPLEXITY IN TIWANAKU 157
158
JOHN W. JANUSEK
from partially reconstructible serving vessels, including particularly high quantities of elaborate
escudilla sherds.
The Late Tiwanaku IV occupation also differed from that in Akapana East in its association with an elaborate, stone-lined drainage
network (Figure 10.10). First encountered by a
French expedition around the turn of the century (Crqui-Monfort 1906), a primary canal of
carved ashlars sealed with clay, measuring 1 m
high and 0.9 m across, descended gently from
south to north about 1 m below the occupation
surface (Janusek and Earnest 1990; Ponce Sangins 1961:22). The canal followed an alignment
6 degrees east of north, similar to that of the wall
foundations. Numerous feeder canals, one of
which originated in an ad hoc basin set into a
Late Tiwanaku IV residential surface, drained
into the primary canal.
CHAPTER 10: RESIDENTIAL DIVERSITY AND THE RISE OF COMPLEXITY IN TIWANAKU 159
[FIGURE 10.10
FIGURE 10.10. Views of the primary subterranean drainage canal under the Putuni complex
160
JOHN W. JANUSEK
TABLE 10.1. Percentages of serving vessel types in distinct Tiwanaku residential sectors
Form
Putuni
AkE 1M
AkE 2
C. J.
Kero
36
4%
203
17%
108
19%
1628
33%
Tazon
88
11%
556
47%
305
53%
2505
50%
Vasija
275
34%
210
18%
86
15%
566
11%
Escudilla
284
35%
159
13%
22
4%
Fuente
102
13%
1%
10
2%
115
2%
Cuenco
18
2%
15
1%
26
4%
148
3%
Other
11
1%
35
3%
21
4%
Total Count
814
100%
1999
100%
578
100%
4962
100%
Putuni
64
AkE 1M
7%
190
AkE 2
14%
124
C. J.
17%
0%
[FIGURE10.2
[TABLE
10.12
ABOUT HERE]
CHAPTER 10: RESIDENTIAL DIVERSITY AND THE RISE OF COMPLEXITY IN TIWANAKU 161
FIGURE 10.11. Map and plans of excavation blocks in Chiji Jawira, including (1)
northeast block with burnt surface and pit kiln (Ar ea A), and (2) north block with
remnant adobe foundations. After Rivera Casanovas 1994:Figures 8.1, 8.12, and 8.27.
clay nodules, misfired wasters, and slumped vessels. Regarding primary firing contexts, excavations revealed only baked open enclosures and
small pit kilns (Franke 1995; Rivera Casanovas
1994), indicating that firing procedures were relatively informal, as they were in other pre-Hispanic complex societies (Balkansky et al. 1997;
Janusek 1999). Archaeobotanical analysis revealed that the remains of fuels preferred by ce-
162
JOHN W. JANUSEK
FIGURE 10.12. View of excavation block in Chiji Jawira (facing the urban cor e), showing adobe building foundations.
Photo by Wolfgang Schuler, courtesy of Alan Kolata.
CHAPTER 10: RESIDENTIAL DIVERSITY AND THE RISE OF COMPLEXITY IN TIWANAKU 163
164
JOHN W. JANUSEK
rated. Unlike assemblages in the Putuni complex, assemblages here were dominated by
tazones and keros (64% of serving ware sherds),
while elaborate escudillas were much less frequent (13% of serving ware sherds).
Residential sectors beyond the principal
moat differed from those inside as one moved
east across Tiwanaku. In Akapana East 2 during
Late Tiwanaku IV, as in Mollo Kontu South and
La Karaa, domestic architecture was similar in
construction and organization to that in Akapana East 1. Waste and water drained into ad
hoc canals leading to outdoor streets. Most serving wares in Akapana East 2 displayed Tiwanaku-style elements, but 70% of them consisted of
keros and tazones, while escudillas were uncommon (4% of serving ware sherds). Further, in
Akapana East 2 approximately 25% of the serving wares were more roughly formed and decorated than those in Akapana East 1. They
included a consistent proportion (5%) of nonlocal wares from valleys southeast of the altiplano
(Janusek 1994), which were absent in Akapana
East 1M and Akapana ceremonial contexts (Alconini Mjica 1995).
Material patterns in Chiji Jawira in Late Tiwanaku IV diverged most significantly from
those inside of the moat. Here, structures rested
on pure adobe foundations, and domestic contexts were surrounded by dense refuse. More
than 80% of serving wares were cruder in manufacture and less elaborate than those in Putuni or
Akapana East 1. Keros and tazones made up 83%
of serving wares, and escudillas were rare. Red
slip, a hallmark of Tiwanaku ceramic style, was
present on only a small portion (approximately
20%) of serving and ceremonial wares. Nonlocal
vessels and vessels with nonlocal influence were
common (Rivera Casanovas 1994, 2003).
Differences in architecture, serving wares,
and sanitation defined a roughly concentric gradation of social status within the emerging city.
High status residential groups concentrated
around the monumental structures of the urban
core (Kolata 1993). The moat, an ancient and
highly visible boundary, emphasized status differences as the settlement expanded. By the end
of Late Tiwanaku IV, the moat demarcated the
civic-ceremonial core from what was now a
sprawling residential periphery. The moat came
CHAPTER 10: RESIDENTIAL DIVERSITY AND THE RISE OF COMPLEXITY IN TIWANAKU 165
FIGURE 10.13. Elaborate escudillas from residential and mortuary contexts in Late Tiwanaku IV occupations
in the Putuni area
166
JOHN W. JANUSEK
[FIGURE 10.14 ]
CHAPTER 10: RESIDENTIAL DIVERSITY AND THE RISE OF COMPLEXITY IN TIWANAKU 167
FIGURE 10.14. Typical serving wares from Akapana East 1M (ac) and
Akapana East 2 (de). Sherds from Akapana East 2 include tazones decorated
with continuous volutes (d) and nonlocal Omereque-style bowls (e) .
168
JOHN W. JANUSEK
rare, and red slip decoration, the hallmark of Tiwanaku ceramic style, was present on less than
20% of serving and ceremonial wares. Llama motifs were common on tazones and ceremonial
vessels from local offerings, and they appear on
slumped vessels (Rivera Casanovas 1994:163;
also Alconini Mjica 1995:198199). These representations appeared on approximately 23% of
vessels without geometric iconography, and
were extremely uncommon, far less than 1% of
such vessels, in any other residential or ceremonial context. Nonlocal vessels and vessels with
nonlocal influence were present at Chiji Jawira
in even higher proportions than in Akapana East
2. Most common were vessels in the derived
Tiwanaku style (Bennett 1936:402; Ponce Sangins 1981; Rydn 1959) typical of the Cochabamba valleys 200 km to the southeast. These
Cochabamba-style vessels composed about 19%
of all serving wares, and included a range of distinctive forms such as challadores, or kero variants
with a tapering body and narrow base, and small
bowls, or cuencos. Cuencos composed 3% of the
assemblages, higher than in any other residential
area.
Therefore, while all residents obtained and
used Tiwanaku-style vessels, each compound simultaneously revealed a distinct assemblage of
serving forms and styles. Key differences included the predominance of escudillas and a variety of other special serving wares in Putuni,
the prevalence of tazones with continuous volutes and nonlocal wares in Akapana East 2, and
the popularity of camelid motifs and Cochabamba-style vessels in Chiji Jawira. Certain domestic rituals also varied between compounds.
The sahumador, a ritual burner common to domestic life in all other residential areas, was virtually absent in Chiji Jawira, where small
figurines of humans and animals were common.
Possibly household rituals stressing abundance
were performed to the near exclusion of an otherwise ubiquitous ritual complex involving
sahumadores.
[FIGURE 10.15 ]
determined that the proportions of tubers, chenopodium, and maize varied significantly between distinct contemporaneous occupations of
Tiwanaku. Chenopodium, or quinoa seeds, were
most frequent (highest density) and best distributed (greatest ubiquity) throughout the site, followed by tubers and maize. Maize, a crop not
widely cultivated in the altiplano, should have
greater density and ubiquity measures among
high status groups, because it was highly valued
in the Andean highlands (Murra 1980:814) but
could be obtained in quantity only through longdistance relations of trade or vertical complementarity (Goldstein 1989; Kolata 1992). In Tiwanaku, however, maize density and ubiquity
measures were highest in Akapana East 2, and its
ubiquity measure was also high in Chiji Jawira.
Therefore, it appears that this crop was not
strictly associated with high-status groups.
Maize was most abundant among groups in the
settlement periphery, who also used relatively
high quantities of vessels from the lower, more
temperate, valleys in which maize grows.
Local Mortuary Practices. The presence of human
burials inside of residential compoundsin the
Putuni area, Akapana East 1M, Akapana East 2,
and Chiji Jawiraindicates that mortuary activity was not entirely relegated to discrete cemeteries. In Tiwanaku, mortuary ritual was closely
linked to domestic life, and appears to have been
variable in practice. In the Putuni area, for example, some burial chambers contained scores of
broken, partially reconstructible escudillas, possibly representing a unique mortuary tradition
common to this high-status group. The desire to
inter certain individuals near living spaces
shared by intimate kindred suggests some type
of local ancestor cult, analogous to mortuary rituals practiced at the time of European contact
(Cobo 1956 [1653]:73, 163165; Rowe 1946:286,
298; Zuidema 1978). In Akapana East 2, the
placement of a stone marker in the patio above
the complex burial supports this idea. Household or residential compound members remembered, made offerings to, and perhaps celebrated
with deceased relatives at auspicious times, reaffirming in ritually charged contexts group solidarity and identity.
CHAPTER 10: RESIDENTIAL DIVERSITY AND THE RISE OF COMPLEXITY IN TIWANAKU 169
FIGURE 10.15. Serving vessels characteristic of Chiji Jawira, including ceremonial keros with llama motifs (a, b) and
Derived-style vessel sherds (c). After Alconini Mjica 1995:Figure 74, and Rivera 1994:Figures 12.112.2.
170
JOHN W. JANUSEK
CONCLUSIONS:
SOCIAL COMPLEXITY IN TIWANAKU
This research demonstrates conclusively that Tiwanaku incorporated extensive residential occupations. During the Tiwanaku IV phase, the
settlement expanded into a sprawling urban
center of at least 6.5 km2, the largest in the Titicaca Basin and one of the largest in the pre-Hispanic Andes. Nonetheless, by no means was
Tiwanaku limited to residential activity. In and
around the monumental core were numerous
temples and courtyards dedicated to various
forms of public and private ceremony (Alconini
Mjica 1995; Janusek 1994:103123; Kolata 1993:
103149; Manzanilla 1992; Ticlla and Vranich
1997). Not unlike many other political centers in
the New World, Tiwanaku was important both
as a religious place of ceremony and as a permanently and densely populated urban center. Bennett came close to understanding Tiwanaku, but
ultimately fell slightly short of the mark. Tiwanaku consists of extensive refuse pits and middens, to some extent the product of large-scale
ceremonial activities, as he suggested. But these
lively feasts and social gatherings both involved
and were sponsored by people living within the
settlement. As in other major Andean settlements (Cobo 1990 [1653]; Kolata 1993; Moore
1996b; Morris 1982; Shimada 1991), ceremony
and urbanism went hand in hand.
This research has begun to illuminate the nature of urbanism and complexity in Tiwanaku.
First, a wide range of evidence demonstrates
clearly that during the Tiwanaku IV period, Tiwanaku incorporated social groups that were
differentiated in social status and in occupation.
Differences in status followed a broadly concentric grade, in which many elite groups resided
inside of the moated core, near the monumental
constructions, while groups of lesser status resided in compounds toward the edge of and outside of the moat. At the edges of the settlement
were groups of lower status, in one case with
strong ties to a distant region. If status was
roughly concentric, it was also marked by a gradient of identity separating Tiwanaku elites
CHAPTER 10: RESIDENTIAL DIVERSITY AND THE RISE OF COMPLEXITY IN TIWANAKU 171
increasing number of bounded residential compounds, each housing a social group that consisted of a number of constituent households.
Each of these larger groups shared similar resources, practiced local domestic and mortuary
rituals, and maintained a common identity, suggesting strongly that they formed kin-based
groups analogous to later micro-ayllus. Status
and occupation were inherently tied to social
boundaries that were expressed in social identity and everyday practice. Urban complexity in
Tiwanaku, as in many pre-Hispanic societies of
the New World, developed out of segmentary
social differences that had profoundly horizontal dimensions.
This examination joins many others in questioning the traditional concept of social complexity (see, for example, Crumley 1995; Joyce
and Winter 1996; Smith 1994; Stein 1994). Relations of inequality and institutions of integration
became increasingly developed as Tiwanaku expanded during the Tiwanaku IV phase, but
these materialized in specific types of social relations and activities. More fundamental historically was the social segmentation out of which
state institutions developed. The state developed out of social inequalities and political institutions already embedded in preexisting kin
groups and communities, and over the course of
the next six centuries state rulers promoted, consolidated, and expanded the roles of social difference and political integration. We return to
Childe, who argued that in preindustrial cities
there is simply no room for skeptics and sectaries (1950:16), or as Max Weber (1958:65120)
put it, for the kin-based totemistic ties characteristic of primitive societies. A growing corpus of research shows that sectarian and
totemistic ties were often the very ground on
which chiefly and state institutions were built.
11.
The Altiplano Period in the Titicaca Basin
Kirk L. Frye and Edmundo de la Vega
THIS CHAPTER PRESENTS data from several seasons of systematic archaeological research conducted in the western Titicaca Basin. It focuses
on the development of the Lupaqa Kingdom,
one of several groups that occupied the Lake
Titicaca Basin during the Late Intermediate, or
Altiplano period (circa AD 1100 to AD 1450). Historically, the Lupaqa have figured prominently
in Andean archaeology because they are portrayed in ethnohistorical accounts as a powerful
pre-Inca polity whose economy included the
control of ecological zones well outside the core
area of the Altiplano (Murra 1968). These ethnohistorical documents provide a graphic account
of the development of a complex and unified
pre-Inca highland polity (Cieza de Len 1984
[1553]; Cobo 1956; Diez de San Miguel 1964
[1567]). However, until recently, many of the
claims concerning Lupaqa political complexity
have not been tested against rigorously collected empirical data. In this chapter, we show
that the collapse of the Tiwanaku state and
changing climatic conditions within the Titicaca
Basin created a political landscape characterized by chronic conflict and competition between separate Lupaqa subgroups. Rather than
leading to centralized leadership, chronic warfare and factional competition prevented the
Lupaqa from becoming politically unified during the Altiplano period. The archaeological
perspective, then, contrasts with the general
picture of Lupaqa complexity portrayed in ethnohistorical documents.
HISTORY OF RESEARCH
Early research in the Lupaqa area consisted of
site reports that included descriptions of the
173
174
mately 200 km2, incorporating three specific ecological zones: (1) a lakeside (38083825 m above
sea level) and hillside terrace (38254100 m
above sea level) agricultural zone, (2) an inland
agro-pastoral zone, defined by its distance from
the lake, and (3) a higher-elevation mountain
zone (> 4100 m above sea level) dominated by
Cerro Atoja. Because most of the important cultural developments in the western Titicaca Basin
have taken place within the agricultural zone
near the lake edge, researchers have tended to focus their investigations there. The ChucuitoCutimbo survey was specifically designed to include a large area of agro-pastoral land away
from the lake edge. Although the inhabitants of
the area today practice some agriculture, the primary economic focus of contemporary inhabitants of this area is animal husbandry. Higherelevation mountain areas are used primarily for
herding animals. This zone is characterized by
some permanent prehistoric and modern occupations, and in higher altitudes, seasonal occupations. The Chucuito-Cutimbo survey shows that
political developments in the agro-pastoral zone
are key for interpreting Lupaqa political history.
175
176
Archaic period cave sites, no Formative or Tiwanaku period sites were found in this ecological zone. Although the agro-pastoral resource
zone was occupied in the Juli-Pomata survey
zone in both Formative and Tiwanaku times, it
was more densely settled during the Altiplano
period.
The population shift from lakeside areas into
agro-pastoral zones during the Altiplano period
likely represents an increased focus on animal
husbandry as an important economic pursuit.
This is consistent with the drought conditions
documented for the Altiplano period from the
ice core data. One of the most interesting questions concerning the Lupaqa political development is the relationship between inland pastoral
and lakeside agricultural groups. It is not clear
whether these new settlements in the inland
agro-pastoral zone represent a movement of the
population from abandoned Tiwanaku centers
and raised field systems, or whether they represent the movement of people into this area from
the western interior region. One possible scenario is that as lakeside economic systems failed
during drought cycles, the populations there
adopted a mixed agriculture and herding lifestyle or abandoned agriculture altogether for
full-time herding. Alternatively, populations
sity. Although much more detailed analysis remains to be done, at least three general
groupings of ceramics in the Lupaqa area are evident so far: (1) the ceramics from the Ccapia
area to the south, whose style is related to the
Pacajes style, (2) ceramics found in the Taraco
Peninsula, and (3) ceramics from the Tiwanaku
valley. Pukarani ceramics occur mostly around
the Juli area, and the Altiplano period ceramics
from the Chucuito-Cutimbo area are more
closely related to Colla-style ceramics than they
are to the ceramic assemblages from the Juli or
Ccapia areas.
[TABLE 11.1
TABLE 11.1. Altiplano and Tiwanaku period site distributions by production zone
Altiplano Period
Late Horizon
52
110
26.93
83.56
0.52
0.79
16
0.63
6.85
0.21
0.63
31
64
15.45
61.23
0.50
0.95
18
30
10.85
15.48
0.60
0.59
AGRO-PASTORAL ZONE
Total habitation sites
177
178
from regions further inland, for whom pastoralism was a primary economic focus, may have
moved into new areas as a result of the Tiwanaku collapse, population pressure, and resource
scarcity, or as result of military action to increase
landholdings in warmer micro-climates nearer
the lake. What is clear is that the largest and
most complex Altiplano period fortified sites
for example, Cutimbo in the northern Lupaqa
area and Tanka Tanka in the southern Lupaqa
arealie within the agro-pastoral ecological
zone. Hyslop (1976) speculated that these two
sites might have been two separate capitals for
two Lupaqa subgroups. That these sites are situated in prime grazing areas suggests that a herding economy was important, an observation
consistent with the ethnohistorical documentation of huge camelid herds under the control of
Lupaqa elites during the Late Horizon (Diez de
San Miguel 1964 [1567]).
The Altiplano period settlement patterns
indicate a relatively fluid political landscape,
characterized by alliance building and warfare
between ever-changing corporate groups (Figure 11.3). Political instability likely created a
situation where people moved their habitations
from one area to another and back again
through time. Along the lake, the typical Altiplano period settlement system consisted of
habitation sites, likely used by individual or
extended family households, and located
within terraced agricultural fields. Forming
small corporate groups, probably on the basis
of lineage affiliation, several individual habitation sites represented the basic social unit. A
defining feature of the Altiplano period settlement pattern and political landscape is the use
of both local refuge sites and large fortified
sites (Figure 11.4). Local refuge sites are located
primarily on hilltops and, in comparison to
large fortified sites, are much smaller, have few
structures other than site containment walls,
and have few surface artifacts. However, although local refuge sites are much less complex architecturally than large fortified sites,
they are found near almost all Altiplano period
settlement clusters and are especially dense
throughout the agro-pastoral zone. Fortified
sites (pukaras) are mountaintop sites surrounded by walls, which in some cases reach
179
these small structures may have served as storage facilities at Pukara Juli and may have looked
like small silos with a stone base and totora siding (de la Vega 1990). On the other hand, medium-sized structuresthose measuring 2 to 4
180
areas. Rather, the different communities that occupied them appear to have operated under a
fairly informal political system.
Burial Patterns
The Altiplano period burial traditions in the
western basin differ from those of earlier periods. Formative and Tiwanaku period burials in
the Lupaqa area usually contain only one individual and are below ground. Altiplano period
burials, in contrast, form three basic categories:
(1) cave and grotto burials (see Chapter 12, this
volume), (2) subterranean cist and slab-cist
181
[FIGURE 11.6
182
FIGURE 11.6. Distribution of Altiplano period above-ground tombs in the Chucuito-Cutimbo and Juli-Pomata survey areas
6). We think that Altiplano period burial practices support the interpretation that Lupaqa political development was characterized by
factional competition. It is expected that if the
Lupaqa had been politically unified, then
chullpa burial towers would have been restricted to either a royal cemetery or used by a
limited number of ruling lineages. Instead, even
though specific groups of chullpas differ in
number, size, and labor cost, and though they
tend to cluster in higher numbers near fortified
sites, they are found throughout the Lupaqa territory. Judging from the masonry techniques
used to construct fancy chullpas, and the ceramics associated with these chullpas, many of the
largest examples were built during the Late Horizon, and some are what Stanish calls redressed
chullpasigloo-shaped chullpas initially constructed in the Altiplano period that were rebuilt
and faced with cut stones after the Inca conquest
(Stanish 2003).
If chullpas are the material expression of the
economic power and the level of political prestige of their builders, then the wide distribution
and high number of chullpas documented
throughout the western lake region indicates
that chullpa building was not restricted to only a
few elite groups. That the majority of large
chullpas constructed with cut stones likely date
to the Late Horizon suggests that political prestige and power afforded to local leaders during
the Altiplano period was more difficult to
achieve and maintain than previously. Mortuary
remains at Altiplano period sites that do not
have a Late Horizon component consist of
poorly constructed igloo chullpas and/or a series of slab-cist or cist tombstypes which do
not appear to convey major status differences
between the individuals or groups of individuals contained in them. It was only later, perhaps
as the result of Inca incursions into the region or
ameliorating climatic conditions, that status differences became more pronounced, and the Lupaqa took on the level of political centralization
indicated in the ethnohistorical documents.
CONCLUSIONS
We offer several important observations in our
conclusion. First, climatic conditions beginning
183
184
one site can be considered a political capital. Instead, as the systematic survey and reconnaissance data show, the pattern of small habitation
sites located near fortified sites is one that is repeated throughout the Lupaqa territory. The
high number of similar fortified and refuge
sites, the nature and wide distribution of
chullpa burials, and the pattern of locally developed ceramic assemblages in the western Lake
Titicaca area indicate the existence of many
small-scale polities. At least during most of the
Altiplano period, these polities, though they
may have been periodically aligned with their
immediate neighbors, do not appear to have
been under the direction of a centralized leader.
However, by the end of the Altiplano period,
major fortified sites had become the focus of ritual feasting and political activity going beyond
their initial function as refuge centers. For instance, Altiplano period decorated ceramics are
found in the highest concentrations in major
fortified sites, sometimes in burials, and almost
never in unfortified habitation sites. The ceramic component at these major fortified sites is
composed of a high frequency of crudely decorated bowl forms, large serving pots, and liquid
containers that probably were part of local
feasting ceremonies. Although we consider fortified sites to have been important political centers, accession to leadership roles in these peer
polities does not appear to have been institutionalized. By the end of the Altiplano period,
large fortified sites probably became the centers
of political and economic activity, most likely
headed by an emerging elite group, which expanded its political and economic influence
through feasting ceremonies and political alliances.
Despite the appearance of large fortified
sites across the Lupaqa landscape, the complexity attributed to the Lupaqa in several ethnohistorical documents appears to be due more to the
effects of Inca imperial expansion into the region
than to internal political development during
the Altiplano period. At present there are no archaeological indicators to suggest the existence
of a Lupaqa king, a Lupaqa capital, or a unified
Lupaqa confederation during the Altiplano period. These findings differ from the view contained in ethnohistorical documents, which
suggest a politically unified Lupaqa. Instead, the
Altiplano period Lupaqa appeared to have been
comprised of a series of small-scale peer polities
who competed among themselves.
12.
The Cave Burial from Molino-Chilacachi
Edmundo de la Vega, Kirk L. Frye, and Tiffiny Tung
185
186
[FIGURE 12.1 ]
Figure 12.1 Map of the Lake Titicaca Basin with Molino-Chilacachi indicated
jor fortified sites (see Frye and de la Vega, Chapter 11, this volume)Cerro Pukara is a small
refuge area, characterized by sparsely distributed architectural remains. Because the cave is
located below the mesa top at the junction of the
hill slope and the volcanic layer, which is away
from standing architecture, the temporal relationship between the two areas is unclear.
The cave was entered through a small opening measuring 90 x 40 cm, but the original entrance is oriented to the northeast and measures
2.2 m in width. The original entrance remains
closed and is concealed by a large and possibly
intentional rock fall. The volcanic tuff forming
the cave walls is an excellent insulator, moderating the high diurnal temperature extremes of the
region. During the month of May, when readings were taken, the relative humidity inside the
cave was between 83 and 88%, with an interior
187
MUMMY BUNDLES
Although lithics, ceramics, wooden tools, and
basketry were recovered, mummy bundles represent the single most impressive artifact category recovered from the site. Bundles contain
individuals flexed in a fetal position, fully encased by Stipa ichu cords of variable thickness, intertwined around the body to form a cocoon-like
bag. A total of sixty-two mummy bundles were
recovered. Of these, forty-eight contained human
remains, while others were cut open by looters,
and still others remained only as fragments.
188
[FIGURE 12.4
189
190
horizontal and vertical cords, the Type 2 technique is less complex and more variable in
quality than the Type 1 technique (Figure 12.5).
Braiding quality was variable; some bodies
were tightly encased with very closely spaced
cords, while others were loosely enclosed with
large gaps between the cords. Future research
may indicate if casement quality can be correlated with gender, age, and/or social status
categories. The term chullpareferred to by
Bertonio (1956 [1612]:93) in his Aymara dictionary as an [e]ntiero o sern donde se metan
sus difuntoslikely represents these Stipa
inchu casements and not the burial towers so
prevalent on the Altiplano landscape.
[FIGURE 12.5]
191
complete or partially complete, ten were still covered in soft tissue and articulated with postcranial
remains, and twelve were fragmented.
The age-at-death profile of the skeletal population shows that more than a third died between birth and seven years of age (Table 12.1).
This frequency is similar to that of contemporaneous (Late Intermediate period) populations
from the nearby Osmore-Moquegua drainage,
where infants and children (individuals 010
years of age) comprise 36% of the skeletal population at both the sites of Estuquia in Moquegua (Williams 1990) and San Geronimo in the
Osmore Valley (Burgess 1999). In contrast, the
contemporaneous Osmore Valley sites of Chiribaya Alta and El Yaral have a higher rate of infant and child deaths (deaths of individuals
aged 010 years), ranging from 43 to 47%, respectively (Burgess 1999). The latter age-atdeath distributions appear to be more characteristic of prehistoric populations, leading Burgess
(1999) to suggest that there may be a sampling
bias at San Geronimo whereby infants and children are underrepresented. Perhaps a sampling
bias is also present among the Molino-Chilacachi sample, resulting from differential preservation, mortuary practices that limited infant/
child burials in the cave, or recovery bias. Conversely, the age-at-death profile could be representative of the once-living population, which
would suggest that Molina-Chilacachi and Estuquia populations shared similar population
structures, both of which differed from those at
Chiribaya Alta and El Yaral.
Skeletal sex was assigned based on the morphology of partial and complete pelvic bones.
Among the thirty adults whose sex could be determined, fourteen (47%) were male and sixteen
(53%) were female/possible female, indicating a
roughly equal sex distribution among the cave
burial population (Table 12.2). There were four
adults whose sex could not be determined.
The presence of criba orbitalia and porotic hyperostosis suggests that some of the population
suffered from iron deficiency anemia during
childhood (see Stuart-Macadam 1987). The anemia may have developed as a result of malnutrition, intestinal parasites that led to diarrheal
disease, or, what is most likely, a combination of
both.
Fifteen percent of the adult population displayed cranial wounds, all of which were located on the frontal parietal bones. The
anteriorly placed wounds suggest that they
were sustained in face-to-face combat. Several
head wounds were healed, indicating that the
victims did not die as a result of the head injury.
In contrast, two adults displayed perimortem
cranial trauma, suggesting that a blow to the
skull may have resulted in death.
Based on an extensive analysis of cranial
deformation styles from cemeteries on the
south central coast, Lisa Hoshower, Jane Buikstra, Paul Goldstein, and Anne Webster
(Hoshower et al. 1995) concluded that models
correlating deformation styles with broad geographic regions (altiplano-coast) are too simplistic. Their study found that the patterns
[TABLE 12.1 ]
[TABLE 12.2
Number of
Individuals
18
22
10
12
27
Adolescent (814)
Sex
Number of
Individuals
Age Group
Male
14
27
Possible male
Female
14
Possible female
Prepubescent
13
25
Adults (2240)
32
40
Sex undetermined
17
12
15
Total
52
100
Total
81
100
192
similar to those found on the Sillustani brownon-cream and black-on-red wares, the Collao
black-on-red ware, and the Allita Amaya wares
described by M. Tschopik (1946:23, 26, 34). Interestingly, the ceramic component bears little
resemblance to the Pukarani styles defined in
the Lupaqa region farther to the south (de la
Vega 1990; Frye 1994, 1997a). Three radiocarbon dates from the site were processed. Two
samples were carbonized material from ceramics and are derived from Units 5 and 7. One
sample is a textile fragment from Unit 8. The
two radiocarbon dates from the ceramic scrapings both date to approximately AD 1300 18
and the date from the textile is AD 1350 55.
These dates place the use of the cave firmly in
the Altiplano period.
The presence of a hallucinogenic kit in the
cave, similar in style to those defined from the
Bolivian site of Nio Korin (Wassn 1972) and
those defined from burial contexts from the
Northern Chilean coastal region (Torres 1985),
represents an extraordinary find. Unfortunately, none of the recovered items was directly
associated with any specific individual or came
from any one collection unit. Hallucinogenic
paraphernalia included cane tubes closed at
one end that contained leather-tipped sticks
and/or spines. One tube contained the residue
of an unidentified, white powder. Wassn
(1972:43-44) reports similar tubes from the
Nio Korin site, which he suggests may have
been used as part of a kit for administering enemas. Other specimens from the complex included leather containers, a bone spoon, quills,
and a small wooden mortar. Also recovered
were an incised wooden snuff tablet bearing
the central figure of a raptorial bird, as well as
snuff tubes, one with polychrome incised designs portraying a bird figure and geometric
motifs (Figure 12.6). Taken as a whole, the materials probably formed one complete hallucinogenic ingestion kit. Stylistically, the snuff
tablet and incised tube are derived from Tiwanaku motifs and were likely produced during
the Middle Horizon. However, their association
with Altiplano period ceramics suggests that
the tray and tubes were curated for use over a
long period of time.
[FIGURE 12.6
193
194
SPATIAL DISTRIBUTION OF
CAVE BURIALS
At present, the distribution of cave burials in
the south central Andes is not well documented.
There are, however, several references to multiple burials in natural openings, including caves,
niches, grottoes, and rock shelters throughout
the region. In the northern area of Omasuyu, Erland Nordenskild (1953), Heath MacBain
(1959), and Stig Rydn (1947) describe burials in
rock openings, while in the Callawaya sector to
the southeast, sites including Nio Korin have
been studied by Enrique Oblitas (cited by
Wassn 1972). Several publications refer to burials in rock cavities in the Mallku territory to the
south (Arellano Lpez and Berberin 1981;
Arellano Lpez and Kuljis 1986; Berberin and
Arellano Lpez 1980). Oscar Ayca (1995) reports
the presence of funerary grottoes from the Colla
site of Sillustani. Limited almost exclusively to
the agro-pastoral zone located well inland from
the lake shore, at least thirty cave burials have
been located in the Lupaqa territory. Nine cave
burials were found in the Chucuito-Cutimbo
survey zone and southward into the Ro Ilave
and Ro Huenque Basins, and Mark Aldenderfer (personal communication) located another
twenty cave burials, some associated with presumably Archaic period cave art. Significantly,
no cave burials have yet been found within the
Juli-Pomata survey area farther to the south.
DISCUSSION
We offer important observations concerning
cave burials in the circum-Titicaca Basin region.
The presence of burials in natural cavities from
Archaic period contexts (Aldenderfer 1990) and
the association of cave burials with cave art support the interpretation that burying the dead in
natural cavities and caves has a long, if not welldocumented, history in the region. Oblitass
(cited by Wassn 1972:14) mention of a cave with
multiple burials containing funerary material
similar to that from Nio Korin may indicate
that a cave burial tradition continued through
Tiwanaku III times; but at present, there is no
clear evidence that the practice was maintained
in Tiwanaku IVV times. Although habitation
195
13.
The Inca Occupation of the
Lake Titicaca Region
Kirk L. Frye
THIS CHAPTER DESCRIBES the effect of the Inca expansion into the Lake Titicaca region, focusing
on the Colla, Lupaqa, and Pacajes ethnic groups.
I review the available archaeological data from
these regions and discuss the relevant ethnohistorical information about the Inca administration. The Titicaca Basin was considered an
important area to the expanding Inca Empire for
several reasons. First, it was important from the
standpoint of labor availability because it possessed one of the more densely settled populations of the central Andes at the time (Cieza de
Lon 1984:353 [1553:Ch. 99]). Second, the economic resources of the Titicaca Basin were considerable. In good years, the agricultural
productivity of the region was quite high, and
the region was considered the breadbasket of
the Altiplano during the Colonial period. The
wealth represented by animal herds was one of
the assets that impressed colonial tax assessors
the most. Perhaps more importantly, the Titicaca
Basin was important to the Inca from an ideological perspective, figuring prominently in the origin myths of the Inca royal lineage, with the
Islands of the Sun and the Moon representing
important pilgrimage sites for the Inca elite
(Arkush, Chapter 14, this volume; Arkush 2000;
Bauer and Stanish 2001; Stanish and Bauer 2004).
As elsewhere in the empire, Inca political organizers took into account military, economic,
and ideological factors while incorporating a
subject region. The Lake Titicaca region is part of
Collasuyu, the southern quarter of Tawantinsuyu. Catherine Julien (1983:28), using ethnohistorical documents, demonstrates that during the
late Inca period the Lake Titicaca area of Colla-
197
198
KIRK L. FRYE
THE COLLAS
Much of what is known about the Collas is the
result of Juliens (1982, 1983, 1993) excellent research at Hatuncolla and from important ethnohistorical documents (especially Cieza de Len
1959 [1553] and Cobo 1979 [1653]). Arkush
shows that the Colla at the time of Inca contact
were probably comprised of a series of smallscale competing groups that were engaged in
constant warfare. Oscar Aycas (1995) work at
the Sillustani provides insight into the sites history and burial traditions. Although there are
many gaps in our knowledge of the Colla region, it is possible to glean from the available resources a basic understanding of how the region
changed in the Late Horizon.
In general, it appears that Inca administration of the Colla region was influenced by
military concerns. Cobo outlines a relatively
straightforward account of the Inca expansion
into the Titicaca Basin. He places this expansion
during the reign of Pachacuti Inca, who wanted
to include the Titicaca Basin into his realm:
THE LUPAQA
More is known about the Lupaqa area during
the Late Horizon than any other province in the
Lake Titicaca region. The Lupaqa area has a long
history of archeological research, in part because
there are many important references to the area
in ethnohistorical documents. Both Pedro Cieza
de Len and Bernab Cobo provide excellent
early accounts of the Lupaqa. Garci Diez de San
Miguels (1964 [1567]) Visita gives invaluable insights into the Lupaqa political and economic
organization in the early Colonial period. The
richness of ethnohistorical documents that describe the Lupaqa has inspired many important
theoretical discussions concerning zonal complementarity (Murra 1968, 1972; Stanish 1992),
the nature of the Inca decimal system (Julien
1982), and labor organization under the Inca
state (Murra 1985; Stanish 1997).
The Lupaqa region also has a rich history of
archaeological research. Early site reports were
published for important Lupaqa sites by Jos
Mara Franco Inojosa and Alejandro Gonzlez
(1936) and by Emilio Vsquez (1939). In the
1940s, Marion Tschopik (1946) defined early Lupaqa ceramic types and burial styles from
around Chucuito. In the 1970s, John Hyslop
(1976, 1977b) conducted an unsystematic survey
of sites located in Lupaqa territory, defined
burial types, and documented where the Inca
road passed through the region (Hyslop 1984).
Beginning in 1988, the Project Lupaqa, under the
direction of Charles Stanish, began systematic
research near the town of Juli, including excavations at the Lupaqa site of Pukara Juli (de la
Vega 1990). Beginning in 1990, the Proyecto Lu-
199
200
KIRK L. FRYE
201
Altiplano period is reported from the JuliPomata region (Stanish et al. 1997:Table 3; Stanish 1997:205). The most profound changes indicated by these data are the dramatic increase in
population, as represented by the change in site
occupation area, and the increase in settlements
on the lakeshore and in the terrace agricultural
zones.
One of the most notable changes in the settlement pattern between the two periods was
the abandonment of Altiplano period fortified
sites. The survey located ten fortified sites, of
which four are categorized as major fortified
sites. They contained standing architecture, fortification walls, and abundant surface artifacts.
Six minor or temporary refuge sites were also
found, characterized by their small size, lack of
standing architecture, and few surface artifacts
[FIGURE13.1
[TABLE
13.3]
ABOUT HERE]
202
KIRK L. FRYE
FIGURE 13.3. The Late Horizon settlement pattern in the Chucuito-Cutimbo survey area
203
TABLE 13.1. Settlement pattern data from the Chucuito-Cutimbo and Juli-Pomata surveys
Chucuito-Cutimbo Area
Juli-Pomata Area
Tiwanaku Period
Altiplano Period
Altiplano Period
22
52
140
11.85
26.93
74.16
0.49
0.52
0.53
3851
3931
3905
16
44
8.67
0.63
21.04
0.54
0.21
0.48
72.58%
2.30%
28.00%
31
75
3.25
15.45
42.64
0.41
0.50
0.57
27.42%
57.30%
57.00%
18
21
10.85
10.48
Percentage of total
TERRACED AGRICULTURAL ZONE
Total habitation sites
Percentage of total
AGRO-PASTORAL ZONE/PUNA
0.60
0.50
Percentage of total
40.4%
14.0%
REFUGE SITES
14.98
6.82
FORTIFIED SITES
43.0
> 50.0
204
KIRK L. FRYE
[FIGURE 13.4 ]
205
206
KIRK L. FRYE
THE PACAJES
Stanishs (1997) discussion of the Inca administrative policies in the Lake Titicaca region demonstrates that in nonmarket economies labor
control is an important focus of imperial administration. Settlement data from the Juli-Pomata
of Pacajes assimilation into the Inca Empire, although Cobo mentions the Pacajes Indians in his
recounting of the Inca conquest of the Titicaca
Basin. He states that for a few days the Pacasa Indians defended the bridge over the outlet
[Desaguadero] of Lake Titicaca or Chucuito, and
in order to win it from them, the Inca sent part of
his army to look for a ford eight leagues downstream (Cobo 1979 [1653]:140141). After crossing the outlet, Pachacuti continued on to
Tiwanaku, where he was so impressed by the
structures there that he commanded that they be
studied so that the architectural techniques
could be incorporated into Inca building practices (Cobo 1979 [1653]:141). Although the site of
Tiwanaku figures prominently in the cultural
history of the region and as such has received a
great amount of study, the nature of the Inca occupation there is less well documented than are
earlier periods of Tiwanaku history.
Our understanding of the Late Horizon in
the Titicaca area of the Pacajes zone comes from
three systematic surveys conducted in the Tiwanaku core areathat of Juan AlbarracinJordan and James Mathews (1990) in the Tiwanaku Valley, a survey in the Catari Basin area (Janusek and Kolata 2003), and the Taraco survey
along the Taraco Peninsula (Bandy 2001b). Other
evidence of Inca influence in the region comes
from the many chullpas, some with cut stones,
dating to the Late Horizon that are found
throughout the Pacajes area (Bellido 1993).
Survey data from the Pacajes region demonstrate that the Inca population resettlement policies in this region mirrored those seen in other
areas in the basin. Local populations were
moved into newly created settlements, especially to areas with the greatest agricultural potential near the lake. A large proportion of the
Tiwanaku Valley population previously living in
dispersed hamlets and small villages appears to
have been relocated to the lakeshore along the
Taraco Peninsula. No clear site size hierarchy describes the Late Horizon settlement pattern in
the Tiwanaku Valley. As reported by AlbarracinJordan and Mathews (1990:161) and by Mathews
(1992b), there were 202 sites with Inca Pacajes or
Late Horizon material, in contrast to 526 sites
dated to the previous Early Pacajes or Altiplano
period. There are no large Inca period sites in the
207
CONCLUSION
The Titicaca Basin was one of the first regions incorporated into the Inca Empire. The diverse nature of its inhabitants, the distribution of
resources in the region, and the existing levels of
political complexity were all factors in how the
Inca state administrated the region. In some areas, military concerns were paramount. In these
cases, a region was divided into many administrative territories. In other areas like the Lupaqa
and Pacajes, economic goals appear to have been
important. The available survey data from the
Lupaqa and Pacajes regions show that the Inca
pursued a maximization strategy designed to increase the productive resources available in the
Titicaca Basin. The strategy involved moving
people from less productive areas into those
with the greatest agricultural potential near the
lakeshore. The surplus production of foodstuffs
was used in state-sponsored festivals and to
208
KIRK L. FRYE
14.
Inca Ceremonial Sites in the
Southwest Titicaca Basin
Elizabeth Arkush
209
210
ELIZABETH ARKUSH
12 Ch. 23]; Valera 1950:145). The Incas also frequently enhanced aboriginal religious sites and
consulted important local huacas (Rowe 1946:
302; MacCormack 1991:141159).
Our archaeological knowledge of Late Horizon ceremonial sites outside the Inca heartland
is incomplete (see, for instance, Van de Guchte
1990:406, illustration 2). Some provincial ceremonial sites, such as Sayhuite in Apurmac and
Vilcashuamn in Ayacucho, are intrusive and
pure Inca in style; others, like Pachacamac or
Wari Wilka (Shea 1969), are Inca additions to important pre-Inca shrines, built with significant
adaptations of local, non-Inca styles and materials. Rarely, there was actual destruction of indigenous huacas (Nielsen and Walker 1999). In
addition, numerous modest, rural ceremonial
sites exista few will be described belowbut
they are little known, giving the erroneous impression that where ceremonial sites were built
in the provinces, they were major sanctuaries directly designed and controlled by the Inca elite.
It is debatable whether intrusive Inca ceremonial installations were placed strictly to influence the conquered masses. For instance, a
proliferation of high-altitude mountain shrines
in Argentina and Chile, a region where the Incas
invested comparatively little in economic or political infrastructure, shows considerable effort
and expense poured into ceremonial sites that
were not very visible or accessible to local populations (Beorchia Nigris 1973, 1985; McEwan and
Van de Guchte 1992; Reinhard 1992; Schobinger
and Constanza Ceruti 2001). Class exclusivity is
clearly apparent in Inca religious practices in
general and at the Island of the Sun sanctuary in
particular. Elaborate regulations controlling access to the island sanctuary reproduced and reflected social divisions. Overall, Inca religious
sites did more, and sometimes less, than signal
Inca dominance.
Factors affecting Inca ideological strategy
may have included the political and economic
investment in the region, local religious practice and the prestige of local religious sites, and,
apparently, military considerations (Hyslop
1990:189). The ideological incorporation of the
provinces was not an event, but an ongoing
process, as Cobo noted:
211
212
ELIZABETH ARKUSH
213
214
ELIZABETH ARKUSH
tive centers, especially those that took advantage of preexisting sacred places. Those sites,
likewise, might not be particularly visible or accessible.
A subsidiary question is whether Inca-style
sites in the region were intended or used as ritual stations on the pilgrimage route from Cuzco
to the Island of the Sun. If they were, they
should have been on the road or near tambos
(way stations), and either Inca in style or related
to the pilgrimage cult in other ways. In either
scenario, huacas that were locally built and used
could also have served the religious needs of the
population.
215
FIGURE 14.1. The Lake Titicaca Basin, with Late Horizon towns, ceremonial sites, and the road
216
ELIZABETH ARKUSH
FIGURE 14.2. The Copacabana Peninsula, with Late Horizon ceremonial sites
217
sector in Late Horizon Copacabana that survives, and the rites related to the pilgrimage at
Copacabana that are mentioned in the documents (e.g., Ramos Gaviln 1988:171 [1621:Ch.
28]) may well have taken place at Intinkala and
Orcohawira.
Both carved-rock sites are so consistent with
Inca stylistic canons that there can be no doubt
they were executed on Inca orders and under Inca
supervision. However, early Colonial documents
suggest that an important local huaca, the idol of
Copacabana, may have been found at Intinkala or
Orcohawira (Sanz 1886). Ramos Gaviln gives a
detailed account of where it was found:
[FIGURE 14.4
4.3
218
ELIZABETH ARKUSH
219
servatory can be safely assigned to the Late Horizon based on its association with nearby Incastyle stairs and walls. One trace of a wall, with
some blocks still in place, cuts across the space
directly in front of the Horca (Figure 14.6). Although this wall could not have been very high
or it would have blocked the trajectory of the
suns rays on the solstice, it probably served as a
boundary separating sacred from profane space,
and may have indicated a social division as well,
such as that between officiators and the public
(Rivera Sundt 1984:98).
[FIGURE 14.6
14.5 ]
220
ELIZABETH ARKUSH
Copacati
Copacati is a carved-rock site on top of a steep,
rocky hill about 4 km south of Copacabana, first
described by Maks Portugal Zamora (1977:299,
see also Rivera Sundt 1978:81). The bedrock is
finely carved in flat stepped shelves (seats and
altars) at several points along the ridge-top in
unmistakable Cuzco-Inca style (Figure 14.7). On
the northwest flank of the ridge (a steep slope
makes the main carvings almost inaccessible)
are two additional worked areas first noted in
this reconnaissance. One is a triangular seat
carved in a rock far down on the side of the hill;
the other consists of two rectangular seats in
the steep rock higher up the hill (Figure 14.8). As
will be shown, a pattern of carved rocks scattered away from the main ceremonial sector is
rather common for the small Inca sites described
here. No Late Horizon habitation site was found
in the immediate area.
As at Intinkala, the Inca carvings at Copacati
were apparently associated with a pre-Inca idol.
According to Ramos Gaviln:
[FIGURE 14.8
14.7 ABOUT
]
HERE]
Besides this Copacabana Idol, the Yunguyos had another which they called
Copacati. The hill where it was located
took its name from this same Idol, which
was on the way out of the town. It was
made of stone, an evil figure completely
curled round with snakes; the people
would resort to it in times of drought,
asking it for water for their crops. Padre
Almeyda, who had charge of the curacy
before the missionaries of my sacred order came into it, heard of this Idol and
had it brought to the town, and when it
was placed in the plaza in the presence
of many people, a snake was seen to detach itself from the Idol and go around it.
The priest seeing this gave them to understand that it was the devil, and that
they should be ashamed to have held
such a vile creature for a god.The
master of the chapel, Don Gernimo
Carvacochachi, seventy-eight years of
age, told us he had seen the snake uncoil
itself from the discarded Idol, which was
221
222
ELIZABETH ARKUSH
223
The site of Ckackachipata, recorded by Stanish et al. (1997:90), was probably the closest associated habitation area to the Playa Chatuma
blocks. Ckackachipata is found on the peninsula
just next to Playa Chatuma (Figure 14.12) and
has a Late Horizon component with fairly modest local ceramics (Stanish et al. 1997). The Island
of the Sun is visible from Ckackachipata, though
not from Playa Chatuma itself.
Other nearby Inca features may be related to
Playa Chatuma (Figure 14.12). In the bay to the
east of the Ckackachipata Peninsula, a large,
square block of stone with a straight groove
carved into one side rests in the water just off the
beach. This is almost certainly from the Late Horizon, but its relationship to Playa Chatuma cannot be determined. Pukara Chatuma is a hill with
a Late Intermediate period and Late Horizon
presence, as evidenced by pukara walls and an
Inca-style square chulpa with fine Inca sherds,
respectively (Stanish et al. 1997:95). On top of the
hill, Stanish et al. noted several carved rocks
(1997:9596). These include an Inca-style carved
rock with a rectangular depression, a large, finely
shaped ashlar block near what may be a slab-cist
tomb, a rock with a single straight groove about
[FIGURE 14.12
224
ELIZABETH ARKUSH
225
FIGURE 14.12. Playa Chatuma and surrounding area. After Stanish et al. 1997:86
10 cm long, and another rock carved with an unusual concentric circle design (Figure 14.13). This
last motif is characteristic of neither Inca nor local carving, but the rock may be Late Horizon in
view of its location near the other rocks at this
site.5 These dispersed ceremonial features scattered across the landscape near Playa Chatuma,
and the relatively disorganized, asymmetrical
layout of the channels (compare, for instance, Samaipata in Bolivia, near the eastern edge of the
Inca Empire), suggest that the site may not have
been organized by a master plan.
Hyslop (1984) was unable to find the Inca
road south of Pomata, but it is unlikely that it
passed closer to Playa Chatuma than the modern road does because of the hilly topography of
this area and the swamp land just to the west.
The site is accessible by foot and requires only
about a half-hours walk to reach it from the
modern road leading to Yunguyu. It probably
required no more effort to reach it in the Inca period. The site could easily have been along a
route from Pomata to Yunguyu and the Copacabana Peninsula beyond.
The canals at Playa Chatuma and Chinchin
Jalave (described below) relate these sites to numerous others across Tawantinsuyu that used
[FIGURE 14.13
226
ELIZABETH ARKUSH
227
CARVINGS AT BEBEDERO
Two examples of Inca ceremonial rock carving
are found at the Bebedero sandstone outcrop,
which is located about 8 km north of Juli and
just west of the modern road. Here the modern
road is probably very close to the Inca road and
may overlie it (Hyslop 1984:123). The Incas
Chair, first described by Squier in 1877 (1877a:
350), is a set of carvings in classic Cuzco-Inca
style on a section of the outcrop near the road
(Figure 14.16). The carvings consist of several
shallow planes or seats, a stairway leading to
the top of the rock, and the thin vertical channel
after which the outcrop is named (bebedero, or
drinking-trough). The rock has been capped
with a tower in recent times.
The vertical channel on the front of the Incas
Chair tie this site, along with others in the area,
to a typical Inca pattern of libation. The quality
of work at this site and the similarity of its
carving to that in the Cuzco heartland make the
site stand out among more modest ceremonial
[FIGURE 14.16]
228
ELIZABETH ARKUSH
229
230
ELIZABETH ARKUSH
FIGURE 14.20. Sketch map of Kenko (Tres Ventanas) (Map and contour lines are
not to scale.)
231
232
ELIZABETH ARKUSH
FIGURE 14.23. Carved stairs on the cliff face at Kenko (Tres Ventanas)
233
234
ELIZABETH ARKUSH
[FIGURE 14.26
14.25
235
236
ELIZABETH ARKUSH
[FIGURE 14.27
237
FIGURE 14.28. The Pachatata sunken court on Amantan Island. The upper wall is a recent addition.
Amantan Island
Amantan Islands two major ceremonial constructions, Pachamama and Pachatata, are problematic, but the site is too important to leave out
of this discussion. Both constructions are sunken
courts on low mountains. Pachatata (also called
Aylicancha) is rectangular and Pachamama is
circular, making it very unusual within the Titicaca Basin (Figure 14.28).13 Both have been restored and modified to an unknown extent in
recent times, and continue to be used for rituals
and offerings (Niles 1987b, 1988; Spahni 1971;
see also Kidder 1943:116 Stanish 2003:188, and
Vsquez 1940). The courts appear to have been
in use in the early Colonial period, for Martn de
Mura reports a famous huaca at Amantan
(Mura 1946:2167 [1590:Bk. 3 Ch. 21]).
[FIGURE 14.28 E]
238
ELIZABETH ARKUSH
Pucara
While Pucara is far to the north of the study area,
it deserves to be mentioned for its ceremonial
role in the Late Horizon. An important Formative center, it was restored and modified in Inca
times. A terrace wall was remodeled with trapezoidal niches, and a stairway with finely dressed
corner stones was built into Pucara walls
(Wheeler and Mujica 1981:589). Fine Inca pottery of both local and Cuzco manufacture can be
found at the site (Wheeler and Mujica 1981:58
70), and Pucara is listed as a tambo in early Spanish colonial documents (Guaman Poma de Ayala
1980 [1613]:1006 [1091/1101]; Vaca de Castro
1908 [1543]:437). Pucara appears in Inca mythology as a site at which Viracocha, in his voyage of
creation, turned disobedient subjects to stone or
called down a rain of fire upon them (Molina of
Cuzco 1947 [1584]):26). The association of Pucara with Viracochas journey outward from the
sacred center of Titicaca makes the site a good
candidate for a ceremonial station on the pilgrimage, which traced Viracochas course in reverse back to the Islands of the Sun and Moon.
and rather than being placed for maximum visibility and use in the Late Horizon, their placement may have been intended as architectural
statements of Inca control over formerly important political centers (Arkush 2001).
CONCLUSIONS
The proliferation of small ceremonial sites along
the southwest shores of Lake Titicacaas well
as the justly more famous Island of the Sun sanctuaryamounted to something of a religious
florescence in the Late Horizon. While the region
does not have nearly the density of ceremonial
sites as the area near Cuzco, the region displays
a surprising variety of small sites available for
ritual. These are mostly carved rock sites, and
they clearly refer to Inca, rather than indigenous,
ceremonial styles, yet they are highly varied in
location, size, level of labor investment, and
technical skill.
The most obvious pattern that emerges from
this group of subsidiary sites is the very lack of a
pattern. Out of fifteen sites,14 six would have required a minor detour from the probable Inca
road, and four sites (Cerro Juana, Kenko and the
carved stone to the west of it, and Amantan Island) would have required a major detour. Nine
sites are on the lakeshore or within view of the
lake, even though the Island of the Sun is not
visible from most of them. However, the pattern
may simply reflect the predominance of lakeside
settlement in the Late Horizon. A cluster of
Cuzco-style sites around Copacabana contrasts
with sites outside the Copacabana Peninsula,
which vary widely in style. Size and labor in-
vestment do not obviously correlate with Cuzcostyle sites or decorated pottery. Several sites
have a dispersed site plan including minor
elaborations or additions, and two (Altarani and
Kenko) show clearly unfinished carving. Two of
the most stylistically authentic Inca religious
sites, Intinkala-Orcohawira and Copacati, are associated with indigenous and possibly nonelite
huacas. The Inca use of Amantan and Pucara
also took advantage of preexisting ceremonial
structures that may well have held religious significance for local inhabitants in the LIP. Meanwhile, in the Late Horizon, the massive chulpa
cemeteries probably provided the main ceremonial sites that were not associated with the Inca
state in the minds of local residents.
While it is difficult to determine exactly what
was going on in the region in the Late Horizon to
produce this heterogeneous mix of sites, comparison with the expectations of the models of Inca
policy outlined above can rule out some possibilities.
The first model, in which Inca administrators used Cuzco-style ceremonial sites to mark
the region as Inca territory and impose imperial
ideology on the Aymara-speaking peoples without accommodation, is untenable in the face of
the data. The styles, locations, and overall pattern of the sites all argue against this scenario.
Several sites feature a mix of styles. Playa Chatuma, for instance, has a Cuzco-Inca style stair
carving, canals which refer to but do not reproduce Cuzco-Inca canons, and the nearby carved
rocks on Pukara Chatuma, whose styles range
from typical Inca to highly unusual and innovative. Kenko (Tres Ventanas) displays Cuzco-Inca
style stairs, masonry, and niches, but an innovative site plan. Furthermore, the locations of the
ceremonial sites are not what would be expected
from a propagandistic building program. Several sites, such as Playa Chinchin Jalave and
even Copacati, are not very visible or accessible.
Many are not close to areas of dense population,
and in consequence are ill-suited as vehicles for
the dissemination of imperial ideology to local
people. Kenko, in particular, has only a small associated habitation sector. Furthermore, there
are no large towns nearby comparable to the cabeceras on the Urcosuyu road. In fact, the Late
Horizon pattern of ceremonial sites in the study
239
240
ELIZABETH ARKUSH
religious autonomy. Meanwhile, their participation in this religious framework helped to reinforce conceptions of social hierarchy in which
the dominance of the Incas was naturalized. The
cosmology of the Incas readily adopted local beliefs, embraced locally sacred places, and accommodated local innovation, while retaining its
essential tenet, the divine solar origin of the
conquering Incas. A recursive relationship for
ideological influence resulted. Inca ideology, religious practices, and sacred sites were continually altered and reshaped by non-Incas and the
nonelite, as well as by their rulers.
Inca religious accommodation contrasts
sharply with the forcible and exclusive imposition of Spanish Catholicism that was to succeed
it (MacCormack 1991). Was there a downside to
an imperial policy of toleration of existing local
or regional huacas? While Inca religious accommodation probably helped to reconcile subject
populations to Inca rule, it may also have weakened the empire by allowing subjects to continue
a strong tradition of identification with smaller
ethnic or regional groups. By the LIP, Andean
ethnic consciousness was anchored in huacas
and pacarinas (mythologized ancestors and origin
places fixed in the landscape), providing a concrete and inalienable sense of ethnicity or grouphood. There is evidence that the Inca rulers were
attempting to weaken regional/ethnic identification with many of their other policies. These policies, enumerated by Rowe (1982), include the
imposition of forcible resettlement or temporary
labor service, the standardization of the arts and
technologies, and the spread of Quechua and
some elements of Inca religion. Yet the Incas
themselves arose out of a traditional Andean conceptualization of self and grouphood and participated fully in it, as evidenced by their own
landscape-based sense of ethnicity (involving descent from a handful of ancestors originating at
Pacariqtambo; see Urton 1990). They could
hardly have done otherwise; but this existing
ideological base may have hindered rather than
helped them in the task of consolidation and unification. The historical documents show that at
the time of Spanish contact the empire was still
thought of as a collection of ethnic units, anchored to place and genealogy by their respective
pacarinas and local huacas. When these groups
splintered off, as they did in the HuascarAtahualpa civil war and at the time of the Spanish conquest, they did so as whole units. The
parts of which the body politic was composed,
tenuously connected by a few empire-wide institutions, came apart easily as soon as the head fell.
NOTES
1.
241
242
ELIZABETH ARKUSH
15.
Archaeological Reconnaissance in the
Carabaya Region, Peru
Lawrence S. Coben and Charles Stanish
243
244
245
246
247
these modifications are the result of an Inka-imposed plan for the exercise of political authority
within this region (Low 1995). Evidence of such
manipulation is present at almost all sites where
sufficient standing architecture remains to make
such a determination. Such evidence includes
platforms (Figure 15.6), plazas (Figure 15.7), and
specially placed worked and unworked stones
(Figure 15.8).
The habitation area of the lower-altitude site
of Kille Kille (Figure 15.9), for example, is about
4.0 ha in area and is located on the main pre-Hispanic road on the first plateau above the river.
Two types of terracing are clearly presenta
finer construction utilizing larger stones and
well-made terraces in the central area of the site,
flanked on both sides by structures using a
cruder construction technique. Structures on the
central terraces are primarily square or rectangular, while those on the sides are primarily
round. A cursory analysis of the surface pottery
[FIGURE
[FIGURE15.8
15.7
15.6]]
[FIGURE 15.9 ]
248
249
250
[FIGURE 15.16]
251
252
253
254
255
256
257
platform is found in one of the larger rectangular rooms in the plaza (Figure 15.29), and two
platforms are found in another (Figure 15.30).
At the northern end of the site, overlooking the
river, is a large multi-roomed structure. An
Inka midden, filled with ceramics that include
the Late Horizon Titicaca Basin Chucuito style
258
259
260
261
262
263
[FIGURE 15.33
15.32 ]
CONCLUSION
Imperial control may be exercised and manifested in a number of ways. While such control
requires military force or the threat thereof, empires throughout history have constructed and
264
265
266
sites and locations throughout the Ollachea Valley. Inka architecture is placed within and upon
the preexisting Late Intermediate layout, often
wedged or jammed therein without regard to,
and often destroying, the preexisting architectural pattern in order to create Inka-type spaces.
We believe that the repeated and extensive nature
of this spatial imposition may result from the importance of this valley as part of the access route
from Cusco to the gold mines to the east. Berthelot (1986) notes that the Inka took special precautions to count and secure this gold, and to send it
directly to Cusco. Unlike many other areas of
Inka governance, the Inka relied on direct control
and oversight of the mining and transportation
process rather than governing through local leaders. It is logical to assume, therefore, that these
precautions would include the safeguarding of
and assurance of control over the transport
routes, and our data support this assumption.
The imposition of Inka architecture in central areas of non-Inka sites plays an integral role
in this strategy of control. It serves, like a flag, as
a distinct and repeated reminder of Inka political authority over the region. Construction of
new structures in the military installations represents a particularly arrogant demonstration of
the Inkas power and the impotence of the local
populaces military forces, while Inka structures
in the centers of other sites reflects the ability to
exercise ideological and political hegemony over
the local region. The repeated and physical reminders of such power and hegemony, reflected
undoubtedly both in the architecture and the
performance of ritual within it, would serve as a
powerful deterrent to any uprising or attempt to
interfere with the gold trade. Such reminders
would of course be strengthened by the geographic proximity of Cusco to this region and by
the placement of some of these ritual sites within
plain view of the roads on which the Inka army
and the gold trade would pass. By impressing
the local populace with architecture and creating
a powerful link between military, ritual, and political control, the Inka reduced the need to
ACKNOWLEDGMENTS
The authors wish to thank Rolando Paredes and
Jos Nez, who accompanied us on this trip.
We likewise thank the anonymous reviewers for
their useful comments.
16.
Settlement Patterns, Administrative
Boundaries, and Internal Migration in the
Early Colonial Period
Matthew S. Bandy and John W. Janusek
267
268
Overall population did decline in this 110year period. The number of adult Indians
in his sample fell overall by 22%.
Snchez-Albornoz concluded that the demographic contraction does appear to have been a
reality, but that it was not as pronounced as had
previously been believed. The dramatic decline
in the number of tributarios that is documented
in so many colonial records can be attributed to
an administrative distinction drawn by the colonial government between originarios and foras-
When we understand the political and economic significance of the colonial revisitas, and
the extent to which population counts were consequently contested and actively manipulated
(not only by indigenous communities but also by
colonial functionaries), the use of administrative
documents to estimate actual rates of population
growth or decline becomes extremely problematic. The results of the recounts, Stern concludes, reflected the relative skills, advantages,
and luck of the interested parties as much as they
did demography (Stern 1982:125). In other
words, there was so much at stake in the process
of generating population measures that the actual documents are very unlikely to be accurate.
If we cannot trust population estimates in
colonial documents, how then can we evaluate
the question of the Early Colonial period demographic catastrophe? How can we begin to evaluate the relative contributions of population
movement, demographic decline, and intentional deception in the Early Colonial period?
Settlement archaeology may help to provide an
answer. Though the archaeological measurement of population growth and decline is a difficult and controversial process (Cook and Heizer
1968; Fish and Kowalewski 1990; Hassan 1981;
Naroll 1962; Parsons 1976; Tolstoy and Fish
1975), the archaeological record at least is free of
the intentional, politically motivated distortion
we have just been discussing. Archaeologists estimate population directly from human occupation debris. Archaeological data can therefore
provide an independent means of addressing
these questions.
270
Saxamar (Pacajes-Inka),
Taraco Polychrome.
We have identified no Urcusuyu Polychrome (M. Tschopik 1946:3233) in our collections. This is rather surprising, given this styles
predominantly southern distribution and possible association with the Inka sanctuaries at Copacabana and the Island of the Sun (Julien 1993:
192199).
By far the most common of the four styles
identified was a local Late Horizon style associated with the Pacajes polity and/or ethnicity.
The style has variously been termed PacajesInka (Albarracin-Jordan 1992; Albarracin-Jordan and Mathews 1990; Mathews 1992a), Saxamar (Browman 1985; Daulsberg H. 1960), and
llamita ware (Graffam 1990). We prefer Saxamar. Examples are illustrated in Figure 16.1p
ae. Saxamar ceramics consist exclusively of
shallow bowls, plates, or basins. These are typically covered with a deep red slip and polished
to a high burnish on both the interior and exterior. The paste is very fine, dense, and well
fired. Bowl or plate rims are usually direct and
rounded, but often have a slight interior bevel.
Saxamar bowl rims are only very rarely flared,
and are never tapered. Therefore, slip color and
rim form alone are normally enough to distin-
FIGURE 16.1. Late Horizon ceramics from the Taraco Peninsula: (ac) Taraco Polychrome, (dn) Chucuito Blackon-Red, (o) Chucuito Polychrome, (pae) Saxamar (Pacajes-Inka), (afam) Inka Imperial
guish Saxamar from Early Pacajes (Late Intermediate period) unrestricted vessels.3
Saxamar modeled decorative motifs include
what appears to be a duck head (Figure 16.1p)
or more commonly single or double nubs (Figure 16.1qr) protruding from the rim. Painted
decoration is always applied using a black pig[FIGURE 16.1 ]
ment. We found only a single example of a Saxamar polychrome (Figure 16.1p). This example
used a white pigment to color the beak and
neck of a stylized duck head. Cross-hatching on
an interior rim bevel is very diagnostic of this
style (Figure 16.1sw), and an undulating line
occupying the same position is not uncommon
272
of the Inka Imperial style seem to have been produced by state specialists and distributed as part
of the functioning of the administrative apparatus (Costin and Earle 1989; Costin and Hagstrum 1995).
The third most common Late Horizon style
is the Chucuito style. Its center of distribution
seems to be the western Titicaca Basin, south of
Puno, and north of Yunguyu. This was the territory of the Lupaqa kingdom (Hyslop 1976;
Murra 1968, 1970; Stanish 1989a). A small number of trade pieces are present in the southern
Titicaca Basin. The Chucuito styles were originally described by Tschopik, and her analysis is
still the definitive source (M. Tschopik 1946).
Chucuito polychromes (M. Tschopik 1946:
2729) are very rare in the southern Titicaca Basin. Only one example was recovered from the
Taraco Peninsula survey; it seems to represent a
catfish (Figure 16.1o). More common is Chucuito
Black-on-Red (M. Tschopik 1946:2931). These
ceramics are very similar to the Saxamar style in
terms of their basic technology. They also are
fine and dense, and are decorated with a glossy
black pigment over a deep red slip. Diagnostic
that is, paintedexamples are exclusively open
bowls. Tschopik documents a wide variety of
decorative motifs for this style. In our area, very
few of these motifs are represented. Some figurative elements are present (Figure 16.1d, gh),
but most examples present some combination of
spirals, undulating lines, and pendant loops
(Figure 16.1ef, il). In this, at least, our sample
resembles that of Stanish et al. (1997:4748, 55)
from the Juli area of Peru.
The fourth Late Horizon ceramic style is Taraco Polychrome, also defined by M. Tschopik
(Julien 1993:190191). The distribution of this
style seems to be centered in the Peruvian town
of Taraco, on the northern side of the lake. It is
exceedingly rare in the southern Titicaca Basin.
Only four examples were recovered by the Taraco Peninsula survey, three of which are illustrated in Figure 16.1ac. Tschopik described the
style as follows:
a fine, compact ware with a white or
cream-colored paste, containing a very
fine temper and occasional reddish-
274
FIGURE 16.2. Early Colonial period ceramics from the Taraco Peninsula
In essence, this argument requires the assumption of the contemporaneous manufacture of two distinct ceramic types,
emerging from two separate cultural traditions, in the Tiwanaku Valley. Specifically, we need to posit that the Inka
ceramic material represented a tradition
that was superimposed over the existing
cultural matrix, rather than one which replaced it completely. Inka-Pacajaes could
have represented an elite or sumptuary
class of ceramic which was available only
to certain elements of the indigenous
Pacajaes society, possibly determined by
position within the Inka administrative
hierarchy of indirect rule. At the same
time, commoners continued to produce
the diagnostic Early Pacajes wares described above, for everyday domestic
consumption. [Mathews 1992a: 194]
[TABLE 16.1 ]
Lower Tiwanaku
Valley (TLV)
Middle Tiwanaku
Valley (TMV)
Early Pacajes
25
88
94
73
Late Pacajes
57
74
88
55
276
we know of pre-Hispanic household compounds. For example, the Pacajes-Inka compound partially excavated by Karen Wise at
Lukurmata consisted of at least three rooms and
seems to have been enclosed by a stone wall
(Wise 1993). It is therefore very similar to modern compounds. The compound at Lukurmata
measures 16 m on one side, though the other dimension could not be determined. Assuming it
was square, the architectural compound alone
covered 0.03 ha. Since this was an isolated
household (dispersed habitation), many of the
activity areas would have been located outside
of the architectural compound itself, as is the
case with modern houses.
A habitation area of 30 x 30 m is, in Bandys
small sample of modern household compounds,
278
associated with an artifact scatter of approximately 50 x 50 m (0.25 ha). That is, the splash
zone extended for approximately 10 m beyond
each side of the habitation zone. He consistently
found this figure to be a good descriptor of the
splash zones surrounding modern houses, so we
will use this figure in this chapter.9
If we measure area in hectares and model
the plan of the artifact scatter as a square, then:
100000 ( A 20 ) 2
H = --------------------------------------------10000
Where A = artifact scatter area
H = habitation area
Figure 16.4b shows the more complicated
case of nucleated habitation in towns and villages. In this case, we have densely packed habitation areas (residential compounds) with
overlapping splash zones in all but the outer
edges of the habitation area. Almost all of the interior of the settlement is habitation area, with a
rim of uninhabited splash zone around the
edge.10 If we view a village in this way, then we
can apply exactly the same correction equation
to it that we applied to the isolated household
compound. Therefore, there is no need to distinguish between different residential patterns in
order to correct for the splash zone.
In practice, of course, things are often not
this simple. Habitation density (persons per
hectare) may vary in contexts of nucleated or
dispersed habitation, so that the relation between population and habitation area is not direct. Other factors can affect the relation, as well.
For example, habitation density will typically be
higher in walled settlements than in nonwalled
ones (Wenke 1975), as people exchange living
space for a reduction in collective labor investment. However, to derive a useful relation empirically would require very extensive
horizontal excavations or excellent surface visibility and architectural preservation (see de
Montmollin 1987 for a Mesoamerican example).
This has been accomplished in some contexts
(see Milner and Oliver 1999:9093 for an example from Cahokia) with excellent results, but we
are very far from this point in Titicaca Basin archaeology. Mississippian or Mesoamerican data
A
H = -----sAh
Where H = number of households
As = corrected site area
Ah = average household compound area
(0.09 ha)
A population index value for the site is obtained by multiplying the sites estimated number of households by an average number of
persons per household. The same set of informal
observations used to establish an average residential compound area also yielded an average
of six persons per household compound. This
concurs approximately with ethnohistoric information. For example, at the time of Viceroy Toledos visit to Tiwanaku in 1574, a total
population of 4329 was recorded for the repartimiento (Choque Canqui 1993:81). Among
these were 868 tributarios (heads of household).
This yields an approximate ratio of 4.99 persons/household.
Therefore:
Ps = HPh
Where Ps = population index value of site
H = estimated number of households in site
Ph = persons per household (6)
This sequence of steps eventually returns a
population index value of 42.67 persons for a 1ha sherd scatter, and a value of 66.67 persons for
1 ha of occupied area. These numbers are on the
low end of the range of numbers typically used
to derive population estimates from site area.
Robert McC. Adams (1965), for example, used a
figure of 200 persons/ha in his study of Mesopotamian settlement. This was based on modern
population densities from the old quarters of
Baghdad and from other local examples. Carol
Kramers ethnoarchaeological study of modern
settlements in Southwest Asia yielded an average of 120 persons/ha (Hassan 1981:6667). Similarly, the classical city of Melos has been
estimated to have contained between 130 and
200 persons/ha within its walls (Whitelaw and
Davis 1991:280). While our numbers therefore
seem low compared to figures employed in the
analysis of Mediterranean and Near Eastern urban settlements, they fall within the 2550 persons/ha range of the Basin of Mexicos High
Density Compact Villages (Parsons 1976: 72).
Our numbers are also considerably lower than
the 100 persons/ha used by Parsons to estimate
the population of Tiwanaku (Parsons 1968). Finally, 4367 persons/ha falls into the middle of
the range of Robert Drennans (1986: Table 13.1)
compact settlement category of Mesoamerican
archaeological sites.
Once the population index value has been
calculated, analyzing population change
through time is simply a matter of tracking the
increases and decreases of this quantity. The
most convenient way to do this is by calculating
an annual rate of population growth relative to
the previous phase. This is done using the following equation (Hassan 1981:139):
Pf
1
r = ---In ------
Pi
T
where r = annual rate of population change
T = number of elapsed years
Pi = initial population index value (at T = 0)
Pf = final population index value
Multiplying the derived value of r by 100
yields an annual percentage rate of population
growth. This is the figure that we will use in this
study to measure rates of population growth or
decline, since growth rates seem often to be reported in terms of annual percentage rates (Hassan 1981:140). Population growth rates can in
this way be calculated for individual sites,
groups of sites, or entire regions.
280
[FIGURE
[FIGURE16.7
16.6]]
10000
Pacajes-Inka
a)
8000
6000
4000
2000
500
1000
1500
2000
10000
Late Pacajes
b)
8000
6000
4000
2000
500
1000
1500
2000
FIGURE 16. 6. Distribution of population by site size: (a) Pacajes-Inka, and (b) Late Pacajes
1977). Together, the log-normal rank size distribution and the three-level site size hierarchy
suggest that the study area was a closely administered and tightly integrated province of the
Inka Empire.
When the different survey areas are considered individually, however, it becomes apparent
that the Pacajes-Inka phase settlement patterns
vary dramatically. Table 16.2 displays a series of
settlement statistics by survey area. First, settle-
282
10000
Pacajes-Inka
Population index
1000
a)
100
10
10
100
1000
10000
Rank
10000
Late Pacajes
Population index
1000
b)
100
10
10
100
1000
10000
Rank
FIGURE 16.7. Rank-size diagrams: (a) Pacajes-Inka, and (b) Late Pacajes
Taraco Peninsula
Katari Basin
Number of sites
292
125
79
3761
3676
7868
12.9
29.4
99.6
Sites/km2
1.46
1.28
0.77
18.8
37.5
77.1
Pop. index/km
lower than the site densities in the other two areas. This means that the Katari Basin contains
fewer but much larger sites than the other two
areas. The Katari Basin has an average site population index value of 99.6, 3.4 times that of the
Taraco Peninsula and 7.7 times that of the Lower
Tiwanaku Valley. The Katari Basin was the demographic and political core of Pacajes-Inka
phase settlement within the study area. The
other two areas were peripheral and subsidiary.
Settlement within the Katari Basin seems to have
been tightly controlled and nucleated, while in
the other areas it was more fluid and dispersed.
[TABLE 16.2 ]
[FIGURE 16.8 ]
Katari Basin
Taraco Peninsula
North
20 km
Tiwanaku
Rivers
Roads
200 m contour interval
International boundary
Administrative boundary
284
Taraco Peninsula
Katari Basin
Number of sites
342
326
105
2793
5021
5155
8.2
15.4
49.1
Sites/km
1.71
3.33
1.03
Pop. index/km2
14.0
51.2
50.5
0.50%
0.52%
0.70%
Katari Basin
Taraco Peninsula
North
20 km
Tiwanaku
Rivers
Roads
200 m contour interval
International boundary
Positive pop. growth
Negative pop. growth
Administrative boundary
Lower Tiwanaku Valley survey area), non-Taraco north (including part of the Taraco Peninsula survey area and all of the Katari Basin
survey area), and non-Taraco south (including
most of the Lower Tiwanaku Valley survey
area). The differences in the population growth
rates are even more pronounced than those between the survey areas themselves. In the Late
Pacajes phase, Taraco experienced population
growth at the very rapid annual rate of 0.76%.
At the same time, adjacent areas experienced
population decline, at rates varying between
0.66% and 1.29% annually.
The baseline population growth rate for the
southern Titicaca Basin throughout prehistory
was on the order of 0.10% annually (Bandy
2001b). This being the case, the very high rate of
Late Pacajes population growth for Taraco
seven times the baseline ratecertainly indicates an influx of population into the area. Such
a dramatic growth rate cannot be accounted for
by internal growth alone. The negative rates in
the adjacent Katari Basin and Lower Tiwanaku
Valley reveal the source of this influx. Very
clearly, people were abandoning their homes in
the Katari Basin and Tiwanaku Valley and were
moving onto the Taraco Peninsula during the
[TABLE 16.4 ]
286
1500
Late Pacajes phase
1000
500
-500
-1000
Taraco
Not Taraco
-1500
505000
510000
515000
520000
525000
530000
535000
540000
545000
UTM East
Non-Taraco (North)
Non-Taraco (South)
Pop. Index
Growth
Pop. Index
Growth
Pop. Index
Growth
Pacajes-Inka
3677
8919
2711
Late Pacajes
5800
+0.76%
5917
0.66%
1251
1.29%
CONCLUSIONS
What then of the Early Colonial period demographic catastrophe? Using the tools of settlement archaeology, we have identified several
trends that bear on this issue.
First, the Early Colonial period in the southern Titicaca Basin was in fact characterized by a
demographic downturn. The Late Pacajes phase
population growth rate in our study area was
0.28% annually. Thus, whether through disease,
warfare, or outmigration, population did decline
during this period. The rate of decline, however,
was not nearly as severe as previous researchers
(Albarracin-Jordan and Mathews 1990; Albarracin-Jordan 1992; Mathews 1992a) have suggested.
Second, we have identified a general shift toward more dispersed patterns of settlement
throughout the study area. In each of the three
survey areas, the Early Colonial period was characterized by an increase in the number of sites
and a decrease in the average population of sites.
This fact would seem to contradict a general ethnohistoric emphasis on reduccines, the forced
concentration and resettlement of populations
that were supposed to have accompanied the
Toledan reforms. If our evidence is at all representative, the reduccin may be more a documentary phenomenon than an empirical one.
Dispersal of population would certainly facilitate
the manipulation of census figures and tribute
ACKNOWLEDGMENTS
The authors would like to acknowledge the assistance of Christine Hastorf, Alan Kolata,
Amanda Cohen, Matthew Seddon, Erika Simborth, Robin Beck, Eduardo Pareja, Felipe
Choque, and Facundo Llusco. The lower Tiwanaku Valley data we are using were, of course,
collected and published by Juan Albarracin-Jordan.
NOTES
1.
288
2.
3.
4.
5.
6.
17.
Archaeological Reconnaissance in the
Northern Titicaca Basin
Charles Stanish, Amanda B. Cohen, Edmundo de la Vega,
Elizabeth Arkush, Cecilia Chvez, Aime Plourde, and
Carol Schultze
MEMBERS OF PROGRAMA COLLASUYU have conducted archaeological reconnaissance throughout the Lake Titicaca Basin in the last several
years. The intent of this work was to assess the
nature and variety of archaeological settlement
in a number of areas in order to formulate future
research designs. As a result, many sites have
been recorded that do not fall into the area of
any ongoing, systematic study. In this chapter,
we provide data from a number of important
sites in the northern and western basin area discovered on such reconnaissance fieldwork.
We make the distinction between intensive
survey, systematic reconnaissance, and reconnaissance (Stanish 2001b). A reconnaissance refers to nonsystematic surface survey with the
intent of finding and recording archaeological
sites. Reconnaissance is particularly useful in areas where we have little knowledge of the range
and nature of the archaeological remains. This
research strategy assists archaeologists in defining broad patterns of settlement and style distribution, and serves to generate hypotheses for
future testing. It is in this context that these sites
were discovered. The data generated from this
work helped us define new projects in several
areas of the region, including the Pukara Valley
(Cohen 2000, 2001), the Huancan-Putina Valley
(Stanish and Plourde 2000), and in the Puno Bay
area (Schultze 2000).
The first modern reconnaissances in the Titicaca region were conducted by Jos Mara
Franco Inojosa and Alberto Gonzlez (1936),
Emilio Vsquez (1937, 1939), Alfred Kidder II
(1943), Mximo Neira (1967), and Marion
Tschopik (1946) in the 1930s through the 1960s.
Vsquez (1940:143150) described a number of
important sites in the Peruvian Titicaca Basin,
such as Sillustani, Cutimbo, Kacha Kacha, Tanka
Tanka, Siraya, Maukallajta, Cheka, Wilakolla,
and Taraco. At Cheka, he described the cut
stones known as El Bao del Inca, and described monoliths in the north near Huancan
and Pukara. M. Tschopik published an article in
1946 that served to define a number of pottery
types, including the Sillustani Series, Chucuito,
and local Inca types. Kidder, likewise, reconnoitered the northern Basin, concentrating on earlier sites. He documented a number of Pukara,
Pukara-related, and Tiwanaku sites throughout
the region. Neiras (1967) report stands as one of
the most important sources of information for
the northern and northeastern sides of the lake.
Likewise, Manuel Chvez Balln, Elas Mujica,
John Rowe, Sergio Chvez, and Mario Nez
discovered a number of sites throughout the
northern area and beyond.
In the mid-1970s, John Hyslop (1976, 1977a)
conducted a more systematic reconnaissance
throughout the western Titicaca region in the
289
290
SURVEYS
In the last decade or so, our program has intensively surveyed or reconnoitered a large number
of areas in the Titicaca Basin (Figure 17.1). The
Juli-Pomata survey covered an area measuring
approximately 360 km2. The reconnaissance associated with this latter project extended from
Pomata to Yunguyu to the east, and southeast to
Desaguadero. These data have been published
in several publications (e.g., Stanish 1994b, 1997;
Stanish et al. 1997). The Island of the Sun was intensively surveyed by one of us (C. Stanish) and
Brian Bauer in 1994, 1995, and 1996, and the final
results have been published (Bauer and Stanish
2001, Stanish and Bauer 2004). Amanda Cohen
surveyed the Pukara Valley in 1998 and 1999
(Cohen 2000, 2001). The Huancan-Putina Valley
was surveyed in 1999 and 2000 by Stanish and
Plourde (Stanish and Plourde 2000). Carol
Shultze has conducted a survey in the Puno Bay
area. All of the regions outside of these areas are
the subject of this chapter, although selected
sites from within these regions have been included as well.
CERAMIC CHRONOLOGY
Numerous seasons of excavations, systematic
survey, and reconnaissance have provided a corpus of discrete ceramic types that were manufactured and used for restricted periods of time in
the region. In the north and northeast Titicaca
Basin, the following pottery types have been
used to date sites discovered on the proposed
survey.
292
Late Huaa
There are huge areas in the northern Titicaca Basin that do not have Tiwanaku occupations.
These areas are replete with sites in productive
zones that have all other pottery types. Some
sites, in fact, have a small scatter of finely made
Tiwanaku pieces, but not in sufficient quantities
or varieties to indicate an occupation dating to
approximately AD 6001100. We hypothesize the
existence of a local Tiwanaku-contemporary pottery style that we call Late Huaa. (Early Huaa
pottery is described below). Late Huaa has
some kero-like forms that are poorly executed in
local pastes. It is also likely that some pottery
types with Collao paste are earlier than the
traditional dates for the LIP, and instead represent Late Huaa types.
FIGURE 17.2. Example of Late Horizon period pottery in the Huatasani style
pottery or Pukara-affiliated styles around the region. However, virtually no sites north of Pukara
toward Ayaviri have Tiwanaku pottery in any
quantity to qualify as a bona fide Tiwanaku-affiliated settlement. The Juliaca area is full of Tiwanaku sites, and the northernmost Tiwanaku site
known to date in this region is that of Maravillas
(see below). Other late Tiwanaku sites are found
up the Ayabacas River Valley. Around Lampa,
one of us (C. Stanish) noted some dense Tiwanaku artifact scatters but did not have permission
to work in the area. Toward the northeast, our
most recent surveys have found substantial
numbers of Tiwanaku sites along the main lake
edge, along the road, and along the edge of Lake
Arapa. However, there are no Tiwanaku sites
away from the roads or lake edge, and there are
none in the Putina area.
In short, there are presently no known Tiwanaku III (Qeya) sites in the area (with the possible exceptions of Isla Estves and Cerro
Chincheros). Tiwanaku Expansive (very late Tiwanaku IV and Tiwanaku V) settlements are restricted to enclaves in a few areas along the two
roads on the sides of the lake, and along the road
to Cuzco. Outside of these areas, there are no
known Tiwanaku sites.
We find numerous sites that have Middle
Formative pottery, Pukara or Pukara-related
pottery, Altiplano period pottery, and occasionally Inca pottery. It is certainly possible that
there was a major abandonment of sites in the
area with the cessation of Pukara pottery production and the dispersal of settlement, with a
subsequent reoccupation of these large sites
with the advent of the Altiplano period.
Can this apparent hiatus be explained with
another model that incorporates these observations? The more parsimonious explanation is that
our ceramic chronology is not accurate, and that
Tiwanaku III (or Qeya) pottery is not a diagnostic
for this time period in the extreme north between
the collapse of Pukara (circa AD 300400) and the
expansion of Tiwanaku (circa AD 600). In other
words, it is more likely that many sites located in
rich agricultural and pasture lands away from
the major road system were continuously occupied and that our ceramic chronology is unable
to distinguish the Tiwanaku III- and Tiwanaku
IV-contemporary occupations.
294
RECONNAISSANCE DATA
Project Site Numbers
We use a simple designation for sites found in
the Titicaca Basin for the purposes of our general
reconnaissance. An identifying letter or letters
that represent a geographical area is followed by
a number. This identification system is intended
to be temporaryuntil such time that the sites
are incorporated into individual research
projects and input into the database of the Instituto Nacional de Cultura, Puno. The letters have
little meaning apart from a general area identification. The numbers are assigned as the sites are
recorded into the Programa Collasuyu master
site register, and are therefore analytically meaningless. The list of letter designations and their
corresponding geographical area is as follows:
P=Puno area; H=Huata area; T=Taraco
area; Hu=Huancan area; Pc=Pukara
Valley; C=Chucuito area; S=Sillustani
area; Ch=Capachica Peninsula; I=Ilave
River area; M=Moho area; J=Juliaca area;
Az=Azangaro area.
The time period designations are as follows:
MF=Middle Formative; UF=Upper Formative; E. Huaa=Early Huaa;
L. Huaa=Late Huaa; Tiw=Tiwanaku;
Alt=Altiplano; LH=Late Horizon.
Canchacancha-Asiruni (Tintiri)
The site of Cancha Cancha-Asiruni was first systematically reported and renamed by S. Chvez
and K. Chvez (1970). They noted, as with so
many previously reported sites in the northern
Titicaca Basin, that the original scientific discoverer was Manuel Chvez Balln. It is located
northeast of Azngaro and across the river from
the hacienda of Tintiri. Tintiri was apparently the
name that Chvez Balln used for the site (e.g.,
[FIGURE 17.3]
[FIGURE 17.5
17.4 ]
296
Size: at least 12 ha
Periods: MF, UF, E. Huaa (?), L. Huaa,
Alt, LH
Programa Collasuyu ID: Az-1
Additional references: Rowe (1963:7),
S. Chvez and K. Chvez (1970)
Capachica
Capachica is today a small town on the peninsula of the same name. In the Toledo Tasa, the
298
Carpa
Carpa was first mentioned by M. Tschopik in her
1946 publication. Likewise, Neira (1967) mentions it in his reconnaissance report. The site is a
beautifully preserved Inca administrative center
located directly adjacent to the modern road
from Huancan to Moho (Figures 17.617.7). It is
located at the neck of the Wilasalto Peninsula.
The modern village is built on the Inca site and
follows the Inca grid pattern. The streets in the
village are in fact original, and there are many
Inca walls still being used. The pottery on the
site is overwhelmingly characterized by wellmade local Inca wares. This site is an important
Inca administrative center on the Omasuyus
highway. It is small by western basin standards,
but is a significant secondary urban center on
this side of the lake.
Size: at least 3 ha
Periods: LH
Programa Collasuyu ID: Hu-2
[FIGURE 17.7
17.6 ]
Cebaduyu (Tinpunku)
This site is located along the south bank of the Ilave River, about 150 to 200 m from the present
course of the river. It is on a low hill with domestic terraces, a kind of site that we have referred
to as type 4 in previous publications (Stanish
et al. 1997; Stanish 2003). Type 4 sites are basically low terraces on hills used as surfaces to
build structures. They may also originally have
been built as agricultural terraces and then later
used as a building site. Conversely, terraces used
for construction are routinely converted to agricultural planting surfaces.
There is a very high density of pottery on
this site, including Upper Formative, Tiwanaku,
Late Intermediate, and Late Horizon diagnos-
Cerro Ccacca
This is a small Tiwanaku hamlet on the northwest side of Cerro Ccacca. It is located on the
low hill that juts into the pampa off of the main
hill. It is a type 4 domestic terrace site about 50 x
50 m in size. The pottery is not very fine. Above
the site to the south is Pukara Ccacca. This latter
site has a substantial Altiplano period component with Inca and pre-Inca chulpas, at least on
the west side.
Size: approximately 1 ha
Periods: Tiw
Programa Collasuyu ID: C-3
Cerro Chincheros
This is one of the most important sites in the
western Titicaca region. This huge type 3 settlement is characterized by one of the best-preserved sunken courts in the area. Type 3 sites are
low hills with terraced sides that lead up to an
artificially flat area with some kind of court or
other corporate constructions. They served as
primary and secondary regional centers in the
Titicaca Basin from the early Formative period
to Tiwanaku times. A description of the site
types is found in Chapter 7 of this volume.
Cerro Chincheros is situated just outside of
Puno, along the Puno-Cuzco highway. The
sunken court is approximately 20 x 30 m in size,
making it the largest such construction outside
of the Pukara and Tiwanaku areas. Surrounding
the sunken court on Chincheros is between 15
and 20 ha of domestic terracing. The existing
walls of the court are up to 3 m high and they
appear to be virtually intact. Likewise, there are
large slabs around the court area, suggesting
that it was faced with cut stone blocks. The pottery on the surface includes Qaluyu or Qaluyulike fragments, Pukara, Tiwanaku III, Tiwanaku
IV/V, and Middle Formative neckless ollas.
Some Altiplano period and Inca fragments are
300
Cerro Cupe
On the road to Hatuncolla is this major Middle
Formative through Altiplano period site. This
type 3 site has a very high quantity of debris, including decorated pottery, bone, and many agricultural implements. There is a substantial
quantity of well-made Tiwanaku pottery, as well
as a few fragments of Pukara incised. At the top
of the site is an area that could have been the location of corporate architecture. There is a
chulpa at the top. It is a large, round, fieldstone
chulpa about 7.0 m in diameter. There are about
six major domestic terraces along the side of the
hill facing the pampa.
The name Cupe or Cupi is mentioned in
documents as a village of craft specialists in the
Titicaca Basin during Inca times (e.g., see Julien
1993:189). Julien locates two such toponyms, one
between Chucuito and Acora and a second near
Ayaviri. This site does not seem to be related to
these documentary accounts, and the lack of Late
Horizon materials on the surface confirms this.
Size: approximately 2.0 ha
Periods: MF, UF, Tiw, Alt
Programa Collasuyu ID: H-2
Cerro Marenza
Cerro Marenza is an Altiplano period site on a
hill of the same name overlooking the Huata
Pampa. There are many slab-cist tombs and possible chulpa bases all along the ridge. It is likely
that M. Tschopik visited this site during her reconnaissance (see M. Tschopik 1946). Virtually
all of the diagnostic pottery is Altiplano period
in date. The pottery includes the very diagnostic
Collao black-on-red and plainwares with identical pastes. Such Altiplano period cemetery sites
are common in the region. The existence of this
site confirms the large post-Tiwanaku occupation in the area. Some of these sherds could date
to the Late Huaa period as well.
Size: large cemetery area
Periods: Alt
Programa Collasuyu ID: H-5
Chacchune
We believe that this site was briefly referred to
by Kidder and Neira in their respective reports
(Kidder 1946; Neira 1967), but the descriptions
are too incomplete to be certain. The site is located on the northeast side of the isthmus that
faces the Huata Peninsula.4 It is a small but major pukara on this side of the lake dating to the
Altiplano period, as indicated by surface collections (Figure 17.1), site plan, and architecture.
There are a number terraces that also seemed to
function as defensive walls up the sides of the
hill. These terraces are very well preserved. In
fact, on the side toward the main access road are
a series of domestic terraces that still preserve
aboriginal houses. The site was densely populated. It is at least 4 ha in size, and probably includes 2 ha more.
The top of this pukara is very important.
There is a plaza-like area on what appears to be
small structures that are best interpreted as storage rooms on either end. We counted about five
structures on each side. Each one was about 2
m2. Behind the storage structures are other
buildings that are most likely residential in nature. This site would make an excellent case
study for the analysis of Altiplano period architecture.
Size: > 4 ha
Periods: Alt
Programa Collasuyu ID: M-1
Checca A
This site was discovered by John Hyslop (1976).
We confirmed that this type 4 site has a moder-
Chimpajara
This site is a small mound on an old oxbow terrace near the Ayabacas River. Today, there is a
house compound built over a raised-mound
area. There is no evidence of corporate architecture. We found two obsidian points, Pukara diagnostic pottery, and a number of plainwares
suggestive of post-Tiwanaku types. The lithics
on the surface include agricultural tools and debris. There are no evident Altiplano period diagnostics or any Tiwanaku pottery.
This is a good example of a small Pukara village. This is significant because it establishes the
fact that small villages with relatively high
quantities of Pukara pottery exist in this area.
This contrasts substantially with the regional
data south of the Ilave River, where no such
Pukara villages are found. In the south, all diagnostic Pukara pottery (or perhaps varieties of Tiwanaku I or Sillumocco Polychrome Incised
see Steadman 1994) are found in small quantities
on large sites. Such data serve to reinforce our
methodological principles of distinguishing between bona fide settlements linked to larger regional polities and settlements with trade wares.
Size: approximately 1.0 ha
Periods: Pukara, E. Huaa (?), L. Huaa (?),
LH
Programa Collasuyu ID: J-5
Cocosane
This small Upper Formative site is located on the
small road west of Isla Escata. The road cuts
through some middens and an area with a number of stone blocks. In 1995, there were a number
Cotos
This site is located on a natural geological rise
near the lake, south of Taraco and north of Pusi.
It is a type 4 site with well-made, broad terraces
up the saddle of the hill. At the crest of this hill
are a number of sharp rock outcrops. From here
there is an exceptional view of the cordillera and
lake. Below this crest and toward the lake is a
flat, marshy area that may have had raised
fields. There is an impressive density of artifacts
on the surface, including pottery, agricultural
implements, and midden remains. There are a
number of Tiwanaku diagnostics. We also found
a small piece of a carved slab, which is typical of
Upper Formative settlements.
Size: 12 ha
Periods: UF, Tiw, LIP, LH
Programa Collasuyu ID: T-1
302
Huajje
Huajje is a huge, U-shaped artificial mound opposite Estves Island. The interior of the U
faces Estves Island. The mound is located on
the road that passes along the edge of the hill
and that ultimately curves around and heads
north to the Paucarcolla Pampa. The northeast
side of the mound has a double platform. We
were unable to examine the top of the site. The
road to Estves Island cuts directly through the
mound on the north and northeast sides. In this
cut there is evidence of at least 2 m of construction fill, including a small canal or sewer that is
exposed in the profile. The fill is composed of
midden material and quarried soil.
The entire south face of the mound, which
faces Estves Island, measures about 450 m in
length. The southeast side of the mound is badly
damaged, but it is clearly part of a single architectural complex. This side of the mound appears to be natural and was probably even
flattened to obtain the desired shape. There is an
internal terrace wall that connects the two arms
of the U and faces a low rise. This mound is
very similar to, though much larger than, the
mound at Tumatumani, near Juli (Stanish and
Steadman 1994). It is worth noting that at Tumatumani the U-shaped mound also was partially
constructed from an excavated hill, and the fill
was used to build up the other side of the
mound in order to obtain the desired shape of
the mound.
The surface artifacts include abundant lithic
debris and pottery sherds, including numerous
Middle Formative, Upper Formative, Tiwanaku,
and later diagnostics. There are numerous agricultural implements. This site is extremely important, as it was clearly part of the Tiwanaku
complex associated with Estves (Schultze 2000).
However, unlike Estves, this mound was very
large in the pre-Tiwanaku periods. We can, therefore, provisionally suggest that Huajje was a major Upper Formative period center in the Puno
Bay prior to the Tiwanaku occupation of the area.
Coincident with the incorporation of the bay area
by Tiwanaku, Huajje continued in importance, as
evidenced by the Tiwanaku occupation of the
site. Huajje, along with Estves, constitutes an
ideal database for modeling the interaction between the Tiwanaku state and local elites.
Size: at least 5 ha
Periods: MF, UF, Tiw, LH
Programa Collasuyu ID: P-5
Huayna Roque
The hill of Huayna Roque dominates one of the
largest hills in the Juliaca area. It was first reported by John Rowe in the 1960s. The site superficially looks like a classic type 3 site with
many, broad domestic terraces around the hill
(Stanish et al. 1997). There is abundant pottery
and other domestic debris on the surface. Agricultural implements and other lithic artifacts
abound on the site. Unfortunately, the site is so
badly disturbed that it is very difficult to determine its extent.
Size: > 5.0 ha
Periods: MF (?), UF, Tiw, Alt, LH
Programa Collasuyu ID: J-2
Huimpuco
This is one of the Ro Ilave mound sites. It consists of two mounds, each about 4 to 5 m high
and approximately 15 to 20 m in diameter. The
small diameter makes for a conical-shaped
mound that we had never before seen in the Titicaca Basin. The site is largely Upper Formative,
with some possible Altiplano material. There is
no Tiwanaku or Inca pottery on the site. It is possibly a walled compound as well, with a high
hilly rise and exterior wall enclosing these two
principal mounds.
These mounds represent a settlement type
not previously defined in the Juli-Pomata region.
These mounds are found on the Ramis River in
the north, suggesting that this is a kind of site
that is typically found on river floodplains.
Size: approximately 1.0 ha
Periods: UF, Alt
Programa Collasuyu ID: I-10
Isla Escata
A small Tiwanaku site was located on the south
center of the former island. It is now a peninsula
and one can drive to the town that is on the island. The site appears to be very small, perhaps
50 x 50 m at most, although we could not survey
the area very effectively. The site is located partially in a modern house compound and we
Kacha Kacha A
Kacha Kacha A is a site named by Hyslop in his
dissertation (1976). This site includes the famous
chulpas south of Acora on the hill first published
by Ephraim Squier in 1877 and mentioned by
naturalists, travelers, and scholars ever since.
Above the chulpa area is a large, classic, type
3 site that has evidence of occupations from the
Middle Formative through the Inca period. The
site covers about 4 ha of area. Although there is
no visible corporate architecture, it is possible
that a small court exists at the top. The artifact
density is particularly high on the top of the hill
and on the domestic terraces along the sides of
the site.
Size: approximately 4 ha
Periods: MF, UF, Tiw, Alt, LH
Programa Collasuyu ID: I-8
Maravillas
Maravillas is located just north of the bridge outside of Juliaca that crosses the Ro Ayabacas on
the highway to Pukara. The entire hill flank to
the west of the road is covered with raised fields
and terraces. The terraces are both agricultural
and domestic. On the northern part of the site is a
large, type 1 mound about 1 ha in size. This is a
gross estimate, however, because there is a modern building on the mound at present, and there
has been substantial damage to the mound. Behind the mound and to the north and south are
domestic terraces with Middle Formative, Upper
Formative, Tiwanaku, Late Intermediate, and
Late Horizon materials on the surface. Furthermore, on the flats between the raised fields and
terraces are large domestic compounds, reminiscent of Formative and possibly Tiwanaku period
constructions in other sites in the region. One of
the better-preserved compounds measures 30 x
42 m. The walls are made with very small stones.
304
Ojerani
This is a site area north of Chucuito that Hyslop
(1976) described as a possible Tiwanaku site. We
were able to walk on the south and west sides of
the valley, as well as walk through the center of
the pampa where the most intensive agriculture
is found. We found no Tiwanaku materials, but
we did find some late diagnostics, particularly
Inca period pottery.
This valley is very rich. It has its own
springs with canals, and some deep wells at the
base of the valley with a fair amount of fresh water. Maize grows here in abundance. At first
analysis, it looks like a nice little Late Horizon
pocket that took advantage of the maize and
other agricultural potential.
Size: dispersed
Periods: LH
Programa Collasuyu ID: C-5
Pachamama Amantan
Pachamama Amantan is the circular structure
several meters in diameter with five concentric
walls inside that is located on the hill of the
same name. There is some resemblance to the
circular structures as Sillustani. There are no
sherds associated with the immediate area of
this site. However, the north face of the hill, the
side on which the structure lies, is full of Inca
pottery and some Altiplano materials. We found
no pre-Altiplano materials in this area. The relationship between the habitation areas and this
structure remain problematic.
Size: < 1.0 ha
Periods: Alt, LH
Programa Collasuyu ID: Ch-4
Pachatata Amantan
Niles (1988) has written an article describing the
Pachatata Amantan, a high hill with a sunken
court at the top. The sunken court is built with
fieldstones and measures about 14 x 14 m in size.
The overall style of the sunken court is Tiwanaku or Upper Formative, albeit with some differences from sunken courts elsewhere. Features
that are typically Tiwanaku and/or Upper Formative on this site include the courts semisubterranean construction, the courts size, a corner
doorway in the court that is virtually identical to
that found at Lukurmata, and a stairway similar
to that at Tiwanaku itself. On the other hand, the
second stairway is not typical of Tiwanaku canons, nor is the fieldstone construction of the
court. The outside wall, that today is an obvious
modern construction, appears to be built on preHispanic foundation stones. If this outer foundation wall is indeed pre-Hispanic, then it is not a
typical Tiwanaku construction. Finally, there is
no evidence of a Tiwanaku period or earlier occupation on the hillside below Pachatata, although Niles reports finding some Tiwanaku
pottery in the region (Niles 1988:Figures 67).
Without intensive excavations, it is difficult to
know the precise architectural history of the
temple construction. At the present time, the
best estimate is that the temple was originally
built as a Tiwanaku construction, with a rebuilding of the area by the Inca or later peoples.
Size: total site > 2.0 ha
Periods: LH
Programa Collasuyu ID: Ch-2
Paucarcolla
Paucarcolla is a fairly large modern village located directly on the Puno-Juliaca highway. Paucarcolla was a moderately large Early Colonial
settlement as attested to by the Toledo Tasa. The
site had 1003 taxpayers and more than 4500 individuals (Toledo 1975:59). The town was divided
into Aymaras and Uros, with the latter constituting about nine percent of the total population. In
the Tasa, it is interesting to note that apart from
the usual tribute items such as meat and wool,
the people of Paucarcolla also contributed dried
fish and salt (Toledo 1975:60). The area therefore
was most likely an important area for salt pro-
Paucarcolla-Santa Barbara
Paucarcolla-Santa Barbara was first excavated
by Lee Steadman (1995). Her work demonstrated that there was a substantial time-depth
to the site, beginning in the earliest Formative
periods and continuing on to the Tiwanaku period. Ilana Johnson (2003) conducted work at the
site confirming these dates. The site is composed
of a series of very large and wide terraces that
encircle the Santa Barbara Hill outside of Paucarcolla. The terraces are littered with domestic
debris on the surface, including high densities of
pottery sherds, agricultural implements, and
other lithic debris. The terraces and associated
habitation area cover as much as 15 ha, making
306
Puente Caminaca
The Puente Caminaca site is located adjacent to
the river on the southern side of the Ro Ramis.
The river at this point is deeply entrenched, and
there is a high and broad river terrace to the
south. The site is located on this natural terrace.
The site limit is about 300 to 500 m away from
the river, a distance that shifts along the northern limit of the site area as the river twists. A series of low mounds and some retaining walls
comprise the site. These mounds are scattered
along a length of the river terrace measuring approximately 250 m and are oriented more or less
parallel to the river along the terrace. This site
type is different from what we see in the JuliPomata area and is more typical of the Ilave
River mounds. In this case, the site appears to be
a series of type 2 mounds that were built in an
unplanned fashion on an ancient river terrace.
There is abundant midden material on the surface, as well as looted tombs and exposed walls.
All indications are that this is a domestic site.
There is no evidence of corporate architecture.
There is a dense scatter of pottery, lithic flakes
(including basalt), bone, and carbon. There is
Pukara Kapalla
This site is located a few kilometers south of
Acora on the main highway (Figure 17.8). It was
first reported by Tschopik in her 1946 report (M.
Tschopik 1946:8, 16). It is a huge mesa formation
with a number of well-preserved, igloo chulpas
leading up a ridge to the major pukara. The site
has several large defensive walls with doors,
and at least 2 ha (probably more) of habitation
area. It is reminiscent of Tanka Tanka. There are
only igloo chulpas on this site and no other
chulpa style. There are also slab-cist tombs. Furthermore, the sherds appear to be early, including several that are kero-like on the base. It is
safe to conclude that this site is early in date. We
made several collections of the site. The only
Late Horizon sherds are found at the very top,
near a well-maintained apacheta.
One of the igloo chulpas is unusual because
it is on a rectangular base about 7.0 x 8.0 m in
[FIGURE 17.8 ]
size. The chulpa itself is about 2.5 to 3.0 m in diameter and sits on top of the base. There are
other clusters of chulpas on the site outside of
the habitation area. These are between 3.0 and
5.0 m in diameter. The pottery in this area is
mainly Altiplano period in date, but there are
some Tiwanaku-like pieces and some Inca pottery, as mentioned above.
The top of Pukara Kapalla has several walls
that were built across the ridge to fully protect
the top of the habitation area. Behind the site on
the opposite side is a very high, natural cliff.
This site is significant for a several reasons. First,
it appears to be one of the rare early Altiplano
period occupations without a later Altiplano period occupation. Pukara Kapalla is therefore
good evidence for the transition from Tiwanaku
to Early Altiplano period. Also, there are some
different post-Tiwanaku period sherds with a
kaolin paste and red slip. There are also many
straight-sided bowls that could be derivatives of
keros. Therefore, this site is evidence for an early
post-Tiwanaku fortified settlement type. Unlike
other pukaras in the region (with the exception
of Huichajaja), this site has good evidence for
transitional Tiwanaku/Altiplano period pottery. This is evidence for population continuity
between these two periods and argues against
the Aymara migration hypothesis.
Size: approximately 2.0 ha
Periods: Alt, scattered LH
Programa Collasuyu ID: I-2
Pukara Totorani
This minor pukara (see Stanish et al. 1997 for a
definition of major and minor pukaras) is located near the town of Totorani in the upper Ilave River area. This is a small pukara, consisting
of two and three walls encircling the hill. Along
the base of the hill are the remains of looted Altiplano period tombs, Altiplano period pottery
fragments, and a few raised canals that extend
into the marshy area near the river. We found no
obvious habitation area, but we only reconnoitered a small area on the roadside about halfway
up the mountain. This pukara is typical of the
hundreds found in the puna away from the lake.
Small, and probably designed to round up camelids and serve as a defensive refuge in raids,
this site was not permanently inhabited.
Size: total complex < 3.0 ha
Periods: Alt
Programa Collasuyu ID: I-12
Punanave
Punanave was first discovered and reported by
Mario Nuez in his reconnaissance of the Puno
Bay (1994). The site is located on a crest high
above Puno near the pass to Moquegua. It is
composed of a series of broad, rectangular terraces with considerable quantities of debris on
the surface. These terraces comprise the normative domestic compound on the site and are usually associated with pre-Altiplano or Late
Intermediate period settlements. One typical terrace, for instance, measures 23 x 33 m. It most
likely housed an extended family or larger social
unit. Disturbed areas on the site indicate that
these terraces were domestic, with large quantities of midden material eroding out of road cuts
and other exposures. The terraces on this site are
particularly broad due to the low, undulating,
natural topography onto which they were built.
The site is huge, possibly containing as many as
fifteen contiguous hectares of domestic residences. Curiously, there is no evidence of corporate architecture on the site. There are no
obvious sunken court areas, nor are there any
cut or shaped stone blocks on the surface. This is
308
Puno Puno
Puno Puno is located directly on the road on the
eastern side of the lake, near the gate to the
Moho province. In fact, the road runs along the
lowest occupational areas of the site on the western side. Puno Puno is a very fine type 3 site located near the lake and the possible raised-field
areas. This, of course, would be significant, because there are few known raised-field areas on
this side of Lake Titicaca. The hill on which the
site sits has about three to four terraces on the
east, and about five terraces on the west. We did
not have permission to go very far to the south,
so the estimate of the site size assumes that the
domestic areas ended where the flat land begins,
as it does in rest of the site.
We found a fair amount of red-slipped pottery fragments that were probably Pukara plainwares. We also found a fair amount of Pukara
incised pieces. Apart from that, there were few
other diagnostics on the surface, in spite of the
fact that there were fairly dense concentrations
of pottery on the surface. This site is a significant
secondary regional center in at least the Upper
Saman Sur
This is a modest type 4 village located near the
Ramis River. The artifact scatter is light, but it
has good diagnostic Tiwanaku and Upper Formative pottery. There are only a few (perhaps
three or four) low terraces that were habitation
Sillustani
Sillustani is one of the most famous sites in the
altiplano. Today, it is a national park and major
Sucano Bajo
This is another small mound located on the ancient river ridge. The ridge more or less parallels
the Ilave River, and the site is located about 1 km
to the north of the river. This site is one of dozens of such large and high mounds on this ridge.
The mound is about 2.0 to 3.0 m high. It is in essence a well-preserved type 2 mound that has a
slightly conical shape.
310
Taparachi
This site is located in the edge of the pampa
south of Juliaca, west of Caracoto. By the standards of the Juli-Pomata survey, this is not an exceptional site. It is a type 4 Altiplano and Inca
domestic terrace site at the edge of the pampa.
However, since this area is so poorly known, it is
important to note this site in this particular publication. The terraces are broad and low, and are
situated along the hillside. On the hilltop, there
are a series of looted tombs, all of a late date, and
all with few fancy objects. Most of the Altiplano
period materials seem to be concentrated in this
higher area, and it is possible that the terraces
are purely Inca period in date. The site also has a
number of lithic artifacts, including agricultural
tools.
Size: approximately 2.0 ha
Periods: Alt, LH
Programa Collasuyu ID: J-4
Tariachi5
Tariachi is the name of the hacienda where this
site is found. The hill is similar to Huayna Roque
or Incatunuhuiri in overall appearance. From
the bottom, near the hacienda house looking
[FIGURE 17.10 ]
[FIGURE 17.11]
Tunuphara
FIGURE 17.11. Terrace on Tariachi with stela
the top are largely Upper Formative and possibly Middle Formative. This therefore gives a Tiwanaku occupation of around 2 ha at most, with
a larger Pukara occupation of at least 8 ha. There
is, of course, a modest Inca occupation as well at
the base of the site. It is curious that one Archaic
point was also discovered on the surface. Unfortunately, we did not have the permission to
check the northwest side of the site. There may
be additional domestic terraces on that side as
well.
We also were kindly given access to a private collection of some artifacts that had been
collected from the site, as seen in the figure below. Of particular interest is a small feline carving and a carved head (Figure 17.12). The figure
is most likely Pukara in style. The owner said
that it came from the area of the site.
Size: at least 8 ha
Periods: Late Archaic (?), MF (?), UF, Tiw,
Alt, LH
Programa Collasuyu ID: J-3
312
political orbit of Tiwanaku between approximately AD 600 and AD 900 (Stanish 1999). The
existence, for example, of Upper Formative
plainwares indicates a Pukara-contemporary
occupation. However, no Pukara sherds were recovered. Likewise, the presence of a substantial
amount of Collao pottery that may date to the
Middle Horizon suggests that the site had an occupation contemporary with Tiwanaku, but
without any evidence of the use or exchange of
Tiwanaku pottery.
Size: 24 ha
Periods: MF, UF, Alt, LH
Programa Collasuyu ID: Ch-1
mound. It is situated on the ridge known as Viscachune, on the hills above the pampa between
Estves and Paucarcolla. There is a low wall
across the site with a platform to the east. This
platform is about 10 x 11 m in dimension. The
site is no more than 1.0 ha in size. The diagnostic
surface artifacts are Formative in date. Inca diagnostics are found to the west, on the other side of
the wall. There is a fair amount of obsidian on
the surface of this site.
Size: 0.51.0 ha
Periods: MF (?), UF, LH
Programa Collasuyu ID: P-7
Wanina
Unocolla
This site at first appears to be a type 3 site by its
location on a low hill with terraces, but in reality
is a type 4 site with a series of domestic terraces
built around a large hill. The site is located directly off of the road. The Ro Aybacas is directly
behind the site. There is not a large quantity of
decorated pottery on this site. There are a few Tiwanaku fragments, but the quantity is too
sparse to call this a Tiwanaku occupation. There
are also a few Pukara fragments, as well as
many Middle and Upper Formative plainware
diagnostics. On the hilltop there is no evidence
of corporate architecture, nor any evidence of
depressions or sunken courts. The site has a
number of agricultural tools and a variety of raw
materials. Obsidian seems to be rare. The top of
the site has a few post-Tiwanaku period tombs.
In short, this is a fairly large nonelite domestic habitation site located in very good agricultural land. We did not get to survey the opposite
side of the site. However, from across the river,
we could see a number of wide and well-built
terraces. If these are habitation areas, then the
site could be two to three times larger than our
estimate.
Size: > 5.0 ha
Periods: MF, UF, Alt
Programa Collasuyu ID: J-6
FIGURE 17.14. View of terraces at Wanina. Looted burial with wall leading to natural hole visible in foreground
314
CONCLUSION
This paper has presented sites from various regions around the Lake Titicaca Basin. Some sites
are published here for the first time, while others
are revisits of previously identified sites in the
basin. As noted, all of the authors implement
full-coverage survey strategies in their own research. Yet, the reader will note that the majority
of these sites have been identified through large
site reconnaissance strategies. We are not suggesting a return to this somewhat outdated
method of survey. Rather, the aim of the authors
is to communicate, through this chapter, the
large number of regions around the Lake Titicaca Basin that remain to be studied through
systematic survey. Reconnaissance, whether
through the use of aerial photographs or roadside survey techniques, can be a highly effective means of identifying regions on which to
focus ones survey efforts.
316
ACKNOWLEDGMENTS
The authors thank members of the Programa
Collasuyu, our friends in the towns of Juli,
Puno, Taraco, and Pukara, Fresia Sardn, Percy
Calisaya, and Rolando Paredes. We gratefully
acknowledge the support of the National Science Foundation, the Cotsen Institute of Archaeology at UCLA, the Wenner-Gren Foundation
for Anthropological Research, the John Heinz III
Foundation, the Field Museum of Natural History, and the Dean of Social Sciences, UCLA.
NOTES
1. We make the distinction between a bona fide
settlement that can be linked to a political
entity versus the occurrence of pottery that
18.
Future Directions in Titicaca Basin Research
Amanda B. Cohen
REGIONAL APPROACHES TO
TITICACA BASIN PREHISTORY
Perhaps the most impressive contribution to archaeology in recent years has been the introduction of regional approaches embedded in
anthropological theory. Beginning in the 1990s,
survey archaeologists began to focus on the
southern and western Titicaca Basin. In particular, the studies of Juan Albarracin-Jordan (1992)
and James Mathews (1992a, 1992b) focused on
the Tiwanaku core area, with one of the main
goals being to document the presence of raised
fields in the region (Albarracin-Jordan and
Mathews 1990). John Janusek and Alan Kolatas
(2003) survey in the Pampa Koani expanded on
this earlier survey and clarified the use of raised
fields through selective excavations. These ar-
chaeologists implemented full-coverage systematic survey strategies and documented all time
periods from the Archaic to the Early Colonial to
the level of chronological limitations. More recent survey work has been completed by Matthew Bandy (2001b) on the Taraco Peninsula in
Bolivia.
Surveys in the Juli-Pomata area of the western basin were conducted by Charles Stanish
and his colleagues in 19901992 and published
in 1997. Since then, Stanish and his students
have surveyed a broad range of areas in the
western basin. Kirk Frye surveyed the area
around Chucuito. Carol Schultze has conducted
surveys and test excavations in Puno Bay. Cynthia Klink and Mark Aldenderfer surveyed the
Ilave River area. In the northern basin, Stanish
and Aime Plourde have surveyed the area
around Huancan and Putina (Plourde 2004;
Stanish and Plourde 2000). I have surveyed the
Pukara River Valley (Cohen 2000, 2001). Finally,
Programa Collasuyu teams have completed a
survey of the Ramis River from Taraco to Arapa.
Future survey work will be moving in two
directions. One, quite literally, is the expansion
of survey areas to the north and to the east. The
first systematic survey in the eastern side of the
lake was completed by Carlos Lmuz (2001)
around Santiago de Huata. Research is currently
planned by Sonia Alconini and Jose Luis Paz
Sora for additional surveys along the eastern
side of the lake.
An admirable example for survey archaeologists can be seen in the current project directed
by Matthew Bandy, working in cooperation with
Lmuz, in the Tiwanaku Valley. Bandy and
Lmuz are revisiting the survey areas of Albarracin-Jordan and Mathews in order to standardize
317
318
AMANDA B. COHEN
CHRONOLOGICAL APPROACHES TO
THE TITICACA BASIN
The Archaic Period
One of the more exciting directions taken by survey archaeologists in this volume is exemplified
in the work of Cindy Klink for the Archaic period (Chapter 2). Indeed, Klinks work is representative of one of the most recent directions for
research in the Titicaca Basinthat of Archaic
period research. Klink notes that systematic
studies of the Archaic did not begin until as recently as 1996 with Mark Aldenderfers
Proyecto Chmak Pacha in the Ilave River Valley.
Klinks rigorous survey with selective sampling
strategy of areas covered by 5-m transects illustrates the challenges presented to archaeologists
as they attempt to effectively identify and record
Archaic period sites.
Indeed, the work of Klink and Aldenderfer
(Chapters 2 and 3) represent some of the first
publications of Archaic period research in the
Titicaca Basin. As indicated by Klink in Chapter
2, the earliest evidence for the human occupation of the Titicaca Basin is during the Early Archaic, with sites dating to between 8000 and
6000 BC. The difficulty with Archaic sites has always been their dating. Chapter 3 is particularly
promising in that it presents a projectile point
chronology that can now be used for the dating
of Archaic period sites around the Titicaca Basin.
Lisa Cipollas work (Chapter 4) represents
the first publication of research utilizing Klink
and Aldenderfers Archaic projectile point chro-
319
320
AMANDA B. COHEN
Tiwanaku
More research has been devoted to Tiwanaku
than to any of the other Titicaca Basin cultures.
As mentioned above, surveys in the core Tiwanaku area (Albarracin-Jordan and Mathews 1990;
Bandy 2001b; Janusek and Kolata 2003; Kolata
2003), as well as in the western basin, have focused on the origins and development of the Tiwanaku state. Settlement pattern studies are
particularly effective in monitoring shifts that
took place between the Formative period and
the Tiwanaku expansion. This shift was addressed in Chapter 7 by Stanish et al. for the
western Titicaca Basin. Through this type of
analysis, they are able to target Tiwanaku subsistence strategies as well as understand modes
of expansion. In particular, they identify a focus
by Tiwanaku peoples on lakeside resources,
even in areas that would have been appropriate
for raised-field agriculture. A similar survey focus has been applied by Carol Schultze for her
dissertation project in Puno Bay.
The settlement surveys in the western basin
have also been followed up with excavations at
a number of sites with Tiwanaku occupations. In
Chapter 8, de la Vega reports on one such site,
Sillumocco-Huaquina, with occupations spanning the Formative through the Late Intermediate periods. The excavations identify terrace
construction that dates to both the Formative period and the later Tiwanaku occupation. Additionally, de la Vega identifies the construction of
ritual and elite residential sectors of the site during the Tiwanaku period. Carol Schultze has
also conducted test excavations in several sites
in Puno Bay that contain both Formative and Tiwanaku occupations. Her dissertation will shed
light on changes in ritual and subsistence after
the Tiwanaku expansion.
Brian Bauer and Charles Stanish conducted
a full-coverage survey on the Island of the Sun,
the results of which have recently been published (Bauer and Stanish 2001; Stanish and
Bauer 2004). In Chapter 9, Seddon reports on the
Tiwanaku sites identified in the course of this
survey in combination with his own excavations
at one of the sites, Chucaripupata. He finds evidence suggesting a pilgrimage route on the site,
as indicated in the layout of the site, as well as
clear ritual activities at the site of Chucaripupata.
These studies in the Tiwanaku periphery
could not have been nearly as effective without
all of the work that has been accomplished in the
Tiwanaku core territories, and in particular at
the site of Tiwanaku. One of the most influential
figures has been the Bolivian archaeologist Carlos Ponce Sangins, who has been actively researching and publishing about Tiwanaku since
the 1950s (i.e., 1961, 1971, 1981, 1990). He has
been joined by many others who have published
their research about the Tiwanaku capital, including Linda Manzanilla (1992) and Javier Escalante (2003). Special notice should also be
given to the work of Alan Kolata (see, e.g.,
Kolata 1986, 1993, 1996b, 2003), who directed a
multicomponent research project at Tiwanaku
and has recently published two superb volumes
summarizing these investigations. Kolata and
his students and colleagues focused on many
varied topics as part of their research, topics including issues of chronology (Janusek 2003),
elite residential architecture (Couture and
Sampeck 2003), ceremonial and public architecture (Couture 2003), diet (Wright et al. 2003), and
human remains (Blom et al. 2003).
Janusek has published on various aspects of
Tiwanaku residential life, including articles
about urbanism and ethnicity (2001b, 2002), and
workshop production (1999). In Chapter 10, he
compares distinct residential sectors at Tiwanaku and identifies variations in ethnicity, introducing a more complex and rich view of the
Tiwanaku urban center.
Other scholars continue to study issues concerning various aspects of the Tiwanaku state.
Following up on her research on the Mollo
Kontu sector of the Tiwanaku capitol, Nicole
Couture has recently completed her dissertation
on elite residences at Tiwanaku's Putuni palace
(Couture 2002, 2003; Couture and Sampeck
2003). Deborah Blom's research focuses, in part,
on the intriguing question of ethnicity in the
state capitol of Tiwanaku and its environs (Blom
2001; Blom and Janusek 2004; Blom et al. 1998,
2003). Blom has analyzed human skeletal material from Tiwanaku as well as collections from
the Moquequa Valley of southern Peru, where
Tiwanaku colonies have been identified, and
from Lukurmata, a Tiwanaku regional center in
the Katari Valley (Blom 1999; Goldstein 1989).
Based on cranial shape modification and other
stylistic archaeological correlates, Blom identified distinct ethnic groups from these areas. Interestingly, each of these ethnic groups seems to
have been represented at the capitol city of Tiwanaku (Blom et al. 2003). Couture and Blom
are directing ongoing investigations on residential and mortuary patterns at Tiwanaku designed to further explore issues of diversity in
the capitol. Alexei Vranich (1999), in his disserta-
321
322
AMANDA B. COHEN
323
References
Abbott, M., M. Binford, M. Brenner, and K. Kelts.
1997. A 3500 14C yr high-resolution record of
water-level changes in Lake Titicaca, BoliviaPeru. Quaternary Research 47:169180.
Abbott, M., G. Seltzer, K. Kelts, and J. Southon.
1997. Holocene paleohydrology of the tropical
Andes from lake records. Quaternary Research
47:7080.
Abercrombie, T. 1986. The Politics of Sacrifice: An
Aymara Cosmology in Action. Ph.D. dissertation, Department of Anthropology, University
of Chicago.
Adams, R. McC. 1965. Land Behind Baghdad: A History of Settlement on the Diyala Plain. University of Chicago Press, Chicago.
Albarracin-Jordan, J. 1990. Prehispanic dynamics
of settlement in the lower Tiwanaku Valley,
Bolivia. In A. Kolata and O. Rivera, editors,
Tiwanaku and Its Hinterland: Third Preliminary
Report of the Proyecto Wila Jawira, pp. 276296.
La Paz.
Albarracin-Jordan, J. 1992. Prehispanic and Early Settlement Patterns in the Lower Tiwanaku Valley,
Bolivia. Ph.D. dissertation, Department of Anthropology, Southern Methodist University,
Dallas.
Albarracin-Jordan, J. 1996a. Tiwanaku: Arqueologa
regional y dinmica segmentaria. Editores Plural, La Paz.
Albarracin-Jordan, J. 1996b. Tiwanaku settlement
system: The integration of nested hierarchies
in the lower Tiwanaku Valley. Latin American
Antiquity 7(3):183210.
Albarracin-Jordan, J., C. Lemz, and J. Paz. 1993.
Investigaciones en Kallamarca: Primer informe de prospeccin. Textos antropolgicos
6:11123.
Albarracin-Jordan, J., and J. Mathews. 1990. Asentamientos prehispnicos del Valle de Tiwanaku,
Vol. 1. Producciones CIMA, La Paz.
Alconini Mjica, S. 1995. Rito, smbolo e historia en la
pirmide de Akapana, Tiwanaku: Un anlisis de
325
326
REFERENCES
327
328
REFERENCES
Brumfiel, E., and T. Earle. 1987. Specialization, exchange, and complex societies: An introduction. In E. Brumfiel and T. Earle, editors,
Specialization, Exchange, and Complex Societies,
pp. 19. Cambridge University Press, Cambridge.
Bruno, M. C., and W. Whitehead. 2003. Chenopodium cultivation and Formative Period agriculture at Chiripa, Bolivia. Latin American
Antiquity 14(3):339355.
Brush, S. 1982. The natural and human environment of the Central Andes. Mountain Research
and Development 2:1938.
Bueno Mendoza, A. 1982. Arquitectura y sociedad
pre-Chavin en los Andes centrales. Boletn de
antropologa Americana 6:119140.
Buikstra, J. E., and D. H. Ubelaker 1994. Standards
for Data Collection from Human Skeletal Remains. Arkansas Archaeological Survey, Fayetteville.
Burger, R., F. Asaro, F. Stross, and G. Salas. 1998a.
The Chivay obsidian source and the geological origin of Titicaca basin type obsidian artifacts. Andean Past 5:203224.
Burger, R., F. Asaro, P. Trawick, and F. Stross.
1998b. The Alca obsidian source: The origin of
raw material for Cuzco type obsidian artifacts. Andean Past 5:185202.
Burger, R., K. Chvez, and S. Chvez. 2000.
Through the glass darkly: Prehispanic obsidian procurement and exchange in southern
Peru and northern Bolivia. Journal of World
Prehistory 14(3):267362.
Burgess, S. D. 1999. Chiribayan Skeletal Pathology on
the South Coast of Peru: Patterns of Production
and Consumption. Ph.D. dissertation, Department of Anthropology, University of Chicago.
Calancha, A. 1972 [1639]. Crnica moralizada del
Orden de San Augustn. In M. Merino, editor,
Crnicas Agustinas del Per. Biblioteca missionalia Hispnica XVII. Facsimile edition. Consejo
Superior de Investigaciones Cientficas, Instituto Enrique Flores, Madrid.
Calancha, A. 1981 [1638]. Crnica moralizada del orden de San Augustn en el Per. Facsimile edition. I. Prado Pastor, Lima.
Cardich, A. 1958. Los yacimientos de Lauricocha. Centro Argentino de Estudios Prehistricos, Buenos Aires.
329
330
REFERENCES
Couture, N. 1992. Excavations at Mollo Kontu, Tiwanaku. Masters thesis, Department of Anthropology, University of Chicago.
Couture, N. 2002. The Construction of Power: Monumental Space and Elite Residence at Tiwanaku,
Bolivia. Ph.D. dissertation, Department of Anthropology, University of Chicago.
Couture, N. 2003. Ritual, monumentalism, and residence at Mollo Kontu, Tiwanaku. In A.
Kolata, editor, Tiwanaku and Its Hinterland: Archaeology and Paleoecology of an Andean Civilization, Vol. II, pp. 202225. Smithsonian
Institution Press, Washington, DC.
Couture, N., and K. Sampeck. 2003. Putuni: A history of palace architecture at Tiwanaku. In A.
Kolata, editor, Tiwanaku and Its Hinterland: Archaeology and Paleoecology of an Andean Civilization, Vol. II, pp. 226263. Smithsonian
Institution Press, Washington, DC.
Crqui-Montfort, G. 1906. Fouilles de la mission
scientifique Franaise Tiahuanaco: Ses recherches archologiques et ethnographiques
en Bolivie, au Chile et dans la Rpublique Argentine. Internationaler Amerikanisten Kongress
2:531550. Stuttgart.
Crumley, C. 1987. A dialectical critique of hierarchy. In T. Patterson and C. Gailey, editors,
Power Relations and State Formation, pp. 155
159. American Anthropological Association,
Arlington, Virginia.
Crumley, C. 1995. Heterarchy and the analysis of
complex societies. In R. Ehrenreich, C. Crumley, and J. Levy, editors, Heterarchy and the
Analysis of Complex Societies, pp. 14. American Anthropological Association, Arlington,
Virginia.
Crumley, C., and W. Marquardt. 1987. Regional Dynamics: Burgundian Landscapes in Historical
Perspective. Academic Press, New York.
Cuentas, Alberto. 1928. Chucuito, lbum grfico e
histrico. Centenario de la ciudad de Juli. Como
capital de la provincia de Chucuito, 18281928.
Juli, Puno.
Custred, G. 1977. Peasant kinship, subsistence and
economics in a high altitude Andean environment. In R. Bolton and E. Mayer, editors, Andean Kinship and Marriage, pp. 117135.
American Anthropological Association,
Washington, DC.
331
332
REFERENCES
333
334
REFERENCES
335
In P. Kaulicke, editor, Huari y Tiwanaku: Modelos vs. evidencias (Segunda parte). Boletn de Arqueologa PUCP 5:251294. Pontificia
Universidad Catlica del Per, Lima.
Janusek, J. 2002. Out of many, one: Style and social
boundaries in Tiwanaku. Latin American Antiquity 13(1):3561.
Janusek, J. 2003. Vessels, time, and society: Toward
a chronology of ceramic style in the Tiwanaku
heartland. In A. Kolata, editor, Tiwanaku and
Its Hinterland: Archaeology and Paleoecology of
an Andean Civilization, Vol. II, pp. 3094.
Smithsonian Institution Press, Washington,
DC.
Janusek, J., and S. Alconini. 1994. Social diversity
and historical change in Tiwanaku ceramics:
Steps toward a Tiwanaku IVV chronology.
Paper presented at the 59th Annual Meeting
of the Society for American Archaeology,
Anaheim, California.
Janusek, J., and H. Earnest. 1990. Excavations at
Putuni: The 1988 Season. Informe preliminar.
Manuscript in possession of J. Janusek.
Janusek, J., and A. Kolata. 2003. Prehispanic rural
history in the Katari Valley. In A. Kolata, editor, Tiwanaku and Its Hinterland II: Urban and
Rural Archaeology, pp. 129174. Smithsonian
Institution Press, Washington, DC.
Janusek, J., S. Alconini M., D. Angelo, and P. Lima.
1997. Asentamiento prehispnico en la region
de Icla, Chuquisaca, Bolivia. Final report of
the Proyecto Icla, 19931996, submitted to the
Universidad Mayor de San Andrs and the
Instituto Nacional de Arqueologa de Bolivia.
La Paz.
Janusek, J., A. T. Ohnstad, and A. P. Roddick. 2003.
Khonko Huancane and the rise of Tiwanaku.
Antiquity 77(296). http://antiquity.ac.uk/ProjGall/
janusek/janusek.html.
Johnson, G. 1977. Aspects of regional analysis in
archaeology. Annual Review of Anthropology
6:479508.
Johnson, I. 2003. The Changing Nature of Political
Power in the Titicaca Basin: The View from Paucarcolla. Masters thesis, Department of Anthropology, UCLA.
Joyce, A., and M. Winter. 1996. Ideology, power,
and urban society in pre-Hispanic Oaxaca.
Current Anthropology 37:3386.
336
REFERENCES
337
338
Mathews, J. 1992a. Prehispanic Settlement and Agriculture in the Middle Tiwanaku Valley, Bolivia.
Ph.D. dissertation, Department of Anthropology, University of Chicago.
Mathews, J. 1992b. Some notes on the early development of the Tiwanaku State. Paper presented at the 20th Annual Midwest Conference
on Andean and Amazonian Archaeology and
Ethnohistory.
Mathews, J. 1995. A re-evaluation of the Formative
period in the southeast Titicaca Basin, Bolivia.
Journal of the Steward Anthropological Society
23:83110.
McBain Chapin, H. 1959. The Adolph Bandelier archaeological collection from Pelechuco and
Charassani, Bolivia. Revista del Instituto de
Antropologa I:980.
McEwan, G. 1990. Some formal correspondences
between the imperial architecture of the Wari
and Chimu cultures of ancient Peru. Latin
American Antiquity 1(2):97116.
McEwan, C., and M. Van de Guchte. 1992. Ancestral time and sacred space in Inca state ritual.
In R. Townsend, editor, The Ancient Americas:
Art from Sacred Landscapes, pp. 347358. The
Art Institute of Chicago.
McGuire, R. 1983. Breaking down cultural complexity: Inequality and heterogeneity. Advances in Archaeological Method and Theory
6:91142.
Menghin, O., and G. Schroeder. 1957. Unyacimiento en Ichua (Departamento de Puno,
Per) y las industrias Precermicas de los
Andes centrales y septentrionales. Acta Prehistrica, pp. 4154.
Menzel, D. 1959. The Inca conquest of the South
Coast of Peru. Southwestern Journal of Anthropology 15:125142.
Menzel, D. 1964. Style and time in the Middle Horizon. awpa Pacha 2:1107.
Mesa, J., and T. Gisbert. 1957. Akapana: La pirmide de Tiwanacu. In C. Ponce Sangins, editor, Arqueologa Boliviana (Primera Mesa Redonda), pp. 141161. Alcalda Municipal, La
Paz.
Milner, G., and J. Oliver. 1999. Late prehistoric settlements and wetlands in the central Mississippi Valley. In B. Billman and G. Feinman,
editors, Settlement Pattern Studies in the Ameri-
cas: Fifty Years since Vir, pp. 7995. Smithsonian Institution Press, Washington, DC.
Mohr, K. 1966. An Analysis of the Pottery of Chiripa,
Bolivia: A Problem in Archaeological Classification and Inference. Masters thesis, Department
of Anthropology, University of Pennsylvania,
Philadelphia.
Mohr-Chvez, K. 1992. The organization of production and distribution of traditional pottery
in south highland Peru. In G. Bey and C. Pool,
editors, Ceramic Production and Distribution:
An Integrated Approach, pp. 4992. Westview
Press, Boulder.
Molina el Almagrista, C. de. 1968 [1552]. Relacin de muchas cosas acaescidas en el Per.
In F. Esteve Barba, editor, Crnicas peruanas de
inters indigena, pp. 5795. Facsimile edition.
Biblioteca de Autores Espaoles 209. Ediciones Atlas, Madrid.
Molina de Cuzco, C. de. 1947 [1584]. Relacin de los
fbulas y ritos de los Incas. Editorial Futuro,
Buenos Aires.
Molina, E., and A. Little. 1981. Geoecology of the
Andes and the natural science basis for research planning. Mountain Research and Development 1(2):115144.
Moore, J. 1996a. The archaeology of plazas and the
proxemics of ritual. American Anthropologist
98(4):789802.
Moore, J. 1996b. Architecture and Power in the Ancient Andes: The Archaeology of Public Buildings.
Cambridge University Press, Cambridge.
Moore, K. 1999. Chiripa worked bone and bone
tools. In C. Hastorf, editor, Early Settlement at
Chiripa, Bolivia, Contributions of the University of California Archaeological Research Facility, number 57, pp. 7394. Archaeological
Research Facility, Berkeley.
Moore, K., D. Steadman, and S. de France. 1999.
Herds, fish, and fowl in the domestic and ritual economy of Formative Chiripa. In C. Hastorf, editor, Early Settlement at Chiripa Bolivia,
Contributions of the University of California
Archaeological Research Facility, number 57,
pp. 105116. Archaeological Research Facility,
Berkeley.
Morris, C. 1982. The infrastructure of Inka control
in the Peruvian central highlands. In G. Collier, R. Rosaldo, and J. Wirth, editors, The Inca
REFERENCES
and Aztec States 14001800, pp. 153171. Academic Press, New York.
Morris, C., and D. Thompson, 1985. Hunuco
Pampa: An Inca City and Its Hinterland. Thames
and Hudson, London.
Moseley, M. 1992. The Incas and Their Ancestors.
Thames and Hudson, New York.
Mujica, E. 1978. Nueva hiptesis sobre el desarrollo
temprano del altiplano del Titicaca y de sus
reas de interraccin. Arte y Arqueologa 5
6:285308.
Mujica, E. 1985. Altiplano-coast relationships in the
South-Central Andes: From indirect to direct
complementarity. In S. Masuda, I. Shimada,
and C. Morris, editors, Andean Ecology and
Civilization, pp. 103140. University of Tokyo
Press, Tokyo.
Mujica, E. 1987. Cusipata: Una fase pre-Pukara en
la cuenca norte de Titicaca. Gaceta Arqueolgica
Andina 13:2228.
Mujica, E. 1988. Peculiaridades del proceso
histrico temprano en la cuenca del norte del
Titicaca: Una propuesta inicial. Boletn del Laboratorio de Arqueologa, pp. 75124. Escuela de
Arqueologa e Historia, Universidad Nacional
San Cristobal de Huamanga, Ayacucho.
Muoz Ovalle, I., B. Arriaza, and A. Auferheide,
editors. 1993. Acha-2 y los orgenes del poblamiento humano en Arica. Ediciones Universidad de Tarapac, Arica, Chile.
Murra, J. V. 1964. Una apreciacin etnologa de la
Visita. In Visita hecha a la Provincia de Chucuito
por Garci Diez de San Miguel en el Ao 1567.
Documentos Regionales para el Etnologa y Etnohistoria Andinas (Lima, Peru) 1: 419444.
Murra, J. 1968. An Aymara kingdom in 1567. Ethnohistory 15:115151.
Murra, J. 1970. Informacin etnolgica e histrica
adicional sobre el reino Lupaqa. Historia y
Cultura 4:4961.
Murra, J. 1972. El control vertical de un mximo
de piscos ecolgicos en la economa de las sociedades andinas. In J. Murra, editor, Visita de
la Provincia de Len de Hunuco en 1562, Vol. 2,
pp. 42976. Universidad Nacional Hermilio
Valdizn, Hunuco, Peru.
Murra, J. 1978. Aymara lords and their European
agents at Potos. Nova Americana 1:231243.
Giulio Einaudi, Torino.
339
340
Parsons, J. 1968. An estimate of size and population for Middle Horizon Tiahuanaco, Bolivia.
American Antiquity 40:259282.
Parsons, J. 1971. Prehistoric Settlement Patterns in the
Texcoco Region, Mexico. Memoirs of the Museum of Anthropology, University of Michigan, Ann Arbor.
Parsons, J. 1976. Settlement and population history
of the Basin of Mexico. In E. Wolf, editor, The
Valley of Mexico: Studies in Prehispanic Ecology
and Society, pp. 69100. University of New
Mexico Press, Albuquerque.
Parsons, J., and R. Whalen. 1982. Prehispanic Settlement Patterns in the Southern Valley of Mexico:
The Chalc-Xochimilco Region. Memoirs of the
Museum of Anthropology, University of
Michigan, Ann Arbor.
Prssinen, M., and A. Siiriinen. 1997. Inka-style
ceramics and their chronological relationship
to the Inka expansion on the southern Lake
Titicaca area (Bolivia). Latin American Antiquity 8:255271.
Pauketat, T. 2000. The tragedy of the commoners.
In M.-A. Dobres, and J. Robb, editors, Agency
in Archaeology, pp. 113129. Routledge, London.
Paul, A., and S. Turpin. 1986. The ecstatic shaman
theme of Paracas textiles. Archaeology 39(5):
2027.
Paz Sora, J. 1999. Excavations in the Llusco Area.
In C. Hastorf, editor, Early Settlement at Chiripa, Bolivia, Contributions of the University of
California Archaeological Research Facility,
number 57, pp. 3136. Archaeological Research Facility, Berkeley.
Paz Sora, J. 2000. La transicin Formativo-Tiahuanaco
en el valle Mesothermal de Tambo Kusi. Licenciatura de Facultad de Ciencias Sociales, Carrera
de Arqueologa, Universidad Mayor de San
Andrs, La Paz.
Pease, F. 1982. Formation of Tawantinsuyu: Mechanisms of colonization and relationship with
ethnic groups. In G. Collier, R. Rosaldo, and J.
Wirth, editors, The Inca and Aztec states, 1400
1800: Anthropology and History, pp. 173198.
Academic Press, New York.
Perrin Pando, A. 1957. Las tumbas subterraneos de
Wakuyo. In C. Ponce Sangins, editor, Arqueologa Boliviana (Primera Mesa Redonda), pp.
172205. Alcalda Municipal, La Paz.
REFERENCES
341
342
Rasnake, R. 1988. Domination and Cultural Resistance: Authority and Power among an Andean
People. Duke University Press, Durham,
North Carolina.
Ravines, R. 1967. El abrigo de Caru y sus relaciones
culturales con otros sitios temprano del sur
del Per. awpa Pacha 5:3957.
Ravines, R. 1972. Secuencia y cambios en los artefactos liticos del sur del Per. Revista del Museo Nacional 38:133185.
Reinhard, J. 1991. Tiwanaku: Ensayo sobre su cosmovisin. Pumapunku (nueva poca) 2:866.
Reinhard, J. 1992. Sacred peaks of the Andes. National Geographic 181:84111.
Reinhard, J. 1995. House of the sun: The Inka temple of Vilcanota. Latin American Antiquity 6(4):
340349.
Revilla Becerra, R., and M. Uriarte Paniagua. 1985.
Investigacin arqueolgica en la zona de Sillustani-Sector Wakakancha-Puno. Bachelors
thesis, Universidad Catlica Santa Mara,
Arequipa, Peru.
Rice, P. 1987. Pottery Analysis: A Sourcebook. University of Chicago Press.
Rick, J. 1980. Prehistoric Hunters of the High Andes.
Academic Press, New York.
Riddell, F., and L. Valdez. 1987. Hacha y la ocupacin temprana del valle de Acar. Gaceta Arqueolgica Andina 16:610.
Rivera, M. 1991. Pre-Tiwanaku developments in
northern Chile. Paper presented at the 56th
Annual Meeting of the Society for American
Archaeology, New Orleans.
Rivera Casanovas, C. 1994. Chiji Jawira: Evidencias
sobre la produccin de cermica en Tiwanaku. Tsis de licenciatura, Carrera de Antropologa,
Universidad Mayor de San Andrs, La Paz.
Rivera Casanovas, C. 2003. Chiji Jawira: A case of
ceramic specialization in the Tiwanaku urban
periphery. In A. Kolata, editor, Tiwanaku and
Its Hinterland: Archaeology and Paleoecology of
an Andean Civilization, Vol. II, pp. 296315.
Smithsonian Institution Press, Washington,
DC.
Rivera Sundt, O. 1978. Arqueologa de la peninsula
de Copacabana. Pumapunku 12:6986.
Rivera Sundt, O. 1984. La horca del Inka. Arqueologa Boliviana 1:91106.
REFERENCES
343
344
Seltzer, G. 1990. Recent glacial history and paleoclimate of the Peruvian-Bolivian Andes. Quaternary Science Reviews 9(2/3):137152.
Seltzer, G. 1993. Late-Quaternary glaciation as a
proxy for climate change in the Central
Andes. Mountain Research and Development
13(2):129138.
Seltzer, G., P. Baker, S. Gross, S. Fritz, and R. Dunbar. 1998. High-resolution seismic reflection
profiles from Lake Titicaca, Peru-Bolivia: Evidence for Holocene aridity in the tropical
Andes. Geology 26(2):167170.
Service, E. 1962. Primitive Social Organization. Random House, New York.
Service, E. 1975. Origins of the State and Civilization.
W. W. Norton and Company, New York.
Shea, D. 1969. Wari Wilka: A Central Andean Oracle
Shrine. Ph.D. dissertation, Department of Anthropology, University of Wisconsin, Madison.
Shimada, I. 1991. Pachacamac archaeology: Retrospect and prospect. In I. Shimada, editor,
Pachacamac: A Reprint of the 1903 Edition, by
Max Uhle. University of Pennsylvania, Philadelphia.
Shott, M. 1997. Stones and shafts redux: The metric
discrimination of chipped-stone dart and arrow points. American Antiquity 62:86101.
Sillar, W. 1988. Mud and firewater: Making pots in
Peru. Report submitted in partial fulfillment
of the requirements for the degree of MSc in
Archaeology, Institute of Archaeology, University College, London.
Silverman, H. 1996. The Formative Period on the
south coast of Peru: A critical review. Journal
of World Prehistory 10(2):95146.
Sitio del Cuzco 1934 [1539]. Relacin del sitio del
Cuzco y principio de las guerras civiles del
Per hasta la muerte de Diego de Almagro
15331539, Madrid [1879]. Facsimile edition
published in Coleccin de libros y documentos
referentes a la historia del Peru 2(10):1133.
Smith, C., W. Denevan, and P. Hamilton. 1968. Ancient ridged fields in the region of Lake Titicaca. Geographical Journal 134:353367.
Smith, M. 1987. Household possessions and wealth
in agrarian states: Implications for Archaeology. Journal of Anthropological Archaeology 6:
297335.
REFERENCES
345
346
Stanish, C., E. de la Vega, and K. Frye. 1993. Domestic architecture on Lupaqa area sites in the
Department of Puno. In M. Aldenderfer, editor, Domestic Architecture, Ethnicity and Complementarity in the South-Central Andes, pp. 83
93. University of Iowa Press, Iowa City.
Stanish, C., E. de la Vega, L. Steadman, C. Chvez,
K. Frye, L. Onofre Mamani, M. Seddon, and P.
Calisaya. 1995. Archaeological survey in the
southwestern Lake Titicaca Basin. Dilogo Andino 14/15:97143.
Stanish, C., E. de la Vega, L. Steadman, C. Chvez
Justo, K. Frye, L. Onofre Mamani, M. Seddon,
and P. Calisaya Chuquimia. 1997. Archaeological
Survey in the Juli-Desaguadero Region of Lake
Titicaca Basin, Southern Peru.Volume 29 of
Fieldiana Anthropology. Field Museum of Natural History, Chicago.
Stanish, C., E. de la Vega, L. Steadman, L. Onofre
Mamani, K. L. Frye, C. Chvez Justo, and M.
Seddon. 1992. Archaeological research in the
Juli-Pomata region of the Titicaca Basin, southern Peru. Final report submitted to the National Science Foundation, Washington, DC.
Stanish, C., and E. de la Vega M. 1991. The Tiwanaku occupation of the southwestern Titicaca
Basin, Peru. Paper presented at the 47th International Congress of Americanists, New Orleans.
Stanish, C., and A. Plourde. 2000. Archaeological
survey in the Huancan-Putina Valley, Peru.
Final report submitted to the National Institute of Culture, Lima.
Stanish, C., and L. Steadman. 1994. Archaeological
Research at Tumatumani, Juli, Peru. Volume 23
of Fieldiana Anthropology. Field Museum of
Natural History, Chicago.
Steadman, D. 1997. Animal bone. In C. Hastorf, M.
Bandy, D. Blom, E. Dean, M. Goodman, D.
Kojan, M. Montao Aragn, J. Luis Paz, D.
Steadman, L. Steadman, and W. Whitehead.
Taraco Archaeological Project: 1996 Excavations at Chiripa, Bolivia, pp. 48-49. Report
submitted to the Instituto Nacional de Arqueologa, Bolivia.
Steadman, L. in press. The ceramics from Quelcatani. In M. Aldenderfer, editor, Quelcatani and
the Evolution of Pastoral Societies in the Titicaca
Basin. Smithsonian Institution Press, Washington, DC.
REFERENCES
347
348
REFERENCES
349
Wright, H., and G. Johnson. 1975. Population, exchange and early state formation in southwestern Iran. American Anthropologist 77:267
289.
Wright, M., C. Hastorf, and H. Lennstrom. 2003.
Pre-Hispanic agriculture and plant use at Tiwanaku: Social and political implications. In
A. Kolata, editor, Tiwanaku and Its Hinterland:
Archaeology and Paleoecology of an Andean Civilization, Vol. II, pp. 384403. Smithsonian Institution Press, Washington, DC.
Zubrow, E. 1990. The depopulation of native
America. Antiquity 64(245):754-765.
Zuidema, R. 1978. Shafttombs and the Inca Empire.
Journal of the Steward Anthropological Society
9(12):133-178.
Zuidema, R. 1990. Inca Civilization in Cuzco. University of Texas Press, Austin.
Index
Note: Page numbes in bold italics indicate illustrations or tables.
AB
Acha-2, 30
Acora, 9, 230, 234, 300, 303,
Altarani, 110, 206, 229, 239
Altiplano period. See Late Intermediate period
Alto Pukara, 319, 68
Amantan, 7, 69, 83, 101102, 237
239, 3, 4, 70, 215, 237, 291
Apachinacapata, 139140
Apurmac, 210
Arapa, 72, 83, 112, 317, 3, 4, 70, 244,
291
Archaic period, 1324, 140, 174, 194,
311, 317, 323
Early, 5, 16, 27, 3032, 34, 3637,
3941, 44, 50, 53, 57, 58, 61, 318,
59
Middle, 5, 16, 27, 3141, 44, 53,
58, 61, 318
Late, 5, 16, 27, 3537, 3941, 44
45, 47, 5354, 5662, 135, 229,
318, 60
Terminal, 6, 16, 27, 44, 4748, 50
53, 6062, 318, 61
Asana, 1617, 25, 2853, 2930, 32
37, 39, 4243, 45, 47, 49, 52
Asillo, 243
Auquipadja. See Peninsula Socca
Ayabacas River Valley, 4, 293, 301,
303, 312
Ayacucho, 210, 268
Ayaviri, 84, 293, 300
Aymara, 95, 100, 107, 156, 214, 239
240, 276278, 288, 304, 307, 276
Azngaro, 4, 83, 198, 295
Bao del Inca, 219
Bebedero, 110, 206, 223, 227, 229,
234
bofedales, 13, 117
C
Cabrauyo, 9596
Cachichupa, 11, 83, 295, 302, 319, 4
Cala Cala, Bolivia, 284
Calahuata, 204
Calvario, 204205
Camarones-14, 39-40
Camata, 71, 73, 84, 108109
351
DE
Desaguadero River Valley, 104, 111
112, 290, 303
Early Intermediate period , 65, 95
El Panten, 44
El Volcn, 56
encomienda system, 10, 101, 286
Escoma River, 7, 114, 289
F
Formative period, 1011, 41, 48, 51,
60, 6594, 110, 116117, 119,
121122, 124128, 130, 132, 140,
175, 177, 179, 318, 320
352
GH
Guaqui, 286287
Hakenasa, 30, 3851, 26, 35, 38, 40,
4243, 45, 4749, 5152
Hatuncolla, 131, 197199, 208, 211,
236237, 280, 291, 300, 305
Horca del Inca, 218219, 222
Hu-14. See Cachichupa
huaca, 95, 98, 209, 210, 214, 217, 222,
237240
Huacaparki. See Viscachune
Huacullani, 111, 292
Huajje, 100, 302
Huajjsapata, 107
Huamanga, 268
Huaa, 292, 294295, 297, 300301,
311
Huancahuichinka, 84
Huancan, 60, 72, 102, 289, 292, 295,
298, 317, 3, 4, 5961, 70, 105,
244, 291
Huancan-Putina River Valley, 5,
5563, 83, 104, 111, 289290, 323
Hunuco Pampa, 131, 240
Huaquina-Sapijicani, 115
Huarochiri manuscript, 210
Huata Pampa clay source, 302
Huata Pampa, 113
Huata, 2, 295, 300, 302, 312, 316, 3,
4, 70
Huatasani, 111, 292
Huayna Roque, 104, 302, 310, 105
Huenque River Valley, 1324
Huichajaja, 178, 307
Huimpoco, 303
I
Ica Valley, 65
Ichu Pampa, 108, 109
Ichu, 104
Ichua, 37, 44
Ilave River Valley, 5, 6, 7, 13, 27, 30
32, 34, 3741, 4448, 5053, 73,
84, 112114, 194, 230, 234, 290,
295, 298, 301, 303304, 308309,
14, 215, 291
Illincaya (Illingaya), 10, 255256,
263
Inca (Inka), 810, 27, 9596, 101, 114,
131132, 135, 138142, 145, 145,
197244, 246247, 250, 255257,
263, 265266, 269270, 274277,
279283, 288, 290293, 298300,
303305, 309312, 322
Inca Cueva-4, 38
Inca Uyu, 173, 204206, 236, 239
Incatunuhuiri, 6566, 7172, 8385,
90, 92, 104, 109, 310, 315
Ingapirca, 225
Inkallakta, 255
Intinkala, 216218, 222, 227, 239
Iskanwaya, 1
Isla Escata, 301, 303
Isla Estvez, 2, 100, 107108, 131,
293294, 302, 312, 316
Isla Quipata, 205
Isla Salinas, 107
Isla Soto, 7, 72, 95102, 319, 96
Island of the Moon (Isla de la Luna),
1, 7, 101,138, 197, 209, 211212,
216, 238
Island of the Sun (Isla del Sol), 1, 5
8, 91, 95, 101, 135142, 174, 197,
204, 206209, 211214, 216218,
222, 226, 229, 234, 236, 238240,
270, 290, 320, 322
Isle of Penintence. See Isla Soto
Ispallauyu, 97, 100, 97
Iwawe, 284
J
Jawira Pampa, 284
Jiskairumoko, 17, 47, 51, 53, 26
Juli, 100, 104, 106, 113, 115116, 199,
225, 227, 272, 302, 14, 106, 215,
291
Juliaca, 7, 104, 112, 293, 295, 302
303, 310
Juli-Pomata survey area, 17, 104,
109111, 113, 138, 174, 176177,
183, 194, 200201, 206, 216, 268,
290, 295, 303304, 308, 310, 317,
321322, 106, 115, 175, 179, 182,
201, 203
K
Kacha Kacha A, 303
Kacha Kacha, 289
Kala Uyuni, 6566, 71, 83, 85, 9091
Katari Basin (Catari Basin), 10, 207,
269, 275, 280284, 268, 287
Kenko, Cuzco, 225, 230
Kenko/Tres Ventanas, 215, 222,
229230, 234, 236, 238240
Kesanani, 218
Kille Kille, 247, 250
Koana. See Cerro Juana
Koani Pampa. See Pampa Koani
Konkowankane, 83
Kunkapata. See Santisima Cruz del
Rancho
Kurinuyu, 236
Kusijata, 219
L
La Paz, Bolivia, 268, 286
La Plata, 268
La Raya, 2, 238
Lacatambo, 229
Lacaya, 283
Lake Arapa, 105, 293
Lake Huiamarca, 7
Lake Umayo, 3, 4, 70, 309
Lampa, 293
Larecaja, 9, 243
Las Cuevas, 38, 4144, 50, 35, 38
Late Intermediate period (Altiplano
period), 810, 5152, 113, 121
122, 128, 174185, 191192, 198,
200201, 207, 212213, 223, 229,
232, 239240, 246247, 255257,
263, 266, 276, 292293, 295, 298,
300307, 309310, 315, 320322
Late Horizon, 113, 173, 178, 183,
185186, 194, 197207, 210217,
219220, 222223, 225226, 229,
232, 234, 236, 238240, 246, 257,
263, 266, 295, 298, 300301, 303
306, 309, 322
lithics, 5152, 95, 101
Lauricocha, 3739
Limatambo Valley, 266
Lukurmata, 1, 83, 8586, 103, 131,
171, 276277, 280, 304
Lupaqa (Lupaca), 910,104, 111,
113, 173174, 176179, 181, 183
184, 194, 197200, 203, 205209,
211212, 236237, 272, 290, 321
322
INDEX
M
Macusani, 244
Majes, 243
Maranuyu, 9798, 9798
Maravillas, 293294, 303
Maukallajta, 289
Mazo Cruz, 307
Mirador. See Viscachune
mita, 267268, 288
mitimaes, 211
Moho, 96, 101102, 292, 295, 298,
308
Molino-Chilacachi, 9, 185195, 322
Mollo, 243
Molloqo Mata, 107
Moquegua, 4, 8, 104, 113, 131, 155,
191, 307
Mount Sapaqollo. See Zapacollo
Moyopampa, 110, 115
Murokato, 141
NO
Nasca (Nazca), 65, 83, 89
Ojerani, 304
Ollachea, 10, 243244, 246, 255, 265
266
Omasuyos, 9, 194, 243, 298
Omereque, 147
Omo, 4546, 131, 192, 46
Orcohawira, 216218, 239
P
PaAjano. See Yayamama
Pacajes style, 710, 111, 113, 177,
197198, 207208, 216, 270, 272
273, 274277, 279288, 322
Pacariqtambo, 240
Pachacamac, 239
Pachamachay, 31, 33, 3640, 42, 44,
48, 50
Pachamama Amanan, 237, 304
Pachatata Amantan, 131, 237, 304
Pajchiri, 1, 83
Paleo-Indian, 5, 55
Palermo, 8384, 92, 113, 116117,
Pallalla, 7
Pampa Colorado, 31
Pampa de Moyo. See Moyopampa
Pampa Huataraque, 108
Pampa Koani, 75, 83, 86, 317
Pampa Pirape, 108109
Paracas Peninsula, 65
Patapatane, 3033, 3637, 2930, 32
Patas, 108
Q
Qaluyu, 5, 7, 55, 6566, 7172, 80
81, 83, 294295, 298, 305
353
R
raised fields, 12, 8, 55, 86, 108109,
112, 166117, 177, 294, 300, 303
304, 308, 312, 320
Raqchi, 238
Republican period, 308
Ro Cala Cala, Peru, 5759
Ro Huancan, 55, 57
Ro Osmore, 41
Ro Ramis, 294, 303, 306, 308, 317
Ro Salado, 110, 116117
Ro Suches, 290
S
Sacred Rock, 8, 74, 140142, 211212
Samaipata, 225, 229
Saman Sur, 308
San Bartolom-Wiscachuni, 69, 320
San Juan del Oro, 95
Santa Apolonia, 225
Santa Rosa River, 15
Santiago de Huata, 6566, 70, 74,
84, 317
Santisima Cruz del Rancho, 309
Sayhuite, 225
Sillumocco style, 75, 83, 90, 110, 116,
117, 130
Sillumocco-Huaquina, 73, 8384,
86, 92, 111, 113, 115133, 227,
320
Sillustani, 1, 108, 194, 198, 213, 289,
292, 295, 304, 306, 309, 4, 105,
291
Siraya, 289
South-Central Andes, definition, 3
Suankata Hills, 116
Sucano Bajo, 309
Sumbay, 30, 36, 40, 45, 2, 26, 35, 40
sunken courts, 3, 78, 69, 7172, 74,
84, 87, 95102, 107, 109, 131,
213, 237, 295, 298, 304305, 312,
319
354
T
Tambillo, 38, 48, 50, 51
tambo, 211, 238, 309
Tanka-Tanka, 1, 178, 289
Taparachi, 310
Taraco, Peru, 7, 72, 83, 88, 111, 272,
289, 295, 301
Taraco. Bolivia, 10, 7072, 177, 207
208, 269, 272, 275, 281, 283287,
317, 319, 323, 271, 274, 275, 282
286
Tariachi, 310
Tawantisuyu, 197, 209210, 225,
Telarmachay, 31, 33, 3640
Tinpunku. See Cebaduyu
Tipon, 238
Tiquina, 218
Titicaca Basin
geography, 24
paleo-environment, 16, 62, 516,
174175, 2
Titicala. See Sacred Rock
Titijones, 41
Titimani, 74, 84, 86
Titinhuayani, 74, 137, 139140
Tiwanaku River Valley, 72, 75, 207
208, 267270, 275276, 281288
Tiwanaku, 1, 3, 5, 710, 6566, 70,
78, 8284, 8794, 96, 100, 103
114, 116122, 124132, 135171,
U
Umasuyu, 198
Unocolla Hill, 309
Urcos, 238
Urcosuyu, 239
Uros, 304
Urosuyu, 198
Urus, 297
V
Vilcanota, 238
Vilcashuamn, 225
Vilquechico, 315
Viscachane, Peru, 108
Viscachune, 312
W
Wacuyo (Wakuyu), 135, 139, 140
Waki, 287
Wakullani, 280
Wanina, 302, 312
Wari, 3, 8
warus warus. See raised fields
Wilasalto Peninsula, 298
Willakolla, 289
YZ
Yara, 41
Yayamama (PaAjano), 6768, 74,
76, 8081, 86, 88, 93, 295, 67
Yumani, 136
Yunguyos, 217218, 220, 222
Yunguyu, 222, 225, 272, 290
Zapacolla, 116117, 227