Arqueologia Del Colca

AN ARCHAEO-HISTORY OF ANDEAN COMMUNITY AND LANDSCAPE:
THE LATE PREHISPANIC AND EARLY COLONIAL COLCA VALLEY, PERU

by
Steven Arlyn Wernke
A dissertation submitted in partial fulfillment of

the requirements for the degree of
Doctor of Philosophy
(Anthropology)
at the
UNIVERSITY OF WISCONSIN-MADISON
2003
To my parents,
Arlyn and Grace Wernke,
for their unwavering support
ii
CONTENTS
List of Figures
List of Tables
xii
Acknowledgements
xiii
Abstract
xviii
PART I: INTRODUCTION AND BACKGROUND
Chapter 1: Community and Landscape in the Andes

Introduction
A Political-Ecological Approach to Community and Landscape
Alternative Hypotheses and Their Correlates
Dissertation Outline
1
1
6
18
32
Chapter 2: Physiographic, Climatic, and Anthropogenic Characteristics of the

Colca Valley Landscape
Location, Geology, and Geomorphology
Soils
Climate Characteristics and Trends
Anthropogenic Landscape Features
34
34
42
44
54
Chapter 3: The Collagua in Regional and Historical Context: Previous Research

Introduction
Previous Archaeological Research
The Collagua in Colonial Documents
67
67
67
80
Chapter 4: Methodology and Research Operations

Introduction: An Archaeo-Historical Approach
Overview of Archaeological Methodology
Selection of Project Area
Archaeological Surveying Techniques in a Montaine Landscape
Site Nomenclature
Field Mapping Techniques
Site Sizes
Site Classification: Chronology
Artifact Collection Strategy
Survey Area Coverage and Total Sites Recorded
Basemap Source Data
Early Colonial Land Tenure Analysis: Overview of the Documentary Sample
92
92
96
98
99
102
102
103
105
106
107
109
111
iii
The Visitas as Text: Issues of Representation
GIS Cartographic Representation and Database Management of
Documentary Data
115
117
PART II: ARCHAEOLOGICAL FINDINGS

Chapter 5: Early Settlement and Agriculture: The Formative and
Middle Horizon
Introduction
Spatial and Chronological Considerations
The Formative Period Settlement Pattern: Overview
Site YA-032: Chiquero
Agricultural Production and Landscape Modification During the Formative
Terminal Archaic and Formative Period Hunting and Pastoralism in the Puna
Discussion
The Middle Horizon Period Settlement Pattern: Overview
Middle Horizon Cemeteries
Agro-Mortuary Wall Sites and Associated Settlements
Discussion
Chapter Summary and Conclusions
126
126
128
129
135
140
143
146
150
157
158
167
169
Chapter 6: The Late Intermediate Period and Late Horizon Occupations

Introduction
Chronological Considerations
The Late Intermediate Period Settlement Pattern: Overview
The Late Horizon Settlement Pattern: Overview
Settlement Sizes
Collagua Domestic Architecture
Inka Imperial Architecture
Cemeteries, Tombs, and Mortuary Architecture
Agro-Pastoral Infrastructure and Production during the LIP and Late Horizon
Pukara Fortifications and Evidence for Violent Conflict
Descriptions of Principal LIP/Late Horizon Settlements
171
171
172
176
182
189
195
217
225
234
251
266
Chapter 7: Spanish Insertions: Early Colonial Settlement and Agro-Pastoral

Production
Introduction
The Early Colonial Period Settlement Pattern
The Doctrinas of the Colca Valley, 1540-1595
Archaeological Evidence for the Doctrinas: Early Colonial Chapels
Toledan Dislocations
Chapter Summary
298
298
299
302
305
312
330
343
iv
PART III: ETHNOHISTORICAL RECONSTRUCTION AND

ARCHAEO-HISTORICAL SYNTHESIS
Chapter 8: Regional Political-Ecological Syncretism under Inka and
Spanish Rule
Introduction
Hybrid Communities: Reduccin, Ayllus and Kurakas
Structural Grafting: Ayllus and Reducciones
A Regional View of Collagua Complemenarity Practices
Kurakas, Craft Specialists, and the Provincial Political Economy
344
344
345
359
369
385
390
Chapter 9: Reconstructing Colonial Land Tenure and Late Prehispanic

Residence Patterns
Introduction
The Local Crop Mosaic
Moiety-Level Land Tenure Patterning
Ayllu-Level Land Tenure Patterning
Discussion and Conclusion
393
393
394
398
400
430
Chapter 10: Summary and Conclusions

Review of the Research Problem
Early Agriculturalist Settlement in the Central Colca Valley
Community Organization and Production during the LIP and Late Horizon
Collagua Communities and Complementarities during Inkaic and Early
Colonial Times
435
435
436
439
443
APPENDICES
Appendix A: Ceramic Sequence
445
Appendix B: Projectile Point Illustrations
538
Appendix C: Archaic Period Sites
541
Appendix D: Survey Artifact Registry
543
Appendix E: Ceramic Distributions by Site
561
Appendix F: Site and Sector Registry
585
References Cited
603
vi
List of Figures
Figure 2.1: Location of the Colca valley
Figure 2.2: The Colca valley and survey area
Figure 2.3: Volcanic features and obsidian sources
Figure 2.4: Geomorphic surfaces in the central Colca Valley
Figure 2.5: Average monthly temperatures in the Colca valley
Figure 2.6: Average annual precipitation, 1955-1982
Figure 2.7: Average monthly precipitation, 1951-1982
Figure 2.8: Monthly precipitation, Chivay, 1973-1972
Figure 2.9: Perspective view of Pampa Finaya
Figure 2.10: Chijra/San Antonio/Chilacota area
Figure 3.1: Overview of the Province of Collaguas, showing internal divisions
Figure 4.1: Survey area and all sites registered (N=169)
Figure 4.2: Map of all toponyms with matching names in the visitas
Figure 5.1: Map of Formative Period sites in the survey area
Figure 5.2: Formative site counts by ecological zone (N=30)
Figure 5.3: Formative site counts by altitude (N=30)
Figure 5.4: Formative ceramics recovered from YA-032
Figure 5.5: Obsidian Type 5D projectile point and scraper from YA-032
Figure 5.6: Andesite hoes from YA-032
Figure 5.7: YA-032 (Chiquero) and surrounding agricultural landscape features
Figure 5.8: sketch map of YA-094
Figure 5.9: Site CO-106, showing cliff overhang (rockshelter) and midden areas
Figure 5.10: Middle Horizon ceramics
Figure 5.11: Painted tablets from YA-169 (Bomboncilla)
Figure 5.12: Middle Horizon sites in the survey area
Figure 5.13: Middle Horizon site counts by ecological zone (N=37)
Figure 5.15: Agro-mortuary walls at site YA-014, Sector E
Figure 5.16: Detail of agro-mortuary wall
Figure 5.17: Ovoid features in agro-mortuary walls at YA-014
Figure 5.18: Gallery tomb at CO-148, Sector A
Figure 6.1: Collagua ceramic sequence (lot numbers indicated)
Figure 6.2: Late Intermediate Period settlement pattern map
Figure 6.3: Late Horizon settlement pattern map
Figure 6.4: Detail of Late Horizon settlements in the central area of the survey zone
Figure 6.5: Late Horizon site size histogram, by area (ha)
Figure 6.6: Late Horizon site size histogram, by house count
Figure 6.7: Late Intermediate Period rank size graph, by site area
Figure 6.8: Late Intermediate Period rank size graph, by house count
Figure 6.9: Late Horizon rank size graph, by site area
Figure 6.10: Circular houses at the site of Laiqa Laiqa, near Tuti
Figure 6.11: Small house of Type 1 masonry with preserved doorway
37
38
39
41
46
47
49
49
61
63
81
108
122
131
133
133
137
138
138
141
144
145
151
152
153
155
155
160
161
164
166
175
181
186
187
190
191
193
194
194
200
202
vii
Figure 6.12: Small house of Type 1 masonry with preserved doorway
Figure 6.13: Detail of doorway
Figure 6.14: Detail of doorway/facade
Figure 6.15: Interior northwest corner of Structure 154 at YA-050
Figure 6.16: Detail of central niche on north wall, YA-050, Structure 154
Figure 6.17: Detail of plaster, YA-050, Structure 154
Figure 6.18: Small house of Type 1 masonry, from the east. CO-164, Structure 28
Figure 6.19: House of Type 2 masonry, from the southeast. YA-050, Structure 107
Figure 6.20: Southwest corner of house of Type 3 masonry. YA-050, Structure 114
Figure 6.21: Facade of large house of Type 4 masonry, from the east.
YA-050, Structure 104.
Figure 6.22: Detail of split-river boulders of Type 4 masonry.
Figure 6.23: Large house of Type 5 masonry, from southwest.
Figure 6.24: Structure 17 (CO-100), from the northeast. Note tenon supports
for second floor
Figure 6.25: Large house of Type 5 masonry, from the south.
Figure 6.26: Wall fragment of Type 6 masonry, from the east.
Figure 6.27: Southwest corner of Structure 31
Figure 6.28: Southeast corner of Structure 31
Figure 6.29: House footprint area histogram
Figure 6.30: Boxplot of house area, grouped by masonry type
Figure 6.31: Boxplot of house area, grouped by site
Figure 6.32: Kallanka structure at YA-050, showing orientations of photos
Figure 6.33: Detail of facade fragment of kallanka (Structure 56) at CO-163
Figure 6.34: Corner remnant with Cuzco Inka style cutstone blocks in Yanque
(YA-041), facing east
Figure 6.35: Overview of the Casa Choquehuanca area, Yanque (YA-041)
Figure 6.36: Colonial house in Yanque (Casa Choquehuanca)
Figure 6.37: Large, three-storey chullpa with two doors (right side),
CO-098 (Fatinga)
Figure 6.38: Two-storey chullpa at CO-098
Figure 6.39: Adjoining two-storey chullpas at CO-098
Figure 6.40: Remnants of red pigment, chullpa at CO-098
Figure 6.41: Two-storey chullpa at CO-098
Figure 6.42: Small chullpa under rock overhang. Site CO-117
Figure 6.43: Cocoon style vegetal fiber mummy encasing
Figure 6.44: Primary feeder canals in the survey area on the north side
of the watershed
Figure 6.45: Primary feeder canals in the survey area on the south side
of the watershed
Figure 6.46: Panorama of abandoned feeder canals and Chilacotacocha reservoir
Figure 6.47: Large maqueta at site YA-162
Figure 6.48: Maqueta at YA-162
Figure 6.49: Large maqueta at YA-162
202
202
203
203
203
203
207
207
208
208
208
209
209
209
210
210
211
213
214
215
220
221
222
224
225
230
231
231
232
232
233
234
239
240
243
248
249
249
viii
Figure 6.50: Maqueta at YA-014
Figure 6.51: Small maqueta at site YA-032 (Chiquero)
Figure 6.52: Map of Pukara fortifications
Figure 6.53: Pukara fortification of site CO-165
Figure 6.54: Pukara fortification of site CO-168
Figure 6.55: Rectangular chullpa on the exterior of fortification wall, site CO-165
Figure 6.56: Pumachiri peak (CO-158)
Figure 6.57: Outer stone wall (denoted by arrows) of CO-158 from above
Figure 6.58: Inner wall encircling Pumachiri peak (CO-158)
Figure 6.59: Pukara fortification to the west of Tuti
Figure 6.60: Fortified settlement to the southeast of Tuti. Note ancient road
in foreground
Figure 6.61: Oblique airphoto of Uyu Uyu (YA-050)
Figure 6.62: Architectural Map of Uyu Uyu
Figure 6.63: Facade of structure 104 at Uyu Uyu, from the east
Figure 6.64: Airphoto of San Antonio/Chijra (CO-100)
Figure 6.65: Promontory of San Antonio (Sector N) from the southwest
Figure 6.66: Airphoto of Llanka (CO-127), including canals
Figure 6.67: Airphoto of Kitaplaza (CO-164)
Figure 6.68: Airphoto of Llactapampa and Tunsa, with major canals
Figure 6.69: Airphoto of Llactarana
Figure 6.70: Architectural map, habitational sector of Llactarana
Figure 6.71: Colonial/Republican era house in Yanque (YA-041)
Figure 6.72: Detail of masonry on gable end
Figure 6.73: Probable LIP/LH house at the southern edge of Coporaque (CO-161)
Figure 6.74: Same LIP/LH house, from southeast, showing doorway
Figure 6.75: Church of Coporaque
Figure 6.76: Detail of masonry, northwest corner of church of Coporaque
Figure 6.77: Circular house and adjoining corral at Jibillea (YA-093)
Figure 7.1: Examples of colonial ceramics
Figure 7.2: Colonial Period settlement pattern
Figure 7.3: View of kallanka and probable chapel at San Antonio (CO-100)
Figure 7.4: Map of the probable chapel at Uyu Uyu
Figure 7.5: View of probable chapel at San Antonio, from the northeast
Figure 7.6: Frontal view of probable chapel at San Antonio
Figure 7.7: Abutting interior join of vestibule wall in probable chapel at San Antonio
Figure 7.8: Facade of the chapel of San Sebastian, Coporaque
Figure 7.9: Perspective of chapel of San Sebastian after restoration
Figure 7.10: Interior of doorway, chapel of San Sebastian, Coporaque
Figure 7.11: Perspective of possible chapel in Yanque, from the east
Figure 7.12: Frontal view of possible chapel in Yanque
Figure 7.13: The reduccin of Yanque, showing community spatial divisions
Figure 7.14: The reduccin of Coporaque
250
250
253
254
255
257
259
260
262
263
264
267
268
272
276
279
282
284
285
289
290
292
292
294
294
295
295
296
301
303
313
316
317
318
323
323
323
324
325
325
332
333
ix
Figure 7.15: Abandoned irrigation features and terraces on Cerro Pallaclle
Figure 7.16: Abandoned irrigation features, Yurac Ccacca
Figure 7.17: Abandoned Inca canal and associated terracing
Figure 8.1: Schematic, ideal sociopolitical organization of Inkaic Province
of Collaguas
Figure 8.2: Schematic, ayllu outliers in herding villages subject to kurakas
in Yanque and Coporaque
Figure 8.3: Crop declarations by village, Yanquecollaguas, 1591/1604/1615-1617
Figure 8.4: Average landholding area per household
Figure 8.5: Landholding area per household
Figure 8.6: Comparison of average landholdings per household amongst
households with maize fields and households without maize fields
Figure 8.7: Proportion of households with livestock, by village
Figure 8.8: Livestock declarations by village, Yanquecollaguas
1591/1604/1615-1617
Figure 9.1:Dot density map of crop declarations in Coporaque, 1604/1615-1617
Figure 9.2: Landholdings by moiety, Coporaque 1604/1615-1616
Figure 9.3: Panorama of the Coporaque area
Figure 9.4: Landholding distribution, ayllu Cupi, Coporaque Hanansaya
Figure 9.5: Landholding distribution, ayllu Aipi/Cupi, Coporaque Hanansaya
Figure 9.6: Landholding distribution, ayllu Calloca, Coporaque Hanansaya
Figure 9.7: Landholding distribution, ayllu Checa Malco, Coporaque Hanansaya
Figure 9.8: Landholding distribution, ayllu Collana Malco, Coporaque Hanansaya
Figure 9.9: Landholding distribution, ayllu Icatunga Malco, Coporaque Hanansaya
Figure 9.10: Landholding distribution, ayllu Yumasca, Coporaque Hanansaya
Figure 9.11: Dot density map of left and right ayllus, Coporaque Hanansaya
Figure 9.12: Landholding distribution, ayllu of official state potters, Coporaque
Hanansaya
Figure 9.13: Landholding distribution, ayllu Collana, Coporaque Urinsaya
Figure 9.14: Landholding distribution, ayllu Pahana Collana Pataca,
Coporaque Urinsaya
Figure 9.15: Landholding distribution, ayllu Pahana Taypi Pataca,
Coporaque Urinsaya
Figure 9.16: Landholding distribution, ayllu Pahana Cayao Pataca,
Coporaque Urinsaya
Figure A.1: Preliminary Colca valley ceramic sequence
Figure A.2: Key to illustrations of ceramics
Figure A.3: Chiquero body sherds
Figure A.4: Rim diameters of Chiquero neckless ollas
Figure A.5: Chiquero neckless ollas
Figure A.8: Chiquero collared ollas
Figure A.9: Chiquero handles and lugs
339
341
342
352
363
371
374
375
379
381
384
397
399
403
408
409
410
414
415
416
417
420
422
426
427
428
429
453
454
456
457
460
461
462
463
464
x
Figure A.10: Chiquero jars (A-C) and ollas
Figure A.11: Rim diameters of Middle Horizon bowls
Figure A.12: Local Middle Horizon red-slipped rim bowls
Figure A.15: Local Middle Horizon bowls
Figure A.16: Local Middle Horizon decorated bowls
Figure A.17: Local Middle Horizon bowls (A-G) and drinking vessels (H-J)
Figure A.18: Rim diameters of Collagua I bowls
Figure A.19: Rim diameters of Collagua II bowls
Figure A.20: Collagua I cumbrous (A-C, E), and a collared (D) bowls
Figure A.21: Collagua I cumbrous bowls
Figure A.22: Collagua I vertical-walled (A-D) and cumbrous (E-G) bowls
Figure A.23: Collagua I cumbrous (A, F) and slightly open (B-D) bowls
Figure A.24: Collagua I vertical walled (A-C, F) and cumbrous (D, E) bowls
Figure A.25: Collagua I vertical walled (A-E) and flaring (F) bowls
Figure A.26: Collagua I flaring bowls
Figure A.27: Collagua II vertical-walled bowls
Figure A.28: Collagua II rounded open (A, B), and flaring (C-E) bowls
Figure A.29: Collagua II rounded open bowls
Figure A.30: Collagua II rounded open (A-C) and flaring (D-F) bowls
Figure A.31: Collagua II cntaros
Figure A.32: Comparison of Firing Quality, Collagua I-Collagua Inka
Figure A.33: Comparison of slip color, Collagua I-Collagua Inka
Figure A.34: Rim diameters of Collagua III bowls and plates
Figure A.35: Rim diameters of Collagua Inka and Inka plates
Figure A.36: Collagua III bowls
Figure A.41: Collagua Inka bichrome bowls and plates
Figure A.42: Collagua Inka bichrome plates
Figure A.47: Collagua Inka bichrome (A-H) and polychrome (I, J) plates
Figure A.48: Collagua Inka (A, E-G, I) and Inka (B-D, H) polychrome plates
Figure A.49: Collagua Inka bichrome (A, B) and polychrome (C) flat-bottom bowls
Figure A.50: Collagua Inka bichrome pitchers
Figure A.52: Collagua Inka arbalo rims and lugs
Figure A.53: Collagua Inka bichrome arbalo handle and body sherds
465
471
472
473
474
475
476
477
483
484
489
490
491
492
493
494
495
496
497
498
499
500
503
504
507
507
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
xi
Figure A.54: Inka bichrome and polychrome arbalo body sherds
Figure A.56: Colonial Period bowls
Figure A.57: Probable early colonial rimsherds.
Figure A.58: Probable early colonial rimsherd from YA-054 (Llactarana)
Figure A.59: Probable early colonial rimsherd from YA-041 (Yanque)
Figure A.60: Probable early colonial rimsherd from YA-050 (Uyu Uyu)
Figure B.1: Projectile Points
Figure B.2: Projectile Points
Figure C.1: Archaic Period sites
Figure F.1: Overview map showing locations of Maps F.2-F.5
Figure F.2: All sites registered, northwest block
Figure F.3: All sites registered, northeast block
Figure F.4: All sites registered, west-central block
Figure F.5: All sites registered, south block
530
531
535
536
536
537
537
539
540
542
586
587
588
589
590
xii
List of Tables
Table 2.1: Dry and wet periods derived from glacial and lake sediment cores
Table 3.1: Visitas to the Colca valley.
Table 4.1: Local sequence of occupational periods
Table 4.2: Modern toponyms in Coporaque that also appear in the visitas
Table 5.1: Formative Period sites (N=30)
Table 5.2: Middle Horizon sites (N=37)
Table 6.1: Late Intermediate Period Sites (N=53)
Table 6.2: Late Intermediate Period Site Counts by Class
Table 6.3: Late Horizon sites (N=72)
Table 6.4: House count by settlement
Table 6.5: Masonry typology
Table 6.6: Masonry type building counts/percentages, by site
Table 8.1: Villages and ayllus of Yanquecollaguas Hanansaya, 1591/1615-1617
Table 8.2: Villages and ayllus of Yanquecollaguas Urinsaya, 1591/1604
Table 8.3: Collagua colonists (mitmaqkuna) living in La Chimba of Arequipa, 1582
Table 8.4: Crop declarations by village, Yanquecollaguas, 1591/1604/1615-1617.
Table 8.5: Non-local household fields, Yanquecollaguas 1591/1604/1615-1617
Table 8.6: Landholdings and livestock of craft specialists, Yanquecollaguas
1591/1604/1615-1617
Table 9.1: Summary statistics, left/right predominance, Coporaque Hanansaya,
1615-1616
Table 9.2: Land tenure summary, left vs. right ayllus, Coporaque Hanansaya,
1615-1616
Table 9.3: Summary statistics, left/right predominance, Coporaque Urinsaya, 1604
Table 9.4: Summary of pre-reduccin ayllu residence patterns
Table A.1: Summary counts and percentages: all ceramics
Table A.2: Chiquero ceramics surface treatment scores (exterior)
Table A.3: Middle Horizon ceramic surface treatment scores by paste type
Table A.4: Collagua I and II bowl surface treatment scores (exterior surface)
Table A.5: Collagua III and Collagua Inka bowl and plate surface treatment scores
Table A.6: Comparison of average surface treatment scores, Collagua I to
Collagua Inka
Table B.1: Projectile point types and date ranges
Table C.1: Projectile point counts, grouped by time period
Table D.1: Survey artifact registry
Table E.1: Site/sector collections sorted by ceramic style
Table E.2: Site/sector collections sorted by time period
Table F.1: Site/sector registry
53
86
106
123
132
154
179
180
188
197
203
203
357
358
364
372
379
388
419
419
425
432
452
455
467
481
504
505
538
541
543
562
574
591
xiii
Acknowledgements
The archaeological field research for this project was carried out with the
authorization of Resolucin Directoral No. 615 from the National Institute of Culture,
Lima.
The field research for this project was made possible through the generous
funding of a Dissertation Research Grant from the Wenner-Gren Foundation for

Anthropological Research (Grant No. 6431). Initial post-field analysis and writing was
funded by a Wenner-Gren Lita Osmundson Grant. I gratefully acknowledge the support
of a Dumbarton Oaks Junior Fellowship in Precolumbian Studies during the 2001-2002
academic year. An Advanced Dissertator Fellowship from the University of WisconsinMadison provided support for the final stages of write-up during the 2002-2003 academic
year.
In completing this project, I am reminded of what a privilege it has been to be

able to conduct research that fascinates me and, even more so, to be able to exchange
ideas and work with so many talented, dynamic individuals.
I warmly thank my
committee members, Frank Salomon, Jason Yaeger, Neil Whitehead, Sissel Schroeder,
and Karl Zimmerer, for all of their advice and always-constructive criticism. My work
benefited tremendously from the input of such a diverse group of great scholars. I am
particularly indebted to my advisor, Frank Salomon, for my scholarly formation. His
intellect, curiosity, and incredible depth of knowledge of the Andean cultural world are
truly inspirational. I am grateful for his unflagging support throughout all stages of my
graduate studies. The early stages of formulating this project developed out of numerous
xiv
stimulating conversations with Neil Whitehead, who urged me to undertake an
intersectional course of study. I could never have anticipated the long and windingbut
ever-more scenicpath this decision would send me down. A long, strange trip indeed.
Jason Yaeger, Co-Chair of the dissertation, was tireless and selfless in providing me with
detailed feedback, often on short notice, through the analytical and write-up phases of the
project, and was always available just to talk. Thanks, friend.
I am grateful to Steve Stern, Florencia Mallon, James Stoltman, Mark Kenoyer,
and Gary Feinman for their support and advice during the early stages of the project. I
owe a particular debt of gratitude to Bill Denevan, who pulled out his maps and airphotos
during an office visit and, with infectious enthusiasm, encouraged me to go to the Colca
to explore research options. Two professors from my years as an undergraduate at the
University of Iowa, Mercedes Nio-Murcia and Laura Graham, had a more lasting impact
on me than they know, and inspired me to pursue graduate study in anthropology.
One incredible individual and dear friend, Maria Benavides, profoundly affected
my ability to undertake this project by so generously providing me with her entire
collection of archival documents related to the colonial history of the Colca valley,
including photocopies and transcriptions of the visitas that are the centerpiece of the
ethnohistorical portion of the study. My warmest thanks to her for her generosity and
good faith in my ability to make a worthy contribution with this treasure of
documentation. I treasure her wisdom and friendship.
In Lima, Kristoff Makowski, who took an interest in my project from the first
time I met him, opened his home to me and assisted me through the National Institute of
xv
Culture permitting process. I look back fondly on our many invigorating conversations
and thank him for his collegiality and generosity. Special thanks to Luis Jaime Castillo
Butters for his support in seeing my proposal through the proper channels at the INC.
My heartfelt thanks go to Willy Ypez Alvarez and Erika Simborth Lozada, who
assisted in all aspects of the archaeological fieldwork and postfield analysis.
We
developed a special bond that only working and living together in the field can create. As
experts in the archaeology of the Department of Arequipa, their contribution to the
project transcended the technicalI hope this dissertation does their hard work justice.
Many thanks to Ericka Guerra Santander, Co-Director of the Colca Valley Regional
Survey Project, for her contributions and work. I am grateful for all of the assistance of
Tamara Flores Ramos, who worked for hundreds of hours entering the visita data into the
database, photographed artifacts, and digitized ceramic diagrams.
Thanks to Evelin
Lpez Sosa for her help in the photo documentation of the artifact collections. Bruce
Owen graciously provided me with a copy of a ceramic data matrix and coding scheme
that proved very useful, and answered my other questions in detail while I was working
in the lab.
The archaeological portion of this project wouldnt have been possible if not for
the goodwill and graces of the villagers of Yanque and Coporaque, who embraced us,
educated us, and often even fed us and slaked our thirst with chicha as we stumbled
ponderously through their fields. My debt to my padrinos in Yanque, Gerardo Huaracha
Huaracha and Doa Luisa Cutipa de Huaracha, is greater than I can ever repay.
Diuspagarasunqui.
I warmly thank the Mayors of Yanque and Coporaque, Ramn
xvi
Cayllagua Cayllagua and William Bernal Huarca, for their support and collaboration with
my efforts. My gratitude to Justino Inka Montalvo, President of the Irrigators Committee
of Yanque Urinsaya, for allowing us to accompany the community during the annual
cleaning of the Misme canal. Sister Antonia Kayser of the parish of the Immaculate
Conception of Yanque always opened her doors to me and cared for me when I fell ill.
In Arequipa, I lived and worked in the research house of the Centro de
Investigaciones Arqueolgicas de Arequipa (CIARQ)a true home away from home in
sunny Sachaca.
I thank Karen Wise and Augusto Cardona Rosas, Co-Directors of
CIARQ, for all manner of logistical and bureaucratic support.

My knowledge of the archaeology and ethnohistory of Arequipa owes largely to
the many stimulating conversations with Mximo Neira, pioneer of Colca valley
archaeology, and with Flix Palacios Ros, Guillermo Galdos Rodriguez, and Augusto
Cardona Rosas. I thank Luis Sardn Cnepa, Director of the National Institute of Culture
in Arequipa, as well as my friends on the staff of the Department of Archaeology of INCArequipaPablo de la Vera Cruz Chvez, Lucy Linares Delgado, Marko Lpez Hurtado,
and Cecilia Quequezanafor helping me through the bureaucratic aspects of the
permitting and inspection process.
After returning from the field, I had the good fortune of spending an academic
year with a vibrant community of scholars at Dumbarton Oaks. I am especially grateful
to Jeffrey Quilter, Director of Precolumbian Studies, for all of his great advice and input
on the project. George Lau, Allan Maca, and Carolyn Tate, my colleagues while at
Dumbarton Oaks, all helped me formulate my ideas in the early stages of writing, and
xvii
contributed immeasurably to my thinking. Warm thanks to Loa Traxler and Jennifer
Younger for their detailed feedback and help.
My family has always been a bedrock of love and support for me, and had more
faith in me than I sometimes deserved. My parents, Arlyn and Grace Wernke, always
encouraged me to explore my interests and find my own path. My sister, Suzanne Bautz,
encouraged me every step of the way. Without their supportpersonal, emotional, and
financialI wouldnt be where I am today.
Finally, my deepest thanks and gratitude go to my wife, Tiffiny Tung, who, more
than any other person, helped me through this process and was there for me at every
critical juncture, even as she was finishing her dissertation at the same time.
Her
formidable intellect and keen observations have fortified this dissertation. Thank you for
always believing in me.
xviii
Abstract
This study investigates relationships between political organization and land-use
from Formative through early colonial times in the southern Andean highlands. The
project combines a detailed synchronic view of agro-pastoral production and exchange
with a diachronic view of how the cumulative effects of distinct land-use practices under
autonomous and imperial rule transformed the built landscape and shaped later land-use
and political economic strategies. The archaeological portion of the project is based on a
90 km2 systematic survey in the Colca valley of Arequipa Department, Peru.
The
ethnohistorical portion reconstructs the areas local- and regional-scale economic

relationships through spatial analysis of landholding and livestock declarations in a series
of colonial censuses (visitas).
The survey findings reveal a marked expansion of settlement and irrigated
agricultural infrastructure during the Late Intermediate Period (LIP; AD 1000-1450),
when a series of villages with distinctive Collagua domestic architecture were
established.
During this period, inter-settlement political relations appear to have
oscillated between competition and cooperation. Defensible settlement locations and

fortifications signal conflict, while hydrological relationships among long primary canals
indicate that water apportionment was coordinated at a supra-settlement scale. Upon
imperial incorporation of the valley, Inka administration hierarchized previously
heterarchical political relations. The Inkas established an administrative center, and
major LIP settlements became secondary administrative sites, where rustic Inka imperial
architecture was prominently situated in association with local elite domestic structures.
xix
Analysis of Spanish visitas provides a complementary view of how local kindreds
(ayllus) had been reordered into a nested administrative hierarchy under Inka rule. The
formal ayllu structure mimicked Cuzco Inka norms unevenly, with startlingly close
matches occurring in the lower moiety. Reconstructed early colonial ayllu land tenure
patterns reveals dispersed household landholdings organized by ayllu affiliation.
Comparison of ayllu land tenure constellations with the terminal prehispanic settlement
pattern provides a basis for retrodicting Inka-era ayllu residence patterns. This analysis
shows that the highest-ranking ayllus resided at the administrative center and at
secondary administrative sites. At the regional scale, ayllus and their leaders articulated
economic flows between high altitude herding populations and specialized maize
production enclaves in low-lying valleys to the south.
PART I: INTRODUCTION AND BACKGROUND
1
Chapter 1
Community and Landscape in the Andes
Introduction
The complex interplay between political organization and land-use patterning in the
ancient Andes has long been a subject of intense scrutiny and debate. Researchers have
illustrated how diverse complementarity relationships between the altitudinally-dispersed
resources of the Andean landscape and the cultural schemata and social forms devised to
exploit it developed throughout Andean antiquity. In the two generations of research
following the paradigmatic formulation of anthropologist John Murra (1964; 1968; 1972),
scholars have traditionally approached these complementarity relationships as a means of
cultural-ecological adaptation, guided by a putatively ancient Andean ideal of community
self-sufficiency. At the same time, household and regional-scale studies have illuminated
connections between changing land-use patterns and changes in, respectively, domestic and
political economy (see Masuda, et al. 1985). The wider temporal frame of archaeological
studies has had a relativizing effect; structured relations between ecology and economy
that in an ethnographic frame appear stable and locked in homeostatic balance appear more
dynamic and historically contingent over archaeological time spans (Erickson 1988, 1999,
2000; Rice, et al. 1989; Stanish 1989a, 1992, 2003; Van Buren 1993, 1996).
In the Colca valley, a major Pacific drainage in the semiarid western flanks of the
southern Peruvian cordillera (Department of Arequipa), researchers have contributed to this
dynamic view through combined archaeological, historical, and ethnographic study. The
Colca valley is one of the most intensively terraced locales in the New World (Denevan
2
2001; Donkin 1979), making it an ideal location for the investigation of prehispanic Andean
political dynamics and land-use patterning. A palimpsest mosaic of approximately 11,000
ha of agricultural terraces covers virtually all geomorphic surfaces below about 4000 masl
in the valley. Previous researchers have identified a sequence of major transformations of
the valley landscape in which unirrigated sloping fields and terraces that were constructed
and used at least as early as the beginning of the Middle Horizon (AD 500-1000) were
subsequently abandoned and replaced by massive complexes of irrigated bench terrace
systems during the Middle Horizon and Late Intermediate Period (hereafter LIP; AD 10001450), a process that probably intensified during the Inka occupation during the Late
Horizon (Brooks 1998; Denevan 1987; 2001:185-205; Malpass 1987; Shea 1987; Treacy
1989).
Explaining such long-term dynamics in production and sociopolitical organization
remains a central issue of debate in Andean archaeology, as well as in other world regions.
Working from a top-down political economy perspective, many archaeologists have
focused on control over labor deployment and its produce as the primary factor in the onset
of agricultural intensification and social inequality, contending that the construction and
maintenance of intensive agricultural infrastructure is predicated on the compulsion of nonelites to produce above their subsistence needs by a stratum of political elites (Earle 1987;
Kolata 1986, 1993, 1996a; Stanish 1994). By contrast, others have argued that households
or other small-scale corporate collectivities can, and have, constructed large-scale, intensive
agricultural systems such as high altitude irrigated terrace and raised-field systems in the
3
Andean highlands (Erickson 1988, 1999, 2000; Gelles 1990, 2000; Graffam 1990, 1992;
Treacy 1989).
Locally, hypotheses have focused on terracing and irrigation in the Colca valley as
technological innovations for adapting to climatic perturbations (Brooks 1998; Denevan
1987; 2001:185-205; Treacy 1989).
Although they identify relevant variables, these
hypotheses appear incomplete, since they do not account for the ways in which such factors
vary in relation to demographic, economic, or political factors. Unlike most other areas of
the Andes, landscape archaeology in the Colca valley has outpaced our knowledge of
settlement organization and patterning. As a consequence, we know comparatively little
about the origins, development, and organization of the two major late prehispanic ethnic
polities of the valley, the Collaguas and Cabanas, and how their historical trajectories relate
to the observed changes in the valleys built landscape.
Document-based studies of
protohistoric Collagua political and economic organization have presented the Collagua as
a classic example of a large-scale ethnic seoro (chieftaincy) with outlier ethnic colonists
settled in neighboring, low-lying valleys to the south, forming a vertical archipelago akin to
the archetypal example of the Lupaqa polity of the neighboring Lake Titicaca Basin
(Benavides 1987b; Galdos Rodrguez 1984, 1987; Mlaga Medina 1977; Pease G. Y.
1977). Test excavations have yielded a preliminary ceramic sequence that identifies a local
LIP/Late Horizon ceramic style associated with the Collagua (Brooks 1998; de la Vera
Cruz Chvez 1987, 1988, 1989; Malpass and de la Vera Cruz Chvez 1986, 1990; Neira
Avendao 1961; Shea 1997), and reconnaissance projects have recorded several major
settlements from these late prehispanic periods with well-preserved, distinctive local
4
domestic architecture (Brooks 1998; de la Vera Cruz Chvez 1988; Guerra Santander and
Aquize Cceres 1996; Neira Avendao 1961; Oquiche Hernani 1991; Shea 1986a, b, 1987).
However, no systematic, full-coverage survey had been conducted in the valley prior to this
project, leaving several critical questions unanswered, such as: What are the culturehistorical antecedents of the Collaguas and Cabanas, and how were these pre-LIP local
groups organized?
How did the Wari and Tiwanaku Middle Horizon imperial states
influence local political and economic organization and, by extension, the configuration of
anthropogenic landforms in the valley? How did local Collagua communities develop out
of the post-Middle Horizon regional balkanization during the LIP, and how were they
organized prior to Inka imperial incorporation?
Ethnohistorical reconstructions of the political organization of the Inkaic Collaguas
province suggest a penetrating imperial presence in which local kindreds (ayllussee
below) were reordered into a formal hierarchy according to Inka categories of rank
(Benavides 1989; Cock Carrasco 1976-77, 1978; Gelles 2000; Guillet 1992:18-19;
Prssinen 1992:362-371; Pease G. Y. 1977; Robinson 2003; Rostworowski de Diez
Canseco 1983:121-123; Zuidema 1964:115-118).
Yet in the absence of systematic
settlement pattern data, archaeological indices for Inkaic influence or presence in the valley
remained poorly defined and debated (Brooks 1998; de la Vera Cruz Chvez 1987; Malpass
1987). Thus, critical questions have been left unanswered regarding how local and imperial
structures of power articulated under Inka rule: Did the Inka significantly reorganize the
local settlement system, or was Inka administration mediated through extant settlements
5
and their agents? How did changes in Late Horizon settlement patterning relate to the
(re)configuration of local agricultural landforms?
In this dissertation, I address these questions through a spatially-integrated
archaeological and ethnohistorical investigation.
This study tracks the origins and
development of the communities in the core area of the Collagua polity and explores how
their later incorporation into the Inka and Spanish empires related to changes in community
organization and land-use patterning.
The project is multi-scalar, both spatially and
temporally, and combines analysis of two primary data sets: first, archaeological data from
a 90 km2 full-coverage survey in the central section of the Colca valley that coincides with
an area described in colonial documentary sources as the seat of authority in the Inkaic and
colonial province of Collaguas; and second, demographic and cadastral data from a series
of detailed early colonial censuses (visitas) of an area that overlaps with the archaeological
survey coverage. I use the archaeological data to provide a long-term view of the origins
and development of the communities in this core area of the Collagua ethnic polity. My
analysis focuses on changes in settlement and agro-pastoral infrastructural patterning that
occurred between the period of autonomous Collagua rule during the LIP and the
subsequent Inka and early Spanish imperial occupations. In complement to this temporallydeep local view, I use the visita data, from the years 1591, 1604, and 1615-1617, to
reconstruct both the regional political and economic organization of the Inkaic and early
colonial province of Collaguas, and the land tenure patterns of local communities within the
archaeological survey area. Within the survey area, I also compare the early colonial land
tenure patterns of local ayllus to the terminal prehispanic settlement pattern as a means of
6
retrodicting their Inka-era residence patterns, thereby affording a detailed, culturallyinformed view of how hybrid local/imperial sociopolitical institutions were mapped onto
the settlement system under Inka rule.
A Political-Ecological Approach to Community and Landscape

The Human/Environment Interface: Landscape as Household and Habitat
Understanding diachronic relationships between communities and the landscapes
they inhabit requires analytical frameworks that can account for both constrictive, limiting
processes and constructive, generative ones.
Recent approaches are working in this
direction, as the reactive, one-way causality implicit in the concept of "adaptation" has been
challenged by models of "interpenetrating," "dialectic," and "recursive" relationships
between humans and their environment. Under various rubrics, such as political ecology
(e.g., Bryant 1992; Zimmerer 2000), historical ecology (e.g., Bale 1998; Crumley 1994;
Kirch and Hunt 1997; Patterson 1994), and landscape archaeology (e.g., Bender 1992;
Crumley 1999; Erickson 2000; Knapp and Ashmore 1999; McGlade 1995; Rossignol and
Wandsnider 1992; Treacy 1994; Wagstaff 1987), recent formulations are collectively
changing underlying assumptions regarding human/habitat relationships, as dynamic
disequilibrium has come into favor as a guiding principal over homeostatic balance (Botkin
1990). Simultaneously, a shift toward understanding landscapes as historically-contingent
and anthropogenic has highlighted the shortcomings of the nature:culture dyad still
prevalent in the literature, which reduces explanations of landscape dynamics to a
comparison of the causal forces of natural versus cultural systems of activity.
The
7
inadequacy of this ecology:economy opposition, however expressed, becomes especially
apparent when considering that both terms derive from the Greek root okow (oikos)
household/habitata fact that points toward an alternative approach that considers
human synergy with the environment as the dynamic process at the interface between
nature and culture (Whitehead 1998).
The paradoxes that arise from this economic:ecological ambiguity are especially
apparent in the Andean region. Discussion of human/environment relationships in the
Andes has been dominated by the vertical complementarity model of Murra (1964; 1968;
1972), which posits that the altitudinally-compressed ecological zonation of the Andes
required societies of varying scale and complexity to adopt mechanisms of access to a
sufficiently diverse resource base. Rather than relying on trade with outside groups or
centralized markets, the uniquely Andean solution, according to Murra, was the direct
colonization of multiple ecological tiers by a single ethnic group, forming a vertical
archipelago of settlements linked through intra-ethnic reciprocal and redistributive
exchange (Murra 1972). Thus, the maximum control of ecological tiers came to be
viewed by many as a pan-Andean cultural ideal of balancing resources and demography,
differentially achieved through space and time. The development of asymmetrical systems
of redistribution was therefore treated as epiphenomenal to the adaptive process of securing
ecologically-diverse resources (see Van Buren 1993; Van Buren 1996).
However, both the ecological and political-economic aspects of the verticality
model have come under increasingly critical scrutiny.
Ecological anthropologists and
geographers have noted that the construction of stable and secure production systems in the
8
Andes is predicated on a simplification of highly localized ecological variability by humans
(Brush 1977:9). Humans transform ecological gradients into ecological tiers (Erickson
2000; Zimmerer 1999), themselves composed of smaller anthropogenic production zones
(Mayer 1985) created through human interventions such as irrigation, augmentation of
soils, and alteration of microclimates. These observations require that a more dynamic
view be developed that is mindful of the ways in which the landscape itself is altered by
humans so as to lessen ecological strictures. As Mayer (1985:47) observes,
When we think of production zones as man-made things, rather than as
adaptations to the natural environment, our attention is directed to how
they are created, managed, and maintained. Then the importance of the
political aspects of control by human beings over each other in relation
to how they are to use a portion of their natural environment will again
come to the fore.
Recent researchers have indeed emphasized the political aspects of control
inherent in vertical complementarity. Archaeological investigations of the prototypical case
study of a vertical archipelago systemthat of the Lupaqas of the Titicaca Basinhave
found that the outlier Lupaqa settlements in the lowland valleys of Moquegua were not of
the great antiquity hypothesized (Stanish 1985, 1989a, b, 1992), and probably never
functioned to provision whole populations (Van Buren 1993, 1996, 1997). Building on
these findings, Van Buren has presented a general critique of the verticality model as overly
functionalist and essentialist, suggesting instead that vertical complementarity is better
viewed as a dimension of social poweri.e., as a mobilization of labor and material by
ethnic lords towards specific political ends (Van Buren 1993, 1996).
In this dissertation, I propose a bridging position that approaches the dynamics of
land-use and sociopolitical change as complex, multi-causal processes both constructed
9
culturally and constricted ecologically. I explore human/environment synergistic processes
through a theoretical framework that posits a recursive relationship between a particular,
Andean oikosthat of the late prehispanic and early colonial Colca valley and surrounding
regionand the ecological and economic practices of the members of local Collagua
communities.
Here, rather than documenting the formal attributes of a vertical
complementarity system as a social formation or structure, the processual variability of

land-use practices is of primary interest. Within my formulation, ecological and economic
practices are structured by, while also constituting through their aggregate effects, the
household/habitat of oikos. This approach therefore conceptualizes complementarity as a
particular kind of agency, related recursively to the structures of landscape and community
organization.
Analysis at several temporal and spatial scales is required to understand the
historical and spatial relationships between built landscapes and the communities that alter
and adapt to them. Diachronically, the construction of built features, such as canals, terrace
complexes, anthropogenic soil regimes, and so on, is an aggregative process that alters the
physical and energetic parameters of production at varying rates and scales, depending on
historical circumstances. Synchronically, anthropogenic landscape features constitute the
congealed labor (Lansing 1991:12) or landesque capital (Blaikie and Brookfield
1987:9-10) that structure production. While political changeas in the case of the Inka
and Spanish conquests of the Andean regioncan occur over short (event- and
conjuncture- level) time spans, landscape-scale infrastructures are generally designed to
be durable and stable (Braudel 1972). Systems of production, qua systems, are organized
10
to produce reliable returns on given inputs of labor and material, and in this sense are
inherently conservative in nature. These contrasting temporal rhythms of political and
economic-ecological change therefore produce a dialectic from which new, hybrid social
formations emerge (Smith 1992a, b). Thus, built landscapes have tangible agentive force
that both constrain and present opportunities for economic and ecological praxis. In order
to understand these varied scales of spatial and temporal interaction, this dissertation
combines analysis of long-term changes in the built landscape of the Colca valley, with a
synchronic view of regional and local-scale patterns of production and exchange during
early colonial times.
Communities, Natural and Imagined: The Case of the Andean Ayllu

My approach to the archaeological and historical study of ancient communities
builds on interactionist frameworks which posit that communities are dynamic, sociallyconstructed institutions that structure and are structured by supra-household interactions
(for a review of this and other approaches, see Yaeger and Canuto 2000). I conceive of
communities as matrices of social interaction that both create and emerge from a sense of
common interest and affiliationa sense of shared identity. This orientation contrasts with
functionalist and behavioralist approaches, which consider community as the natural,
fundamental unit of supra-household social and biological reproduction, necessarily
constituted by co-residence, proximity, and shared economic/ecological praxis (e.g.,
Murdock 1949; Redfield 1955, 1956). According to the behavioralist conception, the
quality of distinctiveness (Redfield 1955, as cited in Kolb and Snead 1997:611) that
11
defines community identity is epiphenomenal to the everyday interactions of living and
laboring in a bounded and defined space (for critique, see Isbell 2000; Wolf 1956). Some
archaeologists have recently advocated this framework as the most pragmatic and testable
approach to reconstructing and comparing prehistoric community organization (Kolb and
Snead 1997).
Clearly, proximity and co-residence affect the patterning and frequency of
interaction between social actors and groups, and community identity often is expressed in
the idiom of settlement, neighborhood, or territory. Patterns of daily interaction constitute a
structure of power and meaning that constrain the parameters of imagination and action
(Bourdieu 1977), although through critical awareness and reflexive monitoring, agents can
question, contest, and alter those structures (Giddens 1979). However, community identity
is not only determined by the propinquity for frequent interaction, but is also imagined
(Anderson 1991). Communities can be composed of individuals who do not frequently
interact or even know one another, but nonetheless share a deep sense of affiliation and
common interest (Anderson 1991). The sense of solidarity in communities emerges from
discourses and practices of affiliation (Yaeger 2000) that essentialize within-group
commonalities and interests (us) in contrast to others (them) (Barth 1969). Within my
conceptualization, therefore, ethnicity is one of many dimensions of community identity
an approach, as I discuss below, entirely befitting of the Andean context. In sum, as a
dynamic process of social identity, community need not be associated with any particular
socio-spatial scale or unit; communities are just as likely to cut across spatial boundaries as
to adhere to them (Goldstein 2000).
12
Ever since the early years following the European invasion (Polo de Ondegardo
1917 [1571]), students of Andean communities have recognized their supra-local,
archipelagic patterns of residence, production, and exchange (Murra 1964, 1968, 1972).
The concept of ayllu was central to the social, political, and economic articulation of
territorially-discontinuous communities in the Andes (Abercrombie 1986; Cock Carrasco
1981; Isbell 1997; Platt 1982; Salomon 1991; Spalding 1984). Ayllu, commonly translated
as clan, was emically-defined as a multi-scalar concept that could reference any segment
along a continuum of socially- or biologically-related collectivities, from the consanguines
of a patrilineage, to clan-like groupings of patrilineages, moieties, and even an entire ethnic
group. For example, Platt (1986) has illustrated how the modern-day Macha of Bolivia
conceive of ayllus in a nested fashion, from minimal ayllus of small patrilocal groups of
neighboring households, to minor and major ayllus made up of groups of related ayllus,
and finally a maximal ayllu encompassing the ethnic group as a whole. The scalar
plasticity of the concept has led to a plethora of definitions that emphasize different aspects
of ayllu membership and organization (for reviews of definitions, see Goldstein 2000:184186; Isbell 1997:101-135; Salomon 1991:21-23; Spalding 1984:28-29). Common to all,
however, are two attributes: ayllus are resource-holding corporate collectivities, and ayllu
membership is reckoned by reference to an actual or fictive focal ancestor.
As resource-holding collectivities, protohistoric ayllus mediated households access
to agricultural land and other immovable assets (Patterson and Gailey 1987; Rowe 1946b;
Salomon 1991:22). While ayllus held these resources as corporate entities, ayllu members
gained access to land and other resources in reciprocity for their labor in collective work
13
projects (e.g., canal cleaning, terrace construction and maintenance) and their participation
in rituals of affiliation, including ancestor-veneration (see below).
However, ayllu
landholdings were not always spatially discrete, and in this sense they were not necessarily
conceived of as bounded units (Murra 1980:30-31).
Ayllus of different scales of
inclusiveness organized agro-pastoral infrastructural systems and the labor to construct and
maintain them, from the daily interactions of cultivation, terrace maintenance, and the
distribution of irrigation water, to larger-scale mobilizations such as canal construction and
cleaning (Gelles 1993, 1995, 2000; Guillet 1978, 1981, 1992; Mitchell 1976; Sherbondy
1982; Treacy 1989). Like their land tenure patterns, ayllus were not residentially discrete;
not only could ayllus occupy several vertically- and horizontally-dispersed settlements, but
members of several ayllus could also share single settlements. For example, early colonial
native testimony from Huarochir province indicates that prehispanic settlements were
composed of multiple ayllus (Salomon 1991:23-24).
Also, the protohistoric Andean
concept of settlement, or llacta, did not necessarily coincide with a nucleated settlement,
but appears instead to have encapsulated a huaca (a shrine, object, or landscape feature that
embodies the superhuman founder ancestor of an ayllu), its territory, and the people within
it (Salomon 1991:23-24). So even the concept of a settlement in the late prehispanic and
early colonial Andes was not territorially bounded in the sense of a discrete cluster of
houses.
As ancestor-focused kindreds, ayllu affiliation was reckoned by reference to a
pointillist landscape of ancestral huacas that were hierarchically-related in space and time.
As Goldstein (2000:185) notes,
14
...even those aspects of ayllu identity that explicitly refer to place refer
not to spatial boundaries, but to the huacas that link an ayllu to its
ancestors...This association of ancestor worship to group identity
suggests that as a community form, the ayllu is more genealogical than
territorial in natureit is bounded by history rather than borders.
Members of minimal ayllus traced their affiliation by reference to a focal, chartering
ancestor, usually embodied in the actual mummified corpse of that individual. The criteria
used in the actual reckoning of inclusion is a perennial topic of debate in Andean
anthropology, but in practice, the precise genealogical relationship appears to have been of
less importance in the reckoning of ayllu membership than a persons social conduct and
political standing as a genealogically-connected individual (Salomon 1991:22; Spalding
1984:28-29).
Ayllu members reaffirmed and reified their community affiliations by
consulting and feting their ancestral mummies, who occupied cities of the deadclusters
of above-ground, multiple-interment mortuary monuments (houses of the dead, or
chullpas in modern archaeological parlance), often situated on prominent hilltops or under
cliffs near settlements (see Dillehay 1995; Isbell 1997). Such ancestral mummies were also
considered huacas (Salomon 1995). As such, they were also conceived of as the proximate
descendants of a hierarchy of superhuman huacaseach the guardian of fertility in its
domain, and each increasingly remote, both spatially and in terms of kinship reckoning
terminating at its apex in the origin-place (place of dawning, or pacarina) of an entire
ethnic group, usually a prominent mountaintop (Salomon 1991; Spalding 1984).
As political entities, I argue that the ayllu represents an example of heterarchical
organization, in which ...each element possesses the potential of being unranked (relative
to other elements) or ranked in a number of different ways... (Crumley 1979:144). A
15
concept originally developed in the field of artificial intelligence (see Crumley 1987),
heterarchical organization is based on lateral connections between elements in an
organization, each of which can take a more dominant role depending on the circumstances.
By contrast, hierarchical organization is based on the vertical integration of elements of
fixed rank.
Examples of complex, orderly, heterarchical systems abound in nature,
including the neural network of the human brain, or the patchiness of plant and animal
communities such as in the case of the ecological tiers of the Andes. Hierarchy represents
only a specific kind or state of heterarchical organization and in this sense is a more
restricted concept (Crumley 1979:145). Crumley (1975; 1979; 1987; Ehrenreich, et al.
1995) has proposed that heterarchy provides an important alternative metaphor and model
for the comparative and diachronic study of social complexity, which has long been
conflated with hierarchy. Thus, rather than charting the presence, absence, or degree of
complexity of a social system by reference only to its degree and strength of vertical
integration, the concept of heterarchy provides a more three-dimensional perspective that
also accounts for flexible, horizontal integration between social groups that can be
variously ranked according to historical circumstances.
Testimony in early colonial texts indicates that the ethnic polities subsumed by Inka
administration were self-defined as bundles of rival ayllus of fluid prestige, rank, and
wealth (see, e.g., Rostworowski de Diez Canseco 1983). The political organization of
protohistoric ayllus, as well as inter-ayllu power relations, were governed according to both
ascribed and achieved criteria of rank. The huacas of clan-level ayllus generally were
conceived of as siblings borne of the origin huaca of the ethnic group (its place of
16
dawning, or pacarina). For example, as we know from the Huarochir testimony, the
origin huacas of clan-like ayllu sets often were ranked according to their birth order from
the apical huaca in the ethnic charter myth. According to this ideology, ...the firstborn or
leading member of a set (e.g., noble heads of a villages component ayllus) functions within
the set as first among equals, but outside the set as the totalizing representative of it
(Salomon 1991:20). In this way, native lords (kurakas), as ayllu representatives, could also
be ranked according to an ideal social structure derived from the relative birth-order
position of their respective huacas. However, crosscutting these ideal-typical criteria of
genealogical inequality, individuals and ayllus could also achieve higher rank through
supremacy in warfare. The thousands of Late Intermediate Period hilltop fortifications
(pukaras) throughout the central and south-central Andes strongly indicate that both intraand inter-ethnic conflict was endemic prior to Inka consolidation.
For example, the
distribution of pukaras in the northwestern Lake Titicaca Basin indicates that the Qolla and
Lupaqatwo major ethnic polities with populations of nearly 100,000 peoplewere much
more politically decentralized (i.e., heterarchical) prior to Inka incorporation than their
leaders early colonial memorial accounts depicted (Stanish 1997a; 2003:209-220). Indeed,
their coherence as political entities may have been most salient primarily when they were
faced with a common external threat such as the Inka army (Stanish 2003:209-220). Such
macro-scale hierarchical political organization appears to have been rather exceptional and
ephemeral in the Andean highlands during the LIP.
Elite manipulation of ayllu ideologymost elaborated by the Inkainvolved the
representation of asymmetrical, hierarchical relationships within and between clan-like and
17
maximalayllus in the same terms as the lateral, symmetrical links between consanguines
of lineage-like micro-ayllus (Murra 1956, 1980). Following in the substantivist tradition
of Murra, archaeologists and ethnohistorians have demonstrated how Inka strategies of
expansion and consolidation were so successful in part because they manipulated key
cultural principles and practices familiar to their subjects, such as vertical complementarity
as a logistical system for articulating populations and resources (see Masuda, et al. 1985),
conspicuous public feasting as a primary forum for displaying state largesse (e.g., Morris
and Thompson 1985), and ancestor veneration as a primary idiom of political discourse
(e.g., Rostworowski de Diez Canseco 1983).
Clearly, ayllus constituted the primary building blocks of imperial administration,
and Inka statecraft relied heavily on the representation of state/subject relations as an
extension of ayllu relations. However, in effect, Inka policies also significantly altered
intra- and inter-ayllu relations, and fomented the formation of new imagined communities
of the state. The Inkas sought to build stable, hierarchical structures of governance out of
heterarchically-organized ayllus by amplifying and codifying latent or extant inequalities of
rank and its means of attainment, while preserving the corporate character of the ayllu
itself. In Cuzco, the royal ayllus (panacas) sealed-off achieved status as a criterion of rank,
and membership rank was narrowly defined according to bilateral reckoning (Julien 2000;
Zuidema 1977). In the provinces, imperial administration submerged local differences as
some ethnic groups and elites were promoted in status relative to others. As a result of
Inkaic consolidation and centralization, the authority and domains of pliant ethnic elites
expanded greatly under Inka rule. The process of consolidating competing ethnic polities
18
into vertically-integrated provincial units has been demonstrated in many cases, from the
northern provinces of Pichincha and Imbabura in modern Ecuador (Salomon 1986), to the
Cajamarca province of northern Peru (D. G. Julien 1993), the Wankas and Xauxas of the
Mantaro valley in the central Peruvian Andes (Costin and Earle 1989; D'Altroy 1987,
1992), and the Qolla and Lupaqa of the Titicaca Basin (Hyslop 1976; Julien 1983;
Lumbreras 1974a; Stanish 1997b). In this study, I illustrate in detail how the autonomous,
heterarchical communities of the Collagua ethnic polity of the Late Intermediate Period
were reordered under Inka imperial occupation according to a formal hierarchy structured
in the image of Inka ideals.
Alternative Hypotheses and Their Correlates

Given the inherently broad temporal and spatial scope of the study, the following
hypotheses are designed to take soundings across the broad historical arc of settlement and
agricultural production in the Colca valley, ranging from the earliest period of sedentary
human occupation associated with unirrigated agricultural production, the intensification of
settlement and production during the LIP and Late Horizon, to the subsequent contraction
of settlement, population, and production during early colonial times.
As I discussed above, most field research in the Colca valley has been focused on
understanding changes in the built landscape of the valley. The findings of previous
researchers can be divided into a sequence of four principal periods of landscape
transformation: 1) an early period, dating at least as early as the beginning of the Middle
Horizon and almost certainly earlier (Brooks 1998; Treacy 1989), in which Colca valley
19
agriculturalists cultivated in unirrigated sloping field and terrace systems, many clustered
around, and augmenting, natural drainages (quebradas), 2) a subsequent phase, dating most
likely to the Middle Horizon and Late Intermediate Period, during which local peoples
abandoned these unirrigated field systems and constructed extensive canals and massive
complexes of rock-faced bench terrace complexes covering virtually all slopes and surfaces
in the lower reaches of the valley (Brooks 1998; Malpass 1987; Treacy 1989), 3) a period
of probable expansion and intensification of irrigated bench terrace production under Inka
rule during the Late Horizon (AD 1450-1532) (Malpass 1987; Shea 1987; Treacy 1989),
and 4) a period of deintensification of production during colonial and early republican
times, including the abandonment of many bench terrace complexes and canals, especially
along the upper margins of the valleys agricultural core (Denevan 1987; 2001:185-205).
Early Agriculture and Settlement in the Colca Valley

Building on these findings, the first hypothesis is aimed at clarifying the nature of
productive and sociopolitical dynamics during the earliest agriculturalist occupation of the
valley, associated with an accretionary pattern of unirrigated sloping fields and terraces.
Given that agriculture is impossible without irrigation in the valley today, previous
researchers have focused on climatological and technological factors for explaining the
functioning of these early rainfed- and runoff-based field systems, suggesting that, without
irrigation canals, early agriculturalists adapted to wetter climatic conditions by situating the
fields around natural drainages and harvesting water by diverting runoff to the fields
(Brooks 1998; Treacy 1989).
20
These explanations, however, only account for factors of ecological constriction (the
absolute moisture requirements of crops) and economic construction (water harvesting
techniques). In accord with the general model reviewed above, these factors are expected
to both structure and be structured by political organization. Treacy (1989:122-126) has
observed that the accretionary pattern of these early fields suggests that they were not built
as integrated, centrally-administered field complexes, but instead formed conglomerations
of segmented terraces, each of which could have been built by individual households or
other small-scale collectivities. The lack of settlement evidence, however, has precluded
any thorough evaluation of the timing of their construction and use, or of the social
organization of local groups that built and used them. In fact, prior to this project, no preMiddle Horizon settlements had been reported in the Colca valley. This is a major void in
the local occupational sequence, since the shift to sedentary, agricultural lifeways in
surrounding localesbest documented in the Lake Titicaca Basin to the east (Aldenderfer
1989, 1998; Bandy 2001; Bauer and Stanish 2001:138-141; Hastorf 1999; Stanish 1997a;
2003:99-109; Stanish, et al. 1994; Steadman 1995)occurred between about 2000 and
1300 years before the Middle Horizon, during the Early Formative (ca. 1500-800 BC).
Before this project, no Formative Period ceramics had been identified in the Colca
valley. Reportedly high (but unspecified) proportions of Middle Horizon ceramics were
recovered from surface reconnaissance of these early fields, but it seems unlikely that these
low intensity, unirrigated field systems would have been built and used as late as the
Middle Horizon, a period in which intensive irrigated agricultural systems spread
throughout the Andean highlands (Moseley 1992:216-230).
By contrast, Brooks
21
(1998:400-401, 405) has recently reported much earlier radiocarbon dates from buried soil
horizons in unirrigated terraces, leading her to hypothesize that terrace construction began
in the valley sometime prior to 2400 BC. However, these dates, derived from pooled
charcoal in bulk soil samples, are also clearly out of sync with the regional culturehistorical chronology, since they fall within the Late Archaic Period (4800-3000 BC)a
time in which groups throughout the region were engaged in hunting and gathering
lifeways.
The systematic survey data and artifact collections from this project provide a basis
for clarifying the social organization and chronological placement of early agriculture in the
valley. Given the agglutinated, low-intensity pattern evident in these field systems, I expect
that political and economic units at the time of their construction and use would have been
small and minimally hierarchical, with low supra-household demands for labor and surplus
production. Such political and economic organization could have been supported by the
low productive potential and acephalous, aggregative construction of the system. This
pattern is consistent with Early Formative social organization in the Titicaca Basin (Stanish
2003:99-109), and I expect that early agricultural settlement in the valley would have
developed roughly coevally with this and other surrounding locales. Trade links with the
Titicaca Basin throughout the Formative are abundantly evident in the distribution of
obsidian from the Chivay sourcelocated just east of the survey area of this project
throughout the Lake Titicaca Basin, where the great majority of obsidian at Formative sites
was procured from the Chivay source (Brooks, et al. 1997; Burger, Asaro, Salas, et al.
1998; Burger, et al. 2000). So prior to this project, it appeared quite probable that a
22
significant Formative occupational component had yet to be identified. I discuss corollary
metrics in further detail below, but in general terms, the settlement pattern correlate of this
Formative Period social organization would be characterized by a dispersed settlement
pattern of small settlements of approximately equal size.
The Middle Horizon: Expansionist States and Their Local Influence

The influence or presence of Middle Horizon expansionist statesthe Wari and
Tiwanakuin the Colca valley also remains poorly understood. Local Middle Horizon
ceramics are clearly derivative of Wari, not Tiwanaku, regional styles, and no Tiwanaku
ceramics have been recovered in the valley (Brooks 1998; de la Vera Cruz Chvez 1989;
Malpass and de la Vera Cruz Chvez 1986, 1990; Neira Avendao 1961, 1990). One large
Middle Horizon settlement (Achachiwa) has been identified in the lower reaches of the
valley (de la Vera Cruz Chvez 1987, 1988, 1989, 1996), although it appears to represent a
local settlement, and not a Wari administrative center (Brooks 1998:87; Schreiber
1992:104).
However, economic links with Tiwanaku are evident in the continued
predominance of obsidian from the Chivay source throughout sites in the Lake Titicaca
Basin, including the imperial center of Tiwanaku itself (Brooks, et al. 1997; Burger, Asaro,
Salas, et al. 1998; Burger, et al. 2000). Thus, the Colca appears to have constituted a
boundary zone between the Wari and Tiwanaku spheres of influence; the former perhaps
more political or ideological in nature, and the latter more economic in nature or based on
more ancient trade networks (Burger, Asaro, Salas, et al. 1998).
23
If this is the case, I suggest that local Middle Horizon settlement patterning could
have changed in either of two general ways.
First, if Wari political influence was
penetrating and included a direct imperial occupation, there should be evidence for
significant expansion and change of the local settlement pattern from the previous
Formative occupation, including the development of a site size hierarchy, and/or the
establishment of a state administrative center, as indicated by diagnostic imperial
architectural traits and high quantities of elaborate imperial ceramics. By contrast, if Wari
political influence was more indirect or ephemeral, there should be more continuity in the
local settlement pattern, and less evidence for marked inter-site hierarchical organization,
and only indirect influence in local ceramic styles.
The Late Intermediate Period: Community Organization and Agro-Pastoral Production

Previous researchers have hypothesized that the stimulus for the initial construction
of irrigation canals and massive complexes of bench terraces in the Colca valley came from
a climatological shift from wetter (AD 760-1040) to drier (AD 1250-1310) than average
conditions during the Late Intermediate Period (Brooks 1998; Denevan 2001:185-205;
Treacy 1989:133-138). Leaving aside the technical difficulties of correlating these periods
with terrace construction phases, even if perfectly aligned, such correlations would not
signify a causal relationship, but instead would provide information regarding the limiting
parameters within which any agricultural production system must operate. The theoretical
framework developed above suggests that models of linear causality from ecology to
24
economy (and from economy to polity) do not account for the recursive effects between the
two or their relationship to political organization.
This project therefore recasts these factors in a broader frame by analyzing how they
were negotiated within a changing social, political and economic context that made
agricultural intensification advantageous, or that otherwise allowed supra-household leaders
to mobilize the labor needed for the construction, coordination, and maintenance of the
large-scale infrastructure of irrigated, terraced agriculture in the valley. While the local
social, political, and economic context during the LIP was poorly understood prior to this
project, I expected that the growth of the Collagua polity followed the post-Middle Horizon
trajectory of regional development common throughout the central and south central
Andean highlands. Thus, I expect continued growth in the overall scale of local settlement
(i.e., demographic expansion) from the Middle Horizon to the LIP, as well as evidence for
increased competition through time.
In short, I hypothesize that Collagua political
organization was heterarchical in nature during the LIP; that is, composed of internallydifferentiated, supra-settlement communities whose ranks relative to one another were fluid
and variable.
Inter-community relations probably shifted between coordination and
conflict.
I evaluate these hypotheses using several lines of evidence. First, I expect the
development of internally-differentiated communities to correlate with the development of
a marked site size hierarchy. However, in accord with a non-centralized, heterarchical form
of political organization, I expect that no single settlement would dominate the settlement
pattern in terms of size or centrality. Rank size analysis, in which each site is compared to
25
the others in terms of its absolute and rank-order size (from the largest to smallest),
provides a means for characterizing the site hierarchy (Falconer and Savage 1995; Haggett
1965:100-107; Johnson 1987; Paynter 1983). I suggest that a heterarchically-organized
polity will be reflected by a convex rank-size distribution, in which the largest sites do
not greatly differ in size. I discuss the implications of my rank size analysis in further detail
in Chapter 6. Also, as political relations oscillated between coordination and conflict, I
expect that most, if not all large settlements would be situated either in defensible locations
or near fortifications, similar to the nucleated hilltop settlement pattern common to many
LIP contexts (e.g., D'Altroy 1987; Hyslop 1976; Stanish 1997a). In Chapter 6, I also
analyze hydrological relationships between the long feeder canals that carry meltwater from
the surrounding peaks to evaluate how water apportionment was coordinated at a suprasettlement, watershed-scale.
Differences in the size and elaboration of domestic
architecture provide a good archaeological index for inequality between households, and
domestic architecture at most LIP and Late Horizon settlements in the Colca valley remains
very well-preserved.
Previous researchers have noted the distinctive features that
characterize the Collagua architectural style (Brooks 1998; Guerra Santander and Aquize
Cceres 1996; Neira Avendao 1961), and my analysis in Chapter 6 provides a view of the
variability of size and elaboration within that style, which I interpret as reflecting
differences in the wealth and status of households within and between sites.
26
The Late Horizon: From Heterarchy to Hierarchy under Inka Rule
Prior to this project, the nature of Inka influence or presence in the Colca valley was
debated among archaeologists.
Portable artifact media suggested considerable Inka
influence: local Late Horizon Collagua ceramics show clear Inkaic stylistic and formal
attributes (de la Vera Cruz Chvez 1988, 1989; Malpass and de la Vera Cruz Chvez 1986,
1990; Shea 1997), but architectural and settlement indices were more ambiguous. In
general, evidence for direct Inka administration was stronger in the lower reaches of the
valley near the village of Cabanaconde, where structures of Inka style cutstone masonry
were reported, and a probable small administrative center (the site of Kallimarka) had been
identified (de la Vera Cruz Chvez 1987, 1988). In the central portion of the valley, where
the survey area of this project is located, no centrally-planned administrative center with
Inka public architecture had been documented in previous reconnaissance, leading some to
suggest that the Inka never established a direct imperial occupation there (Brooks 1998).
Others, noting the overall paucity of pertinent data, made no definitive statements but
suggested that Inka rule may have been administered through local settlements and elites
(Malpass 1986, 1987; Neira Avendao 1961, 1990; Shea 1986b, 1987, 1997).
However, ethnohistorical studies suggested that Inka administration had a much
more profound impact on the political and economic organization of the Collagua province
than these preliminary archaeological indices seemed to suggest. Although the Collagua
are not mentioned in most of the standard chronicles, a description of the province by the
Spanish magistrate (corregidor) Juan de Ulloa Mogolln from 1586 describes how the
ayllus of the Collagua were arranged according to a formal administrative hierarchy based
27
on dualistic, ternary, and decimal principles (Ulloa Mogolln 1965 [1586]). Forty years
ago, Zuidema (1964:115-118) noted the close parallels between the names of Collagua
ayllus as recorded in this and other 16th and 17th century sources and the names of the
ceque lines that mapped the ritual activities of the royal ayllus (panacas) of Cuzco over
space and time (cf. Bauer 1998:35-37). Based on Ulloas description and the names of
ayllus recorded in the colonial visitas of the Collaguas, subsequent research presented the
Collagua province as an example par excellence of Inka social engineering, whereby local
ayllus were reordered in a nested hierarchy of minimal-, minor-, major-, and maximalayllus according to imperial categories of prestige and administrative scale (Benavides
1989; Cock Carrasco 1976-77; Prssinen 1992:362-366; Pease G. Y. 1977; Rostworowski
de Diez Canseco 1983:121-123).
In this project I evaluate the hypothesis that the Inkas established a centralized,
hierarchical structure of imperial administration out of the previously heterarchicallyorganized local ayllus through integrated archaeological and ethnohistorical analysis.
Archaeologically, the correlates of centralized, coordinated imperial rule should be visible
not only in portable artifacts, but also in the form of significant changes in settlement
patterning and organization, either through the establishment of a large administrative
center with elite and public architecture, or the expansion and reorganization of an extant
settlement. Thus, I would expect a corresponding expansion in the site size hierarchy,
composed of four tiers of site sizes, ranging from small hamlets, to larger villages, towns,
and centers. In terms of rank-size distribution, an integrated administrative hierarchy is
often associated with a log-normal distribution, in which the largest site is approximately
28
twice as large as the second largest site, three times as large as the third largest site, and so
on (Falconer and Savage 1995; Haggett 1965:100-107; Johnson 1987; Paynter 1983). By
contrast, if the Collagua maintained relative autonomy during the Late Horizon, there
should be overall continuity from the LIP settlement pattern and no other evidence for
major state investments in the form of an administrative center or imperial architecture.
My interpretation of the settlement pattern data from my survey, however, relies not
only on the analogical reasoning of such archaeological correlates, but is also informed by
historically-and spatially-contiguous socio-structural homologies that I reconstruct through
analysis of early colonial visitas. These visitas, from the years 1591, 1604, and 1615-1617,
record some of the most detailed demographic and land tenure data of any colonial censuses
in the Andes, affording an opportunity to reconstruct patterns of land use and exchange at
regional and local scales.
Previous analyses of the Collagua visitas have illustrated how the ayllus of the
Collaguas, and the authority of their leaders, were distributed over villages in far-flung
production zones, forming a classic example of a vertical archipelago of settlements
linked to the core population in the central Colca valley by ayllu affiliation, ranging from
pastoralist enclaves in the high altitude grasslands of the upper valley, to maize-production
enclaves in the neighboring, low-lying valleys to the south (Benavides 1989; Galdos
Rodrguez 1984; Pease G. Y. 1977). My regional-scale analysis of visita landholding and
livestock declarations in Chapter 8 reconstructs how supra-local networks of ayllu
affiliation and authority synchronized varied complementarity practices between
agriculturalists and pastoralists under Inka and early colonial rule. As I illustrate, these
29
distinct complementarity practices were the result not only of adaptive processes to
ecological imperatives, but also of status and community affiliation.
At the local scale, my analysis of landholding declarations in the visitas of 1591,
1604, and 1615-1617 reveals how households access to agricultural fieldholdings was
mediated by their ayllu affiliation.
In the censuses, each of the fields declared by
households were located using toponyms. As I discuss in Chapter 4, toponyms here and
elsewhere in the Andes tend to be extremely historically durable; thus, I have reconstructed
the locations of visita landholding declarations by reference to their modern toponymic
counterparts. Since the households of the villages of Coporaque and Yanque registered in
the visitas were the direct descendants of the residents of the Late Horizon settlements in
the project survey area, my reconstruction of early colonial ayllu land tenure provides a
means for reconstructing Inka-era ayllu residence patterns. While archaeologists have
developed sophisticated means for modeling prehistoric land use patterning through various
forms of site catchment analysis, my use of the visita data proceeds in essentially the
reverse fashion. That is, I use the detailed site catchment data recorded in the visitas as a
basis of retrodicting prehispanic residence patterns. As I demonstrate in Chapter 9, the
spatial patterning of ayllu landholdings in Coporaque indicates that they continued to
mediate household access to fields at the time of the visitas, despite resettlement into
nucleated villages a generation earlier. By extension, I suggest that, through patterns of
inheritance, these patterns represent vestiges of prehispanic ayllu land tenure patterns.
Assuming that there was a negative correlation between distance from a settlement and the
amount of land cultivated by its inhabitants, the distribution of fieldholdings in relation to
30
Late Horizon settlements therefore can be used as a proxy for reconstructing where the
ancestors of particular ayllus lived during terminal prehispanic times. Thus, for example,
clustering of ayllu landholdings around a Late Horizon settlement would signal that the
majority of the terminal prehispanic ancestors of that ayllu resided at that settlement. In
this way, I reconstruct how the ranked hierarchy of ayllus under Inkaic rule were distributed
over settlements, providing a more culturally-informed interpretation of the political
relationships between settlements.
Restructuration and Reduccin: Colonial Period Adjustments

Finally, this project seeks to clarify questions regarding colonial-era changes in
local community organization and land-use patterning using an observational framework
congruent with the methodology employed for the prehispanic period. How did hybrid
Collagua/Inka structures of community articulate with early Spanish colonial institutions?
What were the effects of changing Spanish colonial policies on local- and regional-scale
Collagua complementarity practices?
Locally, research by Denevan and colleagues (Brooks 1998; Denevan 1987, 1988b;
2001:202; Treacy and Denevan 1986) documented that approximately half (between about
40-60%) of the agricultural fields in the valley are presently abandoned, and that most of
the irrigated bench terraces no longer under cultivation were abandoned during the colonial
period (Denevan 1987, 2001).
These researchers have suggested that colonial
depopulation, combined with forced resettlement of the population into nucleated reduccin
villages during the early 1570s were probably important factors in terrace abandonment
31
(Denevan 1987; 2001:205-206). According to this scenario, households did not attempt to
increase their landholdings as the population declined, but instead stopped maintaining and
cultivating large tracts of terraces, especially those that were located far away from their
new villages.
While such a cost-benefit calculus probably factored in decisions regarding the
maintenance and/or abandonment of canal and terrace systems, I argue that the changing
structures of community organizationboth built and imaginedduring early colonial
times mediated that calculus. Here, the agency of landscape in the wake of rapid political
change is apparent. That is, the radically altered relationship between settlement and agropastoral infrastructure in the valley after reduccin produced a dialectic in which the landuse practices of local communities had to reconfigure and/or adapt to the ancient structures
of the surrounding built landscape. One outcome of this dialectic was abandonment. If the
role of demographic decline and resettlement was decisive in the abandonment of irrigated
terraces, then it can be expected that there would be more terraces abandoned in areas most
remote from a given reduccin than in nearby areas. I examine this hypothesis both
through analysis of the spatial distribution of abandoned terrace complexes, and through my
reconstruction of land tenure patterns from the visitas.
At the regional scale, my analysis of colonial census data reconstructs how ayllus
were remapped onto the radically altered settlement pattern of nucleated villages created by
the forced reduccin resettlement of Collagua populations. My analysis reveals how suprasettlement networks of ayllu affiliation continued to mediate inter-zonal patterns of
exchange and tribute collection after resettlement. By tracking the distribution of Collagua
32
ayllus, and the authority of their leaders, over reducciones in distinct agricultural and
herding production zones, my analysis documents how Collagua elites continued to assert
their specific Inka imperial status as redistributive eco-brokers, even as their
administrative function became almost exclusively extractive under Spanish rule.
Dissertation Outline
This dissertation is divided into three parts: I. Background, II. Archaeological
Findings, and III. Ethnohistorical Reconstruction and Archaeo-Historical Synthesis. In the
remainder of Part I, Chapter 2 first provides an overview of the physical geography and
anthropogenic characteristics of the Colca valley landscape. Chapter 3 reviews previous
archaeological and ethnohistorical research on the Collagua, and situates their origins,
development, and incorporation into the Inka and Spanish empires in a regional historical
and historiographical context. In Chapter 4, I present the spatially-integrated archaeohistorical methodological architecture of this study, and detail the specific research
operations I carried out to collect and structure the data for the project. In Chapter 5, the
first chapter of archaeological findings, I address questions regarding the timing and social
organization of the earliest sedentary agriculturalists of the valley, and present the
settlement pattern data documenting a subsequent expansion and shift in settlement during
the Middle Horizon. Chapter 6 presents the settlement pattern evidence for the origins and
autonomous political and productive organization of the Collagua ethnic polity during the
Late Intermediate Period, as well as the archaeological evidence for the consolidation and
coordination of Inka imperial administration of the central Colca valley during the Late
33
Horizon. In Chapter 7, I provide an archaeological view of the transition from Inkaic to
Spanish imperial occupation, both before and after the forced resettlement of the population
into nucleated villages in the 1570s. In Part III, the first chapter (Chapter 8) reconstructs
the ideal Inkaic administrative structure of the Inkaic Collagua province, and, through an
analysis of landholding and livestock declarations in the visitas, illustrates how that ideal
structure articulated with local communities and mediated regional-scale complementarity
practices. In Chapter 9, I reconstruct early colonial land tenure patterns of local ayllus
within the archaeological survey area using the visita data, providing a means for also
reconstructing their Late Horizon residence patterns and providing a more detailed view of
how community relations were mapped onto the local settlement system under Inka rule.
The study closes with a brief summary of findings in Chapter 10.
34
Chapter 2
Physiographic, Climatic, and Anthropogenic Characteristics
of the Colca Valley Landscape
Location, Geology, and Geomorphology

The Colca River Valley is a semi-arid highland valley located in the western
cordillera of southern Peru, in the Province of Caylloma, Department of Arequipa (Figure
2.1). It is the largest river valley in southern Peru, both in terms of river discharge and
arable land area ([ONERN] 1973). The regional geographical context is dominated by
great expanses of high, cold grasslands (puna), punctuated by a range of volcanic peaks that
reach above 5000 and 6000 meters above sea level. The valley was formed through uplift
and fluvial incision into this plateau during the Pleistocene (Denevan, et al. 1986). After
arcing northwest from its source near Laguna Lagunillas (close to Puno), the river turns
south before trending westerly towards the Pacific. Below the confluence of the tributaries
near the village of Viraco, the Colca River changes names to the Majes, and eventually to
the Caman River just before discharging into the Pacific on the arid coast.
The Colca valley proper can be divided into lower, central and upper sections
according to general physiographic changes over the course of the river. The lower valley,
encompassing the area around the village of Cabanaconde, is set apart from the central and
upper valley areas by a deep section of canyon. The lower valley is generally steeper,
deeper, and warmer than the middle and upper sections of the valley, and arable land there
is situated in the lower kichwa and upper yungas ecological zones (between about 2800 and
3400 masl), where frosts are less common, and maize agriculture predominates. Villages
35
here also cultivate fruit orchards in the lowest reaches of the river gorge (Gelles 2000). The
central valley, which includes the survey area for this project, extends upstream from the
Colca canyon to just upstream of the village of Tuti. In general terms, the central valley
encompasses the agricultural core of the valley, within the kichwa (locally, ca. 3300-3600
masl) and suni zones (ca. 3600-4000 masl), and is surrounded by the grasslands of the puna
(ca. 4000-5000 masl) along the valley rim. The upper valley begins upstream of the village
of Tuti, around 3800 masl. The valley here is more shallow and open than in the central
and lower valley, and is dominated ecologically by broad expanses of puna grassland
steppe.
Over the course of its entrenchment, the river has exposed a cross section of
geologic strata over 2000 vertical meters in height. Within the inner river gorge, the oldest
stratum in the survey area, a thick layer of Jurassic Period (205-135 m.y.a.) sedimentary and
metasedimentary quartzite, sandstone, and shale (known as the Yura Formation), is
exposed. Rhyolitic and andesitic tuffs and flows overlie the Yura Formation, dating to the
late Cretaceous to early Tertiary (135~54.8 m.y.a.). Subsequent vulcanism during the
Tertiary (65 to 1.8 mya) and early Quaternary (1.8 mya to 10,000 yrs) is responsible for
additional andesitic strata, which are overlain by Pleistocene (1.8 mya to 10,000 yrs)
rhyolitic tuff. Stratovolcanoes line both sides of the valley, and Quaternary lava flows in
the survey area, such as those to the west of Pampa Finaya, intrude upon the second alluvial
terrace created from the downcutting river (Figure 2.2). Potassium argon dates conducted
by Sandor indicate that these flows formed during the Middle to Late Pleistocene (64,000
14,000 and 172,000 14,000 b.p.) (Sandor 1992:233-234).
36
Geologically-recent vulcanism also created major sources of obsidian around two
rhyolitic domes (Cerro Ancachita and Cerro Hornillo) 4-5 km east of Coporaque (Burger,
Asaro, Salas, et al. 1998; Burger, et al. 2000, see Figure 2.3). Burger and colleagues report
the primary source to be part of the Barroso Group, specifically at its point of contact with
the underlying Tacaza Formation , where very large (up to 30 cm long) in situ obsidian
nodules can be observed in exposed stratigraphy in the quebrada los Molinos {Burger, 1998
#371:205, see Figure 2.3). Brooks and colleagues report a large quarry to the immediate
east of Cerro Cotallalli (or Cotallaulli), in similar stratigraphic context. The broad age
estimates of the Barroso group are between 6 and 1 m.y.a. (Palacios, et al. 1993). Brooks
(1998:57-58) reports that obsidian from the Cotallaulli source has been dates to 3.52 m.y.a.
150,000 (i.e. during the Pliocene). The sources are compositionally indistiguishable.
While these two investigations have located obsidian sources with the same chemical
signatures independently, the extent of the Tacaza/Barroso interface is much larger than
these two points, and so it can be expected that other obsidian outcrops will emerge through
future research. Archaeologist Nicholas Tripcevich of the University of California, Santa
Barbara is currently conducting intensive survey of the obsidian source area as part of his
dissertation research (N. Tripcevich, pers. comm 2002-2003). During the survey for this
project, another, we identified a much smaller obsidian source on Cerro Caracachi. This
obsidian is of lower quality than that of the Chivay/Cotallalli source, and appears to be from
a distinct stratigraphic context. Trace element analysis is currently being conducted on
samples from this source.
37
38
Figure 2.3: Volcanic features and obsidian sources
39
40
Historically-recent vulcanism is also significant in the valley and region.
The
massive eruptions of the stratovolcano Huaynaputina, on February 19 and 20, 1600,

darkened the skies of the entire south-central Andean region, caused catastrophic damage to
settlements in the immediate vicinity of the volcano along the upper reaches of the Tambo
river, and blanketed the region in ash (Dvila and Thouret 1999). Ashfall caused major
damage to fields, canals, and buildings in the city of Arequipa, and paralyzed agricultural
production as far north as the Vitor valley, where the large vineyards established by the
Spaniards produced no wine for six years afterward (Mlaga Nez-Zeballos 2002). The
effects of the distant eruption were less dramatic in the Colca valley, but clearly local
farming was affectedthe preamble to the 1604 visita of Laricollaguas Urinsaya explains
how that census was undertaken in response to a petition brought before the viceroy by the
protector de los naturales the previous year. That petition described severe food shortages
and ashes covering the valley landscape (Cook 2003:xxiv). Despite their hardships, the
Collaguas were summoned by the municipality of Arequipa to send a large contingent of
laborers to repair damage in the crippled city (Mlaga Nez-Zeballos 2002). The stratum
of ash from the eruption forms an archaeologically useful stratigraphic horizon throughout
the region, and ash accumulation in the Colca valley has been estimated to have reached
about two centimeters (Dvila and Thouret 1999), but the deposit has not been reported in
stratigraphic excavations to date. Since 1986, the stratovolcano Sabancaya, situated on a
saddle between Ampato and Hualca Hualca overlooking the valley, has been active. Major
eruptions occurred between 1990 and 2000, although activity has since subsided.
41
Geothermal activity is also prevalent throughout the valleyseveral hotsprings dot the
valley and have become a major tourist attraction, especially near Chivay at La Calera.
Figure 2.4: Geomorphic surfaces in the central Colca valley
42
The geomorphology of the valley is characterized by a series of seven mesa-like
alluvial terraces of volcanic parent material, divided by higher-angle escarpments from the
puna lands above 4000 masl to the inner river gorge around 3300 masl. These alluvial
terrace surfaces, denominated Qal 1-7 (Quaternary alluvium 1-7) increase in age with
altitude, with the most recent surface (Qal 1) at the entrenched river channel and
surrounding floodplain (Sandor 1987a, 1992). Above this lowest surface, three previous
alluvial terraces (Qal 2-4) consist of gravels, fine alluvium and volcanic ash (Sandor 1987a,
1992). These are obscured in places by intrusive alluvium from tributaries. Next, Qal 5 is
constituted by volcanic flows (Sandor 1987a, 1992). The Qal 1-5 terraces and intervening
slopes encompass most of the agricultural zone in the valley (Denevan, et al. 1986). The
remaining two Quaternary Alluvium surfaces, Qal 6-7, consist of high alluvial terraces and
stranded interfluves (Sandor 1987a, 1992). QT 8 (Quaternary to Tertiary), the highest and
oldest of these surfaces, may predate the formation of the valley itself (i.e. and hence would
have been the first subjected to river downcutting) (Sandor 1987a, 1992). Above QT 8 lie
the glaciated volcanic peaks of the Misme/Sepregina massif to the north, and the
Ampato/Sabancaya/Hualca Hualca massif to the south.
Soils
Soils in the Colca valley are among the most well-documented in the Andes as a
result of research by Jon Sandor and Neal Eash as part of the Rio Colca Abandoned Terrace
Project, as well as Guillets multi-disciplinary project on modern soil ethnoclassification
(Denevan, et al. 1986; Dick, et al. 1994; Eash and Sandor 1995; Guillet, et al. 1995; Sandor
43
1986, 1987a, b, 1988, 1992; Sandor and Eash 1991, 1995; Sandor and Furbee 1996). In
general, soils in the Colca valley can be characterized as very fertile Mollisols of volcanic
alluvium and colluvium parent material with high humic content (Eash and Sandor 1995;
Sandor 1987a, 1992; Staff 1975; For soil taxonomy, see Staff 1990). Characteristic of
continental grassland settings, Mollisols exhibit a thick, dark, humus-rich surface horizon,
and are among the most productive agricultural soils in the world. The high organic
content, good tilth properties, and high water capacity of Colca valley soils compare
favorably in overall fertility to the Mollisols of the midwestern U.S. (Sandor 1986:249).
As the above described alluvial terraces increase in age with altitude, their surfaces
form a soil chronosequence (Eash and Sandor 1995; Sandor 1987a, 1992). Natural soils of
the lower, younger surfaces (Qal 1 and 2) consist primarily of thin Mollisols and lack welldeveloped argillic horizons. Successively older alluvial terraces (Qal 3-5), exhibit thicker,
darker Mollisol epipedons associated with thick argillic horizon development (Eash and
Sandor 1995).1 Intervening slopes also have thick argillic horizons and duripan formation
(Sandor 1987a). Some of the highest surfaces (Qal 6-7) have thinner A horizons, probably
due to erosion (Treacy 1989:68). Indicative of their age, calcium carbonate accumulation
and duripan formation generally increases on higher surfaces as well (Sandor 1987a:181).2
Topsoils of bench terrace complexes that line the valley slopes exhibit various
anthropogenic characteristics. Trench excavations by Sandor, Malpass, and colleagues
have revealed that terrace wall foundations were excavated into the argillic horizons and
1
2
Argillic horizons are distinguished by significant clay accumulations.
Calcium carbonate buildup probably reflects the age of the underlying geomorphic surface or parent
material. However, this is not a product of the parent rock material, but rather, Sandor (1987:181) believes
that carbonates were probably deposited by wind.
44
duripans of the valley side slopes, providing a sound structural foundation (Sandor 1986,
1987a). Sandor (1992:235) attributes the remarkable stability of the rock-walled terracing
of the valley in part to these underlying epipedons. Bench terrace topsoils tend to be loamy,
and exhibit plaggen3 and anthropic (high phosphorous content) epipedons (Sandor 1992).
Bench terrace topsoils are also unusually thick (between 0.3-1.3 m). Good tilth is evident
in strong fine granular to subangular blocky structure, and tubular macropores and worm
casts signal abundant earthworm activity. In sum, the thick, fertile topsoils of Colca valley
bench terraces represent long-term accretions of fertilization and indigenous soil
management (Sandor 1992). The dating of terracing and these anthropogenic soils and are
discussed in further detail later in this chapter.
Climate Characteristics and Trends

The climate of the Colca valley is cool, semiarid, and unpredictable. Climatic
patterning at the latitude of the Colca is most proximately determined by the orographic
effects of the cordillera, but is also the product of continental- and planetary-scale
circulations (Johnson 1976). As opposed to the central and northern Andes (north of about
11 latitude), where moisture transport is influenced primarily by oscillations in the
Intertropical Convergence Zone (ITCZ) (a planetary-scale circulation that determines the
location and orientation of the Trades), precipitation in the south central Andes is produced
primarily by convection of warm, humid air from the eastern slopes of the Amazon Basin,
in combination with local low pressure systems that develop during the austral summer.
3
Plaggen: An artificially made surface layer produced by long-continued manuring (Soil Survey Staff 1975,
1990)
45
From the latitude of the Colca valley and further south (that is, below about 15 south
latitude), precipitation decreases markedly and is more unpredictable than in the central and
northern Andean regions (Johnson 1976).
Precipitation is markedly seasonal and
unpredictable because of high inter-annual variability. Due to the high altitude and tropical
latitude of the valley, mean annual temperatures are low (around 10 C for the central
portion of the valleysee below), and diurnal temperature fluctuations are greater than
seasonal ones.
Figure 2.5 illustrates how differences between the minimum and maximum
temperatures within each month are greater than seasonal changes, reflecting strong diurnal
fluctuations in temperature.4
These data were recorded by the Oficina Nacional de
Evaluacin de Recursos Naturales between the 1950s and 1970s ([ONERN] 1973). Warm
daytime temperatures result from overall sunny conditions throughout the year. However,
at night, frost is a common hazard to crop viability, especially during the months of May
through October, although frosts can occur during any month of the year. Frost is most
problematic on the valley bottom, due to the settling of cool air over the terraced valley
sides. Hail is also a significant risk factor for crops, and occurs most commonly between
December and March, the months of greatest precipitation (see below).
These data represent average values from the following villages and years: Cabanaconde: 1951-1957, and
1959-1971; Madrigal: 1966-1967 and 1971; Yanque:1951-1957, and 1959-1971; Chivay: 1973-1982;
Sibayo:1952-1953, and 1955-1966; Tisco:1965-1971. All are derived from ONERN 1973, vol. 2, Appendix
1, except Chivay, from ONERN 1973 (years 1967-1970) and, for years 1973-82, SENAMHI weather station
data as reported in Denevan et al 1986:51 (Table 1). Skipped years indicate incomplete data.
46
25
Temperature (Celsius)
20
15
10
Chivay
Sibayo
5
0
-5
be
r
be
em
ec
ov
ct
o
em
be
r
be
O
gu
Se
pt
em
st
ly
Ju
Au
ne
Ju
ri l
ay
M
Ap
ch
ar
ry
M
ua
br
Fe
Ja
nu
ar
-10
Month
Figure 2.5: Average monthly temperatures (Average Low, Total Average, and Average High) for the
villages of Chivay (1973-1982) and Sibayo (1952-1953/1955-1956). See Note 4 for details on years that
data was collected in each village.
Elevation greatly influences temperature at these altitudes, and marked temperature

gradients are present even within the agricultural core of the survey area. Denevan and
colleagues report an 8 C difference in minimum temperatures along a 500 meter transect
from the valley floor (3272 masl) over complexes of irrigated bench terraces along the
lower valley slopes to a high group of abandoned fields at Chilacota (3752 masl) (0 and -8
C, respectively).5 This transect lies 1.5 km southwest of the village of Coporaque, within
the survey area for this project.
Temperatures at higher altitudes in the valley are
correspondingly cooler. Figure 2.5 above also includes data from the village of Sibayo,
Temperatures recorded during the cool, dry months of July and August (July 30-August 3), in 1984.
Temperatures reported here are converted from F reported by Denevan et al.
47
situated as at 3810 masl in the upper valley. In 1970, the average minimum temperature in
Sibayo was 3 degrees cooler than in Chivay (-1.7 vs. 1.3 C).
700
600
Precipitation (mm)
500
400
300
200
100
0
Cabanaconde
Madrigal
Yanque
Chivay
Sibayo
Tisco
Village
Figure 2.6: Average annual precipitation, 1955-1982. See Note 8 for details on years data collected
from each village.
While frost and hail pose significant risks to farmers, water scarcity is the primary
environmental constraint to agricultural production in the valley (Gelles 2000; Guillet 1992;
Treacy 1989; Winterhalder 1993). High evaporation rates result in absolute water deficits
for as many as eight months of the year (Treacy 1989:54, 56). Precipitation data collected
by ONERN between the 1950s and 1970s at several villages along the length of the valley
provide baseline figures for the recent past (Figure 2.6).6 The annual precipitation of
around 400 mm in the central and lower valley villages is below the minimal requirements
6
Complete annual precipitation records from ONERN were available for the following: Cabanaconde: 19511957 and 1959-1971; Madrigal: 1966-1967 and 1971; Yanque: 1951-1957, 1959-1966, and 1968-1971;
Chivay: 1973-1982; Sibayo: 1952-1953 and 1955-1966; Tisco: 1965-1971.
48
for rainfed maize or potato production. Potatoes require at least 500 mm water annually,
and maize needs 500 to 1000 mm (Knapp 1991). However, when planted with wide
spacing (70 to 120 cm apart), quinoa, a hardy Andean chenopod, can be cultivated
successfully with as little as 300 mm of rain annually (Tapia et al. 1979:92, as cited in
Treacy 1989:132).
While precipitation generally increases with altitude in the valley, the benefit of
greater precipitation is offset by higher frost and hail risks at the high altitudes of Sibayo
and Tisco, which are primarily pastoralist settlements. Also, the timing of precipitation is
equally critical to the gross quantity of annual precipitation.
Throughout the valley,
precipitation is highly seasonal, with an average of 65% of the annual total occurring
between the austral summer months of January and March (Figure 2.7).
While the
seasonality of precipitation is predictable grosso modo, inter-annual variation is great, such

that standard deviations from average monthly totals between years are very large (Fig.
2.8). This variability makes precipitation very unpredictable on an inter-annual basis. That
is, some years bring adequate rainfall even for dry farming, but others are exceedingly dry
and make agriculture without irrigation impossible.
49
160
140
Precipitation (mm)
120
100
Cabanaconde
Madrigal
80
Yanque
Sibayo
Tisco
60
40
20
m
D
ec
e
ov
e
N
be
r
be
r
er
r
be
ob
ct
O
Se
pt
e
gu
s
ly
Ju
Au
ne
Ju
ay
M
ril
Ap
ch
ar
ry
M
ua
br
Fe
Ja
nu
ar
Month
Figure 2.7: Average monthly precipitation, 1951-1982. See Note 8 for details on years data collected
from each village.
160
140
Precipitation (mm)
120
100
80
60
40
20
mb
er
er
ce
De
mb
er
ve
No
er
Oc
to b
mb
Se
p te
st
Au
gu
ly
Ju
ne
Ju
y
Ma
ri l
Ap
rc h
Ma
ar y
br u
nu
Ja
Fe
ar y
M onth
Figure 2.8: Monthly precipitation, Chivay, 1973-1972, showing standard deviations.
50
Compounding inter-annual unpredictability are the effects of El Nio/Southern
Oscillation (ENSO) events. These effects vary depending on the intensity of ENSO events,
but moderate to severe ENSO events, which bring torrential rains and flooding to the
coastal regions of southern Ecuador and northern Peru, most often affect the southern
Peruvian highlands in the opposite manner, causing drought. This appears related to the
southward displacement of the Intertropical Convergence Zone, disrupting the convection
of moisture normally transported by the westerly Trades (Caviedes 1975, 1984). ENSOrelated droughts are often doubly damaging because they can occur even after heavy
September and October rains, which modern farmers use (along with early plant blooms
and snow mantle depths on the surrounding peaks) to forecast the nature of the coming
rainy season (Treacy 1989). Thus belying local indices of abundant rainfall, the sudden
onset of an ENSO-related drought can lead to especially heavy losses in seed, crops, and
labor investment. Such was the case during the severe ENSO events of 1971-72, 1982-83,
and 1997-98 (Brooks 1998:212-213; Caviedes 1975, 1984; Treacy 1989:57). It is possible,
however, that modern Colca farmers have not retained other proven ENSO forecast indices
documented elsewhere in the central and south central Andes, such as observance of the
clarity of the Pleiades during the winter solstice (Orlove, et al. 2000). Prehispanic farmers
may have been better able to forecast the weather than their modern counterparts.
Nonetheless, the scarcity, seasonality, and unpredictability of precipitation within the
current climatic regime make cultivation virtually impossible without some kind of water
augmentation regime.
51
However, because temperature and precipitation parameters in the Colca valley are
both close to the absolute requirements for cultigens, small changes in either or both
variables would significantly alter the matrix of constraints and opportunities for
agricultural production in the past. The present climatic conditions should not be assumed
to be an average, or even representative, climatic state. As is becoming increasingly clear
through paleoclimatic research from neighboring areas, the chaotic (but not random) interannual variability in temperature and precipitation observed in the recent past occurs within
longer-term climatic oscillations. During wetter periods in the prehispanic past, rainfed
agriculture was possible and widely practiced in the Colca valley (see below). Conversely,
prolonged dry periods may have created an impetus for the elaboration of irrigation systems
(Brooks 1998; Treacy 1989:106, 136, see below).
Because of their proximity, paleoclimatic data from glacial cores from the ice cap of
Mount Quelccaya, located to the southeast of Cuzco, and from sediment cores from
Titicaca and nearby lakes are of most relevance for understanding long-term Holocene
climate change in the Colca valley.
The Quelccaya data, consisting primarily of
measurements of annual layer thickness, dust concentrations, and isotopic analysis, extends
from the mid 6th to the late 20th centuries AD (Thompson 1992; Thompson 1993;
Thompson 1995; Thompson, et al. 1988; Thompson and Mosley-Thompson 1989;
Thompson, et al. 1985; Thompson, et al. 1986; Thompson, et al. 1994). Data from the
Titicaca lake cores, which record palynological, lithological, and lake level data, covers
between 1550 BC and the late 20th century AD (Abbott, Binford, et al. 1997; Abbott,
52
Seltzer, et al. 1997).7 Within the 1500 year overlap between the two (circa AD 540-1985),
there are points of significant incongruence. The Quelccaya results indicate several short
dry and wet periods not recorded in the lake cores (see Table 2.1). Four very dry periods
were recorded in the Quelccaya cores: AD 540 to 610, AD 650 to 730, AD 1040 to 1490
and AD 1720 to 1860 (Thompson 1992:541, Fig. 27.18). The two relatively brief dry
periods recorded in the glacial cores during AD 540-610, and 650-730 do not coincide with
lowered lake level indices in the lake cores. Rather, the lake core data indicate water levels
within a few meters of current heights throughout the 3rd through early 11th centuries AD8
Only the most severe and longest dry periodfrom AD 1040 to 1490, corresponds with
low water levels on Lake Titicaca (Abbott, Binford, et al. 1997). The sediment core data
from the lake indicate a dramatic (about 20 m below current levels) drop in water levels
during AD 1050 to 1450 (Abbott, Binford, et al. 1997). However, the longest and most
severe dry period recorded in the Quelccaya precipitation record, during the subsequent
period between AD 1040 and 1490, does overlap with the most dramatic (about 20 meters
below current levels) and longest (about 400 years) drop in lake level recorded in the
sediment cores between AD 1050 and 1450. This dry period is followed by a wet period
For the glacial cores, dry and wet periods represent years (or decadal averages) in which the thickness of
accumulated ice deviates significantly above (wet years) or below (dry years) from the average accumulated
thickness (L.G. Thompson et al 1985). The periods referred to here (L.G. Thompson 1994) are revised from
the original periods reported in 1985, and are based on decadal averages. Dry years are inferred from the lake
cores through disjunctures in core strata which represent erosional surfacesi.e. surfaces which were exposed
to the air (indicating drops in lake level). While wet years are more difficult to infer from lake cores, and
not explicitly discussed by Abbott and colleagues, the non-dry years represent years in which water levels
were at or near historical averages (see Abbott, et al 1997).
In addition to the two dry periods in the middle of this period, the glacial cores indicate another period of
greater than normal precipitation between AD 760-1040.
53
reflected in both records (AD 1500-1720). Finally, the Quelccaya record indicates a dry
period between 1720 and 1860 that does not correspond with a drop in lake level.
Table 2.1: Dry and wet periods derived from glacial and lake sediment cores
Quelccaya Ice Cores
Lake Cores
Wet Periods
Dry Periods
Dry Periods
AD 1870-1984
AD 1720-1860
AD 1050-1450
AD 1500-1720
AD 1040-1490
50 BC-AD 250
AD 760-1040
AD 650-730
450-250 BC
AD 540-610
950-850 BC
1550-1400 BC
These incongruencies can be explained in part by differences in spatial and temporal

resolution of the two data sources: the ice core data register annual to decadal-scale
variability during the period of overlap that cannot be differentiated in the lake cores, which
are based on radiocarbon dating (Abbott, Binford, et al. 1997).9 Also, while widely used by
archaeologists as a kind of climatological master sequence, the Quelccaya glacial cores
should be considered as point-source data from a high peak on the eastern cordillera, rather
than as broadly representative of regional or supra-regional patterns. By contrast, changes
in water level recorded in the lake cores represent a net index for climatic conditions over
the 57,000 km2 of the Titicaca catchment basin (Abbott, Binford, et al. 1997; Abbott,
Seltzer, et al. 1997).
As noted by Abbott and colleagues (1997:177), given these
differences, the snow accumulation on Quelccaya is not a reliable proxy for water levels on
Lake Titicaca. Given the proximity of Lake Titicaca (100 kilometers to the east) and the
similarity of its climate patterning, the lake cores are probably the more accurate proxy of
The Titicaca cores were dated using 60 AMS dates, derived primarily from shell and totora reed seeds.
Abbott and colleagues corrected for the reservoir effect of their samples by subtracting 250 years (an
empirically-derived value) before calibrating their dates (see Abbott et al. 1997:172).
54
the two for modeling Holocene climate change in the Colca valley. The Quelccaya ice cap
is further away and borders the climatically-distinct (generally wetter) eastern escarpment.
Anthropogenic Landscape Features

Beneficial Functions of Terrace and Field Systems
The above review of climate and physical geography present the broad (and evershifting) physical and energetic parameters of the Colca valley as a human habitat.
However, as discussed below, the Colca valley landscape is equally a product of, not just a
setting for, human interaction, and is a prime example of how ecological zones
themselves are anthropogenically-constructed so as to lessen ecological strictures. The
Colca valley is one of the most intensively terraced valleys in the New World (Donkin
1979). Approximately 14,000 ha of agricultural terraces and fields cover both sides of the
valley, from the Qal 1 terrace of the inner river gorge to abandoned terraces between
approximately 3600 and 4000 masl (Denevan 1987; 1988b:Table 1; Denevan and Hartwig
1986).
Because the river course is deeply entrenched, water from the river is exceedingly
difficult to harness for irrigation using nonmechanized methods. Thus, the vast majority of
irrigation water originates from glacial meltwater, springs, and associated drainages
(quebradas) around the high peaks on both sides of the valley. The glaciated peaks of
Misme, Huillcaya, and Quehuisha supply the meltwater for the irrigation systems of the
north side of the river. The peaks of Huarancate, Sabancaya, Ampato, and Hualca Hualca
fulfill the same function on the south side of the river. Canal systems vary in length
55
according to local hydrological and topographic factors, ranging between less than 1 km to
more than 30 km in length. In addition to serving as insurance against drought, irrigation
offsets the effects of marked seasonality of rainfall in the region, enabling farmers to
regulate the timing and quantity of water. Irrigation makes it possible to stretch the
growing season by planting well before the onset of summer rains, as well as watering
crops after seasonal rains stop (Treacy 1989:276-279).
Given the massive anthropogenic alterations to the landscape, the ecological tiers
of the Colca valley are best understood as composed of intensively managed production
zones (Mayer 1985). Of the total field area in the Colca valley, 62% (8962 of 14,356 ha)
is terraced (Denevan 1988b:22, Table 1). Almost all areas below about 3800 masl in the
valley have been artificially leveled (Treacy 1989:348).
Terracing provides several
productivity-enhancing properties, including the facilitation of water management, soil

thickening and enhancement, and beneficial microclimatic changes. Primary among these
is the hydrological function.
Now-abandoned unirrigated terraces once augmented
precipitation and runoff by pooling water with diversion walls, and were concentrated
around natural ravines (see below). The form of irrigated bench terraces was determined
primarily by the concern to regulate the flow of irrigation water over their surfaces and
between terraces (Denevan 1987; Donkin 1979; Treacy 1989:348). This allows for the
management of water flow over terrace surfaces, while minimizing erosion. As discussed
above, terracing results in much thicker soils, and through soil management practices such
as manuring, fallowing, crop rotation, and tilling techniques, fertile topsoils have resulted.
Also, terraced hillsides shed cold air in the evening better than unterraced slopes and,
56
because they have a high surface area to mass area ratio, terraces radiate heat accumulated
from solar radiation during the day (Brooks 1998:126; Denevan 1980; Treacy 1989:81-82,
349-350). These combined frost-shedding and heat-sinking functions extend the effective
altitudinal limits of frost-sensitive crops such as maize, allowing for an expanded area of
their cultivation.
Building on comparative research by Spencer and Hale (Spencer and Hale 1961),
and Donkin (Donkin 1979), Denevan and colleagues have developed a local typology of
field and terrace forms based on morphological and functional characteristics (Brooks
1998:126-137; Denevan 1980, 1987; Treacy 1989:74-100).
While terminology differs
slightly between investigators (Brooks 1998:127), four general categories can be

differentiated:
1) Bench terraces. Bench terraces have vertical or slightly inclined fieldstone
facing and nearly level (subtly front-sloping) surfaces. Bench terraces form
large staircases on valley slopes, presenting visually-distinctive fractal
patterns, and are concentrated on the lower valley slopes. Bench terrace walls
range from under one meter to over three meters, depending on slope (the
steeper the slope, the higher the terrace wall). As noted above, bench terrace
retaining walls were constructed by excavating through duripans and B
horizons, providing a firm structural foundation. They are often divided by rock
sidewalls. While the vast majority of bench terraces are irrigated, small pockets
of unirrigated bench terraces have been reported around Coporaque (Treacy
57
1989:93) and the Japo Basin in Chivay(Brooks 1998). Denevan (Denevan 1987)
subdivides irrigated bench terraces into three subtypes.
a) Contour terraces. These are long, relatively narrow terraces curved to
match slope (Denevan 1987:22).
b) Linear terraces. These are short, relatively narrow segments running
laterally across slope (Denevan 1987:22).
c) Broad field terraces (also called apron terraces). These are interspersed
among contour and linear terraces on gentler slopes. They are irregular
in shape or rectangular but not as wide as valley bottom terraces, with
low stone retaining walls (Denevan 1987:22).
2) Sloping field terraces. Found primarily along the upper slopes of the valley,
above contour terrace complexes, sloping field terraces are fieldstone-enclosed,
with a thick downslope retaining wall, behind which soil buildup (probably the
result of both intentional slope modification and erosion processes) creates a
more level planting surface. The lower portion of most sloping field terraces
probably served as the most densely planted area, as well as a collection area for
runoff water. Since precipitation tends to occur in short bursts, the thickened
soils and basin form of the fields would have functioned to store moisture
between storms. This property would have been accentuated in areas underlain
by duripan soils, which would serve as a natural barrier to water filtration
(Treacy 1989:131). Thus, the construction and use of sloping fields were not
58
dry farmed in the strict sense, but instead represent a hydraulic half-way house
between complete dry farming and canal irrigation (Treacy 1989:127).
3) Segmented terraces. Segmented terraces fall within the category of isolated,
short, sloping dry field terrace (Spencer and Hale 1961:9, as cited in Treacy
1989:97), and while sharing the basic hydrological characteristics with sloping
field terraces, are formally distinct enough to be considered a separate type. A
segmented terrace consists of a low (0.5-2m high) arcing retaining wall
(composed of surrounding colluvium), behind which a small, nearly level water
catchment and planting area (as in the case of sloping fields) is followed by a
steeper grade. The total area behind retaining walls is usually between 6-17m
long and 2-6m wide. As the name implies, segmented terraces are not aligned,
as in the case of bench terraces, but form discontinuous clusters on a slope. In
general, segmented terraces are most visible along upper, steep colluvial slopes
(but were probably also prevalent in areas now covered by bench terracingsee
below).
In the survey area, the largest concentrations are found along the
southern and eastern flanks of the Pampa Finaya massif (including cross channel
terraces). In these areas especially dense clusters are found around natural
ravines (quebradas), which were modified through the use of diversion walls to
channel runoff to these fields. Treacy identified five areas in Coporaque where
such runoff control features and associated sloping field and segmented terraces
are found (Treacy 1989:127-133), and others have been identified in the Japo
basin to the south of Chivay by Brooks (Brooks 1998). Terraces which straddle
59
the ravines themselves have also been called cross channel terraces, and can
be considered a subtype of segmented terraces (Brooks 1998:127-129).
4) Rock-walled fields. Rock-walled fields generally occur on the flattest terrain of
the valley bottom, and are essentially unterraced fields enclosed by fieldstone
walls. In some cases, the enclosure walls are informal accumulations that result
from clearing the enclosed fields of colluvium. Many of these are very recent in
origin, serving as enclosures for cattle and other livestock, as many valley
bottom fields are dedicated to alfalfa cultivation (Gonzlez 1995). In other
cases, especially amongst the broad expanses of the Qal 4 pampas on which the
modern reduccin villages, large areas of rock-walled fields are enclosed by
complexes of more formally-constructed agro-mortuary wall complexes
massive, thick (up to 3 meters thick) walls that reach up to three meters in
height, with ovoid cyst-like features on top. Agro-mortuary wall complexes are
discussed in further detail in Chapter 4.
Terrace and Field Chronology: Origins, Use, and Abandonment

The diversity of field and terrace forms outlined above is not only a product of
functional difference. Field, terrace, and water management systems also changed over
time, forming a stratigraphic sequenceand in places, a palimpsestof agricultural
complexes covering large areas. Prior to this project, most archaeological research in the
valley focused on terrace origins and abandonment, and considerable progress has been
made in determining the relative sequence and absolute dates for the origins, use, and
60
abandonment of the valleys different terrace and field systems.
Given the inherent
difficulties of dating terraces through absolute means (due to their often intractably turbid
fills and indeterminacy of carbon sample associations), relative dating of terrace and field
complexes has produced more definitive results. Therefore, the sequencing of terrace forms
is better known than the specific timing of these changes.
In general, there is consensus that irrigated bench terracing superceded an earlier
period of unirrigated field systems (Brooks 1998; Denevan 1980, 1987, 2001; Treacy
1989). Irrigated bench terrace complexes are concentrated below about 3800 masl, while
the earlier, abandoned unirrigated sloping fields and segmented terraces extend higher up
the valley slopes. Thus, the general pattern is characterized by an irrigated core of bench
terraces and valley bottom fields surrounded by corona of earlier, unirrigated sloping
fields and segmented terraces along the valleys upper slopes. Using stratigraphic evidence,
Treacy has demonstrated in detail how irrigated bench terrace complexes intrude upon
extant segmented slopes on the southern and eastern slopes of Pampa Finaya (Treacy
1989:122-127). The landscape-scale evidence used to support the sequence is presented in
Figure 2.9, a three dimensional perspective view of Pampa Finaya from the south. Along
the southern slopes of Pampa Finaya, the Inca canal (the canals local modern name)
descends diagonally, demarcating a sharp boundary between unirrigated segmented
terracing above, and irrigated bench terracing below (Figure 2.9). As pointed out by Treacy
(1989:122-127), the segmented terraces must have originally covered the entire slope face
and predate the canal and bench terracing, otherwise one would have to imagine that the
segmented terrace construction abruptly stopped at the canal, forming a wedge-shaped
61
border along the descending line of the canal.
Also, on the eastern side of Finaya,
segmented fields are found both above and below a different canal (the Qachulle canal, now
abandoned) that irrigated valley bottom fields to the south of Pampa Finaya.
The
segmented fields surrounding the canal must predate it, since there would be no point in
constructing unirrigable fields below a canal (Treacy 1989:122-127). Also, the eastern
slopes of Finaya illustrate how the southern slopes most likely appeared before being
modified with irrigated bench terraces below the Inca canal. As shown in Figure 2.9,
segmented terracing extends to the bottom of the slope and onto the valley bottom on the
eastern side (Fig. 2.9) (Treacy 1989:122-127).
Figure 2.9: Perspective view of Pampa Finaya. This view shows stratigraphic relationship between
early segmented terracing and later canals and bench terracing.
62
Although radiocarbon dates and artifact associations provide some clues, the timing
and causes of the terrace origins, and the shift from unirrigated to irrigated agriculture is a
more disputatious matter.
The evidence is derived from reconnaissance and trench
excavations at a series of bench terraces along an altitudinal transect near the site of Chijra
in Coporaque by the Ro Colca Abandoned Terrace Project (Denevan 1986a, 1988a;
Malpass 1988; Malpass and de la Vera Cruz Chvez 1990; Treacy 1989), as well as
reconnaissance and excavations by Brooks in the Japo drainage to the south of Chivay
(Brooks 1998).
63
Figure 2.10: Chijra/San Antonio/Chilacota area.
64
Dating early terracing in the valley is complicated by the difficulty of differentiating
terrace construction, use, and reconstruction episodes, as well as the methods used for
absolute dating. Brooks presents a hypothesized succession scenario of terrace types, and
argues that unirrigated terracing began in the Colca valley sometime prior to 2400 BC
(Brooks 1998:400-401, 405). However, such an early date would place the construction of
the first terracing in the valley in the Late Archaic Period (4000-2000 BC)an unlikely
scenario, since this time period is characterized by preceramic, preagricultural hunting and
foraging lifeways throughout the southern highlands, including the Colca valley (see
Appendices B and C).
Also, the dating methods used to support this argument are
problematic because they are based on radiocarbon dates (uncalibrated) from bulk soil
samples in terrace fill.10 However, the rest of the succession scenario presented by Brooks
is more plausible, in which unirrigated sloping fields, segmented terraces and bench
terraces were replaced by irrigated bench terracing around the 12th century AD (Brooks
1998:409). Based on evidence from reconnaissance around Coporaque, Treacy argued that
local Middle Horizon ceramics are most associated with sloping fields and segmented
terracing, while Late Intermediate to Late Horizon Period ceramics are much less prevalent
in these contexts, but instead are found in heavy concentrations amidst irrigated bench
terrace complexes (Treacy 1989:106-117). These associations, combined with radiocarbon
dates from excavations in bench terraces (using discrete charcoal samples), led Treacy to
hypothesize that dry farming predominated from Middle Horizon and earlier times (that is,
10
Aside from problems of discreteness and lack of association with features, bulk soil sample constitute an
average value of the pooled charcoal in the sample. Contamination by old carbon in the soil results in
skewing towards old values. This becomes especially apparent when comparing well-provenienced discrete
charcoal samples with bulk soil sample dates.
65
prior to the 11th century AD), and was subsequently replaced by irrigated bench terrace
agriculture around AD 1000-1100 with the onset of the Late Intermediate Period (Treacy
1989:122). Based primarily on the bench terrace excavations, Sandor and Eash reach
similar conclusions (Sandor 1986, 1987a, 1992). Malpass presents an alternative scenario,
citing an apparent lapse in the ceramic chronology during the Late Intermediate Period
(Malpass 1988). Malpass (1988) suggests that the construction of irrigated bench terrace
complexes along the slopes surrounding Chijra may pertain to the Late Horizon, but does
not exclude the possibility that they could have been constructed earlier. In sum, while the
history of unirrigated terrace origins in the Colca valley remains debated, there is some
consensus that the initial construction and use of irrigated bench terracing and associated
canal systems, as well as the abandonment of previous unirrigated field and terrace systems,
took place sometime between the 11th and 13th centuries AD.
Previous researchers have favored climatic stress as the primary causal factor
behind the development of irrigated terraced agriculture in the valley, noting that the timing
of the conversion to irrigated bench terracing correlates with the longest period of dry, cool
conditions recorded in the glacial and lake core data (discussed above) between the 12th and
15th centuries AD. As summarized most recently by Denevan (2001:200):
This scenario, drawing on Brooks and Treacy, is an argument that the
development of complex labor-intensive irrigated terraces resulted, in part at
least, from climatic stress. This is contrary to some popular and academic
thinking that climatic deterioration (dryer, colder) was the cause of
agricultural and social collapse at various places in the world. In the Colca
Valley, climatic stress instead seems to have motivated the development of a
more efficient, more productive, and less risky method of food production.
66
Thus, rather than playing the role of the deus ex machina of collapse (Binford and
Kolata 1996; Binford, et al. 1997; Kolata 1996b; Ortloff and Kolata 1993), climate change
in this scenario is instead the prime mover impelling an intensification of agricultural
production through the elaboration of irrigated terrace and canal systems. Although all
have been cautious not to discount demographic, economic, and political factors, previous
research designs, with their focus on terrace and field systems rather than settlements, have
made evaluation of the role of sociocultural processes as causes for these changes difficult
to assess. This project was designed to address this lacunae, and approaches the dynamics
of prehistoric terraced agriculture and sociopolitical change as complex, multicausal
processes both constricted ecologically and constructed culturally.
67
Chapter 3
The Collagua in Regional and Historical Context: Previous Research
Introduction
The Collagua, until recently, have attracted less scholarly attention than the
comparably-sized large late prehispanic polities of the neighboring Titicaca Basin.
Previous archaeological research in the Colca valley has been limited to reconnaissance and
test pit excavations focused primarily on terraces and fields. These pioneering studies have
produced a preliminary ceramic chronology and an outline of the long-term sequence of
changing agricultural landforms in the valley. Prior to this project, most of what we knew
of the political organization of the Collagua province during Inkaic and early colonial times
was derived from document-based research. Archaeological indices for Inkaic influence or
presence in the valley remained provisionally-defined and debated. Archival studies in the
1970s and 1980s unearthed a wealth of colonial documentation, especially from local parish
archives, including a large series of colonial censuses. These studies characterized the
Collagua as a large-scale ethnic seoro of regional import, and provided preliminary
reconstructions of the organization of the Collagua during Inkaic and early colonial times.
Previous Archaeological Research

In the first modern archaeological reconnaissance in the valley, Mximo Neira
described the distinctive local architectural style of the Late Intermediate Period (AD 10001450) and Late Horizon (AD 1450-1532), characterized by large, rectangular domestic
structures with high gabled roofs, some with tenon supports for a second storey, and, most
68
diagnostically, unusually narrow, tall doorways (Neira Avendao 1961:155). Drawing
upon descriptions of the province in colonial documents, Neira interpreted the large
architectonic remains he observed as further evidence that the Collagua played a paramount
role within regional Inka administration (Neira Avendao 1961:78-87).
Neira also
produced a provisional local ceramic typology that built on Kroebers earlier differentiation
of Chuquibamba ceramics from Churajn and other regional styles (Kroeber 1944). Neira
considered ceramics from the Colca valley to be a local expression of the Chuquibamba
series (Neira Avendao 1961:195), and came to use the terms Chuquibamba and Collagua
interchangeably when describing LIP to Late Horizon ceramics from the valley (see
Chapter 4, Neira Avendao 1986, 1990).
Following Neiras groundbreaking investigation, a second generation of research in
the 1980s and 90s focused primarily on the agricultural landscape of the valley (see Chapter
2), but also included reconnaissance and test pit excavations at prehispanic settlements
around the villages of Chivay (Brooks 1998; Guerra Santander and Aquize Cceres 1996),
Coporaque (Malpass 1986, 1987; Malpass and de la Vera Cruz Chvez 1990; Martin 1986;
Neira Avendao 1986), Achoma (Oquiche Hernani 1991; Shea 1986a, b, 1987), and
Cabanaconde (de la Vera Cruz Chvez 1987, 1988, 1989).
Most recently, Mirriam
Doutriaux (pers. comm. 2003) has just completed full-coverage survey and excavations
around Lari and Cabanaconde in the lower valley, and Nicholas Tripcevich (pers. comm.
2003) is presently conducting systematic survey and test pit excavations in the upper valley,
including the area around the Chivay/Cotallalli obsidian source.
69
The Formative Period and Middle Horizon
While a possible pre-Middle Horizon cemetery has been reported at the site of
Auqui in Cabanaconde (de la Vera Cruz Chvez 1987:97; 1988:99-101), no settlements
predating the Middle Horizon had been identified in the Colca valley prior to this project.
Nevertheless, it seemed likely that extensive human exploitation and settlement of this
major valley began prior to the Middle Horizon but had yet to be identified, because
neighboring locales have well-documented Archaic Period (8800 - 1500 BC) components,
such as the rockshelter complex at Sumbay to the south (Neira Avendao 1968, 1990,
1998), and Formative Period (1500 BC - AD 500) components have been identified in
virtually all of the surrounding valleys.11
Before this project, more was known about the distribution of obsidian from the
valley throughout the south central Andes than the human presence in the valley itself prior
to the Middle Horizon (Brooks 1998; Brooks, et al. 1997; Burger, Asaro, Salas, et al. 1998;
Burger, et al. 2000; Stanish, et al. 2002). The well-documented spread of this obsidian
throughout the region reveals a vast trading network with especially strong ties to the
Titicaca Basin since at least Late Archaic times (4800-3000 BC) (Burger, Asaro, Salas, et
al. 1998; Burger, Asaro, Trawick, et al. 1998; Burger, et al. 2000; Stanish, et al. 2002). The
Chivay source is one of two very large obsidian sources in the south central Andes; the
other is located near Alca in the Cotahuasi valley to the north (Burger, Asaro, Salas, et al.
1998; Burger, Asaro, Trawick, et al. 1998; Burger, et al. 2000).
11
Aside from the well-documented Formative styles of the lacustrine and altiplano zones to the east, these
include the lesser-known Formative ceramic styles from the western valleys of the Department of Arequipa,
such as the Ayawala style of the Chuquibamba valley, the Hachas (or Achas) style of coastal Acar, the Tasata
style of coastal Quilca (the lower Chili) the Socabaya style of the valley of Arequipa (see Cardona Rosas
2002:38-60; Neira Avendao and Cardona Rosas 2000-2001).
70
Diachronic change in the distribution of obsidian from these two sources reveal
shifting trade networks during the Formative and Middle Horizon.
During the Early
Formative (1500 800 BC), the exchange of Chivay source obsidian stretched from the
southern tip of Lake Titicaca to Cuzco in the north, where both Chivay and Alca obsidian
appear in equal proportions at that time (Burger, Asaro, Trawick, et al. 1998; Burger, et al.
2000; Jennings 2002a). During the subsequent Middle (800 200 BC) and Late Formative
(200 BC AD 500) Periods, the proportion of Chivay obsidian in the Cuzco area declined
relative to that of the Alca source, but its prevalence in the Titicaca Basin increased and
expanded southward beyond the southern shore of the lake, making up 89% of the sourced
obsidian at sites around Puno, and 100% of the sourced obsidian at Chiripa and Kallamarka
to the south (Burger, et al. 2000:296-323).
During the Middle Horizon (AD 500 - 1000), when political control of the western
Titicaca Basin was consolidated under the Tiwanaku rule, virtually all of the obsidian at
sites there continued to be procured from the Chivay source (Brooks, et al. 1997; Burger, et
al. 2000:324-343). However, despite these strong trade links with Tiwanaku, no sites in the
Colca valley with indices for a Tiwanaku imperial presence have been documented. No
diagnostic Tiwanaku ceramics have been recovered in the valley. Rather, the forms and
decorations of Middle Horizon ceramic assemblages in the Colca indicate stylistic
affiliation with Wari, not Tiwanaku (de la Vera Cruz Chvez 1987, 1989; Malpass and de la
Vera Cruz Chvez 1990).
Local Middle Horizon ceramics were manufactured with
different pastes than later local wares, and consist of two primary stylistic groups: tricolor
(white and black on red) decorated bowls with Wari-influenced motifs similar to decorated
71
Wari-influenced ceramics from the Chuquibamba valley (the Qosqopa style, see below),
and undecorated but well-polished buff-colored cumbrous bowls with a band of red slip
over the exterior and extending as a band below the rim on the interior (de la Vera Cruz
Chvez 1987, 1989; Malpass 1987:57; Malpass and de la Vera Cruz Chvez 1986:159;
1990). These ceramics derive from stratigraphic levels that yielded radiocarbon dates of
AD 429-694, and AD 436-779 (calibrated 2 sigma) (Malpass and de la Vera Cruz Chvez
1986, 1990),12 suggesting that Wari influence had reached the Colca valley during first half
of the Middle Horizon.
Only two other Middle Horizon sites in the Colca valley have been investigated
Chilacota in Coporaque (Martin 1986), and Achachiwa in Cabanaconde (de la Vera Cruz
Chvez 1987, 1988, 1989).13 The site of Chilacota, located at 3760 masl on a small plateau
feature near Coporaque, comprises a small complex of unirrigated sloping fields and
associated rock wall enclosures, but no habitational features (Martin 1986).
Two
radiocarbon samples and ceramics date its occupation to the Middle Horizon.14
12
The sample numbers for these assays are: WIS-1713 (1440 BP 80) and WIS-1714 (1400 BP 80)
(Malpass and de la Vera Cruz Chvez 1986:163, Table 1; 1990). All dates were presented in the original
reports uncalibrated, and calendar dates were originally derived by subtracting the BP values from the year
1950. However, the WIS-1713 sample was originally reported as 1140 80 (Malpass 1987:61, Table 1;
Malpass and de la Vera Cruz Chvez 1986:162-163), when it should have been 1440 80. This is apparent
both in the original calculation of the calendar year of the date (AD 510), and in subsequent reports of the date
(see Shea 1997:54). I calibrated the dates with Calib 4.3, using the probability distribution curve (Stuiver and
Reimer 1993; Stuiver, et al. 1998). The 1 sigma values are AD 537-669 and AD 544-757, respectively.
13
Middle Horizon components have also been reported at Juscallacta in Chivay (Brooks 1998).
Both of the radiocarbon dates from the site are difficult to interpret. One of them was derived from a bulk
soil sample below a terrace wall, and (predictably, given the inherent mixture of old carbon in the bulk
sample) returned a very early date of 352080 BP (uncalibrated). The other (165070 BP, uncalibrated) was
from a charcoal sample excavated from fill inside the wall of an irrigation canal (Malpass and de la Vera Cruz
Chvez 1986:163). This second sample is more pertinent to the dating of the canal as it relates to the irrigated
bench terracing at Chijra, since, as Martin (1986:221) notes, this canal is located 5-10 meters downslope from
the sloping fields of Chilacota, and therefore could not have irrigated fields there. In any case, this sample
14
72
The site of Achachiwa, located 1.5 km to the west of Cabanaconde, provides
preliminary evidence regarding residential patterning and the nature of Wari influence in
the valley. Achachiwa is composed of clusters of domestic and agricultural terracing with
poorly defined architectural remains within two concentric fieldstone walls that enclose a
total area of about 35 ha (de la Vera Cruz Chvez 1987:97; 1988:108-109, 113-116;
Denevan 2001).15 The main occupation was dated to the Middle Horizon Period based on
the predominance of local Middle Horizon ceramics collected from site reconnaissance and
test pit excavations by Pablo de la Vera Cruz (see Chapter 4) (de la Vera Cruz Chvez
1988:108-109, 113-116; Denevan 2001). De la Vera Cruz initially characterized the site as
a fortified settlement with agglutinated residential areas, noting general affinities with some
Wari administrative centers (de la Vera Cruz Chvez 1987:97-102; 1988:123-125). Based
on this description, others have suggested that the site may have functioned as a Wari
administrative center (Sciscento 1989:268-269) and, more recently, de la Vera Cruz made
this assertion explicitly (de la Vera Cruz Chvez 1996:150).
However, the original
characterization of the site as a local fortified settlement appears most plausible, given the
lack of evidence for any affinity to the orthogonal cellular layout common to Wari
administrative centers, or, for that matter, any apparent centralized planning in the
residential sectorspoints initially raised by de la Vera Cruz himself (de la Vera Cruz
also may have been contaminated by old carbon from the irrigation water carried in the canal, and/or been
transported to the context of recovery by the irrigation water.
15
In an overview of the findings from his Bachelors thesis, de la Vera reports the total area enclosed by the
outer perimeter wall as 700 X 500 meters (de la Vera 1987:97), but in the thesis itself, he reports the size as
700 X 300 meters (de la Vera Cruz Chvez 1988:41). In the latest description of the site (de la Vera Cruz
Chvez 2000:146), de la Vera reports the wall lengths as 700 X 500 meters, so I use these later figures.
However, the area enclosed by the walls is much larger than the residential area, which is primarily
concentrated in two clusters of terracing (de la Vera 1988:40-55).
73
Chvez 1988:123-125) and reiterated by others (Brooks 1998:87; Schreiber 1992:104). In
sum, while evidence for direct Wari imperial administration is weak, local ceramic
assemblages suggest that the valley lies within the southern sphere of Wari influence.
The Colca valley appears to represent a boundary area between Tiwanaku and Wari
spheres of influence. While the obsidian data demonstrate long-term economic ties to
Tiwanaku, Wari influence, as seen via artifact media, appears more political or perhaps
ideological in nature. This situation is probably reflective of a broader pattern beginning to
emerge along a porous contact zone between Wari and Tiwanaku in the Pacific drainages of
southwestern Peru. While Wari-Tiwanaku interaction in the Moquegua valley has been
amply described and debated elsewhere (Goldstein 1989; Moseley, Feldman, et al. 1991;
Owen 1993; Rice, et al. 1989; Williams 1997), new data from the western valleys of the
Department of Arequipa suggests that Wari imperial presence is stronger below about 2500
masl, perhaps signaling that Wari imperial expansion primarily moved along a coastal to
upper yungas transect, while Wari influence in the highland valleys of Arequipa was more
indirect. Certainly in the Colca-Majes-Caman drainage, evidence for direct Wari presence
increases towards the coast.
In the Chuquibamba valley, a tributary of the Majes
downstream from Cabanaconde, Neira, Sciscento, and Cardona Rosas have all noted the
presence of Wari-influenced ceramics.
Neira named a local Wari-derived style of
polychrome ceramics, the Qosqopa (or Ccoscopa) style (Lumbreras 1974b:155, 157; Neira
Avendao 1990:131-136), based on a grave lot in Chuquibamba (also documented by
Cardona Rosas 1993:55-57), but this style was never formally described or defined (see
74
Ypez, et al. 2001).16
Based on the presence of Wari-influenced ceramics but no
administrative center, Sciscento suggested that Wari probably never consolidated territorial
control over the valley, and characterized Wari imperial influence as indirect or hegemonic
(Sciscento 1989:264-271). Moving downstream to the middle Majes valley, stronger Wari
influence in the ceramics and textiles of the settlement and cemetery of Beringa have been
identified during preliminary reconnaissance (Garca Mrquez and Bustamante Montoro
1990; Ratti de Luchi Lomellini and Zegarra Arenas 1987), and recent excavations by
Tiffiny Tung produced abundant Huamanga style ceramics and polychrome Wari textiles
dating to AD 650-800 (calibrated 2 sigma) (Tung 2003). Artifacts associated with burials
recovered during salvage excavations at the cemetery site of La Real, located 8 km
downstream of Beringa, included Wari feather mantles and other polychrome Wari textiles,
as well as large quantities of polychrome ceramics in several imperial styles (Tung 2003).
In the coastal zone, after the river changes names to the Caman, Malpass has suggested
that the site of Sonay, dating to the late Middle Horizon, may represent a small Wari
administrative center with possible orthogonal cellular layout (Malpass pers. comm. 2002).
Although research in most other valleys of the Department of Arequipa is preliminary in
nature, evidence for Wari presence appears similarly weak in the highlands and stronger in
16
The Qosqopa style has long bedeviled archaeologists of the Arequipa region. Based on published
illustrations and photographs of vessels from the type site, it appears that the original vessels used to define
the Qosqopa style actually pertain to several Wari styles, including Chakipampa (Neira Avendao 1990:131,
Figs. C and D), Viaque (Neira Avendao 1990:134, Fig. A), and Huamanga (Neira Avendao 1990:132,
Figs. E and H), while there is also a group (Neira Avendao 1990:133, Figs. A-D) that probably constitutes a
local series that derives isolated motifs (such as S and X figures) from various imperial styles (which, in
turn, continue to appear in later Chuquibamba ceramics). This latter group could represent a distinct style, but
it was not differentiated from the other vessels that pertain to other imperial styles just mentioned. Since the
coherency of Qosqopa as a style is in doubt, its preliminary chronological placement during Middle Horizon
III (post-imperial collapse) based on vague stylistic criteria by Lumbreras (Lumbreras 1974b:157)appears
unsubstantiated. Ypez and colleagues have recently made this point and argued for an earlier date of initial
Wari influence in the western valleys of Arequipa (Ypez, et al. 2001).
75
the yungas and coastal zones. In the Cotahuasi valley, the neighboring valley to the north
of the Colca (and bordering the Wari heartland in the Department of Ayacucho), Jennings
suggests that Wari influence was hegemonic in nature and mediated by local elites, perhaps
at the sites of Collota and Netaha (or Netahaha) (Jennings 2002b). To the south, in the
lower reaches of the Chili valley, reconnaissance by de la Vera Cruz, and more recent
systematic survey and excavations by Cardona (2002:72-74, Cardona Rosas pers. comm.
2002) at the large site of Pampa La Estrella, have documented orthogonal architecture and
Wari ceramics.
In sum, the available archaeological evidence regarding the antecedents to the
Collagua prior to this project, while much of it very preliminary in nature, suggests that
groups in the Colca valley figured prominently in regional networks of economic and
political interaction during the Middle Horizon and earlier times. Although no pre-Middle
Horizon component had been identified, at the very least, trading emissaries or caravaneers
frequented the valley and environs to procure obsidian. But there was a high likelihood that
Archaic and Formative components had not yet been identifiedhypotheses since
confirmed by this project (see Chapter 5). During the Middle Horizon, groups in the Colca
maintained strong trade links throughout the lacustrine and altiplano areas, but never came
under Tiwanaku political rule. Rather, Wari political influence appears to have extended
from the coast into the highland zones of the Colca, but this influence appears indirect, and
no unequivocal evidence for Wari imperial administration had been documented.
76
The Late Intermediate Period and Late Horizon
Due to small ceramic collections, poorly stratified deposits, and a gap in
radiocarbon dates for the Late Intermediate Period (AD 1000 - 1450), previous researchers
in the valley faced the perennial problem of differentiating LIP and Late Horizon ceramics.
Thus, the two time periods were lumped together in most previous discussions. In a
preliminary ceramic sequence, Malpass and de la Vera Cruz originally reported little
evidence for a discrete local LIP style (Malpass and de la Vera Cruz Chvez 1986, 1990).
Instead, they noted a variety of post-Middle Horizon decorated wares, some of which
showed obvious Inka formal and decorative influence (Malpass and de la Vera Cruz
Chvez 1986, 1990). Subsequently, de la Vera Cruz amended this sequence based on
stylistic similarities with the better-known Chuquibamba series (Cardona Rosas 1993;
Garca Mrquez and Bustamante Montoro 1990; Kroeber 1944; Sciscento 1989),
subdividing the sequence into Early, Middle, and Late Chuquibamba styles (corresponding
to the Middle Horizon, LIP, and Late Horizon, respectively). More recently, based on
differences in decoration, Brooks has argued that Collagua ceramics should be considered a
separate style from Chuquibamba (Brooks 1998). Building on this work, and aided by a
larger ceramic sample than previous research, I have proposed a Collagua ceramic stylistic
sequence that differentiates LIP from Late Horizon styles (Wernke 2001b). I present this
sequence in Appendix A.
Owing in part to these chronological problems, archaeological interpretations of the
development and organization of the Collaguas as an ethnic polity, and the local impact of
Inka administration have been limited and somewhat contradictory. Most have argued that
77
the strongest archaeological evidence for direct Inka imperial presence is in the lower
valley around Cabanaconde, where Inka cutstone masonry has been documented in the
village itself, as well as a probable small administrative center at the site of Kallimarka (de
la Vera Cruz Chvez 1987, 1988). High above the lower valley, the Inkas performed
human sacrifice on the glaciated stratovolcano Mount Ampato, one of the principal
mountain deities of the region. Johan Reinhard and Miguel Zrate famously recovered a
well-preserved adolescent female who was killed by blunt force to the head and interred
near the peak on a prepared platform in association with fine Inka ceramics and precious
metal artifacts (Reinhard 1998). In the central valley, Neiras preliminary observations
(Neira Avendao 1961) of a number of large sites with especially elaborate domestic
architecture around Coporaque and Yanque suggested that this area represented the political
center of the Collaguas, though the degree and nature of Inka imperial influence or presence
remained unclear (Neira Avendao 1990:152-166).17 More recently, Malpass (1987) and
Shea (1987), largely in agreement with de la Vera Cruz (1987), suggested that Inka imperial
administration was more indirect in the central valley than in Cabanaconde, although both
noted that irrigated bench terracing was probably expanded and upgraded considerably
during the Inka occupation, especially along warm slopes near the valley bottom (Malpass
1987; Shea 1987). The Denevan team mapped and conducted test pit excavations at the site
of Chijra, a large settlement and terrace complex to the east of Coporaque (downslope of
Chilacota, discussed above), and reported no Inka architecture (Malpass 1986, 1987;
17
This view was undoubtedly influenced by a description of Coporaque by the Franciscan Friar Luis Jernimo
de Or, who lived in Coporaque during the late 1580s and early 1590s. Or tells a story (related to him as
part of local oral history) describing the construction of a house sheathed in copper near Coporaque in honor
of the fourth Inka, Mayta Capac, who was said to have married the daughter of a Collagua lord (Or 1992
[1598]:41). See below for further discussion of the historical and historiographical significance of this story.
78
Malpass and de la Vera Cruz Chvez 1986, 1990; Neira Avendao 1986; Treacy and
Denevan 1986). However, terrace excavations there did reveal evidence for agricultural
terrace reconstruction during the Late Horizon (Malpass 1987). Reconnaissance and test pit
excavations were also conducted in the Achoma area (Oquiche Hernani 1991; Shea 1986a,
b, 1987). Shea (1987) characterized the settlement pattern around Achoma as hamlet- and
village-based and found no evidence for major reorganization or resettlement during the
Late Horizon. The layout of one of these settlements, Escalera, approximates a grid pattern
(Oquiche Hernani 1991), but Shea hypothesized that this reflects settlement reorganization
during the early colonial period before the Toledan reducciones were established in the
early 1570s (Shea 1997). While Late Horizon ceramics (described by Shea as Sillustaniinfluenced) were recovered from Escalera, the strongest evidence for Inka presence around
Achoma came from the site of Malata, located low in the valley on the first alluvial fan
above the river (ca. 3200 masl). This site produced the greatest quantities of polychrome
Inka ceramics in the Achoma area, as well as finely dressed stonework (Shea 1986b:367).
Based on the oft-noted importance of maize in the Inka imperial ideology and political
economy, the concentration of Inka ceramics in lower suni and kichwa zone sites, and their
association with irrigated bench terrace complexes, Shea (Shea 1987:82), similar to
Malpass (Malpass 1987), hypothesized that Inka administration focused on intensifying
maize production through the expansion of irrigated bench terracing in the lower elevations
of the valley.
Subsequent researchers, in contrast, have characterized the settlements of the
central valley as local villages constructed and occupied by local Collagua populations.
79
Two investigations at the large settlement of Juscallacta, located to the south of Chivay,
suggested that it was occupied by a local population primarily during the Late Intermediate
Period (Brooks 1998; Guerra Santander and Aquize Cceres 1996). More generally, based
on an apparent absence of an administrative center and the scarcity of imperial Inka styles
in local ceramic assemblages, Brooks suggested that there was never a significant direct
Inka imperial presence in the central Colca valley (Brooks 1998).
Thus, the prevailing interpretation amongst archaeologists held that Cabanaconde
was more centrally administered than the middle and upper reaches of the valley, owing
perhaps to an imperial interest in maximizing maize production (de la Vera Cruz Chvez
1987, 1988; Malpass 1987; Shea 1987).
This view has been repeated in synthetic
discussions of the prehistory of the valley, in which the area around Cabanaconde is
depicted as more important politically and in productive terms than other zones of the
Colca valley (de la Vera Cruz Chvez 1987, as cited in Gelles 2000:27).
The
archaeological data for the Inka occupation of the central and upper sections of the valley
were more ambiguous, although most researchers argued that Inka rule was probably
mediated by local elites.
80
The Collagua in Colonial Documents
During colonial times, the corregimiento of the Collaguas was the wealthiest and
most densely populated part of Arequipa (Benavides 1996; Cook 1982; Mlaga 1977). The
tasa (summary ledger) of the Toledan visitas (Cook, et al. 1975 [1582]) indicates that the
population of the Collaguas, with 33,900 inhabitants (7,922 tributary households),
constituted one third of the total population under the jurisdiction of Arequipa during the
mid-sixteenth century, and produced 35% of the taxes collected there (Mlaga 1977:108).
While the Collagua are virtually absent from the standard chronicles,18 an account from
1586 in the Relaciones geogrficas de indias by provincial magistrate Juan de Ulloa
Mogolln provides an overview of the organization of the Collagua province under early
colonial rule (Ulloa Mogolln 1965 [1586]). Ulloa described how the valley was divided
between two ethnic groups: the Aymara-speaking Collaguas of the middle and upper part of
the Colca valley, and the Quechua-speaking Cavanas of the lower part of the valley, each
with distinctive mythical origins, dress, and body modifications. The charter myth of the
Collagua related by Ulloa depicts the founding hero-ancestors of the ethnic group as
siblings who descended from their origin place (pacarina) on a distant peak to conquer the
Colca valley.19 Specifically, the pacarina of the Collagua was Mount Collaguata, located
near the village of Velille in the modern Province of Espinar, Cuzco, while the Cabanas
18
Cobo (1979 [1653]:119-120) and Garcilaso de la Vega (1966 [1609-1617]:1:153) briefly describe the
Collaguas as having been incorporated into Tawantinsuyu by Mayta Capac, the fourth Inka of the standard
dynastic succession, through his marriage with the daughter of a principal Collagua kuraka. However, both
sources are probably derived from an earlier account by the Franciscan Friar Luis Jernimo Or (1992
[1598]:41), who worked and lived in Coporaque during the last two decades of the sixteenth century.
19
The chartering myths of highland agriculturalist ethnic groups often depict the founding ancestors as
conquering warriors descending from a distant peak, and dominant over local herding groups (Duviols 1973;
Parsons, et al. 1997; Salomon 1995:322).
81
(also called the Condes or Kuntis of Cabana) claimed descent from the pacarina mount
Hualca Hualca, above the village of Cabanaconde in the lower Colca valley.
Ulloa
described the distinct forms of cranial deformation and headdress of the Cabanas and
Collaguas, each hallmarks of ethnic identity in the Inka empire.20
Figure 3.1: Overview of the Province of Collaguas, showing internal divisions.
Second, Ulloa describes the native division of the Collagua into two ranked groups,
each with discrete territories within the valley: the higher-ranking Yanquecollaguas
(Yanque Collaguas), who occupied the upper part of the valley, and the lower-ranking
20
According to Ulloa, the Collagua had elongated heads and hats, while the Cabana had flattened, widened
crania and hats with flat tops (Ulloa Mogolln 1965 [1586]:327).
82
Laricollaguas (or Lari Collaguas) of the middle part of the valley.21 The names Yanque
and Lari are expressive of rank and kinship ties between the two groups. According to
Ulloa, Yanque (or Yanqui) was a locally venerated name that was used as an honorific in
addressing the paramount kurakas of the province, and was also the name of the capital of
the province, where the principal kurakas used to reside and continue to reside (Ulloa
Mogolln 1965 [1586]:329). Ulloa goes on to discuss how the Aymara term lari (or lare)
is also used as an honorific for the lords of Laricollaguas, and means uncle or relative
of the yanquis (more precisely, mothers brothersee Bertonio 1956 [1612]; Zuidema
1964:115-118), and is also the name of the village where the native lords reside in
Laricollaguas.22 Both Yanquecollaguas and Laricollaguas were internally divided into two
ranked moieties, or saya, called Hanansaya (the higher-ranking moiety), and Urinsaya (the
21
Following the colonial spatial correlate of this division, which presumably follows its prehispanic
antecedent, the border between Yanquecollaguas and Laricollaguas runs to the west of the villages of Achoma
on the south side of the river and to the west of Ichupampa lands to the north (see Figure 3.1). Therefore,
Achoma and Ichupampa were the villages furthest to the west in Yanquecollaguas. However, as noted by
Cock Carrasco (1978:30), these intra-ethnic boundaries appear to have been very porous, since there were
ayllus in both Achoma and Ichupampa that were subject to the kurakas of Laricollaguas.
22
Ulloa explains the terms Yanque and Lari in detail (Ulloa Mogolln 1965 [1586]:329, my translation):
Chapter thirteen. The village of Yanque of this province was called this because yanqui is a venerated name
and they say this term to the principal caciques, and just as they [the principal caciques] used to reside and
continue to reside there, its meaning is the village where the Lords [Seores] reside. Lare is the head of this
province [i.e. Laricollaguas]; it is called Lare because it also has this type of meaning: out of courtesy and
respect they say between themselves lare to a principal cacique and nobleman amongst them, because this
means uncle or relative; and because the lares and yanquis consider themselves as brothers who both
emerged from Collaguata, the mountain mentioned previously, they say that they founded these two principal
towns, the one called Yanqui, where the greater Lords [Seores] were and the other Lare, where the Lords
[Seores] are that follow them and are uncles and nephews; and this is the meaning of these names.
Original: Captulo trece. El pueblo de Yanqui desta provincia se llam as, porque yanqui es nombre
venerado y lo dicen a los caciques principales, y como en ste residan y residen los caciques principales y es
cabecera desta provincia, su significado es el pueblo donde residen los Seores. Lare es cabecera desta
provincia; llmase Lare porque tambin tiene su significado desta manera: por cortesa y respeto dicen
entrellos lare a un cacique principal, y no tiene libertad uno de decir esto, si no es procediente de cacique
principal y noble entrellos, porque quiere decir to o deudo; y como entre los lares y yanquis se tienen por
hermanos y salidos de Collaguata, cerro ya dicho, dicen que fundaron estos dos pueblos principales, el uno
llamado Yanqui, donde estuvieron los mayores Seores y el otro Lare, donde estn los Seores que le siguen e
son tos e sobrinos; y ste es el significado destos nombres.
83
lower-ranking moiety). Several previous researchers have discussed Ulloas description of
the tripartite and decimpartite internal organization of each of these divisions, which, in
combination with the names of the ayllus registered in the visitas to the valley, reveal the
ideal administrative structure of the Inkaic province of Collaguas (Benavides 1989; Cock
Carrasco 1976-77; Prssinen 1992:362-366; Pease G. Y. 1977; Rostworowski de Diez
Canseco 1983:121-123; Wachtel 1977:77; Zuidema 1964:115-118).
In Chapter 8, I
reconstruct this ideal organization in further detail and illustrate how it was fitted to local
ayllu organization.
The Cabanas and Collaguas also had distinct economic foci. The Cabanas, settled
in the lower valley at mid- to lower-kichwa zone elevations around Cabanaconde, cultivated
maize. The Collaguas, settled throughout the higher altitudes of the central and upper
portions of the valley and surrounding puna, pastured massive camelid herds in the puna
zone and primarily cultivated quinoa, maize, and tubers within the valley (Ulloa Mogolln
1965 [1586]:332).
The Collaguas were known for their fertile agricultural lands and
especially for their vast herds of alpacas and llamas in the extensive puna surrounding the
upper valley (Crespo 1977; Benavides 1996). A cleric stationed in the herding villages of
Callalli and Tisco in the upper valley estimated that those two villages alone tended a total
herd of more than 25,000 head in 1585 and 1586 (Crespo 1977:54). The pasturage around
these villages represents only a small fraction of the puna surrounding the valley, so the
total herd must have been immense indeed, and probably ranked among the largest herds of
the south-central Andes during Inkaic times.
84
The importance of the Collagua to the Inkas, and later to the Spanish, is also evident
in the recipients of their grants of encomienda.23
Initially, they appear to have been
retained in encomienda by Manco Inca himself along with other valued estates and
provinces with intimate ties to the Inka dynasty, such as Ollantaytambo, Yucay, Calca, and
Andahuaylas (Julien 1998). Contrary to the norm, these encomiendas were not granted to
conquistadores in the repartimientos generales of 1534, 1535, or 1540perhaps in
exchange for other concessions from Manco Inca to Pizarro (Julien 1998:490-502). In any
case, the Collagua were clearly prized as one of the largest and most lucrative encomiendas
in Peru, and were granted to individuals of the highest tiers of the Spanish elite in the years
following the neo-Inka rebellion.
Francisco Pizarro granted Yanquecollaguas first to
Gonzalo Pizarro on January 10, 1540 (Mlaga Medina 1977), and then to the prominent
vecino of Arequipa Francisco Noguerol de Ulloa, in recognition for his role in defeating the
rebellion of Gonzalo, before reverting to Phillip II himself in 1559 (Cook and Cook
1991:29-32, 127-129).24
During the first four decades of Spanish rule, the only permanent Spanish residents
in the valley appear to have been priests. Franciscan friars, with the order of La Gasca to
reinvigorate religious instruction in the provinces, made early inroads in the Colca valley
during the 1540s by establishing a series of rustic chapels at the principal settlements in the
23
24
For general discussion of the institution of encomienda, see Lockhart 1968:11-33.
La Gasca issued the encomienda grant of Yanquecollaguas to Noguerol on September 10, 1548, five
months after the defeat of Gonzalo Pizarro at Jaquijahuana. Awarded in recognition of his military service
against the Pizarrist rebellion, it was apparently unbeknownst to the Crown that Noguerol had supported the
insurgency before switching allegiances (see Cook and Cook 1991:21-32). Laricollaguas was divided into
two encomiendas along moiety lines. Francisco Pizarro granted Hanansaya to Marcos Retamoso and
Urinsaya to his personal secretary, Alonso Rodrguez Picado, both on the same day as the grant of
Yanquecollaguas to Gonzalo (January 10, 1540) (see Mlaga Medina 1977).
85
valley (Tibesar 1953:46, 65). These villages probably grew gradually as they became
officially recognized as doctrinas (parishes) in 1557 by the Viceroy Marques de Caete
(Tibesar 1953:47, 65-66). As I discuss in Chapter 7, I have identified settlements with
colonial components and chapel structures that probably represent such doctrinas.
These provisional doctrinas were short-lived however, as the Toledan regime
forcibly resettled the valley population into 24 nucleated reduccin villages during
Toledos visita general in the early 1570s (Mlaga Medina 1974, 1977). Despite this
reorganization of settlement, colonial administration continued to siphon tribute through
vestiges of the Inkaic order, leaving the collection of tribute to local kurakas (see Chapter
8).
Because tribute was assessed by population, tribute quotas based on the original
Toledan censuses became proportionately greater burdens on households in the face of

demographic decline, brought about by successive waves of European-introduced
epidemics and outmigration (Cook 1982). Undoubtedly faced with growing protests from
their subjects, the Collagua lords were vocal advocates for lowering their tribute levies, and
petitioned several times for population recounts.25 The resulting post-Toledan censuses, or
revisitas, represent one of the largest series for any single locale in the New World (Table
3.1). They are also among the most meticulous visitas known, rivaling modern censuses
25
There is also specific evidence within the visitas that suggests that they resulted from the dogged protests of
their leaders. For example, the 1604 revisitas were undertaken in response to a petition to the viceroy by the
protector de los naturales (almost certainly at the urging of the Collagua kurakas) because the population
could not produce their tribute levies in the face of widespread hunger and poor harvests, in part due to heavy
ashfall from the Huaynaputina eruption of 1600 (Cook 2003:xxiv). A crisis may have precipitated calls for
the previous visitas of 1591 as well, since epidemics of smallpox (sarampion) are mentioned in the 1591
Yanquecollaguas Urinsaya visita, and the head Franciscan priest of the province, Luis Gernimo de Or, was
present to register and confirm deaths declared by the villagers (Verdugo and Colmenares 1977 [1591]:f. 97r
[343]).
86
and cadastral surveys in their level of detail in recording household-level demographic and
economic information .
Table 3.1: Visitas to the Colca valley.
Those used in this project are indicated in boldface.
Repartimiento
(Provincial
Division)
Yanquecollaguas
Yanquecollaguas
Cabanaconde
Yanquecollaguas
Laricollaguas
Laricollaguas
Yanquecollaguas
Cabanaconde
Yanquecollaguas
Moiety
Hanansaya
Urinsaya
Hanansaya
Urinsaya
Hanansaya
Urinsaya
Hanansaya
Urinsaya
Hanansaya
Year
1591
1591
1596
1604
1604
1604
1615-1617
1645
1667
Completeness/Provenience*
Fragment, published in Pease 1977, APY
Large fragment, published in Pease 1977, MNH
Near complete, transcribed**, APY
Large fragment, transcribed**, APY
Large fragment, transcribed**, APY
Complete, published in Robinson 2003, APY
Large Fragment, transcribed**, APY
Fragment, not transcribed, APY
Callalli only, published in Galdos Rodrguez 1984,
ADA
*APY: Archivo Parroquial de Yanque, housed in the Archivo Arzobispal de Arequipa (Benavides 1990a,
1992); MNH: Museo Nacional de la Historia; ADA: Archivo Departamental de Arequipa.
**Transcription by Laura Gutirrez Arbul. Transcription funding provided primarily by W.M. Denevan,
with additional funding from M.A. Benavides and M. de Romaa (see Benavides 1990a).
The Colca valley visitas have attracted considerable scholarly attention. Noble
David Cook (1982) published a monograph outlining the demographic history of the
Collagua based on the visitas. His findings reveal that the population of the province was
declining at an average rate of about 1% a yearnot as precipitous a decline as in some
other regions, but an alarming rate nonetheless (Cook 1982).
By projecting the
demographic trajectory of the province back in time,26 Cook estimated that on the eve of
conquest the population of the province was in the range of 62,000-71,000 (Cook 1982:8488). The Collaguas were clearly a major demographic presence of a scale similar to that of
26
Specifically, Cook projects the rate of change between the counts from the visita general (1572) and the
revisitas of 1604 to arrive at a figure of 62,513 for the year 1530. However, this assumes that the rate of
demographic decline between 1532 and 1572 was similar to that between 1572 and 1604; but data from other
regions generally present a picture of steeper decline in this early period, so the 62,513 figure is minimal.
Using estimated rates of mortality for known epidemics during these first four decades following conquest,
Cook arrived at the second, higher preconquest population estimate of 71,000 (Cook 1982:83-84).
87
the Lupaqas and Qollas of the Titicaca basin, who numbered in the range of 100,000
inhabitants (Murra 1972). By the visita general, the population was therefore only about
half of its projected size before the European invasion.
Franklin Pease published the earliest visitas of Yanquecollaguas, from 1591, in an
edited volume in 1977 (Pease G. Y. 1977).
This volume contributed to the general
methodological shift in focus away from the chronicles and towards the use of more varied
administrative and ecclesiastical documentation in Andean ethnohistory through the 1960s
and 1970s, and particularly on visitas for reconstructing the protohistoric political and
economic organization of the myriad ethnic polities (seoros) of the former Inka realm.
Based primarily on analysis of the visitas, Pease and colleagues provided the first
ethnohistorical syntheses of the Collagua.
Mlaga Medina discussed how Collagua
political and economic organization articulated with Spanish colonial institutions (Mlaga
Medina 1977), and both he and Pease (1977) discussed the evidence for vertical
complementarity in the Collagua economy. Independent of the project headed by Pease,
Galdos Rodrguez was researching the Collagua in the Arequipa Departmental Archives
and published a fragment of a revisita from 1667 (Galdos Rodrguez 1984). Much like
Murras prototypical example of the Lupaqa (Murra 1972), these ethnohistorians presented
the Collagua as a classic example of a large south-central Andean chieftaincy whose central
mechanisms of political economy involved the redistribution of high altitude pastoralist
goods and lower-lying agricultural products via a vertical archipelago of territoriallydiscontinuous settlements and production enclaves (Cock Carrasco 1978; Galdos Rodrguez
1984; Mlaga Medina 1977; Pease G. Y. 1977). Pease and colleagues, as well as Galdos,
88
illustrated how the core Collagua population in the Colca valley constituted the nucleus of
an extensive settlement archipelago that included thousands of individuals living in
Collaguas islands throughout the puna and warm coastal valleys to the west and south
(Galdos Rodrguez 1984; Mlaga Medina 1977; Pease G. Y. 1977). These outlier colonies
also were recognized by the Spanish for their economic importance.
The grant of
encomienda of Yanquecollaguas from Francisco Pizarro to Gonzalo Pizarro specifically

included the mitimae27 Indians that they have distributed in lesser numbers in different
places (as quoted in Mlaga Medina 1977:112). Similarly, the encomienda of Cabana
in 1543 included mitimae populations in the Majes, Caman, Siguas, Vitor, and Arequipa
valleys (Mlaga Medina 1977:113).
Some of these outlier settlements in lower-lying
valleys were quite large. While they were not recorded in the post-Toledan revisitas, they
were included in the summary ledger (tasa) of the Collagua census undertaken as part of
Toledos visita general. The largest of Collagua colonist enclaves were located in the
valley of Arequipa, located 90 aerial km to the south of the Colca valley, where the tasa
registered 565 Collagua mitimaes in the reduccin of San Juan de Yanahuara, and 1,444
Laricollaguas mitimaes were registered in La Chimba and Tiabaya (Mlaga Medina
1977:113).
27
Mitimae is the Quechua word meaning, roughly, colonist, but also refers more specifically to the ethnic
colonists permanently resettled by the Inka from their home territories to other regions of Tawantinsuyu. As is
the case here, colonial authorities often used the term to refer also to ethnic outlier colonies which may or may
not have been moved by Inka authorities. Mitimaes resettled by the Inka were not state dependents; they were
moved to regions of similar ecology to their homelands, and allotted land for subsistence. They were also
ordered to retain their ethnic regalia, even though they only infrequently visited their ethnic homelands. This
institution appears to have been instituted as a means of preventing ethnic insurgency by fragmenting local
solidarity (see Murra 1980).
89
Subsequent to these pioneering studies, historical research on the Collagua and
Cabana has generally focused on one of the three subdivisions (repartimientos) of the
province.
Guillet (1992) and Gelles (2000) have included historical synopses of
Laricollaguas and Cabanaconde, respectively, in ethnographic monographs on modern

agricultural production in those areas of the valley.
Most recently, Robinson, 2003
#12017} has published the complete 1604 Laricollaguas Urinsaya visita along with an
extensive introductory analysis of the demographic and economic data in that census.
Maria Benavides, working in the early 1980s as ethnohistorian for the Ro Colca
Abandoned Terrace Project directed by William Denevan, published a series of synthetic
articles on the political and economic organization of Collagua (Benavides 1987b, 1988a,
1989, 1995) as well as more detailed studies on Yanquecollaguas, focusing on demography
and land tenure patterning in the village of Coporaque (Benavides 1986b, c, d, e, 1987a,
1989, 1990b). Using data from the 1591 and 1604 visitas of Yanquecollaguas Urinsaya and
the 1615-1617 visita of Yanquecollaguas Hanansaya, Benavides (1986e), and later Treacy
(1989; 1994), were the first researchers to reconstruct the locations of landholding
declarations in the visitas by mapping them according to their modern toponyms.
Benavides (1986e) found that moiety-level land tenure patterns were not spatially discrete
in Coporaque, but instead that Hanansaya and Urinsaya fields were intermixed, and
suggested that moiety-level organization, while apparently not mapped onto the landscape
in a literal sense, may have been most salient in water-distribution. Treacy (1989:250-259),
noting that the landholding declarations in the visitas appeared to be concentrated in highlyproductive areas of bench terracing, suggested that Coporaque villagers adapted to the
90
calamities of epidemics and uncertainties of colonial rule by minimizing risk rather than
maximizing production in the increasingly vacant landscape.
As in many Andean locales, the late colonial and early republican periods in the
province of Collaguas are not as well documented or understood as the early post-Toledan
period. The best records for the period between the late 17th and early 19th centuries are
from local parish archives, including the libros de fbrica which registers the economic
transactions of the local churches, some of the activities of the priests, nuns, and sacristans,
as well as the physical upkeep of the church and church lands. The parish archives also
include (somewhat patchy) birth, baptismal, marriage, and funeral registries of parishes, all
housed in the Yanque Parish Archive within the archdiocese archives of Arequipa
(Benavides 1988b, 1991, 1992).
While documentation is relatively sparse, the 17th and 18th centuries were
apparently difficult years of decline and hardship in the province.
Using the parish
registries, Cook estimated that the population of the valley reached nadir in the mid
eighteenth century at around 8000-10,000 inhabitantsa drop of 87 percent from the
projected 1530 (pre-Conquest) populationbefore slowly recovering, to eventually reach
32,000 in 1972, approximately the same size as four centuries earlier (Cook 1982:84-88).
During this period, the parcialidad-level authority of kurakas diminished as ecclesiastical
authorities sought to reign in their power. Such rivalry between Church and cacique claims
to authority is recorded in the 1689-1731 Libro de Fbrica of the parish of Yanque, which
reveals that during this time period, the church obliged kurakas to provide extensive
resources and labor to the parish and its priests (Benavides 1986a:519; 1986f:387; 1991).
91
As in other regions of the Andes (Saignes 1995), a system of parish confraternities had been
introduced that deflected the reciprocal obligations of the villagers from the kurakas to the
church (Benavides 1986a:519). These changes had a curdling effect on kuraka political
power and local constructs of community, as the parish within each village became linked
to political authority. It is also during this time period that kurakas are first seen referring
to ayllus as pueblos (Benavides 1997). Thus, two centuries after reduccin, Andean and
Spanish constructs of community were conflated. The collapse of native kuraka authority
continued in the wake of the reforms after the great Tupac Amaru rebellion of 1780-82,
when it was determined that governing officials of the pueblos would be named by Spanish
provincial authorities (Benavides 1986a:519).
Today, village identity is more salient than any sense of pan-ethnic Collagua
identity, although the villagers of the former areas of the Collaguas and Cabanas
essentialize their identities in other ways, and maintain different styles of elaborately
embroidered dresses and hats (Femenias 1997, 1998).
The primary irrigation canals
originally built during the LIP and Late Horizon continue to carry water to fields from the
snowcapped peaks surrounding the valley, and their trajectories cut across modern village
boundaries. As a result, conflict, at times violent, between villages over rights to water
sources for their canals have been frequent over the last century (Benavides 1997; Gelles
1990, 2000). In this way, the built landscape of the prehispanic past continues to assert
itself in opposition to the nucleated pattern of villages imposed by Toledo over four
centuries ago.
92
Chapter 4
Methodology and Research Operations
Introduction: An Archaeo-Historical Approach

In previous chapters I have argued that the integration of archaeological and
documentary data is critical not only for understanding trans-contact continuity and change,
but also for evaluating the appropriateness of ethnographic and historical analogies and
models, such as vertical complementarity, for more remote time periods in the Andes. In
this chapter, I outline the methodological architecture for a specifically spatial approach to
integrated archaeo-historical inquiry, as well as describe the specific research operations
involved in data collection and analysis for this project.
In a general sense, this investigation follows in a long tradition of combined
archaeological and ethnohistorical inquiry in the Peruvian Andes. Pioneering Peruvianist
archaeologists such as Uhle, Tello, and Valcarcel often used colonial texts, especially
published chronicles, both to formulate questions and to assign specific functions to
archaeological features and sites. While the use of written texts in early modern Peruvian
archaeology developed largely independently of the Direct Historical Approach in AngloAmerican archaeology, both traditions share the elementary logic of moving from the
known to the unknown (Steward 1942:337). Beyond this elementary logic, even the direct
historical approach itself was diverse in application and goals, and it never reached
coherency as a paradigm. Early attempts at the direct historical approach in North America
and Mexico (Bernal 1949, 1958, 1966; Kidder 1915, 1916; Kroeber 1916; Nelson 1916;
Parker 1916, 1922; Parker and Harrington 1922; Ritchie 1932, 1938; Steward 1942; Strong
93
1933; W. D. Strong 1933; Strong 1940; Wedel 1938, 1940) were generally hobbled by a
static, bounded view of culture, leaving them vulnerable to critiques of tautologyi.e., of
presuming the cultural continuity in question. With the advent of the New Archaeology,
and its deductive-nomothetic approach and emphasis on cross-cultural patterning of social
process, the direct historical approach, along with Boasian culture-historical archaeology in
general, fell out of favor in North American and Mesoamerican archaeology.
But processualism, with its emphasis on elucidating nomothetic laws of cultural
evolution and ecological adaptation, never predominated in Peruvian archaeology. From its
inception, modern Peruvianist archaeology was explicitly linked to the construction of an
autochthonous national identity, and the polemical discourses of historical versus
scientific archaeology that came to dominate debates in Anglo-American archaeology
were never salient in Peru (Burger 1989:37). Rather, the intellectual void left by the death
of Tello in 1947, a year after the publication of the seven volume Handbook of South
American Indians (itself a massive, Stewardian culture-historical tome), was filled by a new
generation of Peruvian and North American archaeologists such as Rebecca Carrin
Cachot, Jorge Muelle, John Rowe, Edward Lanning, and Richard Schaedel, all of whom
were strong advocates of an archaeology integrated with ethnohistory and ethnography
(Burger 1989:38). By the 1960swhile critiques of particularist, ideographic research
abounded elsewherePeruvianists were working to integrate new classes of documentary
sources and to devise more explicit, systematic methodologies for combining historical and
archaeological data than earlier ad hoc approaches. Following trends in archival research,
pioneering studies during the 1960s, such as the Hunuco Project, led by John Murra,
94
Donald Thompson, and Craig Morris, examined then under-utilized administrative
documents such as the visitas analyzed in this project as sources for hypotheses that could
be subjected to archaeological testing (see Murra and Morris 1976).
The hermeneutic of using documentary sources as the starting point, and the
archaeological record as a means of checking, fleshing out, or refuting textual
representations, remained the dominant approach until the present (e.g. Bauer 1998; Bauer
and Stanish 2001; C. J. Julien 1993; Murra and Morris 1976; Stanish 1989a, 1992). In
contrast to most direct historical approach studies in North America, which attempted to
match artifact assemblages to particular protohistoric groups and trace those assemblages
backward in time, most combined ethnohistorical and archaeological projects in the Andes
have attempted to identify aspects of prehispanic Andeanusually Inkasocial and
cultural institutions and practices as recorded in historical documents. These range from
specific case studies of particular aspects of Inka imperial infrastructurefor example, the
organization of Inka roads (Hyslop 1984), administrative centers (Morris 1972; Morris and
Thompson 1985), and storage systems (LeVine 1992)to the investigation of pan-Andean
social institutionssuch as the origins of the ayllu and its relationship to state-level
institutions (Isbell 1997).
However, this more explicitly social approach has produced its own set of problems.
Ironically, the initial success of investigations like the Hunuco Project in producing a
seemingly totalized view of past Andean social worlds resonated with broader intellectual
and popular discourses of indigenist identity and contributed to a hardening of theoretical
outlook and methodological approach. For example, as noted in Chapter 1, the ancient,
95
adaptive functions of vertical complementarity originally hypothesized by Murra came to
be widely accepted as social facts and were often invoked as interpretive shorthands to
explain archaeological distributions, rather than evaluated as hypotheses, as originally
intended (for cogent critiques of this approach, see Stanish 1989a; Stanish 1992; Van Buren
1993, 1996).
Again, textual representations and the hypotheses generated from their
analysis were sometimes used as post-facto corroboration, leading to circular reasoning and
hindering critical evaluation of what have become iconic features of lo andinothe set of
essential traits that are uniquely Andeanin Andean archaeology (cf. Isbell 1997).
Cognizant of these potential pitfalls, the methodological approach of this project
integrates two basic archaeological and ethnohistorical components in a common spatial
framework: 1) full-coverage archaeological survey and settlement pattern analysis and 2)
cartographic representation and analysis of landholding and livestock declarations from a
series of detailed colonial administrative surveys (visitas) from an area that spatially
overlaps with the archaeological survey. The way that these components articulate in this
dissertation differs from the dominant methodological approach in the Andes, as well as
most of the commonly-cited direct historical approach studies, in three major ways. First,
rather than starting with the written record as a source for generating hypotheses for
archaeological testing, this inquiry moves dialectically between the two as independent data
sources. Second, the study does not use documents to derive direct historical analogies
with which to interpret temporally-remote parts of the occupational sequence. Instead, it
seeks to define specific socio-structural homologies that can be compared back and forth
across the terminal prehispanic/early colonial divide. Third, while most direct historical
96
approach studies have been regional in spatial scale, this project is multi-scalar, combining
analysis of land tenure patterns at the local, regional, and macro-regional scales, with
micro-regional settlement pattern analysis. I describe the research operations involved in
each of these components in further detail below.
Overview of Archaeological Methodology

As discussed in Chapter 1, this project builds on recent interactionist theories of
community by approaching them as emergent social constructs rather than as discrete social
facts. The archaeological corollary of this approach is that communities are not clusters of
archaeological residues to be recorded, but social processes to be inferred. This has long
been a salient methodological theme amongst researchers of the late prehispanic Andes,
where widely-dispersed, multi-settlement concepts of community have been the rule, not
the exception (e.g. Blom, et al. 1998; Dillehay and Nez 1988; Goldstein 1993, 2000;
Moseley, Goldstein, et al. 1991; Stanish 1985, 1989a, b, 1992; Sutter 2000; Van Buren
1993; 1996, among many others).
The archaeological methodology of this project draws on these and other
community-scale studies that are working to fill a critical methodological gap between the
more established, micro-scale household and macro-scale regional settlement pattern
methodologies (Goldstein 1993, 2000; Joyce and Hendon 2000; Kolb and Snead 1997;
Marcus 2000; Yaeger 2000). These studies generally use a mixed regional and intensive
survey strategy at the scale of a micro-region, defined as larger than an individual site
but smaller than a settlement region (Yaeger and Canuto 2000:10). In this case, it was
97
clear that we would be able to survey only a small portion of the total Collagua settlement
system, but the project was situated in the area that, based on documentary evidence, was
home to the highest-ranking ayllus of the polity, and coincident with the area of bestdocumented in written texts (discussed below). Also, rather than focusing on a longitudinal
area within the valley rim (as most surveys are designed), the survey was designed to
sample a cross section the valley that included large areas of high altitude puna zones on
either side of the valley rim in order to begin assessing archaeological manifestations of
local-scale verticality relationships. The techniques used in the archaeological survey for
this project follow most of the established conventions for regional survey in semi-arid
environments (Adams 1966; Billman 1996; Blanton, et al. 1982; Parsons, et al. 2001;
Sanders, et al. 1979; Wilson 1988), but with more detailed data collection at sites with
architectural remains. The excellent architectural preservation at most late prehispanic sites
in the Colca afford the opportunity to examine intra- and inter-site differences in social
status and identity. These initial architectural observations also lay the groundwork for a
larger regional-scale study of Collagua verticality. As demonstrated by Stanish in the upper
reaches of the Moquegua drainage (Stanish 1989a, b, 1992), domestic architectural style
can serve as a strong indicator for identifying outlier ethnic colonies, and the distinctiveness
of Collagua architecture holds promise for a similar approach in neighboring valleys.
Thus, the methodological goal for the archaeological portion of the project was to
obtain intermediate scale data resolution that would provide more detail on intra- and intersite variability than traditional regional settlement methods, which tend to treat sites as
functionally-defined units in regional settlement hierarchies, while at the same time
98
covering a broad enough area to provide a view of how community organization changed at
the supra-settlement, but sub-polity, scale.
Selection of Project Area

While archaeological survey could be paired with colonial land-tenure analysis in
many locales in the Andes, the area surrounding the villages of Yanque and Coporaque in
the Colca valley offered especially good potential, as discussed in Chapter 2. This area was
the primary center of the prehispanic Collagua polity, Yanque being the capital of the
colonial Collaguas province. Reconnaissance and previous research revealed a high density
of archaeological sites in the area, including large habitational sites with standing
architecture (Brooks 1998; Malpass 1987; Neira Avendao 1961; Treacy 1989). Also, in
part because of its role as provincial capital, the Yanque area is the best historicallydocumented part of the valley. The visitas to the Colca valley constitute one of the most
extensive series for any locale in the Andes, and Yanque and Coporaque appear in these
visitas more than any other part of the valley (see below). Lastly, the area surrounding
these two villages was the focus of the most previous research in the valley (Benavides
1987b; Cock Carrasco 1976-77; Denevan 1986b; Denevan, et al. 1987; Malpass and de la
Vera Cruz Chvez 1990; Pease G. Y. 1977; Treacy 1989, 1994), which provided a solid
empirical foundation for beginning a long-term research program.
The specific survey area was designed as a broad transect across the kichwa (3350
and 3600 masl in the survey area), suni (3600-3900 masl), and puna (3900-4900 masl)
zones. Rather than using the valley rim as a natural boundary, as is often the case in
99
highland valley surveys, I extended coverage over the rim on both sides of the valley in
order to record and analyze sites with different production foci and thus examine the
relationships among them.
The eastern and western survey limits correspond to the
territorial limits of the modern districts of Yanque and Coporaque. Although the survey
area covers only a small portion of the Collagua polity, it encompasses the settlements
within the core area of the highest ranked communities in the colonial province of the
Collaguas. Time and budgetary constraints ultimately limited the survey size. I conceive
of this survey as the first phase of a long-term valley-wide survey and excavation plan.
Archaeological Surveying Techniques in a Montaine Landscape

The natural and built features of the Colca valley landscape present many
opportunities and challenges for systematic archaeological survey. The semi-arid climate
of the valley affords very good ground visibility, especially during the dry season (May to
October) when we conducted the survey. Valley topography and various built landscape
features, however, sometimes hindered strict adherence to survey transects. We adapted a
number of techniques to deal with the obstacles presented by the valleys high topographic
relief, extensive terracing, and field walls.
Such obstacles were especially dense inside the valley rim. We generally surveyed
this core zone of the survey area using transects spaced between 5 and 20 m. The specific
distance varied according to local conditions and considerations. On the relatively flat
alluvial terraces of the valley floor where agricultural field sizes are generally large, the
crew was distributed at 10 m transect intervals in unharvested fields because of reduced
100
ground visibility, and at 20 m transect intervals in harvested fields where surveyors could
readily see the ground surface. Surveyors would advance from field to field, crossing over
the stone walls that divided them. However, the bench and contour terraces of the steep
lower slopes of the valley made strict transect surveying virtually impossible. To ensure
full coverage in these areas, each crew member surveyed a terrace group, walking a single
transect across each terrace unless a terrace exceeded approximately 10 m in width. Canals
and dividing walls formed natural endpoints for each transect, and the crew descended and
ascended terraces in a zig-zag fashion, using canals or dividing walls as boundaries between
terrace groups. Given the intensity of the survey, the density of sites, and the difficulties of
the terrain, average daily survey coverage in the agricultural core of the valley was usually
0.3-0.5 km2 per day.
In contrast, the open expanses of the suni and puna zones allowed for more
regularized use of transects. Because of its distance from modern settlements, we surveyed
most of the puna during several 4-5 day expeditions, using backpacks and camping
equipment. Good ground visibility and the scarcity of built features allowed for wider
transect intervals, ranging from 25 to 30 m, depending on local topography. The slow
pedogenic processes and good ground visibility in the puna permit more ready observation
of features, even relatively ephemeral soil stains and artifact scatters. For example, we
recorded discrete concentrations of lithic debitage of less than a 10 m2. It is possible that
we did not register some scatters of this small size using 25 to 30 m transect intervals.
Given the objectives of the project, I considered these potential oversights acceptable, since
all sites greater than about 10 x 10 m would be observable, and the wider transect intervals
101
allowed for much greater total coverage. In the puna, daily survey coverage averaged
approximately 3 km2an acceptable compromise for balancing the overall scale and
resolution of the survey, and comparable to other survey projects in the nearby lacustrine
and altiplano zones (Bandy 2001; Seddon 1998; Stanish 1997a).
In all areas, I encouraged crew members to leave their transects if they came upon
potential features or artifact concentrations between transects, returning to the transect after
investigating and recording the sites. Sites were defined as significant concentrations of
ceramic, lithic, or other cultural features in a discrete area. In the case of sherd and lithic
scatters, we did not employ formal artifact density thresholds as site definition criteria, but
rather based our decision on our shared judgment that there was a distinct increase in
density of distinguishable extent in contrast to the background noise of low-density
artifact scattering.
Although this approach is judgmental, with coordination between
members of the field crew, our site inclusion criteria remained consistent throughout the
project. In the field, I also used a 100 meter rule in order to maintain a standard for
including features in a site. That is, I registered features or clusters of features more than
100 m from any other features as different sites, and those within 100 m as part of the same
site. While arbitrary, this rule helped standardize the definition of sites and prevented
overly impressionistic site definition in the field.
102
Site Nomenclature
Site names were assigned using the first two letters of the modern district (YA for
Yanque, and CO for Coporaque) followed by the site number (001-169).28 Thus, for
example, YA-032 was the 32nd site recorded in the district of Yanque during the survey. In
cases where I divided sites into more than one sector (whether based on functional,
temporal, or arbitrary criteria), we assigned a capital letter (A, B, C,) to each sector.
Once we identified a site, the crew would rejoined to record observations of the site and its
constituent features on standardized forms, photograph any features of interest, map the
sites perimeter and features, and collect a representative sample of diagnostic artifacts
from the surface. I describe each of these operations in further detail below.
Field Mapping Techniques

We used a handheld GPS receiver (Garmin 12) for macro-level mapping functions.
These functions included determining daily survey coverage areas, site locations (using the
approximate center of a site), and internal divisions of large sites (e.g., YA-014). I always
used position averaging, so that each point represents the average value of least one
hundred GPS position readings. The generally low-angle horizon visibility allowed for
position accuracies of 20 m in most areas. This margin of error is much lower than that
of hand-plotting site locations on the available topographic maps. Using the GPS to delimit
28
In the field and initial report to the Peruvian National Institute of Culture (Wernke 2001a), sites were also
assigned a two letter prefix code for the Province of CayllomaCA. Thus, site one, located in Yanque, was
recorded as CA-YA-001. Since all sites in the survey are in Caylloma, I have left this prefix out in the
dissertation, although it may be used again in future regional-scale studies.
103
daily coverage areas also allowed for monitoring of coverage during fieldwork by
importing the points into Autocad. I later imported the GPS points into ArcGIS 8.1.
At the site scale, various mapping techniques were used, depending on the nature of
the site and the time and resource constraints of the project. The least precise method, used
for sites of less than 0.25 ha in area, was pacing a sites perimeter as a square or rectangular
area. At larger sites composed only of artifact scatters, we mapped the perimeter polygon
using a Brunton pocket transit and pacing or measuring tape. Finally, at selected sites with
well-preserved architecture, we created plan view maps using the pocket transit and tape.
In the cases of YA-050, YA-054, and YA-094, these maps include all visible standing
architecture. In other cases (e.g., CO-100, CO-163), maps of particular structures or groups
of structures were completed. We also completed sketch maps of site layout and features at
several sites.
Site Sizes
Determining the sizes of most LIP and Late Horizon sites with standing architecture
was relatively straightforward. At these sites, we recorded GPS points around the maximal
area with cultural features. In a few cases (e.g., CO-100, CO-164), this total site area,
which could include mortuary features or artifact concentrations, significantly exceeded the
area occupied by standing domestic architecture, and in these cases, I plotted a second,
smaller polygon for residential area, defined by the outermost domestic structures. The
reduccin villages of Coporaque and Yanque presented special problems because even the
late prehispanic features of these sites have been virtually obliterated.
Furthermore,
104
because it was not feasible to gain access to most house patios, we restricted our survey in
the villages to the unpaved streets. However, the formal grid pattern facilitated division of
the villages into arbitrary collection sectors, which provide a degree of provenience
resolution (Chapter 7).
Site abandonment and construction processes have displaced
artifacts from their original depositional contexts, so strict presence/absence criteria were
not rigidly applied for inclusion of a given sector within a prehispanic occupation. In both
cases, a few isolated structures made with Collagua and Inka-style cutstone masonry also
aided in estimating the site area. Size estimates for these sites is more subjective, relying
on the distribution of structures made of prehispanic building materials and areas with the
highest prehispanic artifact concentrations.
Site sizes could not be determined at most Formative and Middle Horizon Period
settlements because of two factors. First, long-term continuity of occupation at most
settlements means that domestic terracing, agricultural terracing, and standing architecture
from later periods (see discussion below) obscures much of these earlier occupations.
Second, partly related to the first factor, but also in part to the intensity of occupation,
surficial Formative and Middle Horizon Period artifact densities are very low at most sites.
These conditions contrast with, for example, sites in the Titicaca Basin, where Formative
occupations generally were more dense and allow for reliable delineation of componentspecific site extents.
In this survey, fewer than ten Formative Period artifacts were
recovered from most agricultural settlements.
Designation of component-specific
boundaries with such scant collections would have been overly impressionistic and
arbitrary, and so site area data are not available for most settlements from these periods.
105
Site Classification: Chronology

The settlement pattern data presented in this dissertation are divided into ten
occupational period. Table 4.1 presents the dates for these periods, but given the scant
radiometric controls for the local ceramic sequence, the bracketing of dates for all periods
should be considered rough approximations. I base the Archaic Period duration on the most
recent regional projectile point chronology (Klink and Aldenderfer in press). Previous
researchers did not identify local Formative Period diagnostics, and I define this component
by the presence of Chiquero ceramics, an undecorated local ware that shares similarities in
vessel form and surface treatment with other Formative styles in the Department of
Arequipa and the region.29 In the absence of further temporal controls, I consider Chiquero
ceramics to be broadly diagnostic of the Formative Period (1500 BC - AD 500). While this
tentative placement spans 2000 years, future excavations should permit more precise
chronological assignment and perhaps stylistic subdivision of Chiquero ceramics.
Likewise, the local Middle Horizon component cannot be subdivided at present and is
chronologically situated according to its traditional range in the pan-Andean chronology of
AD 500-1000 (Rowe 1962). As I discuss in Chapter 3 and Appendix A, my ceramic
sequence differentiates LIP (Collagua I and II) and Late Horizon (Collagua III, Collagua
Inka, and Inka) ceramics, which have been bracketed by their standard dates in the Titicaca
29
As is standard practice in the south central Andean highlands, I use a dual time period terminology that
combines the standard horizon/intermediate period sequence devised by Rowe and Menzel with a regional
developmental sequence developed by Lumbreras. The Lumbreras nomenclature is used for periods prior to
the Middle Horizon because diagnostics differ fundamentally from those used for the Initial, Early Horizon,
and Early Intermediate Periods in the Central Andes. Also, the terms Preceramic and Archaic are used
interchangeably (thus, Early Preceramic refers to the same time period as the Early Archaic Period, etc).
106
Basin chronology (see Stanish 2003). Ironically, Colonial Period ceramics remain largely
undifferentiated as a temporal component, despite some obvious technological, formal, and
decorative changes (Rice 1997). We recovered some examples of probable Early Colonial
transitional wares, but excavations and radiometric evidence are needed to evaluate this
hypothesis. By contrast, some glazed wares and imported porcelains are undoubtedly much
more recent, and have been grouped within a Late Colonial/Republican component.
Table 4.1: Local Sequence of Occupational Periods
Period
Dates
Early Archaic
8800 BC - 6800 BC
Middle Archaic
6800 BC - 4800 BC
Late Archaic
4800 BC - 3000 BC
Terminal Archaic
3000 BC - 1500 BC
Formative
1500 BC - AD 500
Middle Horizon
AD 500 - 1000
Late Intermediate
AD 1000 - 1450
Late Horizon
AD 1450 - 1532
Colonial
AD 1532 - 1821
Republican
AD 1821 - 2000
Artifact Collection Strategy

The basic objective for artifact collection at most sites was to obtain a sample of
diagnostic artifacts from the surface that could be used to date sites. We collected artifacts
in minimal units called lots, (lotes). Lots consisted of collection bags containing artifacts
of the same material (pottery, lithics, textiles, metals, etc.) from a single defined context,
which was defined by depositional context.
Thus, surface collections were bagged
separately from collections from wall fill at sites with architecture. Small sites of only one
sectorfor example, a small undifferentiated lithic scatterusually consisted of a single
lot for each artifact material class. At sites with multiple sectors, whether arbitrary or based
on inferred functional differences, we collected separate lots for each depositional context
107
and material class for each sector. In total, we collected 617 lots with a total of 3806
ceramic sherds, 1673 lithic artifacts, 136 textile artifacts, and two metal artifacts.
The grab bag approach I applied, although not suitable for analyzing intra-site
distributions in any detail, is widely used as a suitable compromise between scale and
representationthat is, as a means of obtaining a general view of inter-site spatio-temporal
variability in artifact distributions (Adams 1966; Blanton 1978; Blanton, et al. 1981, 1982;
Parsons, et al. 2001; Stanish 1997a; Wilson 1988). It is especially well-suited to contexts
such as the Colca valley, where the surface is thoroughly disturbed by cultivation and
terracing. With consistent collection criteria among team members, this method produced a
reliable database for the relative dating of sites and permitted coverage of a much larger
total area than would be possible using more rigid collection techniques.
Survey Area Coverage and Total Sites Recorded

Using the methods described above, between May and December of 1999, our three
person crew in the Colca Valley Regional Survey Project surveyed a 90.4 km2 contiguous
area between 3400 and 4800 masl across the valley to the surrounding puna on either side.
We recorded 169 archaeological sites with 300 temporal components during the survey
(Figure 4.1; Appendix F).
The sites cover virtually the entire sequence of human
occupation in the valley, from Early Archaic Period lithic scatters through late prehispanic
and colonial habitational sites with well-preserved architectural remains.
108
Figure 4.1: Survey area and all sites registered (N=169)
109
Basemap Source Data
I used four principal data sources in combination for the cartographic tasks involved
in this project: 1) Ministry of Agriculture cadastral maps (1:5000), in conjunction with
larger scale (1:25,000) Ministry of Agriculture maps that show important toponyms and
modern district boundaries), 2) 1:100,000 Instituto Geogrfico Nacional del Per (IGN)
topographic quadrats for the area (Hoja 32-s, Chivay), 3) 1:17,000 scale airphotos from the
Peruvian Air Force (Servicio Aerofotogrfico Nacional [SAN] 1974),30 and 4) satellite
imagery and a digital elevation model (DEM) from the ASTER (Advanced Spaceborne
Thermal Emission and Reflection Radiometer) remote sensing instrument of the Terra
Satellite, launched in 1999 as part of NASAs Earth Observing System (EOS)
(http://asterweb.jpl.nasa.gov/).31
A Digital Elevation Model (DEM) generated from ASTER data formed the
foundation for the base map.32 A DEM represents continuous topographic surfaces by
recording elevations in the regular grid of a raster data model. Each pixel in the raster grid
of a DEM is assigned a Z (elevation) value. A relative DEM records relative differences in
elevation without geospatial references, while an absolute DEM is calibrated using ground
30
Project No. 227-73, taken on March 30, 1974. Frame numbers 2534-2542, 2557-2568, 2574-2576, 25992602, and 2614 (29 total frames). All were scanned on a drum scanner at 800 dpi and saved in tif format
(grayscale). Three frames (2653, 2576 and 2577) in the center of the survey area were also scanned at high
resolution (20 micron resolution) directly from the negatives at SAN. These high resolution scans provided
remarkable 49 centimeter ground pixel resolution when georectified.
31
The ASTER Level 1A imagery used provides 15 ground meter pixel resolution. Level 1A data record
imagery in the Very Near Infrared (VNIR) electromagnetic spectra in three bands. Band three is slightly
longer than visual red, or Very Near Infrared (VNIR). The VNIR bands also contain information about
photosynthesizing plant cover because healthy plant foliage absorbs most of the visible light that hits it while
reflecting most of the VNIR wavelengths. Thus, areas with photosynthesizing plant coverage appear as bright
red areas in the images.
32
I thank Nicholas Tripcevich for providing me with a copy of the DEM.
110
control points, thus positioning the DEM in space according to a geographic coordinate
system. In this case, an absolute DEM recording the central and upper portions of the
Colca valley and surrounding uplands was created from ASTER Level 1A imagery at the
Jet Propulsion Laboratory using eight ground control points collected by Nicholas
Tripcevich of the University of California, Santa Barbara. Tripcevich collected each of the
ground control points with a handheld GPS device (Trimble Geoexplorer II), using the
mean X,Y values of 120 readings for each of the points, which he then post-processed to
arrive at DGPS accuracies of 1-2 m (N. Tripcevich pers. comm. 2002).33 The resulting
absolute DEM provides 30 meter ground pixel resolution (i.e. each pixel represents the
mean elevation of a 30 meter square on the ground). I generated contour isopleths and
hillshade modeling from this DEM using ESRI Spatial Analyst and 3D Analyst.
georeferenced all other airphotos and rasterized maps in ArcGIS, using Ministry of
Agriculture maps, IGN maps, and GPS ground control points from the survey as references.
While coordinate systems on the original Peruvian IGN and Ministry of Agriculture maps
are based on the Peruvian Provisional 1956 datum, all were subsequently transformed to the
more standard WGS 1984 datum. Therefore, all coordinates provided in this dissertation
are based on the WGS 1984 datum.
After importing the GPS point data in the GIS, I created polygon features for those
sites with perimeter GPS points. Second, in order to most accurately represent the location
of those large sites, I generated centroids (the geometric center of the polygon) from each of
their polygons. This centroid point theme was then merged with the remaining small sites
33
DGPS post-processing was done via Trimble Terrasync software, and based on data from the Arequipa IGS
(International GPS Service) station (N. Tripcevich pers. comm. 2002). DGPS data from the Arequipa station
are available on the internet at http://igscb.jpl.nasa.gov/network/site/areq.html
111
that were located in the field with single GPS points, resulting in a basemap with one point
for each site (Figure 4.1). Attribute data included site ID number (as described above), and
site name. Site numbers function as the primary index for linking the point and polygon
themes with the tabular data in the survey database, such as the ceramic chronology, to
arrive at the settlement pattern maps for each time period.
Early Colonial Land Tenure Analysis: Overview of the Documentary Sample

As discussed previously, the Colca valley visitas constitute one of the largest series
of colonial administrative surveys for any locale in the New World, and the villages in the
archaeological survey area, Yanque and Coporaque, are the best documented in the series.
These villages appear in the visitas of Yanquecollaguas Urinsaya 1591 (Verdugo and
Colmenares 1977 [1591]), Yanquecollaguas Hanansaya 1591, Yanquecollaguas Urinsaya
1604, and Yanquecollaguas Hanansaya 1615-1617 (Pease G. Y. 1977:407-452). Only a
small fragment of the 1591 Hanansaya visita remains.34 The other three documents are
also incomplete, but the four in combination provide a view of both moieties of
Yanquecollaguas between 1591 and 1617.35 They complement each other in the sense that
34
35
The original is housed in the Museo Nacional de la Historia, Lima, and lacks folio numbers.
The 1591 Visita de Yanquecollaguas Urinsaya was conducted during the tenure of two corregidores,
beginning with Gaspar Berdugo (fs.1-96v), and completed, starting with Coporaque, by Gaspar de
Colmenares. It is transcribed and published in Pease, ed. 1977 (191-406). It is the smaller and more
deteriorated fragment of the two, preserved from fs. 1r-119v, fs. 120r-155v are fragmentary and illegible. The
original is housed in the Museo Nacional de Historia (MNH), Lima. The 1604 Visita de Yanquecollaguas
Urinsaya was conducted by the corregidor Licenciate Juan de Rivero. It is well-preserved from ff. 53r-413v.
The document is unpublished, and housed in the Archivo Parroquial de Yanque (APY) within the Archivo
Arzobispal de Arequipa (AAA). The 1615-1617 Visita de Yanquecollaguas Hanansaya was conducted by the
Corregidor (and Visitador) Capitn Jernimo de Pamanes. It is well-preserved from ff. 303r-643v (some
folios missing), and is also archived in APY. Warm thanks to Maria Benavides for providing transcriptions
112
the 1591 Urinsaya visita, while only a fragment, includes the opening portion lacking in the
1604 Urinsaya visita. This is an important section which includes Yanque, the colonial
provincial capital. Therefore, in the land tenure analysis presented in Chapters 8 and 9, I
have replaced the incomplete Yanque section in 1604 with its complete counterpart from
1591. Given the relatively short span between the two (13 years), this approach provides an
accurate composite representation. Likewise, the Coporaque section of the 1591 Urinsaya
visita is incomplete, and therefore I use the 1604 census for Coporaque. Finally, since only
a small fragment the 1591 Hanansaya remains, I use the 1615-1617 visita for most of
Hanansaya, again with the exception of Yanque, since this section is lacking in the 16151617 census, but is partially documented in the 1591 Hanansaya visita.
Thus, the
combined 1591/1604 Urinsaya and 1615-1617 Hanansaya data together provide the most
complete possible composite synchronic representation available for the communities of
Yanquecollaguas.
The visitas cover nine reduccin villages in the central and upper valley (moving
down-valley from the puna): Tisco, Sibayo, Callalli, Tuti, Canocota, Chivay, Coporaque,
Yanque, and Achoma.36 The census surveys were conducted in the central plaza of each
reduccin. The census role was organized by ayllu, in descending order of ayllu rank.
Within each ayllu, tributary householdsthose with male heads of household between the
ages of 18 and 50were recorded first, followed by widowers, single men of tributary age
(by Laura Gutirrez Arbul) and copies of the original archival documents. I have cross-checked the
transcriptions against photocopies of the originals.
36
The 1615-1617 Yanquecollaguas Hanansaya visita also records data from an ayllu of silversmiths in the
village of Maca, located within the repartimiento of Laricollaguas, who were subjects of the provincial
paramount in Yanque, as well as outlier ayllus in the villages of Huanca and Lluta, located about 55 aerial km
to the south at the headwaters of the Siguas river (see Chapter 8).
113
(referred to simply as indios or indios tributarios), orphan boys (i.e., less than 18 years old),
orphan girls, widows and single women, elderly (i.e., more than 50 yrs old) and disabled
men (viejos e intiles), the rabble of boys (chusma de muchachosi.e., those without
guardians), and the rabble of girls (chusma de muchachas). The ayllu leaders and the
visitador ordered each household head to declare their family members and landholdings to
the visitador, accompanied by the local priest and magistrates.
The visitas to the Colca valley are quite detailed by any measure, and especially so
for post-Toledan times, when the crush of petitions for population recounts by ethnic lords
and colonial magistrates, for opposing motivations, created a flood of often hastily drawnup censuses (Guevara-Gil and Salomon 1994; Stern 1982a:114-137). They recorded both
demographic and cadastral information. The sex, age, and civil status of each family
member was recorded by the scribe through an interpreter, followed by household
landholdings and livestock. Landholdings were located by toponym, and the predominant
crop grown was also noted. Field sizes were declared most commonly using the Andean
measure of topo (see below).
Under Toledan and post-Toledan administration, tax levies were calculated directly
from the demographic data recorded in the visitas, and the visitadores were acutely aware
that the communities interests ran counter to their own. All members of the community
were to present themselves in the central plaza of each village. The decree authorizing the
visitas was published in advance and announced by the town crier (pregonero pblico),
who gave an explicit warning against hiding individuals or households from the counts.
The visitador rewarded some individuals with an exemption from tax levies who revealed
114
other community members attempting to hide or escape the census.37 The visitadores also
used a variety of checks to verify the veracity of declarations. First, it is evident that they
brought along previous visitas and they consulted them alongside the creation of the new
one. Thus, for example, marginal notes beside household declarations in the 1604 visita
indicate that it was used to check against testimony in the 1615-1617 visita.38 The stated
age and civil status of each household member was also checked against local parish
baptismal, marriage, and death registers brought forth by the local priest.39 The visitador
also summoned and consulted native accountants with quipus. The visitador even called
for the quipucamayocs (keepers and readers of quipu) to present themselves in the 16151617 Hanansaya visita, revealing that this native Andean system of accounting was still
deemed accurate enough to be worthy of consultation by colonial authorities nearly a
century after conquest.40 In the 1604 visita of Laricollaguas Urinsaya, the visitador also
37
For example, APY Yanquecollaguas Urinsaya 1604: f. 70r/v: ...Domingo Ayqui Rastrollo, su hijo ligtimo,
de edad de diez y nueve aos; este yndio esta enfermo y es yntil, y en premio de aver manifestado cantidad
de yndios diferentes pueblos y ayllos de esta provinia, como consta por este padrn, se reserva de trivuto y
tassa y serviios personales y mitas, serviios personales y mitas
38
For example, men who in 1604 were in their early 40s (and thus of tributary age) and listed as tributaries in
the main body of the text have marginal notes saying viejo (i.e., greater than 50 years old cutoff for
tributaries), indicating that the note was made after the person turned 50. Also, many individuals listed in the
1604 and 1615-1617 visitas have a marginal note muerto (dead), which indicates that the visitador was
checking against these records looking for a person who was not present or unaccounted.
39
For example, in the opening declaration of the 1591 Urinsaya visita, it is stated that the priest of Coporaque,
Fray Luis de Or, had brought forth the funeral registry to assist in the proper accounting of deaths caused by
recent epidemics of smallpox and measles . None of the baptismal, marriage, and death ecclesiastical
registers prior to the mid 17th century have been identified (For an inventory of the Yanque Parish Archive,
see Benavides 1988b).
40
The passage mentioning quipucamayocs comes at the beginning of a long preamble dictated by the
visitador in the villages of Tisco and Tuti, announcing the scheduling of the visita around the Easter holiday
and other procedural issues: The first example reads: [left margin: Pueblo de Tisco] El capitn Gernimo
de Pamanes, corregidor i justiia maior desta provinia de los Collaguas por su Magestad, hago saver a los
caziques i governadores, pachacuracas i quipucamaios i alcaldes del pueblo de Tute...se a de ir a la conclusin
della para pasado el segundo da de pascua... (APY Yanquecollaguas Hanansaya 1615-1617: f.338v).
115
called upon the quipucamayoc at the conclusion of the census to ensure that no tributaries
had been obscured from the accounting, and indeed they did find a few errors and
corrections were made (Cook 2003:xxix).
The Visitas as Text: Issues of Representation

While the visitas to the Colca valley are among the most detailed of any known in
the colonial Andes, I do not assume that they represent transparent or unselfconscious
registries of past social processes. I do not subscribe to a documentary realism (La Capra
1985; or documentary idolatrysee Pease G. Y. 1976-77:207, as cited in Guevara-Gil
and Salomon 1985:7) that privileges visitas and other administrative documents as
reflective of the efficient, technocratic aspects of state institutions, immune to the
Eurocentric assumptions and ideological overlays present in the chronicles The very need
for visitas (literally, a visit) at once tacitly acknowledged social realms beyond crown
control and the need for reasserting sovereignty and civitas according to state and church
canons (Guevara-Gil and Salomon 1994). The reasons for their frequency throughout the
Spanish colonies were both structural and practical, and all politically-charged.
Ideologically, there was pressure for frequent administrative survey as part of the drive for
buen gobierno ("good government"an omniscient and well-attuned administration) and
policia (social order) (Guevara-Gil and Salomon 1994). There was also constant political
pressure from below for new surveys, and from both sides of the colonial divide: from local
colonial agents who sought to raise revenue, from native lords who insistently petitioned
colonial authorities for a recount that would lower tribute burdens in the face of
116
demographic collapse, and from internal disputes between encomenderos and corregidores
(Guevara-Gil and Salomon 1994; Stern 1982a, b). In the act of their creation, visitas were
part and parcel to making history as much as they were to recording it. As Guevara-Gil and
Salomon (1994:4) note, rather than diligently registering what was, ...visita authors
defined as real a highly artificial rendering of traditional order, recorded a public event
in which actual humans were arrayed in its image, and set a normative model against which
behavior would (until the next visit) be judged as a defective enactment. In other words, I
am mindful of the complex negotiation of expectation and experience (Whitehead 1995)
played out in the orchestration and recording of a personal visit by the Spanish emperors
emissary and his entourage (Guevara-Gil and Salomon 1994).
The compelling ideological, political, and economic pressures from above and
below, as well as the machinations of number fudging inherent in the creation of visitas,
clearly make a documentary realist reading of visita documents untenable.
But this
observation need not lead to the textually-hermetic conclusion that they are useful only for
understanding how apparently native representations were shot through with European
cultural projections, or how the act of their production reified Spanish colonial
domination.41 Distinguishing expectation from experience in these texts requires both close
evaluation of both intra- and inter-textual representations, as well as going outside the
written texts themselves.42 As discussed previously, it was only after detailed scrutiny of
41
Whitehead (1997; 1995) has made this point a fortiori by illustrating how native genres and realities can be
recovered from even the most seemingly fantastic and obvious of European projections, such as the
chronicle of Sir Walter Raleigh
42
The same can be said of the texts of the archaeological record. That is, other sources (written historical
texts, oral history, ethnographic literature, etc) are not only beneficial but also required in the interpretation of
117
the documentary evidence and integration of archaeological research that the hypothesized
existence of a pre-Inka Lupaqa vertical archipelago was called into question. This project
was designed with these concerns in mind, moving beyond the written texts by comparing
the archaeological manifestations of late prehispanic social organization with the
representation of their vestiges in the visitas and by using modern toponym data to
reconstruct land tenure patterning. I describe my methods for the latter of these two
components below.
GIS Cartographic Representation and Database Management of Documentary Data

Toponym Mapping
Mapping colonial toponym data at a regional scale is already de rigueur in Inka and
early colonial research in the Andes (see, e.g.,Grosboll 1987; Julien 1983; C. J. Julien 1993;
D. G. Julien 1993; Morris and Thompson 1985; Murra 1972; Prssinen 1992). However,
this project integrates analysis of colonial land tenure patterns at multiple spatial scales,
mapping local to regional distributions of landholdings declared by Collagua villagers. I
have facilitated this multiscalar approach through the use of GIS. The types of data
recorded in visitas lend themselves to the tabular form of a database, and their standardized
format made database design and data entry relatively straightforward tasks. In the GIS I
created for this project, each village or toponym in the map is linked to attribute data in a
relational database composed of tables with the demographic and cadastral data recorded in
the visitas. Thus, data from the visitas could be sorted and queried in the database and then
archaeological patterning and break an exclusive (and philosophically untenable) dependence on them. This
essential point was made by Binford in early calls for the development of middle range theory in archaeology
(e.g. Binford 1967).
118
viewed as distributions on the GIS map. At the regional scale, I mapped landholding
declarations in the visitas by settlement. At the local scale, I mapped out the locations of
individual fields by reference to the toponym used to locate them in the visitas. I have
limited my local-scale analysis to the village of Coporaque, which has the most welldocumented modern toponym mosaic in the valley.
Due to their continued economic and cosmological significance, toponyms have
proven extremely historically durable in this and most other areas of the highlands. This
continuity permits many of the fields claimed in the visitas to be located quite precisely by
matching them with their modern toponym areas. The unique specificity with which fields
are located in the Colca valley visitas reflects the link between fields and the irrigation
network that continues to modern times (Robinson 2003:lxv). Today fields in the Colca
valley are located by reference to the dendritic branching of the canal network at two scales
of specificity: first, by irrigation sectorthe general area of fields irrigated by a primary
irrigation canal, and secondly, by irrigation clusterthe smallest group of fields irrigated
by a distribution canal at the distal end of the canal network (Guillet 1987; Robinson
2003:lxv). Thus, each small cluster of fields irrigated by a distal-end distribution canal has
a name; these names continue to be important today as spatio-hydraulic references for
coordinating water apportionment at irrigation scheduling meetings (see Gelles 2000;
Treacy 1989; Valderrama and Escalante 1988). Almost certainly for the same reason, every
field was specifically identified by toponym in the visitas as well. Most of the toponyms
used to locate fields in the visitas appear to represent the more specific, cluster-level of
fields as their modern counterparts. The ancient origins of contemporary toponyms in
119
Coporaque are clearly attested by the fact that most are Aymara words, the language spoken
by the Collaguas in late prehispanic and early colonial times, but since superseded
probably through early colonial missionary effortsby Quechua.
In Coporaque, four modern toponym surveys were conducted by Crdova and
colleagues (Crdova Aguilar, et al. 1986), Treacy (Treacy 1989, 1994), Izaguirre
(Benavides 1986e; Izaguirre Urbano n.d.), and by the Peruvian Ministry of Agriculture as
part of a general cadastral survey for the Agrarian Reform of the early 1970s. I also
recorded toponym data as part of my archaeological survey through opportunistic
interviews with farmers as we encountered them in their fields and during follow-up
verification of some toponym areas during April, 2000, and July, 2002. Of the four
sources, the Ministry of Agriculture cadastral survey provides the broadest coverage
because they surveyed every village in the valley, but the level of toponymic detail in these
maps tends to be rather coarsethese maps tend to lump irrigation clusters together under
single names.43 The toponym mosaic maps from the other three ethnographic sources
provide more fine-grained coverage, and these generally were the most useful for locating
lands claimed by Coporaque villagers in the visitas that today lie within the modern
community boundaries.
In the visitas, Coporaque villagers also claimed fields beyond the modern
community boundaries of Coporaque. These fields were located both within the modern
limits of neighboring Yanque to the adjacent southwest and in more far-flung locales, such
43
These maps are available at the Ministry of Agriculture in Arequipa. However, a new digitized cadastral
survey by the Ministry of Agriculture (in MapInfo format) is presently near completion, and the old 1974
paper maps will soon be archived.
120
as the villages of Huanca, Huambo, and Lluta, all of which are located in lower-lying
valleys near the headwaters of the Siguas river about 50 aerial km to the southwest. Some
were as far as the valley of Arequipa, 90 km to the south. Although the fields of these
locales also were located by toponym, I mapped them only by reference to their settlement
in the regional-scale maps.44
In the case of fields claimed in Yanque by Coporaque
villagers, I was able to locate some by toponym, using the more coarse-grained toponym
coverage of the 1:5000 Ministry of Agriculture cadastral maps.
45
Some of these fields
declarations state explicitly that they were located in Yanque, and then specify the
toponym sector, but many were declared by a toponym that does not occur in modern
Coporaque but does occur in modern Yanque. Thus, while some of the fields in Yanque
claimed by Coporaque villagers could be mapped, an unknown proportion of these fields
could not be located due to the coarse resolution of the toponym mosaic in the cadastral
maps.46
The production of a map displaying toponym sectors for which matches could be
found involved five steps:
1) All maps from previous toponym surveys were scanned and georeferenced in the
GIS so that they could be tiled in relation to one another and inspected.
44
Galdos Rodriguez (1984; 1987) documented mitimae lands from several ethnic groups, including the
Collaguas, in the area surrounding the city of Arequipa. He established the approximate locations of the
Collaguas lands in La Chimba, located on the west bank of the Ro Chili just south of the city.
45
I had expected that community boundaries to have been more porous during the early colonial period, and
that this was indeed the case became clear when Coporaque villagers claimed fields within toponyms that I
knew from archaeological survey in Yanquesuch as the site of Uyu Uyuseveral times in the visitas (see
Chapter 9).
46
That is, because fields claimed within the modern limits of Yanque were not usually identified as being in
Yanque, an unknown proportion of the fields which could not be mapped according to modern toponyms
actually lie with the modern limits of Yanque.
121
2) A complete list of all of the toponyms from these sources was compiled.
3) After constructing the database and entering the visita data, toponym data from
the visitas were sorted alphabetically47 and compared to a table with the modern
toponym data (also sorted alphabetically).
4) In each case of correspondence, an identification code was assigned.
5) Those toponyms with matches (i.e. that both occur in the visitas and were
documented in modern toponym surveys) were mapped in the GIS. I traced the
toponyms as polygon themes, using the georeferenced modern toponym survey
sources as base maps. The identification code for each toponym (generated in
steps three and four) was entered as attribute data for each polygon in the map,
thus serving as the means for linking the spatial data to the database with the
visita data.
The resulting map (Figure 4.2) locates 76 modern toponyms (67 within the modern
limits of Coporaque and 9 within the modern limits of Yanque) that Coporaque households
in 1604 and 1615 also used to locate their landholdings in their visita declarations. The
fields located within these 76 toponyms account for 366.5 out of 1066.25 topos declared by
both moieties. Therefore I can map out 34% of the total field area declared in the visitas.
In terms of number of landholdings, this corresponds to 1020 out of 3115 (33%) of fields
claimed in the visitas. The remaining 2/3 of the declared fields are located in toponyms that
have either changed names over time, or that remain undocumented.48 So the proportion of
mapped field area is a sizeable and presumably representative sample of the overall land
tenure pattern of Coporaque in 1604 and 1615. Future systematic toponym survey could
47
Some additional sorting was required due to orthographic incongruencies between the modern and colonial
sources, and orthographic inconsistency within the visitas. For example, the modern Waykiri toponym was
spelled Guayquiri in the visitas. Other cases required some subjective judgment as to whether the modern
and colonial referent were the same. These are identified where relevant in the discussion. Orthography of
the toponyms presented in the maps follows the ethnographic sources (which do not follow a common
orthographic standard).
48
Subdividing these figures by moiety, my maps account for 570 out of 1522 fields (37%) claimed by
Urinsaya households in 1604. In areal terms, this corresponds to 41% of total Urinsaya declarations (208.5
out of 514.75 topos). Amongst Hanansaya declarations from the 1615 visita, 450 out of 1593 fields (28%)
could be mapped, or 29% in terms of area (158 out of 551.5 topos).
122
perhaps increase the sample size of visita declarations that could be mapped, but this kind
of focused survey was beyond the scope of my project.
Figure 4.2: Map of all toponyms with matching names in the visitas.
123
Table 4.2: Modern toponyms in Coporaque that also appear in the Visitas
Code
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Name
Alcalli
Anchocllo
Ancollaya
Antacala
Apo*
Apopampa*
Aquerana
Avantira*
Bombomcilla
Cachulle**
Canaque
Canteria*
Caque**
Cayra
Ccaya
Chacco*
Chacopata*
Chaquire
Chijra (Chishra,
Chirsha)
Chilcarani*
Chocpayo
Chuankaya
Churani
Churqui
Chururani
Code
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
Name
Cocawire
Collpa
Compuyo
Cusqui
Cuyo
Fallero
Kello*
Kellolucri
Korinapampa
Kundurparara
Kusipampa
Lama
Llactapampa
Llanka
Machingaya
Malcapi
Malco
Malcopampa
Nasama*
Ocolle
Pasnaya
Pataha
Patarana*
Pisnolla
Pucjio
Quelqata
Sahuara
Code
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
Name
Sallihua
Saymana
Sumo*
Supowiri*
Suqilla*
Suripampa
Taapaque
Taqllapukio
Taqowire
(Toccohuire)*
Tocco*
Tocolle
Tunsa
Ullullu
Umaoso
Umaro
Ururani
Utarana
Vilcarana
Waykiri*
Waykirilucri*
Waykiripata
Wichokata
Yawiso (Llahuiso)*
Ycani
Sources: (Benavides 1986e; Crdova Aguilar, et al. 1986; Izaguirre Urbano n.d.; Treacy 1989); Peruvian
Ministry of Agriculture Cadastral Survey 1974/1979; Fieldwork 1999-2000.
*In the Waykiri basin, a low-lying, bowl-shaped basin of contour bench terraces and valley bottom fields
(the dark line outlines the extent of the toponyms within the Waykiri area).
**Ministry of Agriculture cadastral maps indicate that this sector is called Cachulle. However, Treacy
(1989:257) maps a toponym called Caque that is in the same area, but Treacys maps represent toponym
locations as points, so its specific outline could not be mapped. I have thus included Caque within the
toponym outline of Cachulle according to the ministry of agriculture maps, since it probably represents a
subdivision within that large toponym sector.
Treacy (1989:257) is the only source locating these toponyms. They are small toponym sectors in the
Waykiri basin, but their specific location could not be determined from Treacys maps.
124
Units of Measure: Topos

Estimating field sizes declared in the visitas is a matter of approximation, since a
variety of measures were used. By far the most common measure used was the topo.
Topo (or tupu) is a Quechua term that was used both as a unit of distance and of area in
the Inka empire (D'Altroy 2002:246-247). As a unit of area, it is unclear if it was
standardized in terms of spatial area or if it referred to a relative measure that varied
according to terrain, fertility, or other factors. Murra (1980) argues that the topo was
more likely a relative measure referring to an area, depending on soil type, slope, etc.,
that was sufficient to feed a married couple for a year. Cobo indicated that a topo was 50
fathoms long by 25 fathoms wide (91.4 x 45.7 m), or about 0.42 ha (as cited in Rowe
1946a:324). Garcilaso provides an estimate of 1.5 fanegasa much less precise unit of
measure, or about 0.96 ha according to the modern standard for the fanega (as cited in
Rowe 1946a:324). Rowe (1946a:324) favors the estimate of Cobo, since he provides the
most precise measure, and it is closest to the modern day equivalencies. These vary
slightly between modern Departments, but they are all very close to a third of a hectare.49
For the purpose of visualization and comparison to modern field areas, its modern
standardized equivalent in Arequipa of 3496 m2 provides a reasonable estimate.
However, for most analyses in this dissertation, I have left the units of measure in topos.
Given the highly dispersed land tenure pattern here and elsewhere in the Andes,
each households total landholdings generally were distributed among many small,
49
For instance, Glave and Remy (1983) state that a topo in Ollantaytambo was equal to 96 varas by 48
varas, (about 82.9 x 41.5 m), or 3440.4 m2. Gade identifies a topo in Vilcanota as approximately 0.32 ha.
Its modern standardized equivalent in the Department of Arequipa is 3496 m2 (Benavides 1986:450).
125
widely-dispersed fields, often in distinct microclimates and production zones, rather than
as a single large plot. Therefore, most fields were declared as fractions of a topomost
as a quarter or a half topo. However, some small plots were declared as a pata (patch),
chacara (field), pedao (piece), pedaillo (small piece), anden (terrace),
andenillo (little terrace), and, rarely, solar (patio). Although the areas of these
fields will never be known with certainty, they presumably were somewhat smaller than
even a quarter topo (the smallest fraction of a topo declared), but, consistent with the
previous work of Benavides (1986b; 1986c; 1986d; 1986e) and Treacy (1989), I have
equated them with a quarter topo in order to have a common unit of measure in reporting
field areas.50
50
Exceptions to these equivalencies occur in two cases. First, fields were sometimes listed as a fraction of
a chacarai.e., a fraction of a field. In these cases, the term topo appears to have been substituted by
the generic term for field (chacara), and I have calculated the area based on the area of a topo (e.g. 1/4
chacara = 1/4 topo). The second exception occurs when only an indication of a fraction is given, with no
unit of area. This usually occurs in houses with many fields, where a long listing of fields is dictated to the
scribe. For example, one or more of the fields may omit the areal unit: In Tacoviri (i.e. in the toponym
Tacoviri), a quarter of maize. Here, it is clear that the referent is a topo, and I would assign the size of the
field to 0.25 topo.
PART II: ARCHAEOLOGICAL FINDINGS
126
Chapter 5
Early Settlement and Agriculture: The Formative and Middle Horizon
Introduction
As discussed in previous chapters, before this project virtually nothing was known
about human settlement in the Colca valley prior to the Middle Horizon. In the absence
of Formative Period sites, previous researchers initially suggested that the first significant
agricultural occupation of the valley may have occurred during the Middle Horizon (de la
Vera Cruz Chvez 1987; Malpass 1986, 1987). Radiocarbon dates from excavations
provided some indication that agricultural terracing originated prior to the Middle
Horizon, but no settlements with occupations corresponding to these dates had been
identified.
Considering the importance of the Formative Period for understanding the longterm historical trajectory of the south-central Andes, and the importance of the Colca
valley in regional economic networks during that time, this is a vast and critical gap in
knowledge. The now well-documented spread of Chivay source obsidian throughout the
Titicaca Basin and beyond has demonstrated that the Colca valley had constituted a key
node in regional trade networks since the Archaic Period, and was probably home to
significant Archaic and Formative populations. Throughout the altiplano and western
slopes, a series of major social transformations occurred during the Formative Period,
including the onset of agriculture, sedentism, and ceramic production in the Early
Formative, as well as the initial development of marked social inequalities and political
centralization during the Middle and Upper Formativeprocesses which coalesced (in
127
tandem with developments in the central Andes) in the formation of the first state in the
region at Tiwanaku.
Regarding the local influence of Tiwanaku, previous research
presented an unresolved paradox: on the one hand, the ubiquity of Chivay source
obsidian in the altiplano demonstrated strong trade ties between Tiwanaku and the Colca
valley, but on the other hand, no evidence existed for political control or influence in the
valley by Tiwanaku (see Chapter 3). Rather, ceramic collections from local Middle
Horizon sites were devoid of Tiwanaku styles and were composed exclusively of local
wares, some derivative of Wari imperial styles. Resolving, or at least clarifying, the
nature of this paradox is critical both for understanding how Wari and Tiwanaku spheres
of influence articulated in the south, and for illuminating the proximate origins of the
Collagua ethnic polity.
In this chapter, the first section presents evidence for a significant Formative
Period occupation in the central Colca valleywhat most likely represents the earliest
period of agricultural settlement in the valley. Given that this is the first documentation
of a Formative Period component in the Colca, discussion is primarily descriptive and
conservative in terms of interpretation. Much research remains before the timing and
causes of the onset of agriculture, sedentism, and social inequality will be understood.
The second section presents the survey findings which document a Middle Horizon
Period occupation characterized by a local outgrowth of the Formative Period settlement
pattern that lacks indices for direct imperial administration by either Wari or Tiwanaku,
but clearly within a more general sphere of Wari influence.
128
Determining site sizes for many Formative and Middle Horizon Period sites was
complicated by two factors.
First, domestic and agricultural terracing and standing
architecture from later occupations probably obscures much of the Formative occupation
at large settlements.
Second, also probably as a consequence of later occupations,
surficial Formative and Middle Horizon artifact densities are very low at most sites.
These conditions contrast with, for example, sites in the Titicaca Basin, where Formative
occupations are generally more dense and allow for reliable delineation of componentspecific site extents (Bandy 2001; Seddon 1998; Stanish 1997a; Stanish, et al. 1994). In
this survey, less than ten (and, in many cases, less than five) Formative and Middle
Horizon Period artifacts were recovered from most sites, making component-specific
boundaries at multi-component sites impressionistic.
Therefore, with a few notable
exceptions (discussed below) area estimates have not been determined at most sites.
Chronological control also remains coarse at this early stage of Formative Period
research in the valley. Ceramics broadly diagnostic of the Formative Period, which I
have named Chiquero ceramics (see below) were identified in the survey artifact
collections. However, the Formative Period, which is subdivided into three to four
phases in many areas of the altiplano, spans about two thousand years, from around 1500
BC to AD 500. Chiquero ceramics are undecorated domestic wares, making crossreferencing with the better-known decorated Formative styles in the region impossible at
present (see Appendix A). No projectile points in the regional sequence have been
temporally isolated to the Formative Period; however, there are projectile points
129
diagnostic of the Terminal Archaic to Formative Periods. We recovered examples of this
point type from five sites in the survey, and these are included in the following
discussion.
Therefore the Terminal Archaic to Formative Period settlement pattern
discussed below represents a palimpsest of occupations spanning more than two

millennia, and not a contemporaneous state of occupation between all sites.
Likewise, local Middle Horizon ceramics have been grouped into a single
temporal component, spanning approximately 500 years, between about AD 500 and AD
1000. Local Middle Horizon diagnostics are confined to the bowl form, and include
decorated, undecorated, slipped, partially slipped (rim-slipped), and unslipped wares. As
discussed in Appendix A, this variability may allow for future phase subdivision. For
example, the polychrome decorated type similar to the Qosqopa style may pertain to an
earlier phase (Middle Horizon Epoch 1), and Collagua I ceramics may prove ultimately to
consist of a terminal Middle Horizon to early Late Intermediate transitional type, but
Collagua I ceramics have been provisionally grouped together with Collagua II ceramics
as a Late Intermediate Period style (see Chapter 4).
The Formative Period Settlement Pattern: Overview

With these caveats in mind, the survey data unequivocally identify a significant
Formative Period occupation in the central portion of the Colca valley. A total of 30
Formative Period sites were documented, including a sizeable agricultural settlement
(YA-032), as well as a series of smaller sites on both sides of the valley and in the puna,
both in rockshelter and open air contexts. Of these 30 sites, 21 were classified as
130
settlements (including mounded midden-type sites), five as agro-mortuary complex
sites, three as rockshelters, and one as a cemetery (Figure 5.1, Table 5.1, Figure 5.2).
131
Figure 5.1: Map of Formative Period sites in the survey area
132
Table 5.1: Formative Period sites (N=30)
Site No.
YA-002
YA-008
YA-009
YA-014
YA-030
YA-032
YA-034
YA-041
YA-043
YA-045
YA-046
YA-054
YA-055
YA-059
YA-082
YA-083
YA-087
YA-094
CO-100
CO-104
CO-105
CO-106
CO-121
CO-150
CO-152
CO-159
CO-163
CO-166
CO-168
CO-169
Sector(s)
A/B
C
B
B/C
B/C
Altitude
3495
3403
3478
3533
3575
3665
4187
3484
3417
3556
3494
3572
3485
3808
4623
4467
4337
4254
3666
4362
4348
4397
3537
3513
3884
4419
3528
3545
3547
3568
Ecozone
Kichwa
Kichwa
Kichwa
Kichwa
Kichwa
Suni
Puna
Kichwa
Kichwa
Kichwa
Kichwa
Kichwa
Kichwa
Suni
Puna
Puna
Puna
Puna
Suni
Puna
Puna
Puna
Kichwa
Kichwa
Puna
Puna
Kichwa
Kichwa
Kichwa
Kichwa
Site Class
Midden/Rubble Mound
Small Cemetery
Small Scatter
Agro-Mortuary Wall Complex
Large Scatter
Large Scatter
Small Scatter
Midden/Rubble Mound
Small Scatter
Small Scatter
Small Scatter
Small Scatter
Small Scatter
Small Rockshelter
Small Rockshelter
Small Scatter
Small Rockshelter
Small Scatter
Small Scatter
Small Scatter
Large Rockshelter
Midden/Rubble Mound
Midden/Rubble Mound
Small Scatter
Small Scatter
Small Scatter
Site Area
0.06
0.04
Unknown
Unknown
Unknown
3.88
2.19
Unknown
0.01
1.51
0.34
1.75
0.03
0.25
0.01
0.01
0.09
0.04
Unknown
0.01
0.06
0.10
0.01
0.15
0.18
0.36
Unknown
Unknown
Unknown
Unknown
133
18
16
14
12
Rocks helter
Agro-Mortuary Com plex
Settlem ent/Scatter/Midden
6
4
2
0
Kichw a (3353-3600)
Suni (3600-3900)
Puna (3900-4900)
Ecological Zone (masl)
Figure 5.2: Formative site counts by ecological zone (N=30)
20
Count
Count
Cem etery
10
10
0
3300 - 3600
3600 - 3900
3900 - 4200
4200 - 4500
4500 - 4800
Altitude
134
All major physiographic and ecological contexts in the valley were occupied or
exploited during the Formative; however, settlement clearly was more concentrated in the
lower elevations in the survey area (Figures 5.1-5.3). Of the 21 sites classified as
settlements, 14 (11 in the kichwa zone, and 3 in the suni zone) are located amid the broad
pampa of the Qal 4 alluvial terrace above the river and surrounding slopes. While most
sites in the lower reaches of the valley were concentrated on the north side of the river,
the largest Formative settlement registered during the survey (YA-032, Chiquero) is
located on a promontory overlooking the valley floor on the south side of the river (see
below). In the puna, seven open air sites and three rockshelter sites with Formative
ceramics and/or Terminal Archaic to Formative points were identified. Of these ten puna
sites, all but three (two small open air scattersCO-152 and CO-159, and one small
rockshelterYA-082) produced either Late or Terminal Archaic Period projectile points,
indicating considerable continuity from earlier hunting and foraging patterns. Puna sites
are concentrated around bofedales and quebradasthe two principle water and resource
patches of the zone.51
Aside from site sizes and artifact densities, no features differentiate Formative
Period sites. There is no evidence for corporate architectonic features (e.g. formal mound
features) or other indices for intra- or inter-site inequalities in economic or political
status. No site size hierarchy is apparent amongst sites for which occupational areas
could be established. Local Chiquero ceramics are undecorated, and only one decorated
51
This patchiness is clearly a factor in the overall continuity of occupation revealed at most sites in the
puna. Presently- or recently-occupied estancias (herding households and hamlets) are present at most sites
in the puna as well, and the suvey crew came to expect the presence of earlier components as we
approached them during our high altitude survey.
135
Formative sherd in the Pukara style was recovered (see below). In sum, the Formative
Period settlement pattern in the agricultural elevations of the survey area indicates that
the population was distributed between a series of small agricultural hamlets and villages
concentrated on the valley bottom.
Relationships between agriculturalist settlements in the lower reaches of the
valley with sites in the puna remain unclear. On the one hand, evidence for hunting, and
possibly pastoralism in the puna during the Formative could represent the residues of
groups engaged in a mixed pastoralist/hunting economy, or of hunting forays by
agriculturalists, a mix of both of scenarios, or a diachronic shift between these patterns.
On the other hand, the presence of projectile points at the large settlement of YA-032 (see
below) suggests that agriculturalists also engaged in hunting, although the points
recovered from YA-032 are broadly diagnostic of the Terminal Archaic to the Late
Horizon, so they could pertain to any or several of these periods of occupation. The
general lack of chronological control precludes further clarification of these surely
complex and changing interzonal relationships.
Site YA-032: Chiquero

The site of Chiquero (YA-032) dominates the settlement pattern in terms of size
and quantity of diagnostic artifacts recovered.
The site is located at the end of a
prominent spur that descends from the southern puna rim, providing views over the broad
pampa to the west of Yanque (Figures 5.1 and 5.7). A modern chapel (dedicated to the
Virgin de Chapi) and enclosed churchyard occupies the top of this hill. While the chapel
136
and churchyard largely obliterated whatever features were present in the what was
probably the central portion of the site, its function as an agriculturalist settlement is
apparent from the abundant domestic and agricultural artifacts present throughout the site
surface, including utilitarian ceramics, as well as grinding (batanes) and lithic cultivation
implements. A dense ceramic and lithic scatter covers the 3.88 ha area of the site.
Aside from its promontory position, which gives clear views up and down the
valley, the ecological context of the site is also significant. At 3665 masl, the site is
situated in the suni zone, an ecotone between the kichwa and puna zones, and lies
adjacent to a formal path that follows the major quebradas over the valley rim to connect
with networks of paths to Arequipa and Yura to the south, and Lluta, Huanca, and
Huambo to southwest. Assuming that ancient paths followed these same routes of least
resistance out of the valley, the site would have been well-situated as a point of
articulation between diverse production zones.
137
Figure 5.4: Formative ceramics recovered from YA-032 (lot numbers indicated)
Chiquero is the type site for local Formative Period ceramics, having produced
64% of the total diagnostics collected for the period (N= 97 of 152). As discussed in
Appendix A, Chiquero ceramics assemblages are dominated by spheroid, neckless
ollasa hallmark vessel form of the Formative in the regionbut also contain small
amounts of other ollas and constricted-orifice forms (Figure 5.4, A-J). All are unslipped,
138
with distinctively striated surfaces (interior and exterior) resulting from a coarse vegetalfiber brushing technique. One diagnostic Pukara Classic style sherd with a post-fire
incised zoomorphic motif (probably a camelid foot) was also recovered (see Appendix A,
Figure 5.4, K). This sherd provides a measure of temporal control, indicating that the site
was occupied between 200 BC and AD 400 (L. Klarich pers. comm. 2003), but the
duration of occupation probably extends considerably on either side of these dates.
Figure 5.5: Obsidian Type 5D projectile point and scraper from YA-032 (lot numbers indicated)
Figure 5.6: Andesite hoes from YA-032 (lot numbers indicated)
139
Lithic artifacts recovered from the site surface include grinding implements
(batanes), scrapers, projectile points, and hoes, indicating that the sites inhabitants
engaged in hunting, cultivation, and domestic tasks. Lithic debitage and projectile points
were found in high concentrations throughout the site. The five projectile points (Type
5D points in the regional chronology) are only broadly diagnostic of Terminal Archaic to
Late Horizon times, but are most prevalent in Formative (particularly Late Formative)
assemblages in the region (Klink and Aldenderfer in press). The large sample (N=69) of
stone agricultural implements (hoes and hoe fragments) recoveredmore than any other
site in the surveysignals agricultural cultivation by the sites inhabitants.
The
prevalence of hoes at the site may in part reflect opportunistic exploitation of the
abundant andesitic clasts found over the site surface. However, obvious evidence for
use-weardistal-end polishing and striationsindicates that the implements were used,
and not just manufactured locally. These andesite hoes are very similar in size and form
to those recovered from Formative Period contexts throughout the southern Peruvian
Pacific drainages and circum-Titicaca Basin (Seddon in Bandy 2001; Stanish 1997a;
Stanish, et al. 1994:65-71). The occurrence of stone hoes drops off dramatically in postFormative contexts throughout the south-central Andes, as other (presumably perishable)
materials replace this lithic technology, and such appears to be the case in the Colca as
well. During the survey, no stone hoes were recovered from sites lacking Formative
Period ceramics, making them a temporally-diagnostic artifact class.
140
Agricultural Production and Landscape Modification During the Formative
There is indirect evidence about the type of agriculture practiced by the
inhabitants of Chiquero and the agricultural infrastructure of the time. First of all, the
irrigated bench terrace complexes that surround the site clearly postdate the Formative
occupation. This is apparent in the distribution of Formative artifacts at Chiquero, which
is truncated along the sites north perimeter by irrigated bench terraces, indicating that
they overlie the Formative occupation of the site. More precisely, some of these irrigated
terraces are abandoned, and it is only in these areas of the site perimeter, where the
collapse and erosion of the bench terraces has revealed the underlying Formative
occupation, that Formative materials continued to be recovered in significant quantities.
Thus, irrigated bench terraces were not present on the surrounding slopes at the time of
the sites occupation. The site apparently followed the original slope surface of the
hillside, or was composed of domestic terracing that was later obscured by the
agricultural bench terraces.
141
Figure 5.7: YA-032 (Chiquero) and surrounding agricultural landscape features
142
The site inhabitants probably cultivated unirrigated sloping fields, such as those
visible on the ridgetop immediately above the site, where 14 ha of abandoned rock-lined
sloping fields are present on a small mesa called Senja Pata. These fields are similar to
unirrigated sloping fields which have been demonstrated to predate irrigated bench
terracing in Coporaque and Chivay (Brooks 1998; Treacy 1989). Their orientation would
have collected sheetwash from the ridgeline and slopes above, augmenting the natural
drainage pattern of the local topography. Given the close association of this particular
complex with Chiquero, it is reasonable to suggest that the site inhabitants engaged in
this type of augmented dry farming in these and other fields in the surrounding slopes,
many of which were subsequently replaced by bench terrace complexes.
On the valley bottom, aside from artifact scatters, identifying direct evidence for
early agricultural production was hindered by post-Formative landscape modification, but
the survey did identify a class of agricultural field features that appear to have origins in
the Formative. Specifically, large field-dividing wall complexes cover large areas of the
relatively flat Qal 4 and 5 alluvial terrace surfaces on both sides of the river. These
agro-mortuary wall complexes are composed of massive walls with rock-lined, ovoid
tombs and probable storage features at regular intervals along the tops of the walls. The
walls enclose elongated fields that vary in size between 20-50 m long by 10-25 m wide. I
describe these walls in more detail in the Middle Horizon section below. We recovered
Chiquero ceramics in association with ovoid features on top of the walls at five such sites
(YA-014, YA-030, CO-166, CO-168, and CO-169). The largest of these complexes, site
YA-014, is located on the broad pampa directly downslope from Chiquero. As discussed
143
in the Middle Horizon section below, agro-mortuary wall complexes agglutinated slowly
over time, and during the Formative their areas were probably only a fraction of their
eventual extent, but the presence of Chiquero ceramics associated with collared tombs at
YA-014 suggests that agriculture on the valley bottom was widespread during the
Formative. These fields are currently irrigated by the same canal systems that irrigate the
bench terraces on the slopes above, but since the bench terraces postdate the Formative
occupation, they must have been either dry farmed or irrigated by canal systems that were
later replaced.52
Terminal Archaic and Formative Period Hunting and Pastoralism in the Puna
The survey also recorded ten sites with diagnostic Terminal Archaic and/or
Formative occupations in the puna. Of these ten sites, five (YA-083, YA-087, YA-094,
CO-104, AND CO159) were dated based on the presence of projectile points diagnostic
of the Terminal Archaic to Formative Periods (Type 5C in the regional point chronology,
see Klink and Aldenderfer in press). Of the remaining five, three (YA-082, CO-105, and
CO-152) were dated based on the presence of Chiquero ceramics, and two (YA-034 and
CO-106) produced both Type 5C points and Chiquero ceramics.53 Terminal Archaic to
52
More speculatively, short feeder canals could have syphoned water away from the quebradas which
would have carried much more water prior to the construction of the present canal system (which currently
syphons water from these drainages at higher altitudes).
53
However, more high altitude sites were probably used during the Formative, since nine other puna sites
from which only Type 5D projectile points were recovered have been excluded from these counts. As
discussed in above, Type 5D points have a long use history in the region, appearing in Terminal Archaic
through Late Horizon contexts, but appear to be most prevalent during the Formative (Klink and
Aldenderfer in press). An unknown number of these nine sites can therefore be expected to have Formative
occupations as well. (YA-040, YA-073, YA-078, YA-084, YA-089, CO-102, CO-111, CO-156, and CO159).
144
Formative Period sites in the puna show remarkable continuity with previous Archaic
Period occupations. All but three (two small open air scattersCO-152 and CO-159,
and one small rockshelterYA-082) of these ten sites also produced either Late or
Terminal Archaic Period projectile points.
Figure 5.8: Sketch map of YA-094
We registered four puna rockshelter sites for the period (YA-082, YA-083, YA094, and CO-106). 54 Among these, three (YA-082, YA-083, and YA-094) consisted of
small artifact scatters associated with overhanging boulders which would have served as
windbreaks or short-term logistical camps. One small rockshelter site (YA-094), located
adjacent to a quebrada, presented features suggestive of a dual hunting/herding function
54
Nearly all of the rockshelter sites also bear evidence of pictographic ochre-based rock art, and a great
variety of representational and non-representational motifs were recorded, but at present these features
cannot be dated with any certainty.
145
during the Formative (Figure 5.8). The area in front of the rockshelter at this site (Sector
A) may have functioned as a small corral (Figure 5.8).
The fourth rockshelter siteCO-106is much larger, consisting of a series of
five small caves formed by an overhanging cliff adjacent to a major quebrada.55 At this
site, a 20 m long profile with visible stratigraphy up to 1.8 m thick has been exposed by
erosion near the edge of the quebrada in front of the rockshelters. In addition to the two
Terminal Archaic to Formative points recovered, we also collected diagnostic points from
the Early, Middle, Late, and Terminal Archaic Periods, as well as one Chiquero sherd.
Figure 5.9: Site CO-106, showing cliff overhang (rockshelter) and midden areas
55
At lower elevations this quebrada (quebrada Aquenta) becomes the Ro Cantumayo, which runs through
the village of Coporaque before spilling into the Colca. The site is also adjacent to the Waynaqorea Pampa,
a long glacial outwash plain that skirts the Mount Mismi Massif, giving access to the high puna and
continental divide to the north, as well as to major thoroughfares that lead eventually to the southern
Province of Espinar in the Department of Cuzco (Fieldwork 1999-2000).
146
We recorded six lithic and ceramic scatters in open air contexts in the puna, five
of which (YA-087, CO-104, CO-105, CO-152, and CO-159) consist of lithic and ceramic
scatters of 0.25 ha or less. However, one open air site (YA-034) stands out for its size,
density of material, and long continuity of occupation. The site occupies a sloping planar
surface beside a small quebrada, and also lies adjacent to the path discussed above that
runs near Chiquero (YA-032)only 2.8 aerial km away (within one hours walk from
Chiquero to YA-034, and about 30 minutes down from YA-034 to Chiquero). With an
area of 2.47 ha and densely covered with lithic debitage and artifacts, YA-034 produced
the most diagnostic projectile points of any site in the survey (35 complete or near
complete points). Deep cultural deposits are also evident in exposed stratigraphy at the
site. Projectile points recovered from the surface include diagnostics from the Early,
Middle, Late, and Terminal Archaic Periods, as well as Terminal Archaic through Late
Horizon Periods. Chiquero ceramics were also present. Thus, the site appears to have
been used or occupied throughout virtually all of the human occupation of the valley.
Discussion
While spatial controls for the early agro-pastoral settlement of the valley remain
rudimentary, this survey provides a baseline from which to work, having identified
ceramics broadly diagnostic of the Formative Period, as well as 30 sites distributed
throughout all major physiographic and ecological contexts of the central portion of the
valley. The largest settlement (YA-032, Chiquero) is situated in the transitional suni
zone, giving easy access to both the valley bottom and surrounding slopes. While this
147
site stands out in terms of size, the Formative Period is characterized by an
undifferentiated hamlet/village settlement pattern, with no indices for marked
sociopolitical inequalities within or between sites. This may result from sampling error,
since other Formative Period settlements have been obscured to an unknown degree by
subsequent landscape modification and settlement.
Assuming that the survey area
constitutes a representative sample for the valley as a whole, the observed pattern would
result in a convex rank-size settlement distribution of the valleysignaling several localscale sociopolitical groups. Expanded survey coverage will increase the sample size and
allow for further evaluation of this hypothesis.
At this early stage in knowledge of the Formative in the valley, temporal controls
are equally rudimentary. The current chronology precludes distinction of diachronic
processes within this vast stretch of time. Major social, political, and economic changes
almost certainly occured during this time, and the valley was clearly integrated into
regional economic flows, as demonstrated by the widespread distribution of Chivay
source obsidian throughout the Titicaca Basin and altiplano. In these neighboring areas,
profound social transformations occurred during the Formative, as the first chiefdoms
came about in parallel with the development of agriculture and pastoralism, the
elaboration of corporate art styles and monumental architecture, and the spread of a
regional mytho-religious tradition (the Yaya Mama Religious Tradition). Of course,
knowledge of these processes is undergirded by decades of excavation and nearly 2000
km2 of systematic survey coverage (see Stanish 2003). Future survey and excavations
148
will certainly refine our understanding of the earliest phases of agro-pastoral settlement in
the Colca valley.
Compared to previous research in the valley, which focused on agricultural
terracingand by extension, on valley-side slopesthe documentation by this survey of
Formative Period settlements and agricultural features on and around the valley bottom
presents a distinct view of early agricultural production in the Colca valley. Treacy
(Treacy 1989:93-99, 123-133), Brooks (Brooks 1998), and Denevan (Denevan 2001:
172-173, 192-201) have suggested that early farming was concentrated primarily along
the upper slopes of the valley in the form of segmented and sloping-field terrace systems
that augmented natural drainage patterns through cross-channel terracing, check dams,
and other runoff harvesting features.
According to this scenario, these early water
harvesting systems were superseded by irrigated bench terracing (Denevan 2001: 172173, 192-201; Treacy 1989:93-99, 123-133), perhaps with an intervening phase of
unirrigated bench terracing (Brooks 1998:374-386). The timing of these changes was
unclear and debated prior to this project. Treacy (Treacy 1989:122-126), based on
radiocarbon dates from trench excavations in bench terraces and reconnaissance of
segmented terraces, argued that segmented terraces were used during the Middle
Horizon. Brooks, based on radiocarbon dates from more recent terrace excavations
places the first agricultural terraces in the valley much earliersometime prior to 2400
BC (Brooks 1998: 400-401, 405).56 By contrast, this survey documents widespread
56
However, as discussed in Chapter 2, these early dates are unreliable since they were derived from bulk
soil samples (which could easily be contaminated with old carbon), and such an early date would place
agricultural terrace use during the Late Archaic, an preceramic, preagricultural period of hunting and
foraging throughout the south central Andes.
149
settlement and agricultural features on the valley bottom and surrounding slopes during
the Formative Period. While the Formative encompasses 2000 years, it does establish a
general time frame for the onset of agricultural production in the valley. This range is
also more consistent with regional developments than either of the previous scenarios,
since sedentism, agriculture, and ceramic production all begin during the Early Formative
in the Titicaca Basin, altiplano, and neighboring valleys.
Also, while sloping fields, and perhaps segmented terraces were in use during the
Formative, the virtual absence of any Formative artifacts or settlements in and around
them suggests that they were used intermittently, and in any case were only part of a
larger agricultural production system that also covered the valley floor. Perhaps sloping
fields and segmented terraces were used according to short- or long-term fluctuations in
climate. As reviewed in Chapter 2, Titicaca lake core data indicate that four major dry
periods (between 1550-1400 BC, 950-850 BC, 450-250 BC, and 50 BC-AD 250)
occurred in the two millennia between 1500 BC and AD 500 that make up the Formative
Period (Abbott, Binford, et al. 1997). Precipitation at or above modern levels occurred
during the intervening periods. Given that dry farming is virtually impossible with the
quantity and periodicity of modern precipitation in the valley (see Chapter 2), it is highly
unlikely that these fields could have been cultivated during these dry periods. Certainly,
local farmers would have been attuned to changes in the precipitation regime, and
cultivation may have shifted or expanded to include the sloping fields and segmented
terracing along the upper slopes during times of plentiful precipitation.
Of course,
current chronological controls are too coarse to evaluate this scenario, so it must be
150
considered hypothetical. But the concentration of settlements and early agricultural
features on the valley floor and lower slopes suggests that early cultivation was more
broadly distributed than previously hypothesized, and probably focused on and around
the valley bottom and lower slopes, as opposed to the high slopes and quebradas.
The Middle Horizon Period Settlement Pattern: Overview

The survey registered 37 sites with Middle Horizon componentsseven more
than during the Formative Period (Figure 5.11, Table 5.2). However, 54% (20 out of 37)
of these sites were not occupied during the Formative, signaling a significant shift in
settlement. As was the case for the Formative, determining component-specific site sizes
at many settlements was hindered by overlying occupations (Table 5.2). Only a handful
of Middle Horizon sherds were collected from most sites in the kichwa zone. Fewer than
five Middle Horizon sherds were collected at each of 28 of the 37 sites registered.
151
Figure 5.10: Middle Horizon ceramics
Nonetheless, the diagnostic Middle Horizon ceramic sample (N=217) is 42%

larger than the Formative collections (N=153). Only 29 sherds (13%) from the Middle
Horizon collections were decorated. These decorated bowl sherds, generally black-andwhite on red slip (applied to a buff base), include motifs arranged in external horizontal
design fields found in Qosqopa and other regional Wari variants, such as X figures, Scurves, and rhomboids (Figure 5.10, K-N).57 Undecorated local Middle Horizon sherds
are either unslipped buff colored wares, or bear a distinctive red slip over this buff base,
extending just below the interior rim. All are finely executed, polished, and well-fired.
57
For more detailed discussion, see Appendix A.
152
The hallmark vessel form of the Middle Horizon is the slightly closed, ellipsoid bowl
with a symmetrical or slightly reinforced lip (Figure 5.10, A-J).
Figure 5.11: Painted tablets from YA-169 (Bomboncilla)
Another class of distinctive artifacts, painted tablets, begins appearing during the
Middle Horizon in the Colca and neighboring valleys, and a large sample (N=80) was
collected during the survey. Most of these (84%) were recovered from a single site
CO-169 (Bomboncilla). We recovered these tablets from the vicinity of four small
subterranean stone-lined boxes built into the base of an agro-mortuary wall, which had
been disturbed by looting.58 We also recovered the largest sample of Middle Horizon
ceramics (N=61, or 28% of the total collections for the period) from this site.
58
This painted tablet tradition is unique to the Arequipa region. At present, their known distribution
extends from the Chili/Vitor/Quilca drainage in the south to the Cotahuasi/Ocoa drainage in the north.
They have been recovered in a variety of contexts, but most commonly in house wall foundations
(especially under the corners of walls) and in mortuary contexts (especially on top of the roofs of tombs).
153
Figure 5.12: Middle Horizon sites in the survey area
154
Table 5.2: Middle Horizon sites (N=37)
Site No.
YA-001
YA-004
YA-006
YA-007
YA-008
YA-009
YA-012
YA-014
YA-030
YA-032
YA-034
YA-041
YA-042
YA-046
YA-050
YA-054
YA-057
YA-059
CO-061
YA-090
YA-093
YA-094
CO-098
CO-100
CO-121
CO-147
CO-148
CO-149
CO-150
CO-151
CO-154
CO-163
CO-164
CO-165
CO-167
CO-168
CO-169
Sector
B/C
B
C
A/D
A/C/E/N
B
A
G/J/K/M/N
B/C/D
A/C
A
B/C/D
Altitude
3444
3413
3471
3489
3403
3478
3460
3533
3575
3665
4187
3484
3448
3494
3527
3572
3431
3808
4360
4300
4292
4254
3763
3666
3537
3579
3573
3552
3513
3822
3767
3528
3556
3543
3537
3547
3568
Ecozone
Kichwa
Kichwa
Kichwa
Kichwa
Kichwa
Kichwa
Kichwa
Kichwa
Kichwa
Suni
Puna
Kichwa
Kichwa
Kichwa
Kichwa
Kichwa
Kichwa
Suni
Puna
Puna
Puna
Puna
Suni
Suni
Kichwa
Kichwa
Kichwa
Kichwa
Kichwa
Puna
Suni
Kichwa
Kichwa
Kichwa
Kichwa
Kichwa
Kichwa
Site Class
Midden/Rubble Mound
Midden/Rubble Mound
Small Scatter
Midden/Rubble Mound
Small Cemetery
Small Scatter
Small Scatter
Small Scatter
Large Scatter
Small Scatter
Midden/Rubble Mound
Large Scatter
Small Scatter
Small Scatter
Small Scatter
Small Scatter
Small Scatter
Small Scatter
Small Scatter
Small Scatter
Large Cemetery
Small Scatter
Small Scatter
Midden/Rubble Mound
Small Scatter
Small Cemetery
Small Scatter
Small Scatter
Large Scatter
Site Area
0.51
0.61
0.04
0.01
0.04
0.73
0.08
Unknown
Unknown
Unknown
2.19
Unknown
0.01
0.34
Unknown
Unknown
0.14
0.25
Unknown
Unknown
Unknown
0.04
0.05
Unknown
0.01
25.72
7.44
Unknown
0.19
0.4
0.015
Unknown
Unknown
1.71
Unknown
9.46
1.86
155
30
25
20
Agro-Mortuary Complex
15
Cemetery
Settlement/Scatter/Midden
10
0
Kichwa (3353-3600)
Suni (3600-3900)
Puna (3900-4900)
Ecological Zone (masl)
Figure 5.13: Middle Horizon site counts by ecological zone (N=37)
30
20
Count
Count
Fortification
10
0
3300 - 3600
3600 - 3900
3900 - 4200
4200 - 4500
Altitude
Figure 5.14: Formative Period site counts by altitude (N=37)
156
Although settlement areas or other site attributes cannot be compared between the
Formative and Middle Horizon, a more general characterization of settlement pattern
change can be inferred from the number, types, and distribution of sites over the
landscape with respect to geomorphic and agro-ecological context. First, there is an
increase in the number and concentration of sites in the lower reaches of the valley during
the Middle Horizon than during the Formative. 76% (28 out of 37) of Middle Horizon
sites are located in the lowest 250 m of the survey area, as opposed to 57% (17 out of 30)
during the Formative (Figures 5.12-5.14). Second, the increase in the number of kichwa
zone settlements accompanies an expansion of valley-bottom agricultural production, as
is evident in the increase in the number (from five to eight) of agro-mortuary complex
sites on the pampas and rolling hills to the west of Yanque and to the southeast of
Coporaque (see below). Also, five sites that later become major settlements during the
LIP and Late Horizon (specifically, YA-041, YA-050, YA054, CO-100, CO-150) were
established during the Middle Horizon.
While most settlement and agricultural production was focused in the lower
reaches of the valley, three settlements with Middle Horizon ceramics in the suni zone
(that is, between 3600 and 3900 masl) were also registered (Figures 5.12-5.14). Two of
these represent new Middle Horizon occupations, consisting of a ceramic scatter
associated with a small unirrigated terrace complex (CO-059, 0.25 ha area), and a small
group of unirrigated sloping fields (CO-151, Chilacota, 0.4 ha area).59 The third Middle
Horizon suni zone settlement is YA-032, Chiquero, the largest Formative Period
59
Ceramics and radiocarbon dates from test pit excavations by the Ro Colca Abandoned Terrace Project at
the latter of these sites (CO-151, Chilacota) also date this site to the Middle Horizon Period (see Martin
1986)
157
settlement in the survey (discussed above).
However, only three Middle Horizon
sherdsas opposed to 97 Formative sherdswere recovered from the site, indicating

that the site was nearly abandoned by that time.
In the puna, the number of sites decreases by half (from ten to five sites) as
compared to the Formative Period, and all but one (YA-094) are open air sites that also
have major LIP and Late Horizon pastoralist occupations with domestic architectural
remains and corral features. Three of these four pastoralist settlements (CO-061, YA090, and YA-093) produced no Formative diagnostics, suggesting that they were
established during the Middle Horizon. Thus, while hunting in the puna doubtless
continued during the Middle Horizon,60 there appears to be a shift away from
rockshelters, windbreaks, as herding hamlets were established, marking the onset of more
intensive herding activity in the puna.
Middle Horizon Cemeteries

The first evidence for differential treatment of the dead in the form of distinct
tomb styles also occurs during the Middle Horizon. The largest cemeteries recorded in
the surveyat the necropolis sites of CO-098 (Fatinga) and CO-154where large
groups of multi-storey mortuary towers (chullpas) are situated under overhanging cliffs
were first used during the Middle Horizon. These cemeteries lie directly upslope and
north of the largest settlement in the survey area (CO-100, San Antonio/Chijrasee Fig.
60
As in the case of the Formative Period, more high altitude sites were probably used during the Middle
Horizon than the five reported here, since nine puna sites from which only Type 5D projectile pointsa
type only broadly diagnostic of the Terminal Archaic through the Late Horizon were recovered are
excluded from these counts. Any of these sites therefore could have been occupied during the Middle
Horizon. There are no point types in the regional sequence chronologically isolated to the Middle Horizon.
158
5.12). Like CO-100, LIP and Late Horizon ceramics dominate their assemblages, and
these sites will be described in more detail in Chapter 6. But the recovery of Middle
Horizon ceramics from both of these sites indicates that a tradition of building mortuary
structures under rock overhangs (a tradition common in the regionsee below)
originates in the Middle Horizon. Two other types of tombsstone-lined collared tombs
and gallery tombswere also documented in direct association with Middle Horizon
ceramics. Collared tombs were registered in agro-mortuary wall complexes (at sites YA014, CO-147, CO-148, CO-149, and CO-169), and in one case (YA-009), on top of a
large boulder next to the Colca river.61 Also, one example of a gallery tomb complex
connected by a narrow tunnel was built into an agro-mortuary wall at CO-148 (see
below).
In the absence of well-preserved residential sites (that is, Middle Horizon
settlements without overlying occupations), this variety of tomb architectureranging

from simple, single-interment cysts, to mortuary tower complexesconstitutes the
earliest (albeit tentative) index of social inequality and the emergence of local elites.
Agro-Mortuary Wall Sites and Associated Settlements

As I discussed above, agro-mortuary wall complexes become more widespread in
the lower reaches of the valley during the Middle Horizon. The complexes are located in
three geomorphic contexts: 1) on the pampas of the Qal 4 alluvial terrace to the west of
Yanque (sites YA-014 and YA-030), 2) on the rolling hills of the Qal 5 Quaternary lava
61
These tombs are similar in masonry and execution to the impressive hanging silo type tombs of the
site Chininea (YA-097), situated amidst sheer cliffs in the inner river gorge. The site consists of a tightlyclusterered beehive pattern of some 20-40 ovoid tombs. Given the general affinity between these tombs
and those of site YA-009, they may pertain to the Middle Horizon, but because of the difficulties of the site
setting, we could not gain access to it. In any case, this style of tomb is widespread throughout the valley,
and appears to represent a pre-LIP mortuary tradition.
159
flows to the southeast of Coporaque (sites CO-147, CO-148, CO-149, CO-165, CO-168),
and 3) at the base of the southwestern corner of the Pampa Finaya massif (site CO-169).
At a broader scale, agro-mortuary wall complexes appear to be distributed throughout the
valley, but they have not been identified as such by other researchers. Similar walls and
burial features are reported at Achachiwa, a large Middle Horizon settlement near
Cabanaconde (discussed in Chapter 4, see de la Vera Cruz Chvez 1987), where de la
Vera reported cyst tombs built into wide field-dividing walls. Also, Miriam Doutriaux
has recently registered large field division walls with cyst tombs during survey around
Lari (M. Doutriaux pers. comm. 2002).
160
Figure 5.15: Agro-mortuary walls at site YA-014, Sector E (Airphoto Source: SAN 1974, Frame
2577)
161
Figure 5.16: Detail of agro-mortuary wall
The growth of these wall complexes over very large areas is undoubtedly a
complex process, and only the broad outlines of this development can be hypothesized
with this initial survey. Their size and extent attest to massive labor inputs, but their
irregular configuration suggests a gradual aggregative process, rather than a centrallyplanned or quickly executed building episode. This bottom-up hypothesis also stems
from the fact that ceramics from the entire prehispanic sequence have been recovered
from agro-mortuary walls. By the Middle Horizon, component-specific site areas could
be defined for five (of the eight sites dating to the period) wall complex sites (CO-147,
162
CO-148, CO-165, CO-168, CO-169).62 All of these are situated on the volcanic hills to
the southeast of Coporaque, which surround a cluster of settlements. The largest of these,
CO-147 and CO-148 (with areas of 25.72 and 7.44 ha respectively) together with CO149, CO-165, CO-168, are situated on the hills directly northeast of the settlements CO150 (Llactapampa) and CO-163 (Tunsa), both of which become major settlements during
the LIP and Late Horizon. A small sector with two agro-mortuary walls, CO-169,
produced the most diagnostics of any site for the period,63 and is adjacent to the
settlement CO-164, also overlain by later domestic architecture. So there is a cluster of
settlements (of unknown size) associated with large and growing areas of irrigated
agricultural fields and agro-mortuary walls by the Middle Horizon Period in this zone. In
all cases, LIP and Late Horizon site areas expanded considerably (see Chapter 6). About
half of the area covered by agro-mortuary walls in Coporaque was abandoned during the
Colonial Period, but in Yanque, the walls continued to be used as field dividers until the
present. In short, agro-mortuary walls appear to constitute a long, local tradition with
origins in the Formative, but expanding throughout the Middle Horizon, LIP and Late
Horizon, and subsequently partially abandoned during the Colonial Period. As discussed
below, the particular characteristics and configuration of these walls appear unique to the
Colca valley.
62
These estimates should be considered provisional, since they represent the total area of judgmentallydefined site sectors from which Middle Horizon ceramics were recovered.The remaining three sites (YA014, YA-030, and CO-149) produced three or fewer Middle Horizon sherds, making component-specific
site area definition impossible.
63
61 Middle Horizon sherds were recovered from CO-169, 29% of all diagnostic ceramics for the period,
as well as 67 painted stone tablets. See discussion in this chapter.
163
On the valley floor (i.e. on the Qal 4 terrace), agro-mortuary wall complexes form
a cellular pattern of interconnected walls when seen in plan view (Figure 5.15). Small
irrigated agricultural fields (currently cultivated) occupy the cells enclosed by the
walls. In the volcanic hills southeast of Coporaque (i.e. amidst the Qal 5 surfaces), the
walls follow hilltop ridges and descend along slope fall-lines, acting as dividing walls
between irrigated terraces.
The walls are integrated into the terrace retaining wall
masonry, (i.e. they are not abutting, but interlocking joins), and therefore they either
precede the construction of the terraces or were coevally constructed with them. In all
contexts, the walls are generally very tall, reaching as high as 4-5 m (but more commonly
1.5-2.0 m high), and are very thickon average approximately 2 m (e.g. Figure 5.15),
but spanning up to 10 m wide at intersection points (e.g. Figure 5.16), where they are
better described as large platforms. The walls are constructed of unworked fieldstone
facings with smaller gravel- to large cobble-size fill with mud mortar. Some sections
have double coursings of large fieldstones along one or both facings (Figure 5.16). In
other words, the walls are composed of colluvium that would have littered the
surrounding ground surfaces, and they clearly functioned in part as a means to clear field
areas of rocks in order to facilitate cultivation. However, the reason for their construction
cannot be explained in strictly agro-functional terms.
164
Figure 5.17: Ovoid features in agro-mortuary walls at YA-014. Upper

Right photo is of the same feature as in the foreground at lower left.
In all geomorphic contexts, circular and ovoid rock-lined features of varying size
are present at regular intervals along the tops of the walls (Figure 5.17). Where features
have been disturbed by looting, they are visible approximately every 0.5 to 4.0 m, and
their depths range from 0.5 to 2.0 m, forming cyst features. They are more densely
concentrated at wall intersections, as seen in Figure 5.16. Based on the presence of
disturbed human skeletal remains associated with ceramics in and around several of these
features, they have been identified as collared tombs.64 The smaller features would have
accommodated single, upright, tightly flexed interments,65 while larger tombs may have
64
Less frequently, similarly-sized features project slightly above the surrounding surface, resulting in a
small silo- or chullpa-like appearance.
65
Flexed burials predominate in the south central Andean highlands throughout the entire prehispanic
sequence. At many cave and rock overhang cemeteries where soft tissue and textiles were preserved, the
bundle-type interments were directly observed, in which the individual is wrapped in single or multiple
165
housed multiple interments or more elaborate burials for single individuals. Also, larger
circular to ovoid features protruding or bulging from the sides of walls were observed
less frequently. These averaged approximately 2 m in diameter (Figure 5.17, lower
right). Access would have been from the top, since niches or doors were not present on
the sides of any of these features. No human remains were observed with this type of
feature, and so their function remains unconfirmed, but they may represent larger tombs
for multiple individuals or more elaborate interments for high status individuals.
Alternatively, they could have functioned as storage silos (colcas).
As mentioned above, a unique gallery type tomb was found in an agro-mortuary
wall at site CO-148. Here, the walls act as dividers between irrigated contour bench
terraces (presently abandoned). The tomb is situated on a small promontory, where the
wall widens to form a large platform-like area approximately 10 m wide. The tomb has
been exposed by looting, revealing two long, narrow galleriesthe upper one measuring
4.8 m x 0.8 m, and the lower one 8.0 m x 0.8 m. Both were covered with large stone
slabs, most of which had been removed by looting, and were approximately 0.75-1.00 m
deep (see Figure 5.18). The exposed sections of the galleries were thoroughly looted and
no human remains were visible, but it almost certainly contained multiple, flexed
interments. Most intriguingly, the burial chambers were connected by a 2 m long rocklined tunnel measuring about 50 cm square.
The function of this tunnel remains
enigmaticit is clearly too narrow for a person or mummy bundle (at least, of an adult)
to fit through. One possible explanation is that it was used to move skeletal remains
textiles, and finally placed inside a tightly-wound and cinched cocoon-like vegetal-fiber capsule. See
Chapter 6 for further discussion.
166
between galleries, perhaps as new individuals were interred. However, the exposed
sections of the galleries were thoroughly looted and no human remains were visible.
Figure 5.18: Gallery tomb at CO-148, Sector A
Beyond the field-dividing function, the agricultural utility of the walls is also a
matter of speculation. On the one hand, it appears that the thickness of the wallswell
beyond that necessary to simply divide fieldswas designed to accommodate the tomb
features described above. But their great mass and extent may also have produced a
subtle microclimatic benefit by acting as wind breaks during the day, while radiating heat
167
at night (stored up through solar radiation during the day), similar to the warming effect
of bench terracing (see Chapter 2). In other words, they may have effected a marginal
increase in temperature sufficient to offset some of the frosts that settle to the valley
bottom during the night.
Discussion
The Middle Horizon marks a period of considerable expansion and reorganization
of settlement and agricultural production in the central Colca valley. Within the survey
area, the number of sites increased only modestly during the Middle Horizon as
compared to the Formative (from 30 to 37), but 54% of these sites represent new
occupations, marking a significant shift in settlement patterning. Settlements became
more concentrated in the lowest 250 m of the survey area, and agro-mortuary wall
complexes expanded in number and area on and around the valley bottom. Differential
treatment of the dead, as reflected in a variety of tomb forms, provide the first provisional
indices for social inequality. However, since architecture and other residential features
from this period were obliterated by later occupations, and Middle Horizon ceramics
were scant at most settlements, little more can be said about local sociopolitical
organization aside from some notable absences. Unlike the Formative, no settlement
stands out for its size or organization. No site size hierarchy is apparent, and there is no
evidence for direct imperial administration by Wari or Tiwanaku.
Clearly, no site
registered by the survey would qualify as an imperial administrative center (by any proxy
index: size, layout, quantity of imperial ceramics, etc.).
168
However, as discussed in Chapter 3, it is equally clear that local Middle Horizon
ceramics are derivative of Wari, not Tiwanaku, styles. Despite the ubiquity of Chivay
source obsidian throughout the Titicaca Basin and at Tiwanaku itself, no Tiwanaku (or
Tiwanaku-influenced) ceramics were recovered during the survey, or by any other
investigation in the valley. Therefore, assuming that this obsidian trade is reflective of
broader economic flows, it appears that local groups maintained strong and long-standing
trade connections with the Tiwanaku heartland while remaining politically independent
of Tiwanaku. There is no evidence for Tiwanaku political influence or presence in the
valley.
As for Wari political influence, the survey data are equivocal. On the one hand,
there is a significant shift in settlementover half (54%) of the sites for the period
represent new occupations. But in the absence of any site that could be identified as a
Wari administrative center here or elsewhere in the valley, it appears that the form of this
influence was rather indirect rather than territorial in nature. It is possible that Wari
influence is stronger in the lower reaches of the valley. Ceramics reported from the site
of Achachiwa, located lower in the valley near Cabanaconde, are stylistically similar to
Qosqopa and other regional Wari variants, while only 13% of the Middle Horizon
ceramics collected during this survey are decorated, and these share only a few isolated
motifs in common with Wari imperial styles. However, as discussed in Chapter 3,
Achachiwa probably represents a local fortified Middle Horizon settlement, and not a
Wari administrative center. Clearly, evidence for Wari imperial presence (in terms of
ceramic assemblages and the presence of planned settlements) strengthens progressively
169
towards the mid- to coastal elevations along the Colca/Majes/Caman drainage and other
Pacific valleys of Arequipa (also discussed in Chapter 3). In general, Wari imperialism
appears to have focused primarily on the littoral to mid-valley elevations in the south.
Within this regional context, the central Colca valley represents a terminus of Wari
influence along the coastal-highland axis of the drainage, constituting a porous boundary
or buffer zone between Wari and Tiwanaku spheres of influence.

The survey findings presented in this chapter establish that the earliest period of
agricultural settlement in the valley occurred during the Formative Period (1500 BC-AD
500). Ceramics broadly diagnostic of the Formative were identified, and 30 Formative
Period sites were registered during the survey. Settlement was concentrated on the valley
floor and lower slopes. The largest settlement was just under 4 ha in size, markedly
larger than any other settlement for which site area could be determined. However, no
indices for the development of social inequalities were identified. The settlement pattern
for the Formative is characterized by undifferentiated hamlet- and village-size
habitational sites.
The concentration of settlements and agricultural features (agro-
mortuary wall complexes) on and around the valley bottom indicate that early agricultural
production was more concentrated in these areas than previously hypothesized. Hunting
in the puna continued during the Formative, and logistical and residential encampments
show continuity of occupation from the Archaic Period.
170
During the Middle Horizon, the number of sites increases slightly (from 30 to 37),
but 57% of these sites represent new occupations, marking a significant shift in
settlement. Habitational sites were more concentrated in the lower elevations of the
survey area than during the Formative, and evidence for valley bottom agriculture in the
form of agro-mortuary wall complexes also expanded. Most of this expansion occurred
on the pampas and rolling volcanic hills on the north side of the river, to the southeast of
modern Coporaque, in association with a cluster of settlements. While site areas for these
settlements could not be determined, they later become large villages and towns during
the LIP and Late Horizon. Overlying occupations at most settlements hinders inference
of social organization during the Middle Horizon, but the diversity of mortuary features
dating to this period provides the first tentative indices for power asymmetries in the local
prehispanic sequence. There is no evidence for direct imperial administration by either
Wari or Tiwanaku, although evidence for Wari influence, as reflected in local ceramics,
is stronger. The observed shift in settlement during the Middle Horizon may therefore
relate to more general geopolitical changes occurring throughout the region, and the local
population was clearly engaged in regional exchange networks, but there is no evidence
for a major influx of population in the form of a direct imperial conquest or occupation of
the valley. Thus, prior to the ethnogenesis of the Collagua polity, the Colca valley
appears to have been situated at the edge of Wari political influence and Tiwanaku
economic influence.
171
Chapter 6
The Late Intermediate Period and Late Horizon Occupations
Introduction
Major transformations in the scale and organization of local communities
occurred during the five centuries spanning the Late Intermediate Period (AD 1000-1450)
and Late Horizon (AD 1450-1532).
In this chapter, I present the survey data that
document a shift from heterarchical, non-centralized community relations during the LIP
to more centralized, hierarchical sociopolitical organization under Inka rule.
I first
present a summary of the settlement pattern data, which signal a marked expansion of
settlement during the LIP, when the primary settlements in this core area of the Collagua
ethnic polity were established, as well as the subsequent Inka occupation, when imperial
structures were built at each of the large settlements with local elite architecture and an
administrative center was established at the site of the reduccin of Yanque. I then
discuss the distinctive features of local Collagua domestic architecture, as well as the
significance of intra- and inter-site disparities in the size and elaboration of houses, the
distribution and types of Inka imperial structures, and the elaboration of large complexes
of mortuary monuments associated with ancestor veneration near the principal
settlements in the survey. The infrastructure of agricultural and pastoral production
clearly expanded markedly with settlement during the LIP and Late Horizon, and I
describe the ways in which the system of hydraulic and agricultural landscape
engineering that developed during these times reflects a coordinated regime of multisettlement communal water and land allocation. I also suggest that most of the canals
172
and bench terraces in the survey area were built and used during the LIP, and that Inka
efforts at increasing (surplus) agro-pastoral production were focused on intensifying
camelid production in the puna and improving extant agricultural infrastructure while
regimenting agricultural inputs and building terraces in a few previously uncultivated
areas. I also discuss how multi-settlement community relations extended to, and were
reinforced by, common defense against external threats. I argue that defensive site
locations and the elaboration of hilltop fortifications signal a period of frequent violent
conflict during the LIP and probably included the Inka conquest of the valley. Finally, I
provide summary descriptions of each of the principle settlements of these time periods.
Chronological Considerations
The ceramic sequence that I developed divides local Collagua ceramics into four
stylistic categoriesCollagua I, II, III, and Collagua Inkathat I group into two
chronological components: LIP (Collagua I and II) and Late Horizon (Collagua III and
Collagua Inka).
The sequence builds on the preliminary chronology developed by
Malpass and de la Vera in the 1980s (de la Vera Cruz Chvez 1987, 1988, 1989; Malpass
and de la Vera Cruz Chvez 1986, 1990), and the work of Brooks (1998:317-356), who
differentiated Collagua ceramics from Chuquibamba (Cardona Rosas 1993; Kroeber
1944; Morante 1939; Sciscento 1989), Churajn (Kroeber 1944; Lumbreras 1974b:208),
and other regional styles.
Radiocarbon dates associated with local Middle Horizon
ceramics excavated by Malpass and de la Vera Cruz provide ante quem dates for the
beginning of the Collagua sequence (Collagua I). These dates, AD 429-694, and AD
173
436-779 (calibrated 2 sigma)66, roughly coincide with the first half of the Middle
Horizon. Thus, Collagua I ceramics may overlap slightly with the terminal Middle
Horizon. The end of the sequence, marked by clear Inka stylistic influence, is positioned
according to the generally-accepted dates of Inka imperial occupation of the region (AD
1450-1532), which are undergirded by radiometric dates from the neighboring western
Titicaca Basin (e.g., Bauer and Stanish 2001:251-255; cf. Stanish 1997a:47-48).
Currently, there are no local radiocarbon dates from undisturbed contexts associated with
decorated Collagua I or II ceramics, so the LIP portion of the sequence must be
considered provisional.
Future excavations and chronometric data should allow for
evaluation and refinement of the current sequence.

The sequence derives from a similiary seriation approach (Rowe 1961:326-327)
based on differences in formal and decorative elements, but is also bolstered by crossdating with related styles in surrounding locales (see Appendix A). Bowl and plate forms
dominate the diagnostics, and the general trend over the sequence in terms of form is a
change from more constricted, globular bowls of Collagua I, to more open, flat-bottomed
bowls and shallow plates of Collagua III and Collagua Inka from the Late Horizon
(Figure 6.1). Accompanying this morphological change is a shift in the placement of
decoration from the external to internal surface between Collagua II and Collagua III.
Collagua I decorations, executed in black on red and black and white on red on the vessel
exterior, are organized in horizontal design fields and show continuity in design motifs
66
I also discuss these dates in Chapter 3. The sample numbers and uncalibrated values for the assays as
originally reported by Malpass and de la Vera Cruz are: WIS-1713 (1440 BP 80) and WIS-1714 (1400
BP 80) (Malpass and de la Vera Cruz Chvez 1986:163, Table 1; 1990:). I calibrated the dates with
Calib 4.3, using the probability distribution curve (Stuiver and Reimer 1993; Stuiver, et al. 1998). The 1
sigma values are AD 537-669 and AD 544-757, respectively. See Chapter 3.
174
from local Middle Horizon ceramics, such as Qosqopa and other regional Wari variants.
Collagua II bowls are intermediate in form between the constricted forms that dominate
Collagua I, and open forms of Collagua III and Collagua Inka. Collagua II decorations
are executed in black on red only, and are generally not delineated as horizontal design
fields. The thick-lined curvilinear motifs of Collagua II are broadly similar to Altiplano
Period (local LIP) styles of the western Titicaca Basin, such as Pucarani Black on Red
(Stanish 1997a:47-48, 153 [230.001-3]), Kelluyo (Stanish 1997a:46-47, Figure 22) and
Tanka Tanka Black on Orange (Hyslop 1976:431-435). Collagua III and Collagua Inka
ceramics exhibit clear indices of Inka influence.
Collagua III bowls show greater
continuity in slip color, surface treatment, and design motifs with Collagua I and II, but
with formal details, such as rim protuberances typical of Inka bowls (e.g. Stanish
1997a:47, Figure 99). Collagua Inka vessels are local variants of Cuzco pottery, executed
in bichrome and polychrome, and are well-crafted in terms of firing, surface treatment,
and decoration (see Figure 6.1). For detailed discussion and further illustrations of
Collagua ceramics, see Appendix A.
175
Figure 6.1: Collagua ceramic sequence (lot numbers indicated)
176
The Late Intermediate Period Settlement Pattern: Overview
Clearly, a major expansion of settlement occurred during the LIP, both in terms of
the number of settlements and the size of the LIP components at sites previously
occupied during the Middle Horizon (Figure 6.2). There is strong evidence for increasing
political and economic inequalities during the LIP, such as the development of a threetiered settlement size hierarchy, inter- and intra-site disparities in the size and elaboration
of domestic architecture, and major differences in tomb elaboration and mortuary
treatment. However, as I discuss below, no settlement stands apart in terms of size,
centrality or elaboration of architecture during the LIP. Instead, an elite class appears to
have inhabited several large settlements, and it is not until the Late Horizon that the
settlement now occupied by the reduccin of Yanque became a primary political center.
Fifty three sites have LIP components, including 19 settlements, nine ceramic
concentrations (including middens and rubble mounds), nine agro-mortuary complex
sites, 10 cemeteries, and three fortifications.
The settlement pattern reflects both
continuity and growth from the preceding Middle Horizon occupation. On the one hand,
several new settlements were establishedalmost half (14 of the 31) of LIP settlements
lacked Middle Horizon components. On the other hand, two thirds (17 of 25) of Middle
Horizon settlements (including scatters, middens, and rubble mounds) continued to be
occupied during the LIP. Given the small Middle Horizon collections and overlying LIP
occupation, I cannot directly compare component areas at these settlements, but ceramic
collections provide a general proxy measure for intensity of occupation. Only 75 Middle
177
Horizon sherds were collected from these 17 sites, averaging four sherds per site, while
LIP collections averaged 16 sherds per site, for a total of 270a 260 percent increase.
While the settlement pattern is not centralized, Uyu Uyu (YA-050), San
Antonio/Chijra (CO-100), and Tunsa/Llactapampa (CO-150) stand out collectively in
terms of the number, size, and elaboration of domestic structures at these sites (see
below).
Uyu Uyu and San Antonio are located on promontory/hillside settings
overlooking the pampas on the alluvial surface above the valley bottom, and are
surrounded by extensive contour and linear bench terrace complexes in warm
microclimates.
During the survey, we registered the sites of Tunsa (CO-163) and
Llactapampa (CO-150) separately because they are located over 100 m apart, but they
probably functioned as distinct housing sectors of a single, dispersed settlement. As a
single settlement, Tunsa/Llactapampa is of comparable size to Uyu Uyu and San
Antonio, both in terms of area and number of houses. The largest LIP settlements with
elite architecture in the surrounding areas beyond the survey zone, such as Juscallacta
near Chivay to the east (Brooks 1998; Guerra Santander and Aquize Cceres 1996), and
Malata and Achomaniy near Achoma to the west (Oquiche Hernani 1991:131-140; Shea
1997), are in the same size range or smaller than the largest settlements within the survey
zone, so it appears that this decentralized pattern holds for the entire central valley.
Two of the second tier village-size settlements in the kichwa and suni zones, CO127 (Lama) and CO-164 (Kitaplaza), are clustered near Tunsa/Llactapampa on the
pampas to the southeast of Coporaque. One village-size settlement, Chiquero (YA-032),
is located in a promontory setting. After being virtually abandoned during the Middle
178
Horizon, this site was resettled during the LIP and continued to grow during the Late
Horizon. In the puna, two village-size LIP pastoralist settlements were registered for the
LIP, one on the north side of the river (YA-093) and one on the south side (YA-034).
Architectural remains were preserved only at YA-093, where we recorded 12 circular
houses and associated corrals, while at YA-034, occupation continued from earlier
Archaic, Formative, and Middle Horizon times.
The remaining twenty hamlet-class settlements and ceramic scatters make up the
bulk of the site count for the LIP, and are dispersed throughout most physiographic and
ecological contexts in the survey area. Three agriculturalist hamlets (YA-045, YA-054,
and CO-103) with domestic architectural remains are located amidst agricultural fields
near larger villages and towns. At the site of Llactarana (YA-054), located 500 m
southwest of Uyu Uyu, seven small houses were registered and mapped amidst irrigated
bench terracing. Six prehispanic houses dispersed throughout bench terraces in the upper
kichwa zone were registered at YA-045, located just under a kilometer from both Uyu
Uyu and San Antonio/Chijra. We registered twelve houses at the hamlet of Lama (CO103), which is situated on the pampa to the south of Coporaque, within a kilometer of the
cluster of four larger villages (CO-127, CO-150, CO-163, and CO-164) on the valley
bottom. Some of the nine hamlet-size ceramic concentrations, middens, and rubble
mound sites may represent domestic residues where later agricultural field clearing and
wall building has obliterated the original domestic context, but other functions are
possible at these highly disturbed sites. In the puna, one hamlet-size open air ceramic
scatter was recorded at CO-105. Three small puna sites are rockshelters (YA-066, YA-
179
082, CO-156), which were probably used as temporary refuges or encampments for
pastoralists.
Table 6.1: Late Intermediate Period sites (N=53)
Site No.
1
2
4
7
8
9
14
29
30
32
34
41
44
45
47
48
50
52
53
54
56
59
66
82
88
93
98
100
103
105
114
118
119
120
121
123
127
128
147
148
149
Sector
B/C
A/B
C
B-D
A-E
A-X
A-D
A-E
A/B
Altitude
3444
3495
3413
3489
3403
3478
3533
3615
3575
3665
4187
3484
3510
3556
3497
3554
3527
3559
3674
3572
3521
3808
3864
4623
4286
4292
3763
3666
3558
4348
3665
3697
3759
3706
3537
3757
3563
3606
3579
3573
3552
Ecozone
Kichwa
Kichwa
Kichwa
Kichwa
Kichwa
Kichwa
Kichwa
Suni
Kichwa
Suni
Puna
Kichwa
Kichwa
Kichwa
Kichwa
Kichwa
Kichwa
Kichwa
Suni
Kichwa
Kichwa
Suni
Puna
Puna
Puna
Puna
Suni
Suni
Kichwa
Puna
Suni
Suni
Suni
Suni
Kichwa
Suni
Kichwa
Suni
Kichwa
Kichwa
Kichwa
Site Class
Midden/Rubble Mound
Midden/Rubble Mound
Midden/Rubble Mound
Midden/Rubble Mound
Small Cemetery
Midden/Rubble Mound
Large Village
Large Village
Small Scatter
Small Cemetery
Hamlet
Small Cemetery
Isolate
Town
Scatter
Isolate
Small Village
Scatter
Scatter
Rockshelter
Rockshelter
Small Village
Small Village
Large Cemetery
Town
Hamlet
Isolate
Small Cemetery
Small Cemetery
Small Cemetery
Small Cemetery
Midden/Rubble Mound
Small Cemetery
Town
Site Area
0.51
0.06
0.61
0.01
0.04
0.73
91.75
0.02
23.00
1.86
2.19
unknown
0.01
1.51
0.01
0.04
4.26
0.03
0.25
1.75
0.02
0.25
0.06
0.01
0.70
0.75
0.05
8.65 (12.35)
0.25
0.06
0.01
0.01
0.04
0.01
0.01
0.01
4.23
0.01
27.94
35.46
23.24
180
150
151
154
156
161
163
164
165
166
167
168
169
A/B/D/F
3513
3822
3767
4448
3626
3528
3556
3543
3545
3537
3547
3568
Kichwa
Puna
Suni
Puna
Suni
Kichwa
Kichwa
Kichwa
Kichwa
Kichwa
Kichwa
Kichwa
Small Village
Hamlet
Small Cemetery
Rockshelter
Small Scatter
Town
Town
Small Fortification
Small Fortification
Small Fortification
Table 6.2: Late Intermediate Period site counts by class

Site Class
Count
Settlements
19
Agro-Mortuary Complexes
9
Cemeteries
10
Rubble Mounds/Middens/Scatters
9
Fortifications
3
Rockshelters
3
Total
53
1.40
0.40
0.02
0.01
unknown
5.75
3.76 (5.70)
1.71
5.29
7.28
9.67
1.86
181
Figure 6.2: Late Intermediate Period sites in the survey area
182
The Late Horizon Settlement Pattern: Overview
The local settlement system reached its apogee in terms of the number of
settlements and occupied area during the Late Horizon. There are 72 sites with Late
Horizon components19 more than the LIP. These include 29 settlements, 14 ceramic
concentrations, 12 agro-mortuary complex sites, eight cemeteries, five rockshelters, and
four fortifications. The Late Horizon settlement pattern is clearly a direct outgrowth of
the hamlets, villages, and towns established during the LIP. Eighty seven percent (46 of
53) of LIP sites also had Late Horizon components. Amongst settlements (including
ceramic scatters and rubble mound sites), the continuity is even more striking: 27 of the
28 LIP settlements continued to be occupied during the Late Horizon.
Because LIP and Late Horizon components are commingled at most settlements,
it is not possible to quantify the expansion of settlement in terms of settlement area or
house counts. Again using ceramic collections as a gross proxy measure, it appears that
settlement expanded considerably during the Late Horizon.
Of the 27 settlements
(including ceramic scatters and rubble mound sites) with both LIP and Late Horizon
ceramics, over three times as many Late Horizon sherds were collectedthat is, 1468
Late Horizon sherds versus 433 LIP sherds. Although some of the predominance of Late
Horizon ceramics may reflect their more recent deposition, this is offset to a degree by
the fact that the Late Horizon represents a much shorter time span than the LIPless
than 100 years (ca. AD 1450 to 1532) versus over 400 years for the LIP (ca. AD 1000 to
1450).
Moreover, LIP and Late Horizon components are thoroughly intermixed in
surface contexts due to continued cultivation, construction, looting, and other
183
disturbances, so these differences are probably generally reflective of settlement intensity
(density of population, houses, etc.).
While there is strong locational continuity of settlement during the Late Horizon,
three important changes signal significant sociopolitical changes brought about by Inka
administration. First, the primary locus of political power probably shifted from the
roughly coequal elite LIP settlements of Uyu Uyu (YA-050), San Antonio/Chijra (CO100) and Tunsa/Llactapampa (CO-163/150) on the north side of the river to the site of
Yanque (YA-041) on the south side of the river, which became the largest settlement in
the survey (Figure 6.4).
While the layout of the prehispanic site of Yanque was
obliterated by the construction of the reduccin, the extent and concentration of Late
Horizon ceramics (including an unusually high proportion of elaborate polychromes)
clearly signal the presence of what was the largest settlement in the survey area. Also,
the only Inka cutstone masonry registered in the survey is found in Yanque, and while the
masonry is not in its original structural context, its presence itself strongly suggests that
an important administrative center was present at Yanque. Cutstone masonry was quite
rare even in the Inka heartland outside of Cuzco, and was reserved for administrative
centers and special ceremonial sites (Niles 1993).
Second, as Yanque became the
primary administrative center, the major LIP settlements on the north side of the river
appear to have functioned as secondary centers. These sites bear the stamp of Inka
occupation in the form of prominently located, large, multi-door kallankas, or great halls
(at YA-050, CO-100, CO-163, and CO-164). The largest kallankas are found in close
association with elite Collagua houses at the largest LIP settlements (see section on Inka
184
architecture below). Third, a keen imperial interest in raising camelid production is
reflected in a marked expansion of settlement in the puna. Nine of the 16 Late Horizon
settlements that lack LIP components are pastoralist settlements located in the puna.67 As
I discuss below, these settlements are associated with major hydraulic features, signaling
an intensification of pastoralist production under Inka rule. Thus, the survey findings
provide strong evidence for centralized, coordinated Inka rule in the central Colca valley,
with a primary administrative center in Yanque.
These findings differ significantly from the previous document-based view of the
Inka presence in the central valley. The oft-cited reference in the literature to Coporaque
as the Inkaic capital of the province (e.g., Cook 1982, 2002; Guillet 1992; Neira
Avendao 1961, 1990) can be traced to a single story related by the Franciscan friar Luis
Jernimo de Or, who led the Franciscan mission in the Colca valley during the 1590s
(1992 [1598]:159 [41], my translation):
In the service of Mayta Capac Inka, who had as his wife Mama Yacchi, a
native of the Collaguas, the Indians made in that province a great house all
of copper to accommodate the Inka and his wife, as if coming to their
homeland they came to visit, and I have experience in that province, and I
diligently searched for the copper and found a quantity in the possession
of an old Indian in his depository, and four large bells were made and still
copper remained left over, and asking where the remainder was, they said
that they gave it to Gonzalo Pizarro and his militia to make horseshoes, for
fear that they would suffer the fate of a principal cacique who the tyrant
had burned for not revealing its whereabouts.68
67
Other Late Horizon sites lacking LIP components include five small Late Horizon ceramic
concentrations in the kichwa zone (YA-010, YA-012, YA-042, YA-046, and YA-055) and one in the puna
(CO-153), three cemeteries (YA-003, YA-049, and CO-122) and three agro-mortuary wall complexes (YA015, YA-026, and YA-028).
68
"En fervicio de Mayta capac Inga, que tuuo por muger a Mama Yacchi natural de los Collaguas, hizieron
los indios de aquella prouincia vna grande cafa toda de cobre para apofentar al Inga y a fu muger, que como
a patria la vinieron a vifitar, de lo cual tuue relacion en aquella prouincia, y con diligencia que pufe en
descubrir el cobre, halle cantidad en poder de vn indio viejo depofitario del, y se hizieron quatro campanas
185
In this passage, Or never refers to Coporaque as the capital of the Inka
province of Collaguas, but the allusion to a copper palace has been interpreted as
such. The reference to the marriage between Mayta Capac and the Collagua
princess most likely refers to a high ranking member of his commemorative
corporation (panaca) rather than to Mayta Capac himself, since he is only the
fourth Inka in the dynastic sequence, and his reign predates the expansionist
imperial period by at least four generations (cf. Pease G. Y. 1977:140-141).69
Such marriage alliances between the panacas and provincial aristocracies were a
key strategy in Inka expansionism, as well as a primary currency of the internal
politics of the Cuzco elite (see D'Altroy 2002:86-108). The story most likely
represents a (probably somewhat garbled) translation of local oral historical lore
that relates a marriage alliance and the construction of an elite Inka structure in
the Coporaque area. My archaeological findings suggest there was a significant
Inka imperial presence in the Coporaque area, but subordinate to the primary
center established at Yanque.70
grandes y aun fobro cobre, y preguntando por lo demas que faltaua, dixeron que lo auian dado a Gonalo
Piarro y a fu exercito para hazer herraduras de cauallos, con temor de que a vn cacique principal que no lo
quifo descubrir hizo quemar el tirano" (original orthography).
69
As pointed out by Pease(1977:141), it is quite possible that the head of the panaca of Mayta Capac
would bear his name.
70
By contrast Brooks (Brooks 1998:428) has questioned the historicity of the account, and attempted to
debunk the story as a hoax, arguing that the Inkas never established a strong direct imperial presence in
the central Colca valley. In any case, my findings are certainly not consistent with Brooks accompanying
assertion, based on brief site walkovers around Coporaque, that Coporaque and the nearby ruins of
prehispanic villages were never more than a cluster of sleepy mountain agricultural hamlets throughout all
of prehistory (Brooks 1998:428).
186
Figure 6.3: Late Horizon settlement pattern map
187
Figure 6.4: Detail of Late Horizon settlements in the central area of the survey zone
188
Site No.
1
2
3
4
6
7
8
9
10
12
14
15
25
26
28
29
30
32
34
40
41
42
45
46
48
49
50
52
53
54
55
56
59
61
66
78
89
90
93
94
98
100
102
103
105
106
Sector
A
A/B
A/B
B
C
B-D
Table 6.3: Late Horizon sites (N=72)

Altitude Ecozone
Site Class
3444 Kichwa
Isolate
3495 Kichwa
Midden/Rubble Mound
3414 Kichwa
Small Cemetery
3413 Kichwa
Midden/Rubble Mound
3471 Kichwa
Isolate
3489 Kichwa
Midden/Rubble Mound
3403 Kichwa
Small Cemetery
3478 Kichwa
Ag. Feature/Scatter
3482 Kichwa
Midden/Rubble Mound
3460 Kichwa
Scatter
3533 Kichwa
3509 Kichwa
3666 Suni
Isolate
3658 Suni
Agro-Mortuary Complex (Fragment)
3621 Suni
3615 Suni
3575 Kichwa
3665 Suni
Large Village
4187 Puna
Large Village
4149 Puna
Isolate
3484 Kichwa
Town
3448 Kichwa
Midden/Rubble Mound
3556 Kichwa
Hamlet
3494 Kichwa
Scatter
3554 Kichwa
Isolate
3480 Kichwa
Small Group Below Ground
3527 Kichwa
Town
3559 Kichwa
Scatter
3674 Suni
Isolate
3572 Kichwa
Small Village
3485 Kichwa
Scatter
3521 Kichwa
Scatter
3808 Suni
Ag. Feature/Scatter
4360 Puna
Large Village
3864 Puna
Large Rockshelter
4259 Puna
Small Scatter
4233 Puna
Isolate
4300 Puna
Hamlet
4292 Puna
Small Village
4254 Puna
Isolate
3763 Suni
Large Cemetery
3666 Suni
Town
4068 Puna
Small Rockshelter
3558 Kichwa
Hamlet
4348 Puna
Isolate
4397 Puna
Large Rockshelter
Site Area
0.51
0.06
0.01
0.61
0.04
0.01
0.04
0.73
0.03
0.08
91.75
0.19
0.03
0.33
0.01
0.02
23.00
1.86
2.19
0.15
17.96
0.01
1.51
0.34
0.04
0.04
4.26
0.03
0.25
1.75
0.03
0.02
0.25
2.83
0.06
0.35
0.26
1.00
0.75
0.04
0.05
8.65 (12.35)
0.01
0.25
0.06
0.01
189
109
111
114
121
122
123
127
128
147
148
149
150
151
153
154
156
158
159
161
163
164
165
166
167
168
169
4340
4355
3665
3537
3769
3757
3563
3606
3579
3573
3552
3513
3822
3858
3767
4448
4564
4419
3626
3528
3556
3543
3545
3537
3547
3568
Puna
Puna
Suni
Kichwa
Suni
Suni
Kichwa
Suni
Kichwa
Kichwa
Kichwa
Kichwa
Puna
Puna
Suni
Puna
Puna
Puna
Suni
Kichwa
Kichwa
Kichwa
Kichwa
Kichwa
Kichwa
Kichwa
Hamlet
Isolate
Small Cemetery
Midden/Rubble Mound
Small Cemetery
Small Cemetery
Town
Small Village
Hamlet
Scatter
Large Cemetery
Small Rockshelter
Large Fortification
Isolate
Town
Town
Town
Small Fortification
Small Fortification
Small Fortification
1.00
0.07
0.01
0.01
0.63
0.01
4.23
0.01
27.94
35.46
23.24
1.40
0.40
0.01
0.02
0.01
1.90
0.36
4.85
5.75
3.76 (5.70)
1.71
5.29
7.28
9.67
1.86
Settlement Sizes
Site Size Hierarchy
Based on larger ceramic samples and the presence of standing domestic
architecture at most settlements, the LIP and Late Horizon are the first periods for which
I estimated site sizes for all settlements. Due to the commingling of the two components,
I have not attempted to differentiate component-specific LIP/LH areas at sites where both
are present. Nonetheless, comparison of the two time periods is still useful, since there is
an expansion in the settlement size hierarchy between the LIP and Late Horizon. In areal
terms, the Late Horizon settlement site size histogram is markedly skewed, and there are
190
no clear breaks in frequency between what I have termed, for descriptive purposes,
hamlets (0.01-0.5 ha), villages (0.5-3.0 ha), and towns (3.0-9.5 ha).
However, the
addition of a much larger (17.95 ha) settlement that was almost certainly the primary Inka
administrative center for the central valley marks a clear break in scale from local LIP
settlements (Figure 6.5).71
30
25
Count
20
15
10
0
0.010.5
0.5-1.5 1.5-2.5 2.5-3.5 3.5-4.5 4.5-5.5 5.5-6.5 6.5-7.5 7.5-8.5 8.5-9.5
...
17.518.5
Site Size (ha)
Figure 6.5: Late Horizon site size histogram, by area (ha)
71
A small settlement was apparently present at Yanque during the LIP, but I could not determine the extent
of the site based on the small and widely dispersed sample of only 26 LIP sherds over the total extent of the
reduccin, as opposed to the 209 Late Horizon sherds we collected there.
191
7
6
Frequency
5
4
3
2
1
0
14
13
12
11
10
90
80
70
60
50
40
30
20
-1
-1
-1
-1
50
40
30
20
10
00
-1
-1
-9
-8
-7
-6
-5
-4
-3
-2
-1
10
House Count
Figure 6.6: Late Horizon site size histogram, by house count
Calculated in terms of number of houses, the sample size decreases, but a threetiered site hierarchy is apparent (Figure 6.6).
This excludes the primary center of
Yanque, however, because reduccin has obliterated the original LIP/Late Horizon
architecture. Given its areal extentdouble the size of the next-largest settlement, it
almost certainly occupied its own top tier. Thus, it appears that the settlement system
expanded from three- to four-tiered system between the LIP and Late Horizon.72
72
In terms of the site-size nomenclature above, hamlets as defined by house counts correspond to sites with
1 to 30 houses, villages correspond to sites with 60 to 90 houses, and towns correspond to sites with 120 to
150 houses (Figure 6.6). These area/house count measures are congruent (i.e. the same sites make up each
category using both measures), with the exception of the settlement of Uyu Uyu (YA-050), which had
especially dense housing and therefore registers as a village by areal measurement but a town by house
count. In this case, given the excellent architectural preservation at the site, the number of LIP/Late
Horizon houses (N=139) is the more accurate index of its size, and by this measure is the largest settlement
in the survey by three houses.
192
Rank-Size Analysis
Rank-size analysis provides a complementary means of characterizing
hierarchical organization in the settlement pattern.
Rank-size analysis measures the
relationship between the size of settlements and their ranki.e. the largest settlement is
ranked first, the second largest, second, etc.and is predicated on the assumption that a
systemic relationship exists between the size (whether measured by area, houses, or
population) of a settlement and its functional role and importance in a political-economic
hierarchy. In this sense, rank-size analysis rests on some of the same assumptions as
central place theory (for detailed discussion, see Falconer and Savage 1995; Haggett
1965:100-107; Johnson 1987; Paynter 1983).
Based on analysis of site-size distributions from several world regions, the ranksize rule predicts that a well-integrated and efficient hierarchical system will be
configured such that the second largest site is 1/2 the size of the largest site, the third
largest site is 1/3 the size of the largest site, etcetera. Plotted on a log-log graph, this
distribution forms a straight line with a slope of -1 from the largest to smallest site
(referred to as the log-normal distribution). The log-normal distribution serves as a
convenient reference for characterizing the degree of centralization in a settlement system
(Falconer and Savage 1995; Haggett 1965:100-107; Johnson 1987; Paynter 1983).
Deviations from the log-normal distribution may signal either a highly centralized
or decentralized system, depending on the direction of the deviation. In a primate (also
called concave) rank-size distribution, the largest site is larger than a log-normal
distribution, or stated inversely, the second through nth largest sites are smaller than a log-
193
normal distribution.
Primate distributions are generally interpreted as reflective of
centralized, extractive command hierarchies in which a near monopoly of administrative

and economic resources are concentrated in a large center. By contrast, in a convex
rank-size distribution, the largest site is smaller than in a log-normal distribution.
Convex distributions are generally interpreted as indicative of a weakly hierarchical, or
poorly integrated settlement system. However, as pointed out by Crumley (Crumley
1975, 1979, 1987), complexity and integration, need not be conflated with
hierarchy. Thus, I argue that convex rank-size distribution do not necessarily signal
weak integration or the absence of hierarchy, but can instead reflect heterarchical social
organization (Crumley 1979:144).
Site Size (ha)
10
0.1
0.01
1
10
100
Rank
Figure 6.7: Late Intermediate Period rank size graph, by site area
194
Size (House Count)
1000
100
10
1
1
10
100
Rank
Figure 6.8: Late Intermediate Period rank size graph, by house count
100
Site Size (ha)
10
0.1
0.01
1
10
100
Rank
Figure 6.9: Late Horizon rank size graph, by site area
The rank-size graphs for these two periods show a shift from a convex
distribution during the LIP to a log-normal distribution during the Late Horizon. This
trend would almost certainly be more marked if house counts could be used to compare
the two time periods, since the convexity of the LIP distribution is underrepresented
195
using site areas. That is, the two largest LIP settlements in terms of house counts, Uyu
Uyu (YA-050) and San Antonio/Chijra (CO-100) are virtually identical in size (Figure
6.8). A comparable graph for the Late Horizon cannot be done because of the lack of
preserved architecture at Yanque (YA-041) the largest settlement of the period. In any
case, Uyu Uyu and San Antonio/Chijra, together with the sites of Tunsa (CO-163) and
Llactapampa (CO-150) were clearly important political centers in the central valley
during the LIP, but there is no evidence that any site was politically or economically
dominant.
Instead, as I discuss in more detail below, political rule and economic
decision-making appears to have been coordinated between settlements. While there is a

general expansion of settlement during the Late Horizon, the establishment of a large
settlement (and probable administrative centersee below) at Yanque (YA-041) is the
decisive factor in the shift toward a log-normal distribution during the Late Horizon. I
therefore interpret the change in rank-size distribution as a shift from a heterarchicallyorganized political-economic system under autonomous Collagua rule during the LIP
towards a more hierarchical, centralized administrative structure under Inka rule.
Collagua Domestic Architecture

Since Neiras early reconnaissance (Neira Avendao 1961), researchers working
in the Colca valley have commented on the distinctive features of local domestic
architecture, such as the high gabled rooflines, unusually tall, narrow doorways, and
finely worked tabular-style masonry of many buildings (Brooks 1998:429-430; Guerra
Santander and Aquize Cceres 1996; Linares Mlaga 1993; Malpass 1987; Neira
196
Avendao 1990; Shea 1986a, 1987). However, these observations have been either
preliminary in nature or based on observations from a single site, and previous
researchers generally ignored intra- and inter-site architectural heterogeneity in the
interest of describing a normative Collagua architectural style.
During the survey, we collected architectural data from 654 structures (excluding
mortuary structures and recently abandoned buildings), including size, layout,
orientation, masonry, and other architectural details (such as wall thickness, windows,
niches, doors, gables, plaster). I categorized most (91%, or 593 cases) as houses. I use
the term house to refer to undivided, single room architectural spaces that could have
served primarily as residential structures.73
In most cases, houses are free-standing
structures, although some houses are adjoined on their short axis. This judgmental
definition probably lumps together buildings that served distinct domestic functions (such
as sleeping versus cooking or storage spaces), but given the scant amount of excavation
in domestic contexts in the valley, it is not possible to determine these differences from
survey data.
Also, it is not currently possible to reliably differentiate LIP from Late Horizon
periods of construction and/or use for local houses due to continuity of occupation and
commingling of the two components at settlements.74 Some of the observed variation in
house morphology, masonry style and other structural details may represent change
73
Not included in this category are small annexes that adjoin some large houses, as well as mortuary
structures (chullpas) and obvious examples of public or ceremonial architecture such as Inka kallanka
buildings and probable colonial chapels. I categorized structures with only one corner or side preserved as
unknown.
74
Houses that probably date only to the Late Horizon are limited to thirteen circular houses at two
pastoralist sites that produced no LIP ceramics (eight at CO-061 and five at CO-159) and one house from
Uyu Uyu (YA-050) with a diagnostic Late Horizon sherd in its wall fill.
197
through time, and it may be possible to seriate local domestic architecture in the future.
Likewise, some houses undoubtedly continued to be occupied through the first four
decades of Spanish colonial rule, until the population was forcibly resettled during the
early 1570s into the reducciones. Of the 593 houses recorded during the survey, I
categorized 580 as LIP/Late Horizon. Despite this functional and temporal imprecision,
there is clear patterning in the layout and construction of local domestic architecture.
These patterned attributes, which I discuss below, signal shared local conceptions
regarding the proper organization of residential space and constitute the Collagua
architectural style.
Table 6.4: House count by settlement a
Site No.
House Count
6
1
45
6
48
2
50
139
54
7
61
8
93
12
100
136
103
12
127
88
150
20
159
5
163
70
164
74
Total
580
a
Excludes Colonial and Republican Period houses
House Forms
Collagua agriculturalists and pastoralists built distinct houses during the LIP and
Late Horizon: agriculturalists built quadrangular houses and pastoralists built circular
houses. All 535 quadrangular LIP/LH houses registered in the survey are located in the
agricultural core of the valley, and all LIP/LH houses in the puna are circular, which
198
account for 70% (29 of 42 cases) of all circular houses in the survey.75
Of the
quadrangular houses, most are square to slightly rectangular in floor plan. This is evident
in the length to width ratio of houses, which ranges between 1:1 and 1.8:1 (1:1 being
square), with nearly three quarters (74%) in the 1.0-1.3:1 range.76 There is no statistically
significant relationship between house size and length to width ratio (Pearsons
correlation coefficient=0.101, p=0.077).
The remaining 30% of circular houses are
interspersed with quadrangular houses in agriculturalist settlements.
They could
represent seasonal homes for pastoralist households, special function domestic structures,
or perhaps houses of ethnic colonists (mitmaq) resettled there by the Inka.
As discussed above, there is strong evidence that settlement in the puna expanded
considerably during the Late Horizon, probably as a result of efforts to intensify camelid
production by the Inkas. Nine new pastoralist sites were established in the puna heights
of the survey area during the Late Horizonthe most striking change between the LIP
and Late Horizon settlement patterns. But the overall patterning of basic house form
probably reflects the maintenance of separate architectural traditions amongst local
agriculturalists and pastoralists rather than an influx of population from other ethnic
groups or the establishment of ethnic enclaves by the Inka. Circular houses appear to
predominate throughout the upper valley, beginning around the altitude of the site of
Laiqa Laiqa near Tuti, and they share with the agriculturalist houses the distinctive
75
We also registered three ovoid structures in agriculturalist settlements, but these appear to be variations
on rectangular buildings with very rounded interior corners.
76
Based on the 308 out of 535 quadrangular LIP/Late Horizon houses that were complete enough to
measure both axes.
199
Collagua trait of having very narrow doorways (Figure 6.10).77 At the intra-ethnic level,
distinct house forms probably served as markers of herder/cultivator identities as
production intensified and became more specialized amongst both groups over the course
of the LIP and Late Horizon. At a regional scale, the general circular house form
amongst Collagua pastoralists may reflect their long-standing inter-ethnic ties with the
agro-pastoralist peoples to the east, since circular domestic structures are common to the
Qollas, Lupaqas, and other Aymara-speaking peoples of the Titicaca Basin and altiplano
(Hyslop 1976; Stanish 2003).
In sum, circular houses probably represent a local
expression of a regional puna/altiplano domestic architectural tradition during the LIP

and Late Horizon, but these local herding populations were also moved around
considerably as part of an Inka effort to intensify pastoralist production in the vast
expanses of puna around the valley.
77
Like the quadrangular houses of Collagua agriculturalists, these round houses also have very narrow (4050 cm) doorways, further suggesting that they were constructed by local Collagua pastoralists.
200
Figure 6. 10: Circular houses at the site of Laiqa Laiqa, near Tuti. Photo courtesy
of Augusto Cardona Rosas.
Common Construction Attributes

All Collagua houses have a single door, and in the case of rectangular houses, the
door is always situated on (or very near) the center of the long axis. Doorways are
slightly trapezoidal, but are much narrower than the trapezoidal doorways of Inka
architecturemost would require an adult to turn sideways to enter. Basal widths of
doorways range between only 40 and 80 cm with an average of 60.7 cm (standard
deviation = 7.6), with top widths tapering by 10-30 cm compared to basal width (with an
average of 16 cm). Doorways are also quite tall, averaging 1.8 m in height (standard
deviation = 0.29).
The two basic forms of Collagua houses have distinct roof types. Circular houses
have hip roofs, and quadrangular houses have gabled roofs; gables are always situated on
the short axis of rectangular houses (i.e. are side-gabled). The only structures I observed
201
with shed roofs were on recently abandoned houses. Windows are rare, but where
present, are located in the center of the short axis in the middle of the gable.
House walls are always double coursed with an internal fill of mud with gravelsize inclusions. The average thickness of the most house walls range between 60 and 80
cm, but some large houses with second floors can be as thick as 1.2 m. Most wall
profiles are tapered, such that their crowns are usually 5-10% narrower than their bases.
Most house walls also have a slight, inward-leaning batter. The degree of batter may
have been partly a stylistic choice (i.e. to give the building an greater or lesser trapezoidal
profile), and/or a functional necessity in order to counteract the outward force pressing
down from the weight of the roof (cf. Protzen and Batson 1993:211-212, 228-233).
Interior wall facings were plastered with a 2-3 cm thick clay matrix mixed with coarse
sand, pebbles and vegetal fiber (probably Ichu grass). The plaster was applied in at least
two layers: the first for filling uneven spaces between masonry stones, and subsequent
layers for creating a smooth plastered facing. Reddening on the exterior face of the
plaster in some houses, especially along the walls lower halves, suggests a burn episode,
perhaps when the population was forcibly resettled to the reducciones. Interior corners
are usually rounded, not sharply angular, while external corners are either made of
dressed stone or made with blocky fieldstones so that they form sharp angles. Wall
niches are common and are generally square or rectangular, ranging in size between
about 25 and 40 cm on a side, and 20-30 cm deep. Niches can be placed on any interior
wall face, generally along the same horizontal axis, and niches on exterior wall surfaces
are very rare.
202
Figure 6.11: Small house of Type 1

masonry with preserved doorway, from
the east. CO-164 (Kitaplaza), Structure 4.
Figure 6.12: Small house of Type 1

masonry with preserved doorway, from
the south. CO-163 (Tunsa), Structure 33.
Figure 6.13: Detail of doorway, from the east.

Structure 84, YA-050 (Uyu Uyu). Basal width
of doorway: 50 cm, top width: 40 cm, 220 cm
height above present surface.
203
Figure 6.14: Detail of doorway/facade, from the east. Structure 103,

YA-050 (Uyu Uyu). Type 2 masonry. Basal width of doorway: 60 cm,
top width: 50 cm, 220 cm height above present surface.
Figure 6.15: Interior northwest corner of Structure 154 at

YA-050. Note thick plaster, regular placement of niches,
tenon supports for second floor.
Figure 6.16: Detail of

central niche on north wall,
YA-050, Structure 154.
Figure 6.17: Detail of plaster, YA-050, Structure 154.
204
House Masonry
Masonry style and quality range widely amongst houses, from small, very rustic
dwellings to grand two-story constructions of well-worked cutstone mortared masonry.
During the survey, I have categorized the spectrum of masonry styles and qualities into
seven types (see Table 6.5).
At the low end of the spectrum, Type 1 masonry is
composed of uncut fieldstones of various sizes, shapes, and colors set in large amounts of
mortar (Opus incertum). Wall headers and blocky corners were left undressed and lack
header and stretcher rows to tie the walls together. Houses of Type 1 masonry are very
common: they make up 37% of the 454 observable cases (Table 6.6). Type 2 masonry,
while uncoursed like Type 1, is composed of fieldstones of more uniform shape and color
(Opus poligonale). Also, the doorways, wall heads, and corners are made with dressed
stones that are laid up in alternating header and stretcher rows. Corner stones can be
either blocky or tabular in form (Figures 6.14 and 6.15). Type 2 masonry is also the most
common: 40% of houses were built of Type 2 masonry. Type 3 masonry, making up
12% of the cases, is similar to Type 2 in the quality of wall heads, doorways, and
cornerstones, but stones of more uniform size and shape were laid up together to form
roughly coursed wall facings.
205
Table 6.5: Masonry typology
Type
1
2
3
4
5
6
7
Description
Unworked fieldstone of varying size, shape, and color, no
coursing, rough corners and heads
Some worked, selected fieldstone, no coursing, dressed
corners and heads
Worked, coursed fieldstone, dressed corners and heads
Coursed split river boulder facade, dressed corners and
heads
Coursed tabular worked slabs, finely dressed corners and
heads
Coursed rectangular worked blocks, finely dressed
corners and heads
Worked fieldstone with belt courses of alternating color,
finely dressed corners and heads
Labor Input:
Shaping
Low
Labor Input:
Construction
Low
Medium
Medium
Medium
Medium
High
High
High
High
High
High
High
High
Table 6.6: Masonry type building counts/percentages, by site

1
Site
No.
45
48
50
54
93
100
127
150
163
164
Total
Count
%
Count
%
Count
%
Count
%
Count
%
Count
%
Count
%
Count
%
Count
%
Count
%
Count
%
1
16.7%
47
34.1%
1
100.0%
5
100.0%
7
8.0%
43
55.8%
2
12.5%
15
25.0%
48
77.4%
169
37.2%
2
4
66.7%
1
50.0%
47
34.1%
46
52.9%
31
40.3%
10
62.5%
31
51.7%
12
19.4%
182
40.1%
Masonry Code
4
Total
1
16.7%
1
50.0%
27
19.6%
14
16.1%
1
1.3%
2
12.5%
8
13.3%
53
11.7%
7
5.1%
1
1.1%
8
1.8%
6
4.3%
19
21.8%
1
1.3%
2
12.5%
6
10.0%
29
6.4%
4
2.9%
1
1.3%
2
3.2%
9
2.0%
4
.9%
6
100.0%
2
100.0%
138
100.0%
1
100.0%
5
100.0%
87
100.0%
77
100.0%
16
100.0%
60
100.0%
62
100.0%
454
100.0%
Excludes 127 unobservable/no data cases
At the high end of the masonry spectrum, fine masonry houses were built of
selected fieldstones of uniform size and color that were dressed, carefully chinked and
laid up with minimal amounts of mortar. Doorways, wall heads, and corners were
constructed of dressed stones of interlocking headers and stretchers, firmly tying the
structure together. Types 4-7 are all finely executed in these respects, but each is
206
stylistically distinct. House of Type 4 masonry make up 1.8% of the cases, and all but
one of them are located in Uyu Uyu (YA-050).
Houses of this masonry type are
distinguished by their facades, which are composed primarily of rolled river boulders that
were split in half and laid up in courses with their cleavage planes facing out (a variant of
Opus mixtum) (Figures 6.21 and 6.22). The other walls of these houses were executed in
an uncoursed masonry similar to Type 2. While rolled river boulders are plentiful in the
quebradas near the site, the fact that only the facades of houses were dressed and coursed
in this careful fashion indicates that this masonry style was not simply employed as an
expedient use of local materials, but instead conveyed a uniform aesthetic and projected
the prestige of their residents. Type 5 masonry is very distinctive, presenting a laminar
visual effect through the coursing of thin, tabular-shaped slabs (Opus mixtum). No Type
5 masonry houses are found at Uyu Uyu. They are all located in sites to the east; all but
one within modern Coporaque, and most of these (19 of 29) in San Antonio/Chijra (CO100). Like the other finely-worked masonry types (Types 4-7), Type 5 masonry houses
are uncommon, making up 6.4% of the cases. This masonry type predominates amongst
the large, well-crafted houses at the settlement of Juscallacta, located east of the survey
area in modern Chivay (Brooks 1998; Guerra Santander and Aquize Cceres 1996).
Type 6 masonry is composed of coursed rectangular to ovoid blocks that are worked on
all sides of the stone facing, giving it a more regular, brick-like external appearance
(Opus quadratum). The worked surfaces resemble those of Inka cutstone masonry, but
the blocks are longer and narrower than the quadrangular blocks of late Inka masonry.
Six of the nine houses of this type are found in Uyu Uyu (YA-050). Not surprisingly,
207
most are poorly preserved, probably due to colonial and republican-era mining of their
blocks for building material. Type 7 masonry is similar to Type 3 (i.e. worked, coursed
fieldstone, dressed corners and heads), but with the addition of belt courses of white
stones, creating decorative horizontal bands around the structure. All four cases of this
masonry type are found in Uyu Uyu (YA-050).
Figure 6.18: Small house of Type 1 masonry, from the east. CO-164, Structure 28.
Figure 6.19: House of Type 2 masonry, from the southeast. YA-050, Structure 107.
208
Figure 6.20: Southwest corner of house of Type 3 masonry. YA-050, Structure 114.
Figure 6.21: Facade of large house of Type 4 masonry, from the east. YA-050, Structure 104.
Figure 6.22: Detail of split-river boulders of Type 4 masonry. YA-050, Structure 104.
209
Figure 6.23: Large house of Type 5 masonry, from southwest. YA-100, Structure 17.
Figure 6.24: Structure 17 (CO-100), from the northeast. Note tenon supports for second floor.
Figure 6.25: Large house of Type 5 masonry, from the south. CO-163, Structure 10.
210
Figure 6.26: Wall fragment of Type 6 masonry, from the east. YA-050, Structure 85.
Figure 6.27: Southwest corner of Structure 31, the second largest house at YA-050.
Western gable and southern facade are visible, as well as the interior of the eastern
gable. Note white belt courses (Type 7 masonry). Cornerstones robbed out.
211
Figure 6.28: Southeast corner of Structure 31, YA-050. Note belt courses on gable
and facade, and dressed cornerstones at base.
House Sizes
House size provides another important index of architectural variability, as well as
a proxy index for differences in social status within and between sites.78 Of the 580
LIP/Late Horizon houses in the sample, 335 (both rectangular and circular) were
complete enough to determine house sizes. For the following analyses, I use building
footprint areathat is, the total area enclosed by the exterior of a house foundationas
the measure for house size instead of interior area because we could not measure interior
78
Ulloa (1965 [1586]:332) states that Collagua commoner houses were small, thatched houses, while the
houses of kurakas were larger, and distinguishable by the great thickness of their roof thatching. He also
mentions that the timbers used to support the roof were brought from far away, including the valley of
Arequipa and from within the Colca valley gorge downstream of the village of Tapay in the lower valley
(Ulloa Mogolln 1965 [1586]:332).
212
dimensions of many houses due to danger of wall collapse and dense cactus growth
inside the structures.79
House footprint areas varied markedly, with the largest houses over ten times as
large as the smallest houses. A histogram of house sizes (Figure 6.29) shows that nearly
three quarters (73%, or 246 out of 335) are between 10 and 40 m2. This size range
probably represents the small to medium houses of the bulk of the population; that is, of
low-ranking and commoner households, although some were probably small outbuildings
of commoner or elite households. Another notable feature of the house size histogram is
the sharp drop in the number of houses greater than 40 m2, suggesting that these represent
a special class of large houses. Houses greater than 40 m2 represent the top quartile of
houses in terms of area, and based on their size, position, and masonry I identified many
of these as elite houses in the field.
79
The majority of house walls ranged in thickness between 60 to 80 cm, so the figures I present generally
exceed interior area by about 1.44-2.56 m2. Small irregularities in building shape (e.g. corners that are not
square, slightly undulating walls) are also not accounted for in the footprint area, resulting in minor
inaccuracies in the footprint area figures.
213
100
90
House Count (N=335)
80
70
60
50
40
30
20
10
0
0 - 10
20 - 30
10 - 20
40 - 50
30 - 40
60 - 70
50 - 60
80 - 90
70 - 80
100 - 110 120 - 130
90 - 100
110 - 120 130 - 140
Footprint Area (m2)
Figure 6.29: House footprint area histogram
House masonry and size appear to be reliable indices of the social status and
economic position of their inhabitants. As the boxplot in Figure 6.30 illustrates, houses
of fine masonry tend to be larger than those at the low end of the spectrum, and those of
intermediate masonry quality (Type 2 and Type 3) tend to be intermediate in size.
80
Type 6 masonry is an exception to this rule, since houses of this type are generally
smaller than those of the other fine masonry types. Two alternative scenarios could
account for this lack of correlation: their small size may be the result of a compromise
80
I use box and whisker plots to provide a visual summary of the dispersion house sizes around a central
tendency. The "box" represents the middle 50% of the data, referred to as the interquartile or midspread
range. That is, 25% of cases fall below and above the edges of the box; these values are referred to as the
lower and upper hinges, respectively. The line in the box marks the median. The lines or whiskers
extending beyond the lower and upper hinges represent the upper and lower (or first and fourth) quartiles
the lowest and highest 25% of the cases. Beyond these, outliers and extremes are individually marked.
Outliers are those cases 1.5-3 box (midspread) lengths away from the upper or lower hinge. Extremes are
those greater than 3 times the box length away from the hinges. (see Shennan 1997:44-45).
214
between masonry quality and quantity, or instead they may represent an expeditious use
of worked masonry that was quarried from the corners and doorways of other buildings to
create a small house with minimal labor investment. If the latter scenario is true, these
Type 6 houses are late additions and may have been constructed during the early colonial
period, prior to resettlement in the reduccin villages.
Future excavations could
potentially differentiate these alternative scenarios. Nonetheless, the overall pattern holds
that houses of worked masonry tend to be larger than houses of unworked masonry, thus
reflecting the ability of their inhabitants to mobilize labor and material in their
construction.
140
120
House Area (m2)
100
80
60
40
20
0
N=
119
118
30
22
Masonry Type
Figure 6.30: Boxplot of house area, grouped by masonry type.

Outlier cases marked by circles, extremes marked with asterisks.
215
140
93
House Footprint Area (m2)
120
100
17
65
31
80
10
58
60
69
40
20
0
N=
91
12
41
69
14
46
52
45
50
54
93
100
127
150
163
164
Site No.
Figure 6.31: Boxplot of house area, grouped by site. Outlier

cases marked by circles, extremes marked with asterisks.
Comparing the size range of houses by site (Figure 6.51), it is clear that large
settlements have a greater range of house sizes than small settlements, and their ranges
tend to be on the larger end of the house size spectrum. The two largest settlements, Uyu
Uyu (YA-050) and San Antonio/Chijra (CO-100) stand out together for their high
midranges, upper quartiles, and outlier values. As was the case for settlement size, these
two sites are nearly equivalent in terms of house size ranges, further indicating that they
were of approximately equally high status. The settlements of Tunsa (CO-163) and
Llactapampa (CO-150) also stand out for their high third and fourth quartile ranges and
outliers. Although these are two separate settlements according to the site boundary
criteria outlined in Chapter 2, they lie within 200 meters of one another and almost
certainly functioned as a single settlement with two clusters of houses. The houses of
Tunsa (CO-163) are more diverse in size range, and the largest houses there are in the
216
size range of the third and fourth quartiles of San Antonio/Chijra (CO-100) and Uyu Uyu
(YA-050). As discussed below, Llactapampa (CO-150) appears to represent an elite
sector of generally large, finely-constructed houses near an Inka kallanka structure on the
west edge of Tunsa. This is reflected in the absence of houses smaller than 20 m2 and the
high fourth quartile of Llactapampa.
The two remaining large settlements (towns),
Kitaplaza (CO-164) and Llanka (CO-127), have lower house size ranges and appear to
have been composed mainly of commoner agriculturalist households. Medium-sized
houses at Llanka (i.e. those in its midspread range) are smaller and more uniform in size
than at any other settlement. Houses at Kitaplaza are more diverse in size, but the largest
houses at each of these settlements reaches only the median house size of Uyu Uyu (YA050) and San Antonio/Chijra (CO-100), and their first quartiles are among the lowest in
the survey. The circular houses of Jibillea (YA-093), a small pastoralist village in the
puna on the north side of the river, range by an order of magnitude, but are all within the
small to medium range compared to the rectangular structures of the agriculturalist
settlements. The remaining two small agriculturalist villages of YA-045 and YA-054, are
exceptional given their small areas and numbers of houses, but overall large house sizes.
One large outlier house at YA-054, measuring 11.4 x 6.7 m, may represent an elite house,
but was too poorly preserved to observe masonry, doorways, or other details. In sum, the
house size data indicate that the largest settlements (YA-050, C0-100, and CO-150/163)
were inhabited by diverse populations of commoners and elites, where there were major
differences in the size of houses, while commoner houses at smaller settlements were
more homogeneous in size and smaller on average than houses at the large settlements.
217
Inka Imperial Architecture

No intact Inka architecture had been documented in the central valley prior to this
project, but houses of Inka cutstone masonry have been reported in Cabanaconde, and the
nearby site of Kallimarka, with a large central plaza and ushnu platform, appears to have
functioned as a small Inka administrative center in the lower valley (de la Vera Cruz
Chvez 1987, 1988). During the survey, we registered three great halls or kallankas
very long, rectangular buildings with multiple trapezoidal doorwaysas well other
public or administrative structures. These Inka buildings are situated adjacent to open
plaza spaces in prominent locations at each of the major settlements in the survey: Uyu
Uyu (YA-050), San Antonio/Chijra (CO-100), and Tunsa (CO-163). Also, a smaller Inka
structure with at least two (and possibly three) trapezoidal doors is located at San
Antonio/Chijra, and a structure of similar size and proportions (but poorly preserved) is
situated in front of an open plaza at Kitaplaza (CO-164).
Long, multi-door structures of this type are similarly situated along the sides of
open plazas at most major Inka settlements (Hyslop 1990; Morris 1966; Niles 1984,
1987). Large, central plazas were a key feature of Inka site layout, setting an open,
public stage for state rituals that enacted the Inka ideology of state redistributive largesse
(Moore 1996). Plazas created an central, structured ceremonial space for the ostentatious
display of state generosity through ritual feasting (Hyslop 1990; Morris and Thompson
1985). The long, multi-door structures such as the ones documented during the survey
may have functioned as feasting areas for elites, as opposed to the more public space of
218
the plaza itself. The great kallankas that line the huge plaza at Hunuco Pampa are
archetypical in this regard, and excavations there revealed high concentrations of serving
wares, especially large amphora (arbalos) for serving chicha (Morris and Thompson
1985). The kallankas documented during the survey, all situated adjacent to central, open
plaza areas, appear to follow this pattern at a much smaller scale.
The dimensions and construction details of the kallankas I documented are very
distinct from Collagua houses. Those at Uyu Uyu (YA-050) and San Antonio (CO-100),
the two well-enough preserved to measure building footprints, are much longer and
narrower than Collagua houses, measuring 29.3 x 6.8 m and 24.3 x 7.5 m, respectively.
The long axis of the kallanka at Tunsa has been partially obliterated, but four doorways
are visible, and it appears to be of similar dimensions, measuring 7.2 m wide. There are
other design similarities common to all three.
The kallanka at Uyu Uyu, the best
preserved example, has seven doorways ranging in width between 1.6 and 1.7 m wide at
the basemuch wider than Collagua doors. Their trapezoidal form is evident, although
none is preserved to its full height. This kallanka was later modified, probably during the
colonial period, dividing the interior into two rooms, with an adjoining, external room
built against the facade (Figure 6.32). Abutting wall joins clearly show that these were
not part of the original layout of the structure (Figure 6.32-D). The kallanka at San
Antonio is poorly preserved, but sections of all four walls remain standing. It probably
had six trapezoidal doorways, but none of them has both sides preserved. Worked stones
are visible on one side of the three northernmost doorways. The other three have been
looted, but the base of the wall segments between them are still present. Thus, I inferred
219
these door locations from the absence of wall foundations in the positions one would
expect the doors to be based on the spacing between the other doors. Also, trapezoidal
windows are situated between doorways in all cases where the walls are well enough
preserved, as is the case at Uyu Uyu and Tunsa (compare Figure 6.32-C and Figure 6.33).
Given the general emphasis on symmetry and regularity in Inka architecture, windows
may have been placed between each of the doors. Likewise, in cases where the gable is
preserved (at Uyu Uyu and Tunsa), a trapezoidal window was situated along the central
axis of the gable, with its base at the height of the roof hip. Trapezoidal niches, another
hallmark of Inka architecture, were placed at regular intervals in the interior of the
kallankas. Two complete niches and one partial niche are present in a preserved section
of the back wall (long axis, southern) of the kallanka at San Antonio, and two are present
on the back wall of the kallanka at Uyu Uyu. While the size and morphology of these
structures follow Inka architectural canons, their masonry in all cases is of uncoursed,
roughed-out fieldstone with dressed stones in the corners, doorways, windows and niches
(Type 2 masonry). Clearly, these buildings are not the product of specialized state
stonemasons. However, fitted cutstone masonry is very rare in general, even in the
imperial capital and amongst the royal estates and major administrative centers in and
around the Cuzco heartland (Niles 1993).
In short, these buildings were probably
constructed using local labor with some form of state oversight. Given their size and
locations, they were probably used as public halls, or perhaps the reserve of elites during
state-sponsored feasts.
220
Figure 6.32: Kallanka structure at YA-050, showing orientations of photos.
221
Figure 6.33: Detail of facade fragment of the kallanka at CO-163 (Structure 56).
Note trapezoidal window, similar to kallanka at YA-050. Doorways on either side.
Another long, multi-door structure (structure 64), smaller than the kallanka
described above, lies adjacent to a major path along a prominent ridgeline that divides
San Antonio and Chijra (CO-100). Like the other kallankas, this building is of markedly
different proportions than Collagua houses, measuring 13.5 x 5.4 m.
Two wide
trapezoidal doors (basal width: 90 cm, superior width: 70 cm, preserved to 1.3 m above
present surface) on either side of the eastern-facing long axis are visible, and a third door
may have been present in the center, in the position of a void in the front wall. Like the
other kallankas, this structure opens to a terraced plaza space (26.5 x 15.0 m) just off a
major path through the site. Another structure of the same dimensions (13.4 x 5.0 m) is
located on a terrace in front of a large patio at Kitaplaza (CO-164). Based on its
dimensions and location within the site, it appears to be another example of Inka public
222
architecture, but this is a more tentative identification, since the facade of this structure is
almost entirely obliterated, and no doorways were definable.
Figure 6.34: Corner remnant with Cuzco Inka style cutstone blocks in Yanque (YA-041), facing east.
The most elaborate Inka buildings were probably located in the site now occupied
by the reduccin of Yanque (YA-041), the colonial provincial capital. As seen in Figure
6.34, remnants of Inka cutstone masonry are present in a wall two blocks east of the
central plaza. However, the original buildings were demolished by the construction of
the reduccin, and these masonry blocks are clearly not in their original context.
223
Nonetheless, it seems likely that at least one Inka cutstone structure once stood in
prehispanic Yanque. Adjacent to this wall, colonial structures locally known as the house
of the cacique Choquehuanca are built with well-worked cornerstones that were probably
mined from prehispanic structures (Figure 6.35). Bas-relief animal motifsperhaps
representations of vizcachas81are also carved in stones adorning the exterior north wall
of the easternmost structure of the Choquehuanca house (Figure 6.36). The mixture of
LIP/LH masonry types in these structures reveal their colonial period of construction.
For example, in Figure 6.36, a thin, tabular Type 5 masonry slab is intermixed with
Cuzco-Inka like blocks in the southwest corner of the structure.
The architectural
evidence, along with the size of the settlement and concentrations of Inka polychrome
ceramics there (see below) suggest that Yanque was the site of a major Inka settlement,
and probably functioned as the primary administrative center in the valley.
81
Vizcachas (Lagidium viscacia) are large rodents that inhabit rocky areas in the puna.
224
Figure 6.35: Overview of the Casa Choquehuanca area, Yanque (YA-041).
225
Figure 6.36: Colonial house in Yanque (Casa Choquehuanca).

Note bas-relief zoomorphic motifs and mixed masonry types in
corner. From the northwest.
Cemeteries, Tombs, and Mortuary Architecture

During the LIP and Late Horizon, peoples throughout the central and south
central Andes interred their dead in above-ground chullpas, or houses of the dead,
often in large groups of chullpas forming cities of the dead. As we know from colonial
226
written sources (e.g. Guaman Poma de Ayala, et al. 1987), these cities of the dead were
conceived of as a kind of parallel settlement system for the deceased members of local
ayllus, whose living members regularly feted and consulted the mummified remains of
their focal ancestors (see Isbell 1997; Salomon 1995). Given the centrality of ancestral
mummies as focal points in ayllu membership reckoning, and the importance of ancestor
veneration and consultation in ayllu political and economic life, considerable effort was
expended in the elaboration of chullpas and the preparation of the dead for interment.
In the survey area, clusters of chullpas are situated in the rocky scarps
surrounding each of the major settlements. The largest groups of chullpas are found in
the overhanging cliffs directly upslope of San Antonio/Chijra (CO-100) at the sites of
CO-154 and CO-098 (Fatinga). The latter site is the largest cemetery in the survey,
where we registered 40 elaborate two- to three-storey chullpas (Figures 6.37-6.40). This
impressive site was probably the cemetery of the elite residents of San Antonio/Chijra.
Site CO-154 is also a large cemetery composed of 16 chullpas overlooking San
Antonio/Chijra (see Figure 6.64 below), but they are heavily looted and poorly preserved.
Each of Sectors A through I at San Antonio/Chijra, scattered throughout the rocky scarps
to the west and south of the residential area, consist of individual or paired chullpas.
Another large dispersion of chullpas built under overhanging boulders is located on the
western and southern flanks of Pampa Finaya, just above the abandoned Inca canal (see
Figures 1.2 and 1.3 above).
These may pertain to the inhabitants of the nearby
settlements of Kitaplaza and Llanka on the pampas below. We observed other poorly
preserved, rectangular, free-standing chullpas scattered throughout the rolling hills near
227
Tunsa/Llactapampa (CO-163/150), including some within nearby hilltop fortifications
(discussed below) and others within the large agro-mortuary wall sites of CO-148 and
CO-149.
Clearly, chullpas were reserved for high status individuals. Most commoners
probably continued to be interred in the subterranean, rock-lined collared tombs that are
scattered throughout settlements. These collared tombs are generally round to slightly
ovoid in plan view and cylindrical in profile, ranging in size between 50-125 cm in
diameter, and 75-100 cm deep, and were covered with long, flat stone slabs. To the
adjacent south of Uyu Uyu (YA-050), about 12-15 ovoid to rectangular subterranean
collar tombs are scattered around the top of a small hill (YA-049). However, these tombs
are severely looted and only a few isolated skeletal elements were visible. During
reconnaissance in 1998, I observed a small group of poorly preserved chullpas on the
slopes to the immediate northeast of Uyu Uyu, but these have since been completely
obliterated by modern terrace construction.
Chullpa Architectural Characteristics

The freestanding LIP and Late Horizon chullpas we documented are virtually all
quadrangular in floor plan. Most are roughly square in proportion and measure 3-5 m on
the exterior. Chullpas under boulders and cliffs are set in slightly from the drip line with
flat facades and lateral sides that enclose irregular spaces, depending on the shape of the
overhanging area they enclose.
The placement of chullpas under the cool recesses of
rock overhangs and cliffs was almost certainly explicitly designed as a means of
228
preserving the mummified remains.
As I discuss below, textile and soft tissue
preservation is generally very good despite extensive looting.

Their masonry of most chullpas is composed of roughly coursed rubble, primarily
of tabular or rectangular slabs with varying amounts of mortar (Figures 6.37-6.42). Walls
are double coursed and 45-65 cm thick. It appears that most or all exterior wall surfaces
were also smoothed over with a mud and pebble plaster of the same consistency as the
mortar. This plastered surface may also have been finished with a fine plaster and
decorated with paint in a manner similar to the interior wallspatches of red pigment
(probably ochre, applied over a second, smooth layer of white plaster) remain on the
interior walls of the best-preserved chullpas of Fatinga (Figure 6.40). Most are of two
storeys with a semi-subterranean first floor, although a few examples with as many as
three storeys remain (e.g. Figure 6.37). Small doors provided continued access to the
dead after interment, as well as a means of interring additional individuals. The doors
themselves are generally rectangular to slightly trapezoidal in form, ranging in size from
35-60 cm in width and 55-75 cm in height, and usually constructed of three long stones
two upright members and a lintel (Figures 6.37-6.39, 6.41 and 6.42). Second floors (and
any subsequent floors, where present) are marked by distinctive belt courses of projecting
tabular header stones (Figures 6.37-6.39, 6.41 and 6.42). Some multi-storey chullpas
may have originally had only one storey, and the original roofs, marked by these beltcourses, then functioned as floors for the storey above. The base of the corbel vaulted
roof on most multi-story chullpas is marked by a cornice identical to the belt course
marking the lower floor(s). Thus, additional storeys of some chullpas may have been
229
built in a vertically modular fashion during separate construction episodes, perhaps as
each level filled with interments and more space was required.
Where in groups,
chullpas are also horizontally modular, occurring in abutting rows of self-contained

structures. The abutments are clearly visible as seams on the exterior (e.g. Figures 6.36
and 6.38), and their interiors were maintained as wholly separate spaces. As mentioned
above, some chullpas at Fatinga still have remnants of red paint on the interior walls,
which was applied over a smooth layer of white plaster that covered the coarse plaster
that covered the masonry (Figure 6.40).
230
Figure 6.37: Large, three-storey chullpa with two doors (right side),
CO-098 (Fatinga). This chullpa was heavily looted, containing only
isolated skeletal elements.
231
Figure 6.38: Two-storey chullpa at CO-098. Note

cornice at roofline, obliterated adjacent chullpa.
Figure 6.39: Adjoining two-storey chullpas at CO-098. Note cornice between first and
second storeys.
232
Figure 6.40: Remnants of red pigment applied to plaster on internal walls of a

chullpa at CO-098.
Figure 6.41: Two-storey chullpa at CO-098.
233
Figure 6.42: Small chullpa under rock overhang. Site CO-117.
Interments
Most interments in rock- and cliff-overhang chullpas are exceptionally wellpreserved in terms of soft tissue and textile preservation, but are often heavily disturbed
and disarticulated by looting. The interior of most chullpas are filled with commingled
skeletal elements from many individuals. I have termed the interment style a cocoon
bundle, since the tightly flexed mummy bundle is bound in a cocoon-like capsule of thick
vegetal fiber cable (Figure 6.43). Similar vegetal fiber bundles have been reported by
Justin Jennings (2002b) in the neighboring Cotahuasi valley, and in cave cemeteries in
the western Titicaca Basin (E. de la Vega, pers. comm. 2000). While we only observed
one completely intact bundle, it appears that inside the vegetal fiber enclosure, the
individual was wrapped in several layers of textilesboth plain and polychrome
234
decorated. Other grave goods we observed in the tomb interiors include both decorated
and utilitarian ceramics, including storage and serving vessels.
Figure 6.43: Cocoon style vegetal fiber mummy encasing. This bundle contained
the remains of a 3-4 year old individual.
Agro-Pastoral Infrastructure and Production during the LIP and Late Horizon
In parallel with the expansion of settlement, irrigated agriculture in the central
valley also reached its maximum extent during the LIP and Late Horizon. By the Late
Horizon, irrigated bench terracing covered nearly all slopes below the 3800 masl contour.
The agro-mortuary wall complexes established during the Formative and Middle Horizon
also expanded to encompass a total of 209 ha on both sides of the valley bottom LIP and
Late Horizon. As I discussed in Chapter 2, irrigated bench terracing overlies earlier
unirrigated fields along some slopes, and virtually all unirrigated fields were probably
abandoned by the LIP (and probably during much of the Middle Horizon as well) in favor
of the more productive and less risky irrigated terracing. Dating the construction and use
history of agricultural features is always tenuous, but associations between settlements,
235
canals and terracing suggest that most of the current canals and bench terraces were built
during the LIP and were probably modified and slightly expanded during the Late
Horizon. All LIP/Late Horizon settlements are located along a major canal or at a key
node in the canal network. Given the strong association between settlements and canals
and the overall continuity of settlement between the LIP and Late Horizon, it does not
appear that there was a major infrastructural overhaul or expansion under Inka rule.
Rather, I argue that Inka efforts at increasing agricultural production focused on
improving extant canals and terracing while bringing some (relatively small) steeplysloping and otherwise marginal areas into production. In addition, the Inkas probably
instituted more centralized and regimented organization of labor by coordinating
irrigation scheduling, crop rotation, fertilizing, and maintenance projects. However, as I
discuss below, there is stronger evidence for intensification of settlement and
infrastructure for pastoralist production in the puna under Inka rule. Both the number
and size of pastoralist settlements in the puna increases significantly during the Late
Horizon, and several are associated with large reservoirs and canals that were used for
controlling pasturage irrigation.
Canals, Reservoirs, and Terracing

The extensive canal networks built and used by the Collaguas during the LIP and
Late Horizon attest not only to large-scale mobilizations of labor and resources for their
construction and maintenance, but also to complex systems of supra-settlement water
management. Given the dispersed ayllu land tenure patterns that I document in detail
236
during the colonial period in the following chapters (a pattern that continues to the
present and is the norm throughout the Andes), individual households, and ayllus and
settlements as collectivities, almost certainly cultivated fields and terraces irrigated by the
waters of several primary and secondary canals during the LIP and Late Horizon. In this
sense, irrigation water was a primary currency of supra-settlement community relations
and organization. Local communities coordinated the apportionment of this scarce and
precious resource at several hydrological scales, as canal systems radically altered natural
drainage patterns. Indeed, after the establishment of the reduccionesas community
eventually became synonymous with villagethe breakdown of this multi-settlement,
river basin-scale hydraulic management led to near-constant and sometimes violent
conflict over water rights between neighboring villages in the valley (see Benavides
1997; Gelles 2000).
Upstream-downstream relationships between micro-watersheds and canals map
out these community water management negotiations. Given that the Colca river is
deeply incised, its waters were inaccessible for prehispnaic irrigation. Thus, nearly all of
the irrigation water in the valley originates from long feeder canals that divert meltwater
from the surrounding glaciated peaks (Brooks 1998; Denevan 2001; Gelles 1990, 2000;
Treacy 1989; Treacy and Denevan 1986; Waugh and Treacy 1986). The main intakes
(boquerones or bocatomas) of most feeder canals are situated at base of high altitude
snowpack or glaciers (many situated above 5000 masl), but a major portion of their
water supply also comes from the many rivulets and ravines (quebradas) that intersect
with their courses. That is, the feeder canals cross many micro-watersheds along their
237
courses, and gather this water rather than allowing it to continue flowing downstream
along its natural course. Collecting water from several sources with large feeder canals
in this way requires less engineering and labor than constructing multiple small canals
from each quebrada, while also raising water discharge and velocity and minimizing loss
to seepage (Treacy 1989:147).
However, this form of water management requires
negotiation and coordination regarding which water sources should supply which canals
and fields.
This kind of watershed management is apparent in the hydrological relationships
between feeder canal networks on both sides of the river basin. On the north side of the
valley, the 25 km Misme canal irrigates the fields around the LIP/Late Horizon
settlements within modern Yanque Urinsaya such as Uyu Uyu (YA-050) and Llactarana
(YA-054). However, the Misme canal collects water from several quebradas that would
otherwise flow into the Sahuara river, which supplies about 80% of the irrigation water
for the fields around the LIP/Late Horizon settlements in Coporaque (Figure 6.44)
(Treacy 1989:147).82 The Sahuara itself is a hybrid natural/anthropogenic river that has
been artificially channelized in many areas along its course. Its waters derive primarily
from two streams bypassed by the Misme canalthe Waynaqorea and the Aquenta. The
courses of the Waynaqorea and Aquenta have also been altered in order to force their
confluence. If left to drain naturally, the Waynaqorea stream would flow easterly into the
Qallachapi river, away from Coporaque. Instead, the Waynaqorea has been channelized
to flow more westerly toward the Aquenta, forming the Sahuara river at the point of their
82
Treacy also provides an overview of the feeder canal network for Coporaque (Treacy 1989:145-146,
Figure 35); however, his map was apparently based on interviews with Coporaque villagers, and he did not
include or discuss the Misme canal.
238
union (cf. Treacy 1989:144-147). Moving downstream from this point, the large stream
of the quebrada Chachayllo is shunted into the Misme canal83, while the next two major
quebradas downstream, the Yanchawi and Laqraqe are bypassed by the Misme and left
to flow to the lands surrounding Coporaque. After flowing under the Misme canal
(which forms an aqueduct at their intersection), the Laqraqe drains into the Sahuara, but
the Yanchawi stream is diverted from its natural drainage with a small crossover canal
into the neighboring stream of the Chilliwitira River, which then forms a feeder canal for
the western agricultural sectors of Coporaque. After these two quebradas, the Misme
canal arcs westerly and captures more water from the springs of the quebrada Umajala
and other minor streams before dropping into the agricultural sector of Yanque Urinsaya
to the west of Coporaque. There, it flows over a waterfall above the site of Uyu Uyu
(YA-050) and irrigates the fields surrounding it and the neighboring settlement of
Llactarana (YA-054). On the south side of the river, the 15 km Huarancante canal is the
primary feeder canal for the fields around the site of Yanque (YA-041) in modern
Yanque Hanansaya (Figure 6.45). This long feeder canal similarly draws water from
several quebradas which, if left to flow naturally downslope, would instead irrigate fields
to the west of the LIP/Late Horizon settlement of Juscallacta in modern Chivay (Brooks
1998; Brooks and Olivares Ayala 1998).
83
The waters of the Chachayllo, for example, have been the focus of a long-standing conflict between
Yanque and Coporaque, since the Misme canal, which irrigates Yanque Urinsaya, runs through what is
now the puna lands of Coporaque and draws the water from the Chachayllo stream, rather than allowing it
to flow downslope to irrigate Coporaque. The two villages have actually battled each other at the water
source, and a Yanque villager was killed there in 1971. The village irrigation councils appealed to the
Ministry of Agriculture in Chivay, and Yanque villagers, having claimed rights to the water since time
immemorial, retained those rights (Benavides 1997).
239
Figure 6.44: Primary feeder canals in the survey area on the north side of the watershed.
Figure 6.45: Primary feeder canals in the survey area on the south side of the watershed.
240
241
Assuming that canal volumetrics in the past were similar to those of today,
coordination would also have been required at each subsidiary node in the canal network.
As primary feeder canals divide dendritically into secondary and tertiary distribution
canals, water usually has to be switched back and forth between them since there is not
enough to supply several at once (for detailed discussion, see Waugh and Treacy 1986).
Reservoirs partially mitigate problems of seepage and lack of volume by storing and
enabling controlled release of water with greater velocity and discharge rates, and are
situated at several nodes in the canal network. However, there probably would not have
been enough water to supply more than one secondary canal at once from most primary
canals. Thus, for example, the fields surrounding the settlements of Kitaplaza (CO-164),
Llanka (CO-127), and Tunsa (CO-163) are irrigated by secondary canals that share a
single primary canal, and it would have been necessary to alternate water between them
according to some kind of rotational scheduling regime.
Given the constraints of gravity and topography, as well as the large labor
investments that would be required to reengineer canals, the courses of current feeder
canals closely approximate their ancient courses. Minor deviations between ancient and
modern courses are also observable. For example, while mapping the current course of
the Misme canal, I also mapped sections of relict canal running closely parallel to its
current course high on Waynaqorea pampa (Figure 6.44). Near the source of the Misme
canal at over 5000 masl, I also observed (but did not map due to time constraints) an
abandoned canal section that arcs around a hillside toward the opposite side of the
242
continental divide.84 This is probably the same canal reported (but not observed directly)
by Treacy (1989:150-151) as the Carhuasanta inter-valley canal, which crosses back into
the Pacific/Colca watershed and flows easterly toward the village of Tuti, following the
course of the Qallachapi River. This relict canal further illustrates the supra-local scale
of water management practiced by the Collaguas during the LIP and Late Horizon.
The survey crew mapped other abandoned canals and their associated abandoned
terrace complexes that clearly were used during the LIP and Late Horizon.
In
Coporaque, a large, approximately 1 m wide stone-lined canal (called the Inca canal
today) originally drew water from the Sahuara/Cantumayo and skirts the western and
southern flanks of Pampa Finaya, irrigating a band of linear bench terraces just above the
present canal course, arcing around the south side of the Finaya massif to irrigate the
linear bench terraces just above the large LIP/Late Horizon settlement of Kitaplaza (CO164) (Figure 6.44). Although we did not take collections, the survey crew observed
abundant LIP and Late Horizon sherds on the surface of these abandoned terraces. To the
west, another large abandoned canal system originally drew water from the Chillihuitira
and Wasamayo rivers, flowing to a large reservoir (Chilacotacocha) above Chilacota
(Figure 6.46). This canal and reservoir were previously registered and mapped by Treacy
and Denevan (Treacy and Denevan 1986). A secondary canal diverted water to the east
to irrigate a group of terraces on the steep slopes of the Chunancaya toponym sector
(Figure 6.46). From this reservoir, another canal flowed along the western margin of
Chilacota, and becomes an aqueduct in the center of Chijra, irrigating the upper sector of
84
The hydrological origin of the Amazon river is located on the north side of the Misme massif, to the west
of the origin of the Misme canal.
243
bench terraces there and in the Alto Ccayra toponym sector on the slopes below (cf.
Denevan 2001; Treacy 1989:150; Treacy and Denevan 1986:202-203). Also, along the
eastern edge of the survey in southeastern Coporaque, large complexes of abandoned
irrigated terraces line the undulating volcanic hillsides. Although I did not map it, the
long, meandering feeder canal for these terraces captured water from the
Qachulle/Qayachapi rivers to the northeast of Pampa Finaya.
Figure 6.46: Panorama of abandoned feeder canals and Chilacotacocha reservoir.
We also documented two relict stone-lined canals (RC026 and RC028) within the
boundaries of modern Yanque Urinsaya. Like the abandoned canals of Pampa Finaya
and Chijra in Coporaque, these two canals irrigated high sectors of bench terracing. Both
captured water from the quebradas and bofedales around the pampa Ccasi in the puna.
Canal RC026 skirts the southern slope of Calacachi peak (the site of a small obsidian
source) and would have irrigated a small sector of high bench terraces and sloping fields
above the Misme canal (Figure 6.44).
Llactarana (YA-054).
Canal RC028 descends the ridgeline above
Secondary distribution canals branch downslope to irrigate a
narrow band of abandoned irrigated bench terraces that surround the settlement, above
the present course of the Misme canal. The main canal continues down the ridgeline into
244
a small (8 m diameter) reservoir, and the outgoing canal would have irrigated the
abandoned terraces on the high slopes above the Misme canal to the east and south.
On the south side of the river above the reduccin of Yanque, another abandoned
canal complex and reservoir irrigated the steep eastern terraced slopes of Cerro Pallaclle
(Figure 6.44).
This canal drew its water from the Huarancante feeder canal.
The
reservoir was probably used for rotational irrigation turns using the distribution canals
that radiate out from the reservoir to each of the terrace sectors. I surveyed these terrace
complexes and conducted test pit excavations at two houses there in July 1997.
recovered Inka and colonial ceramics from my survey of the terracing, and one of the test
pit excavations produced Inka and early colonial ceramics in situ on a house floor
(Wernke 1997). This sector was therefore occupied and used during the Late Horizon
and the colonial ceramics provide a terminus post quem for their abandonmenti.e.
during Colonial or Republican Periods.
Reservoirs and Canals in the Puna

In the puna, there is evidence that the Inkas sought to intensify the production of
pasturage for camelids by establishing new settlements and augmenting the natural
hydrological regime. Nine new herding settlements appear to have been established
under Inka rule, and most of these sites are situated adjacent to large reservoirs, canals,
and dam features. Reservoirs in the puna were probably used primarily for controlled
inundation in low-relief areas in order to augment the pasturage growth of natural
bofedales. For example, a large reservoir was constructed about 250 meters east of the
245
herding village of Jibillea (CO-093, see site description below). This reservoir consists
of an 86 m long earthen berm that encloses a natural depression. The berm measures 3-4
m high on average, and is 5-10 m thick. A stone sluice valve near its southwest corner
would have regulated the water level and discharge. If filled the reservoir would flood a
large area upslope, and its outflow could have been distributed with small distribution
canals throughout the pampas below. Two similarly-sized reservoirs lie adjacent to the
neighboring Late Horizon herding settlement of YA-061 to the west.
Two larger examples of reservoir/lagoon-type features are located at the north end
of Pampa Finaya and the slopes to the west. On Pampa Finaya, a stone-lined canal that
diverts water from the Waynaqorea stream flows into a large, shallow basin. Although
no settlements are located around this feature, we did observe a handful of diagnostic
Collagua Inka plate sherds near the canal. Today this canal and reservoir are in poor
condition, but are still used. The reservoir is apparently designed to create an artificial
pasturage lagoon on the otherwise dry plain of Pampa Finaya, and not for regulating
downstream flows of irrigation water, since there are no outflowing canals from it.
However, a secondary canal did divert water from the feeder canal above the reservoir to
join the Qachulle canal at the base of the eastern flanks of Pampa Finaya. When full, this
lagoon would cover 20 hectaresa significant patch of dense pasturage.
To the
northwest of Pampa Finaya, another dam-like feature (RC-032) runs encloses a shallow,
50 m wide draw next to the Sahuara river. This feature would have inundated a few
hectares on its upstream side. Half a kilometer to the east, a canal siphons water from the
Sahuara river to irrigate a 7-10 ha size pasturage patch.
246
Carved Stone Sculptures of Terracing and Hydraulic Systems

The central importance of agricultural infrastructure in the cosmology of local
agriculturalists is reflected in the detailed, large-scale carved stone representations of
terracing, canals, and reservoirs on boulders and rock outcrops.
These impressive
sculptures dot the landscape throughout the valley, and several others have been
documented, most notably a locally-famous maqueta near the village of Maca that is
shown without fail to all tourists on the bus ride between Chivay and Cabanaconde
(Brooks 1998:287-293; Romaa, et al. 1987).
The regional distribution of these
sculptures appears to extend throughout most of the Pacific drainages in the northern half
of the Department of Arequipa. Similar examples are found in the Chuquibamba valley
to the west, as well as at LIP/LH sites in and around the valley of Arequipa, including
Tambo de Len in Chiguata (15 km east of the city) and Maucallacta in Pocsi (20 km
southwest of the city) (A. Cardona Rosas, pers. comm. 2003).
Locally referred to as maquetas (models), the sculptures have hydraulicallyfunctional miniature reservoirs and canals that may represent specific canal and terrace
systemsmore detailed study could shed light on this question. Whether or not they
were literal models of the surrounding landscape, they were probably important huacas85
used in fertility rituals, probably involving libations of chichaa key ingredient in all
agro-pastoral fertility rituals in the Andean highlandswhich could have been poured
85
The literature on huacas and their role as markers of cosmology, identity, and mediation between the
natural and supernatural worlds is vast, but as a shorthand, I refer to van de Gucthes (1999:155)concise
definition: a material object or location which received ritual attention, and the force which inhabited
that object or location.
247
into the reservoirs and through the canal courses of the sculptures (an interpretation also
suggested by A. Cardona, pers. comm. 2003, and Brooks 1998:293).
Given that the maquetas are situated far from settlements, the rituals enacted at
the maquetas may have involved the inhabitants of several settlements, and, much like
the construction and use of their full-scale counterparts, would have reified and
reaffirmed supra-settlement relations and constructs of community. Nine of the ten
maquetas we documented in the survey are located away from large settlements, and
instead are associated with terrace complexes and valley bottom fields.
It seems
reasonable to imagine that the same groups of farmers that cultivated fields irrigated by
particular canals, or those charged with distributing water to them, enacted fertility rituals
at their respective maquetas (perhaps at the beginning of each cycle of cultivation).
A cluster of four maquetas (site YA-162) is located above the terraces on the
western slopes of Cerro Pallaclle, below the Huarancante canal and above the terraces
and fields it irrigates to the south of Yanque (Figures 6.47-6.59).86 Three of these four
maquetas are very large and especially well-made. They range in size from 3-4 m on
their long axes, and mimic their full-scale counterparts with long primary (feeder) canals
and subsidiary canals that distribute water to their terraces. Several reservoirs, which
appear to have been created by augmenting natural depressions in the rock, are also
present along the tops of the boulders and feed distribution canals for the miniature
terraces. the top of several canal courses. Another large example is found on the valley
bottom amongst the agro-mortuary wall complexes of site YA-014 (Figure 6.50). This
maqueta wraps around both sides of a large boulder and also has several reservoirs that
86
Brooks (1998:287-288) earlier reported one of the three maquetas of CO-162.
248
feed miniature distribution canals. The largest example in the survey, and thus far
documented in the valley in general, covers intermittent areas of a 40 x 30 m section of
exposed bedrock on a steep slope below the Misme canal on the opposite side of the
river. In this case, there are several discrete clusters of terracing that were probably made
at several different times, and they are more eroded and not as well executed as the other
examples.87 The other examples (at sites YA-008, YA-032, YA-057, and YA-074) are
smaller (less than 2 m) but are also hydraulically functional.
Figure 6.47: Large maqueta at site YA-162. Note distribution canals descending between terrace
groups
87
Unfortunately, a recently constructed canal cuts through the center of this maqueta and it has been
severely damaged.
249
Figure 6.48: Maqueta at YA-162. Note reservoirs with distribution canals.
Figure 6.49: Another large maqueta at YA-162, showing extensive contour terracing.
250
Figure 6.50: Maqueta at YA-014. Modern wall built on top of boulder. The sculpture continues
around the end of the boulder (left side of frame) to the opposite side.
Figure 6.51: Small (approx. 1.4 m wide) maqueta at site YA-032 (Chiquero). Note distribution
canals.
251
Although dating these sculptures with any precision is virtually impossible, they
were presumably elaborated coevally with the construction and use of their full-scale
counterparts, and so they probably date to either the LIP or Late Horizon. Some carved
boulders in the Inka heartland, such as the famous Sayhuite stone (van de Guchte
1999:162) and the Piedra Cansada (see, e.g. Bauer 1998:59), which was a shrine in the
ceque system of Cuzco, also show landscape features such as terraces and canals.
However, they are stylistically quite distinct and often incorporate representations of
animals, humans, or supernatural beings (see van de Guchte 1990). These types of
maquetas appear to represent a separate tradition unique to the Pacific drainages in the
northern half of the modern Department of Arequipa.
Pukara Fortifications and Evidence for Violent Conflict

While hydraulic relationships between canals indicate coordination of resource
apportionment between settlements, evidence for violent conflict in the central Colca
valley greatly increases during the LIP. Historical accounts of warfare, weaponry, and
fortifications by both Ulloa Mogolln (1965 [1586]) and Or also suggest frequent
conflict prior to, and perhaps during, Inka imperial incorporation.
Regarding
fortifications, Ulloa (1965 [1586]:332) states only that [t]here are some fortresses on top
of some hills of little or no import. Elsewhere, he describes in detail the weaponry used
by the Collagua in warfare, which included copper-headed maces, copper axes, slings,
and bolas made of camelid nerves (tendons?) weighted with heavy copper balls (Ulloa
Mogolln 1965 [1586]:330). A passage by Or (Or 1992 [1598]:39) states that hilltop
252
fortresses were maintained by ayllus, suggesting that conflict between ayllus was
common.
Or links pre-Inkaic warfare to competition for agricultural fields, and
describes warfare in terms of defense against field raids (Or 1992 [1598]:39). Ors
account also suggests that status could be achieved through warfarein particular, he
relates an account of a man who displayed to him an antique shirt covered with the
fingernails of the men that his ancestors had killed in battle (Or 1992 [1598]:39).
Within the survey area, it appears that at least three hilltop concentric wall
fortifications (CO-165, CO-167, and CO-168) were constructed and used in the survey
area during the LIP. The three LIP fortifications (CO-165, CO-167, and CO-168) we
registered share characteristics common to the pukara forts that are ubiquitous
throughout the Andes during the Late Intermediate Period. I have identified the sites as
pukara fortifications based on their position and layout, which consists of a series of
concentric fieldstone walls around prominent hilltops in the local agricultural landscape.
All three are located on hilltops in the Qal 5 alluvial surface to the southeast of
Coporaque. The hills fan out finger-like over the surrounding pampas, providing views
up and down the valley.
All three pukaras are within 600 meters of four major
settlements (CO-127, CO-150, CO-163, and CO-164) and of one hamlet (CO-103) on the
pampas to the north and south (Figure 6.52). Settlements located away from these forts
(YA-032, YA-045, YA-050, YA-054, CO-100) are located in defensible promontory or
hillside contexts. With the exception of Yanque (YA-041), which was probably a very
small settlement at the time, all other LIP settlements were located less than 1 km from a
hilltop fortification or were situated in a defensible promontory setting.
253
Figure 6.52: Map of pukara fortifications in relation to major LIP/LH settlements, canals, and trails.
254
It is unlikely that any of the pukaras were built in response to conflict between the
residents of these settlements, since the pukaras are located virtually equidistant between
them and no settlement would have had a strategic advantage in terms of access to the
forts. For example, the villages of Kitaplaza (CO-164) and Llanka (CO-127) are both
located about 350 meters from the pukara fort CO-168. Likewise, the next pukara to the
south (CO-167), situated on a promontory adjacent to pukara CO-168, is located midway
between the large settlements of Llactapampa (CO-150) to the south and Llanka (CO127) to the north. The pukara furthest to the southwest (CO-165) is located between the
large settlement of Tunsa to the south and Llanka to the north. In short, all of settlements
on the valley bottom to the south of Coporaque are within shouting distance of a
fortification, and none holds an advantage over another in terms of terrain or line of site.
Therefore, it seems most likely that the pukaras were used in defense of threats that were
external to the these settlements collectively.
Figure 6.53: Pukara fortification of site CO-165, from the north. Arrow indicates location of
doorway in outer concentric wall.
255
Figure 6.54: Pukara fortification of site CO-168, from the northwest. Arrows denote locations of
outer and inner concentric walls.
At least two concentric walls have been identified at all three sites. Pukaras CO165 and CO-168 are typical pukaras in the sense that they are true concentric ring
fortifications composed of two walls that enclose irregular (but generally ovoid) areas
around hilltops (Figures 6.53 and 6.54). The fortification walls at these sites, composed
of double coursed, fieldstone masonry with pebble and mud mortar fill, are up to four
meters high and 1.0 - 2.2 meters in basal thickness. Pukara CO-167 is distinct in both
size and layout. The outer fortification wall of CO-167 actually encloses two hilltops
within a massive V shape measuring 520 meters in total length (about 200-250 meters
on each side). The point of the V is oriented to the northeast, while the open end to the
southwest coincides with steep scarps and cliffs that form natural barriers to the hilltops.
256
Less steeply sloping areas in this southwestern side have been protected by fortification
walls near each hilltop. The outer fortification wall of CO-167 is also generally thicker
and taller than the walls of the other two pukarasbetween 1.9 and 2.7 meters thick (at
the base) and up to 4.5 meters high (measured from the lower, exterior slope surface).
Access points were minimized at all three pukaras, and doorways appear to have
been positioned to create maze-like kill alleys. There are formal doorways still preserved
at CO-165 and CO-168. At CO-165, a full doorway including the top lintel is located on
the north side of its outer concentric wall, and a doorway (with only one side preserved)
on its inner wall is located on its south sidethat is, on the opposite end from the door in
the outer wall, thereby forcing those entering to circumambulate the hill to gain access to
the hilltop (Figure 6.53). Similarly, at CO-168, a narrow (92 cm wide) doorway is
preserved up to 80 centimeters above present surface on the northeastern side of the inner
concentric wall. While no doorway in the outer concentric wall was identified, an eroded
section of wall on the opposite side of the hill from the inner wall doorway is the most
likely area where this access would have been located. In both cases, groups attempting a
siege would have been exposed to attack as they traversed the intervening strip of hillside
between the two walls. No doorways were observed in the long, outer wall of CO-167,
and while a robust defense would have been required to prevent breaches along its length,
this potential vulnerability was perhaps offset by its great height (up to 4.5 meters).
257
Figure 6.55: Remains of a rectangular chullpa on the exterior of fortification wall,

site CO-165.
The concentric walls of the pukaras did not enclose habitational areas. They are
devoid of domestic structures, but clusters of collared tombs (both subterranean and
incorporated into the fortification walls) and chullpas are present within and around the
concentric walls. For example, at pukara CO-168, we observed seven subterranean cyst
tombs and three probable chullpa foundations. Two of the three chullpa foundations are
located on the hilltop inside the inner concentric wall, which encloses an 27 x 18 m ovoid
area. These chullpas were highly disturbed by looting, probably because they would
have been conspicuous in the local terrain. Due to the intense looting, the hilltop consists
of a large rubble mound, of which only these three chullpas could be positively identified
and, based on the amount of rubble littering the hilltop, several more chullpas were
probably once present. Five small (40 to 70 cm in diameter) circular rock-lined features
(probably collared tombs) were also present in an agro-mortuary wall remnant at this site,
258
but these features predate the fortification walls, since the outer concentric wall of this
pukara intrudes upon the agro-mortuary wall at their intersection. At pukara CO-165,
two well-preserved rectangular single-storey chullpas lie just outside of the southeast
corner of its outer concentric wall, as well as a more rustic ovoid chullpa built into an
adjacent terrace wall. Within the 40 x 17 m ovoid area enclosed by the inner concentric
wall at CO-165, a large rectangular chullpa, measuring 2.5 x 1.6 meters, as well as two
looted circular collared tombs (both measuring one meter in diameter) are situated on the
hilltop itself. Similarly, at CO-167, three large rectangular chullpas are situated on the
northwestern hilltop within a second enclosure wall. Two of these three were wellpreserved, measuring 4.0 x 2.5 and 3.1 x 2.5 m on the exterior, both with 70 cm thick,
double-coursed fieldstone masonry walls with mud and pebble fill. Part of corbel-vaulted
roofs and small doors remained intact in both structures, and we observed patches of clay
plaster painted with red pigment (probably ochre or cinnabar) in the interior of one of
these chullpas.88
Despite major disturbance by looting, preservation within these
chullpas was remarkable, and fragments of the exterior vegetal fiber wrap (fardo) from
the mummy bundles were observable. Large fragments of undecorated ceramic jars, as
well as diagnostic Collagua I and II sherds were present in and around these chullpas.
88
Given time constraints and the difficulty of gaining access to the interior of the chullpas, as well as
disturbance to the interments, a proper MNI was not possible to determine, but five adult crania were
visible in one, and eight in the other. But these figures probably represent only a small part of the actual
MNI in the tombs.
259
Figure 6.56: Pumachiri peak (CO-158), from the northeast. Arrows

denote outer stone wall around the eastern flank.
260
Figure 6.57: Outer stone wall (denoted by arrows) of CO-158 from above, facing east.
Another possible pukara (CO-158) is situated high on the 4696 masl peak of
Pumachiri, which dominates the northern horizon in Coporaque, and is a principal apu
(mountain deity) of the village. A massive fieldstone wall (one to two meters thick and
one to three meters high) runs approximately 350 meters across the steep northeastern
slope of Pumachiri (Figures 6.56 and 6.57). This wall ends at a prominent ridge on the
eastern corner of the peak on one side, and a series of scarps and cliffs on the opposite
side, thus blocking access to the peak from the north. The peak itself is protected on its
north side by cliffs, and on its south side by two more concentric fieldstone walls. The
innermost wall around the peak, averaging about 2 m thick, is generally poorly preserved,
but one formal doorway (2.2 m wide) opens to the south, giving a spectacular view of
Coporaque and the valley bottom below (Figure 6.58). The outer wall, 10 to 20 meters
downslope, also averages approximately two meters in thickness, and unlike the other
261
pukaras, this wall has three doorways: one to the southeast (2.4 meters wide), one to the
south (2.0 meters wide), and another to the southwest (2.2 meters wide). Also distinct
from the other pukaras, there were no observable tombs on the rocky peak or elsewhere
at the site.
Neither were there any residential structures at Pumachiri; however,
temporary shelterssmall (ranging from about six to ten m2), generally ovoid
unmortared fieldstone structures, often built against bouldersare scattered throughout
the site, and a dense cluster of over 100 shelters is situated downslope from the outer
concentric wall near the prominent ridge on the northeast side of the peak. Thus, large
groups probably camped at the site for short periods. Aside from their small size, the
ephemeral nature of occupation at the site is also reflected in the very low artifact
densities at the siteonly five diagnostic sherds, all from the Late Horizon, were
recovered.
262
Figure 6.58: View to the south, through doorway (center) of inner wall encircling Pumachiri peak
(CO-158). Village of Coporaque is center-left, Yanque in the background to the upper right.
These last points raise questions regarding the function of Pumachiri and its
relationship to the other pukaras: its overall layout suggests a defensive function, but the
site is also of a wholly different scale and setting than the other three pukaras. These
may pertain to two distinct size classes of pukaras that formed local- and regional-scale
defense networks.
This pattern has been documented in the Qolla region of the
northwestern Titicaca Basin, where small, local pukaras tend to cluster around larger
major fortification pukara sites such as Pukara Juli and Tanka Tanka during the LIP
(Altiplano Period) (Stanish 1997a:66, 74, 117-118; 2003:211). The three smaller pukaras
discussed above are very similar to the minor refuge pukaras of the Qolla region,
which are also located on small hilltops surrounded by agricultural fields, have no
263
evidence for habitation, and tend to enclose clusters of chullpas and cist tombs (Stanish
1997a: 36, 38, 117-118). Especially intriguing is the fact that small pukaras appear
earlier and major fortifications later in the Qolla region (Stanish 2003:210), which, while
my collections are small, appears to parallel my findings. Stanish has argued that the
small pukaras were used primarily during local-scale internecine conflict earlier in the
LIP as competing factions within the Qolla ethnic group vied for power, but the major
fortifications were used later as the Qollas unified to defend themselves when faced with
a major external threat such as invading Inka forces (Stanish 2003:210).
Figure 6.59: Pukara fortification to the west of Tuti.
264
Figure 6.60: Fortified settlement to the southeast of Tuti. Note ancient road in foreground.
This may be the case in the Colca valley as well. While the sample size of
pukaras registered by the survey is small, they clearly reflect a broader pattern of hilltop
fortifications throughout the valley.
Just beyond the southeastern survey boundary,
another probable pukara fortification is situated on a high ridge above Yanque (within
the District of Chivay). This site consists of two concentric walls that ring a hilltop along
this ridgeline. This site may have been used by inhabitants of Yanque (YA-041), the
closest major LIP/Late Horizon settlement. Pukaras are also present in the upper valley.
I have seen three probable pukaras near the village of Tuti: a large concentric ring site on
a hilltop above the valley bottom to the southeast of the village, another associated with a
settlement, also to the southeast of the village, and third on a hill to the west of the
village.
In Achoma, to the adjacent south of Yanque, Shea (1987) and Oquiche
(1991:143-152) report a fortification (Auquinikita) along a ridgeline to the south of the
265
village. Oquiche also reports a smaller fortification called Koricancha on this same ridge
(Oquiche Hernani 1991:143-149). Shea (1987), confirming earlier observations by Neira
(1961), describes fortification walls at the large LIP/LH hilltop settlements of Pilloniy
and Achomaniy in Achoma. Neira has photographed and described the large pukara of
Pachamarca (or Pachamarquilla) in the district of Maca, to the adjacent west of Achoma
in the central valley (Neira Avendao 1990:181). In the lower valley, no pukaras have
been identified to date, but de la Vera Cruz describes three fortified hilltop settlements:
Achachiwa (discussed in the previous chapter), Umawasi, and Antisana.
Although
Achachiwa dates primarily to the Middle Horizon, the primary occupations at Umawasi
and Antisana, (based on surface ceramics) date to the LIP and Late Horizon (de la Vera
Cruz Chvez 1988:74-79, 103).
In sum, violent conflict appears to have been endemic in the Colca valley during
the LIP, perhaps reflecting warfare between competing factions within the valley.
Fortifications and fortified settlements continued to be used during the Late Horizon,
indicating that, in a manner similar to the Qolla, the Collagua may have unified when
faced with the exterior threat of Inka conquest of the valley, as higher, larger
fortifications such as Pumachiri (and possibly Pachamarca in Achoma) were built and
used.
Alternatively, given the remote, high-altitude setting of Pumachiri, and its
prominence as a local apu, it may represent a ceremonial or ritual center. But these
interpretations need not be considered mutually exclusive or contradictory in terms of
tactical/strategic versus a ceremonial/ritual functions, given the well-documented
practice of ritual warfare (tinku) amongst neighboring groups such as the Canas y
266
Canchis and Chumbivilcas (Orlove 1994). Structured violent conflict could have been
considered generative, not destructive, within certain sociohistorical contexts, and in this
sense constitutive of community constructs and vital to their reproduction. While the
ephemeral nature of occupation at Pumachiri and most of the other pukaras holds little
promise for clarification of their function through further site-specific study, future
systematic survey will improve our understanding by placing them in a broader regional
context.
Descriptions of Principal LIP/Late Horizon Settlements

YA-050, Uyu Uyu
Uyu Uyu89 (YA-050) shares the top tier of the settlement hierarchy with San
Antonio/Chijra (CO-100) during the LIP and the second tier during the Late Horizon. No
previous excavations or systematic survey have been conducted at Uyu Uyu. In 1959,
Neira visited Uyu Uyu for part of a day during his brief reconnaissance of the valley
(Neira Avendao 1961), and based on its impressive domestic architectural remains, has
subsequently described the site as the original capital of the Collaguas (Neira
Avendao 1990:172).
Uyu Uyu is located on a prominent mesa-like promontory on the north side of the
river, within the agricultural sector of modern Yanque Urinsaya (Figures 6.61 and 6.62).
The residential area of the site measures 4.26 ha in area and is composed of a dense
concentration of 161 architectural spaces situated on low, broad domestic terraces. Based
on differences in the size, distribution, and elaboration of domestic and Inka public
89
Alternate spellings: Ullullu, Uyo Uyo, also called Yanque Viejo.
267
architecture at the site, I argue that Uyu Uyu housed a diverse population of both
Collagua elites and commoners, and was a secondary administrative center under Inka
rule.
Figure 6.61: Oblique airphoto of Uyu Uyu (YA-050) from the 1931 ShippeeJohnson expedition. Courtesy of the American Museum of Natural History.
268
Figure 6.62: Architectural map of Uyu Uyu.
269
Several houses are grouped together on most terraces at Uyu Uyu, forming
residential compounds such that multiple structures were probably used by individual
households, or that several, perhaps related or extended family households shared a patio
space. While their layout does not appear rigidly planned, two configurations of these
compounds can be distinguished. In some cases, several houses are situated side-by-side,
with doorways opening in the same direction. This arrangement is common amongst
small to medium houses, as in structures 21, 22, and 23, (Figure 6.62). In contrast to this
more open, linear arrangement, other compounds form an L-shaped configuration,
usually with large and elaborate houses forming the short axis of the L across the width
of the terrace, as in structures 30-35 and 121-126 (Figure 6.62), both part of compounds
that include large elite houses.
I categorized 143 of the 161 structures at the site as houses, 139 of which were
dated to the LIP/Late Horizon, and of these, 91 were complete enough to measure
building footprints. Variability in house sizes provide one index of social inequality
within the settlements inhabitants. The LIP/Late Horizon houses at the site range in size
from small 15 m2 dwellings to much larger and more elaborately constructed houses of
up to 91 m2, many of which have tenon supports for a second floor or attic. The
midspread (interquartile) range of house sizes (footprint areas) at the site falls between
31.6 and 53.7 m2, with a median size of 41.7 m2. Only the midspread range of house
areas at San Antonio/Chijra (CO-100) is this high (see below). Some of the largest and
most elaborately constructed houses in the survey area are found at Uyu Uyu. Based on
their size and quality of masonry, I identified twelve houses (structures 13, 24, 26, 31, 42,
270
65, 66, 104, 112, 114, 125, and 126) as elite residences at the site. All but two of these
houses are in the top quartile at the site in terms of size, and stand apart for their fine
masonry. Most have especially impressive facades and tenon supports for a second floor.
For example, the facade of structure 104, a large house with an adjoining annex near the
center of the site, is composed of well-fitted split river boulders (see above, Figure 6.22).
Structures 24, 31, and 125 all have belt courses of white stones forming decorative
horizontal stripes around the building (e.g. Figures 6.26 and 6.27 above). The largest
houses at the site are also unusually tall structures. For example, the gable of structure
31, the second largest house at the site, is 6 m high (Figure 6.27). In short, based on these
very large and elaborate houses present, I infer that an elite class was present at Uyu Uyu,
and residents at the site, together with those from San Antonio/Chijra, were generally of
high social status.
Furthermore, eight of the twelve elite houses (structures 125, 126, 65, 66, 42, 114,
112, and 104) are clustered in the center of the site, situated either adjacent to one another
or on the same terrace. These elite houses surround the north and west sides of a plaza,
which is enclosed on its western side by an Inka kallanka structure. This close spatial
relationship between local elite residences and Inka architecture associated with public
ritual and display of state largesse is suggestive of how local elites gained in status by
their participation in Inka administration of the central valley. Therefore I argue that Inka
rule, while centered primarily at an administrative center at Yanque, was coordinated and
mediated by local Collagua elites at Uyu Uyu, San Antonio/Chijra, and probably to a
271
lesser extent Tunsa (CO-163) and Llactapampa (CO-150)all sites with local elite
houses in close association with rustic kallankas or Inka masonry.
Surrounding the site, the flanks of this mesa are covered by spectacular contour
bench terraces which, due to their warm microclimate and frost-shedding properties, are
especially valued by Yanque Urinsaya farmers for their high maize productivity. The
overall site layout does not appear centrally organized, but instead structured around the
contours of domestic terraces that follow the local topography. The site forms a visually
integrated whole with the agricultural terracing, and in this sense the site does not appear
intrusive. Instead, the surrounding agricultural terracing accommodate the site, and their
contours are followed by the domestic terraces in the settlement (Figure 6.61). The initial
construction of the current configuration of the surrounding bench terraces therefore
probably coincides with the LIP occupation of the site. The primary canal for Yanque
Urinsaya, the Misme canal, passes over a waterfall (Ccayra Cucho) just upslope of the
northern end of the settlement before flowing westward to irrigate the bench terraces and
valley bottom fields near Ichupampa to the west. A subsidiary canal branches from the
Misme canal and runs alongside a path through the center of the site. Other feeder canals
branch around the site to irrigate the surrounding terraces. Yanque villagers today plant
crops on the former domestic terraces of Uyu Uyu, and crops are even cultivated inside of
some large structures.
272
Figure 6.63: Facade of structure 104 at Uyu Uyu, from the east. Note tenon
supports for second floor, and clay plaster on interior wall surface.
CO-100, San Antonio/Chijra

Covering 12 hectares, San Antonio/Chijra is the second largest settlement (after
YA-041, Yanque) in the survey by areal measure, and the second largest (after YA-050,
Uyu Uyu) by prehispanic house count (N=136). The site spans much of the eastern and
southern flanks of Yurac Ccacca, a prominent peak reaching 3817 masl on the north side
of the Colca river. The two toponym areas of the site are divided by a prominent ridge.
San Antonio includes the terraces and buildings on a promontory and slopes on the
eastern of the ridge, while Chijra encompasses the houses and terracing to the west of the
ridge on the broad, basin-shaped southern slopes of Yurac Ccacca (Figure 6.64). I have
included both habitational sectors as a single site because all features are within 100 m of
each other and the areas were occupied coevally.
273
Unlike Uyu Uyu, most of San Antonio/Chijra is situated on steep, terraced
hillsides that today are divided between upper, abandoned bench terraces and lower,
currently cultivated bench terraces. The lower terraces are irrigated by two canals, called
the San Antonio and Ccayra. Virtually all of the architectural remains at the site are
within the upper band of abandoned bench terraces. Prior to their abandonment during
the colonial period (see following chapter, Cf. Denevan 1987), these upper terraces were
irrigated by water from a reservoir to the northeast of the peak of Yurac Ccacca, flowing
past Chilacota (site CO-151see Chapter 4) before running through Chijra as an
aqueduct. This aqueduct splits near the center of Chijra to irrigate the terraces above the
other two canals (Denevan 2001; Treacy and Denevan 1986). The terraces below these
two canals are currently used primarily for maize cultivation, as they were during the
colonial period (see Chapter 9). Downslope of the site lies the Waykiri area, a large basin
of contour terraces and valley bottom fields situated in the warmest microclimate in
Coporaque, and particularly valued today for their high productivity (Crdova Aguilar, et
al. 1986:64-65; Treacy 1989). Immediately upslope of Waykiri are massive complexes
of linear bench terraces (in the Alto Ccayra and Ccayra toponyms areas) that span the
southern flanks of Yurac Ccacca just below Chijra. Vertical end walls divide these
terraces into regularly-sized irrigation sectors.
Their layout was clearly centrally-
planned, perhaps dividing the irrigation sectors into regularly-sized production units.90
These bench terraces continue upslope through San Antonio/Chijra, and serve as
domestic terraces for houses. But domestic terraces are not set apart as such; rather,
90
Some of the terraces in Ccayra were reconstructed in the 1980s and 90s by DESCO (a Peruvian N.G.O.),
but these are easily distinguishable from the surrounding ancient terraces.
274
houses were built on the surrounding agricultural terraces. As discussed previously,
findings from excavations conducted by Malpass, de la Vera, and Neira in Chijra indicate
that these terraces probably originate in the LIP, but their current configuration is most
likely a product of subsequent modification during the Late Horizon (Denevan 2001;
Malpass 1987; Treacy 1989). Ceramic collections also indicate a larger occupation
during the Late Horizon. Out of the total sample of 294 sherds we collected at the site,
Late Horizon (Collagua III, Collagua Inka, and Inka) sherds outnumbered LIP (Collagua
I and II) sherds by three to one136 versus 46 sherds, respectively. Thus, I suspect that
most of the standing houses at the site were built during the latter part of the LIP or the
Late Horizon.
During the survey, I divided Chijra into three sectors, registered as sectors J, K,
and L (Figure 6.64).91 Sector J encompasses a cluster of eight houses to the west near a
relict aqueduct.92 Sector K, located in the center of the slope, is composed of agricultural
terraces with 18 dispersed (generally poorly preserved) houses and looted tombs. As an
arbitrary measure, I demarcated the eastern side of sector K just east of the massive
structure 17, the second largest house at the site (Figure 6.64; see also Figures 6.18 and
6.19 above). Sector L, where we registered five houses, is the furthest sector to the east
in the Chijra area, and borders the ridgeline that separates it from the San Antonio area.
Sector M is comprised of a dense cluster of 91 houses on the eastern-facing slope, and
91
Sectors A-I are composed of isolated tomb features and chullpas on the surrounding boulder-strewn
ridges. See below.
92
This is the area where the Denevan team focused most of their excavations in the early 1980s (see
Malpass 1987).
275
sector N the promontory topped by a colonial structure that is probably an early
Franciscan chapel (see following chapter).
Previous research at the site consists of a brief initial visit to the San Antonio
sector of the site by Neira in 1959, and a much more intensive investigation of the Chijra
area as part of the Ro Colca Abandoned Terrace Project. Neira initially described the
large structures on and surrounding the promontory at the southeastern corner of the site.
Apparently referring to the large structure on the top of the promontory, Neira has
referred to San Antonio as the most important sanctuary in the valley (Neira Avendao
1990:152).93 In Chijra, research focused on understanding the construction, use, and
abandonment sequence of agricultural terraces. Reconnaissance and test pit excavations
were conducted at one house and five terraces in Chijra (Malpass 1986, 1987; Malpass
and de la Vera Cruz Chvez 1986, 1990; Neira Avendao 1986; Treacy and Denevan
1986). Malpass (1987:51) reports seven houses on and around the terraces excavated by
the Denevan team. Although they did not systematically survey the rest of the settlement,
Treacy and Denevan report a total of 51 structures in the combined San Antonio/Chijra
area.
93
In the following chapter I present evidence indicating that this structure was an early colonial chapel,
probably built by Franciscans sometime between their arrival in the valley during 1540s and prior to the
forced resettlement of the populations to the reducciones in the early 1570s.
Figure 6.64: Airphoto of San Antonio/Chijra (CO-100).
276
277
In general, architectural preservation was not as good as at Uyu Uyu, in large part
from more intense looting of worked cornerstones from structures, leaving their walls
vulnerable to collapse.94 We recorded 177 structures at San Antonio/Chijra, 136 of
which I defined as LIP/Late Horizon housesthree less than at Uyu Uyu. Of these, 105
houses are densely packed on the terraced slopes and promontory of San Antonio, and 31
are dispersed throughout the terraces of Chijra.
The architectural data from San Antonio clearly signal the presence of an elite
class and marked social inequalities within the resident population. Forty one (30%) of
the 136 houses at San Antonio/Chijra were complete enough to measure building
footprintsa lower percentage than at Uyu Uyu, but a large sample nonetheless. House
sizes vary widely between 14.0 and 136.5 m2 . This range is broader than at Uyu Uyu,
and as discussed above, the two largest houses at the site, structures 93 and 17, are the
largest in the survey area. As a group, houses at San Antonio are very largeequaled
only by Uyu Uyu. The midspread of house sizes ranged from 25.0 to 52.0 m2 and median
house area of 37.5 m2 is matched only by houses at Uyu Uyu, reflecting the generally
94
I infer that cornerstones have been looted in part because the natural collapse pattern is the reversethat
is, because the corners are the strongest part of the building, especially if well executed with alternating
headers and spanners, they are typically the last portion of a structure standing. Thus, walls normally
collapse inward first due to the batter (inward-leaning angle) of the walls, and the corners are left standing.
This point is also illustrated for Inca architecture by Protzen (Protzen and Batson 1993:234, Fig. 13.21).
Also, many of the cornerstones from San Antonio and other settlements around Coporaque were probably
used to build the church there, which is the only 16th century church in the reducciones of the Colca valley
that has remained largely intact. In contrast to the other churches of the valley, which, in the wake of
earthquakes during the 17th and 18th centuries were rebuilt mostly or entirely of ashlar masonry of volcanic
tuff (called sillar), brought to the valley from the puna heights to the south, the walls of the Church of
Coporaque are constructed of thin, tabular masonry of the same characteristics as the Type 5 masonry of
prehispanic houses at San Antonio, Tunsa, and Llactapampa in Coporaque. See the following chapter for
further discussion.
278
large size of houses and high status of their inhabitants relative to the other sites in the
survey.
Elite houses are concentrated on the eastern-facing terraced slopes of San Antonio
in sector M, as well as the promontory of sector N. All of these elite houses are in the top
house-size quartile (greater than 52.0 m2)95 and are constructed of coursed, tabular (Type
5) masonry. However, these represent only structures complete enough to measure area
and observe masonry style; amongst less-complete structures there are almost certainly
more elite houses in sectors M and N. In general, houses are larger in sectors M and N
than in the rest of the site (with a mean house area of 47.8 m2 vs. 38.1 m2 for sectors J-L).
Sixteen of the 19 houses in the third quartile (between 37.5 and 52.0 m2) are located in
sectors M and N. Small and medium-size houses are also interspersed throughout this
sector, but it is clearly a high status area of the site. Six of the seven houses I identified
as elite residences are also located there. The largest house at the site (and in the survey),
structure 93, is situated on a 4.5 meter high terrace on the upper eastern slope of sector
M, overlooking the promontory and valley bottom below. Other elite houses, such as
structures 84, 85, and 87, are closely aligned side-by-side lower down on the hillside.
The second largest house, structure 17, stands out as an isolated structure in the center of
the southern slope of the Chijra area, near its eastern end (Figures 6.23 and 6.24). This
14 x 12 m building is especially impressive for its tabular masonry on both exterior and
interior wall facings and especially well-executed corners. It was also probably a very
95
Or, in the case of structure 93, the largest house at the site (and in the survey) is an upper outlier in terms
of area, measuring 13.0 x 10.5 m (136.5 m2). Upper outliers are those 1.5-3 times the midspread range
above the median.
279
tall structure, since there are tenon supports for a second floor on all four walls (Figure
6.24).
Figure 6.65: Promontory of San Antonio (Sector N) from the southwest.
As at Uyu Uyu, these local elite houses are closely associated with at least one
and possibly two prominently-situated Inka kallankas. As discussed above, the larger of
the two (structure 154), occupies the saddle between the promontory of sector N and the
terraced hillside of sector M, thus straddling the point of access between the two sectors
(Figure 6.65). Also similar to its counterpart at Uyu Uyu, structure 154 is fronted by a
probable plazain this case, a level 18 x 28 m terrace that spans the remaining width of
the saddle. A second rustic Inka structure (structure 64), is situated just west of the
ridgeline separating sectors Chijra and San Antonio. Like the other kallankas, this
280
structure opens to a terraced plaza or large patio space (26.5 x 15.0 m) just off the main
path that runs along this ridge.
CO-127, Llanka
Llanka (CO-127) is a large, dispersed village to the southeast of Kitaplaza, on the
Qal 5 pampa to the north and east of the pukara fortifications discussed above. The
settlement is composed of 88 small- to medium-size houses amidst agricultural fields and
terraces.
The relative homogeneity of house sizes, and the predominance of Inka
polychrome ceramics suggest a strong state presence in the establishment and occupation
of the site. The primary occupation of the site clearly dates to the Late Horizonwe
collected only six LIP sherds, as opposed to 64 Late Horizon sherds, 16% of which were
polychromesthe highest percentage of Late Horizon polychromes of any settlement in
the survey.96
Houses at Llanka are distributed in various orientations at the site and there is no
centralized planning apparent in its overall layout. Many of the houses adjoin one
another, suggesting that households probably used more than one structure, forming
small domestic compounds similar to those of Uyu Uyu. While the overall layout of the
settlement suggest household-level planning, house sizes are much more homogeneous
than at other sites, and tend to be quite small. The midspread of house areas ranges
between only 16 and 25 m2 (see Figure 6.31 above). All but three are built of uncoursed,
Type 1 or Type 2 masonry. The larger structures at the site are discretely clustered on the
96
Excluding sites with small (less than 20) Late Horizon sherds.
281
south end of the site, but even these are only in the median size range of houses at Uyu
Uyu and San Antonio.
Overall, Llanka appears to have been a settlement of commoner agriculturalists
established (or at least considerably expanded) during the Inka occupation of the valley.
Given the homogeneity of house sizes and high percentage of Inka polychrome ceramics,
I suggest that this site may represent a settlement established with direct state
management and oversight.
282
Figure 6.66: Airphoto of Llanka (CO-127), including canals. Path in upper right leads to
Kitaplaza (CO-164) to the northwest, and to terracing and the settlement of Tunsa (CO-163)
to the south.
CO-164, Kitaplaza
Kitaplaza (CO-164) is located about 350 meters northwest of Llanka (CO-127),
along the modern road between Chivay and Coporaque. A major path diverges from the
283
road and passes to the north of a central open space between the two distinct residential
sectors of the site, leading eventually to Llanka and surrounding agricultural fields to the
southeast (Figure 6.66). To the adjacent north of the site are linear bench terraces that
cover the southern flanks of Pampa Finayathe Mosoqchacra (new field) toponym
sector. Surrounding the other sides of the site are valley bottom fields and terraces on the
slopes between the Qal 5 and Qal 4 alluvial surfaces. Although the site produced a
handful (N=4) of Middle Horizon sherds, the primary occupations of the site are from the
LIP and Late Horizon and, based on the predominance of Late Horizon ceramics, I
suspect that the site grew considerably during under Inka occupation. We collected 23
LIP (6 Collagua I and 17 Collagua II) sherds and 54 Late Horizon sherds (19 Collagua
III, 32 Collagua Inka, and 3 Inka sherds).
Most of the standing architecture probably dates to the Late Horizon. Houses are
not aligned by any overall site layout design, but form residential compounds, with two
or three houses oriented around small patio spaces. The 74 LIP/Late Horizon houses at
the site are mostly in the small- to medium-size rangethe largest houses only reach the
median house size at San Antonio/Chijra (CO-100) and Uyu Uyu (YA-050). All but two
are of low quality Type 1 and Type 2 masonry. The two houses of higher quality
masonry are made of coursed, rectangular and ovoid worked blocks (Type 6 masonry).
Together with these more elaborate houses in the southern residential sector is a possible
Inka ceremonial or public structure at the site. As discussed above, a large (13.4 x 5 m)
rectangular structure with windows on either side of its facade is not consistent with
Collagua domestic architectural canons and may be an Inka structure, similar to the
284
smaller of the two kallankas at San Antonio/Chijra. As in the better-documented cases of
Inka public architecture in the survey, this building fronts a small plaza-like space near
the center of the site.
Much like the nearby site of Llanka, Kitaplaza appears to have been a large
commoner agriculturalist settlement occupied primarily during the LIP and Late Horizon.
The site represents a typical mid-size agriculturalist village within the local LIP/Late
Horizon settlement pattern.
Figure 6.67: Airphoto of Kitaplaza (CO-164). Note two housing sectors separated by major
path from the main road.
285
CO-150, Llactapampa, and CO-163, Tunsa
Llactapampa (CO-150) and Tunsa (CO-163) are both situated on the Qal 4 river
terrace above the river gorge to the south of Coporaque, directly downslope and to the
south of the three pukara fortifications discussed above. They clearly housed a diverse
population of agriculturalists, including local elites, and probably functioned as distinct
sectors of a single large settlement, since they are less than 150 m apart and were
coevally occupied. Combined, Llactapampa (20 houses) and Tunsa (70 houses) would
form the fourth largest settlement in the survey.
Figure 6.68: Airphoto of Llactapampa and Tunsa, with major canals.
With 70 houses dispersed over 5.75 ha, Tunsa is much larger than Llactapampa.
The midspread range of house sizes at Tunsa (20-37 m2) is lower than that of Uyu Uyu
and San Antonio (21-42 m2 and 25-53 m2, respectively), but there are also clearly elite
286
houses at the site. As was the case at Uyu Uyu and San Antonio, the elite houses of
Tunsa are situated near a kallanka facing an open space that probably functioned as a
plaza at the entrance to the site. This plaza area divides the settlement in two. Most of
the houses are located to the east of the plaza, in valley bottom fields and surrounding
low hillsides (Figure 6.68).
Houses span the full range of size and quality of
construction, including a very large (78 m2) house of fine Type 5 masonry on all sides.
To the west, a smaller area with 15 houses occupies a low hill behind the kallanka. Here,
another very large house (72 m2) is situated along the north edge of the hill.
About 130 meters to the northwest of the kallanka at Tunsa, the 20 houses at the
site of Llactapampa (CO-150) are generally large and well-constructed, and the site as a
whole may have functioned as an elite sector associated with the kallanka of Tunsa.
Only two houses at Llactapampa are of uncoursed, unworked Type 1 masonry, and the
smallest house is 24 m2considerably larger than the low-end range of other sites in the
survey. Four houses are especially large and well made: two of coursed, tabular masonry
(Type 5), and two of coursed, dressed fieldstone (Type 3). These structures are of similar
quality and size as the elite houses of San Antonio/Chijra (e.g., Figures 6.23 and 6.24
above).
In sum, Tunsa and Llactapampa together probably formed a large town with an
internally differentiated population that included both elites and commoners. The central
placement of an Inka kallanka between the elite houses of Tunsa and Llactapampa
suggests the centrality of state public ritual in local Inka administration, as well as the
importance of local elites as mediating agents between local communities and the state.
287
YA-054, Llactarana
Llactarana is a small hamlet and associated ceramic scatter situated amongst
irrigated bench terraces about half a kilometer to the southwest of Uyu Uyu (Figure 6.69).
The site is located just above the present course of the Misme canal, in a band of
abandoned linear bench terraces that were previously irrigated by the canal when its
course was slightly higher on the hillside. The present course of the canal thus forms the
line between presently abandoned and cultivated terraces around the site. The site is
covered by a dense ceramic scatter, and 254 diagnostics were recovered, including 54
LIP (15 Collagua I and 39 Collagua II) and 129 Late Horizon (23 Collagua III, 101
Collagua Inka and 15 Inka) sherds. We divided the site into four sectors (sectors A-D)
based on surface features and artifact concentrations. Sector A is defined by a small
rubble mound feature with possible cyst tombs (all badly eroded) in association with a
dense ceramic scatter. The ceramic concentration continues downslope into abandoned
terraces in Sector B.
Sector C, to the adjacent north of sectors A and B, is the
architectural core of the site. We were able to map seven houses at the site (Figure 6.70),
but because of terrace wall collapse and heavy cactus growth, preservation of these
structures was generally poor, and I suspect that other structures are no longer surficially
visible. Six of the seven houses (structures 1-5, and 7) are small to medium in size (2037 m2), and the seventh (structure 6) is poorly preserved but considerably larger. Several
houses share single terraces at the site, and all face downslope to the east, but no
residential compounds as such could be defined. Overall, the architectural evidence
288
suggests that Llactarana was an undifferentiated hamlet with a handful of households that
probably maintained close contact with residents of the neighboring major settlement of
Uyu Uyu. Sector D consists of a ceramic scatter that continues across the abandoned
terraces to the north of the residential area of sector C.
289
Figure 6.69: Airphoto of Llactarana with surrounding terracing and location relative to Uyu
Uyu (YA-050). Note abandoned terraces above modern canal course.
290
Figure 6.70: Architectural map, habitational sector of Llactarana
YA-041 and CO-161, Yanque and Coporaque

Both reduccin villages in the survey, Yanque (YA-041) and Coporaque (CO161), are situated on top of LIP/Late Horizon settlements. Especially in the case of
Yanque, the provincial capital of the Collaguas province during the colonial period, the
presence of what appears to have been the largest settlement in the central valley was
probably the decisive factor regarding where to locate the reduccin. Both are also
located on the pampa of the Qal 4 alluvial surfacethe broad, gently sloping terrace
above the river gorgeand are surrounded by large expanses of valley bottom fields.
The settlement at Yanque was almost certainly the largest prehispanic settlement
in the survey area. It also appears to have been the primary Inka administrative center for
the central valley and possibly for the Collaguas province as a whole. The results of the
street survey indicate that the prehispanic settlement of Yanque was concentrated in the
northern half of the present reduccin, where sherd densities are markedly higher.
291
Collagua I (N=7) and Collagua II (N=19) sherds were too few and too dispersed
throughout the reduccin to reliably estimate a site area for the LIP. However, the site
produced the largest collection of Late Horizon ceramics of any site in the survey, with
209 sherds (6 Collagua III, 168 Collagua Inka, and 35 Inka). Another indicator of a
strong Inka presence, 15% of Late Horizon sherds were polychromessecond only to
Llanka in terms of percentage of polychromes in Late Horizon collections.97
Also, as I discussed above, architectural remains at Yanque include the only
examples of Inka cutstone masonry in the survey. Assuming that these blocks were not
moved a great distance from another settlement, their presence indicates that the
settlement was an important political center under Inka rule. Dispersed throughout the
northeastern quadrant of the village, other colonial and republican-era buildings
constructed of a conglomerate of unworked fieldstone and tabular and rectangular
cutstone blocks reveal the presence of what were probably large elite Collagua houses
(see Figures 6.34-6.36 above). Many are identical to Type 6 masonry of Collagua
houses, and other thin, tabular blocks interspersed in some house walls were probably
from the corners, doorways and wall heads of prehispanic houses.
97
This excludes sites with small samples (less than 20) Late Horizon sherds. The only site with a higher
percentage of Late Horizon polychromes is Llanka (CO-127).
292
Figure 6.71: Colonial/Republican era house in Yanque (YA-041) constructed with

probable prehispanic (Type 6) masonry
Figure 6.72: Detail of masonry on gable end
By contrast, Coporaque appears to have been a much smaller settlement, and the
late prehispanic occupations are much more ephemeral. These findings are contrary to
293
expectations, since as discussed above, Coporaque, based on the account of Or (1992
[1598]:159 [41]), is widely cited in the literature as the Inkaic capital of Collaguas (Cook
1982, 2002; Mlaga Medina 1977; Neira Avendao 1990). Our ceramic collections were
scant in Coporaquethe survey crew collected only 15 LIP and Late Horizon sherds.
Architectural evidence of the LIP and Late Horizon occupations of Coporaque is also
limited and somewhat ambiguous. On the one hand, no Cuzco Inka stonemasonry is
present, and we identified only one house of probable prehispanic origin. This large
house of tabular (Type 6) masonry is positioned diagonally relative to the street grid in
the southeast corner of the village, breaking with the otherwise regular orientation of
houses in relation to the streets (Figures 6.73 and 6.74).98
It was modified after its
original construction with an internal wall that separates the structure into two separate
rooms. A handful of other houses, concentrated in the southern half of the village,
incorporate tabular (Type 6) and rectangular (Type 7) worked blocks of probable
prehispanic origin, but were clearly constructed and used during colonial or even
republican times. Also, the church of Coporaque, built in the 1570s along with the rest of
the reduccin, is built of tabular (Type 6) and rectangular (Type 7) masonry. Given the
paucity of ceramic collections in Coporaque and the widespread evidence of cornerstone
mining at the nearby sites of San Antonio/Chijra (CO-100), Tunsa (CO-163) and
Llactapampa (CO-150), the vast majority of the blocks used to construct these buildings
appear to have been transported from these sites.
98
This structure has recently been encroached upon and damaged by the construction of a hotel complex.
The hotel uses this Inka house as its centerpiece, and is now surrounded by guest rooms.
294
Figure 6.73: Probable LIP/LH house at the southern edge of Coporaque (CO-161),
from the north. Note thin tabular (Type 5) masonry. An internal dividing wall was
added later, most likely during the Colonial Period.
Figure 6.74: Same LIP/LH house, from southeast, showing doorway.
295
Figure 6.75: Church of Coporaque from the north. This structure is built of thin
tabular (Type 5) masonry that was probably mined from the surrounding LIP/LH
sites. See Figure 6.76 below.
Figure 6.76: Detail of masonry, northwest

corner of church of Coporaque
296
YA-093, Jibillea, and CO-061
Jibillea (YA-093) and CO-061 are both small herding villages located in the puna
to the north of the river. Jibillea is composed of 12 circular houses (4-7 m in diameter)
and associated rock-walled corrals, and we registered 8 circular structure foundations of
similar size at CO-061, also in association with corral features. These two sites are
representative of a broader pattern of expanded settlement and intensification of
pastoralist production in the puna under Inka rule. As discussed above, nine of the 16
Late Horizon settlements lacking LIP components are puna pastoralist settlements, but
several other settlements with only small LIP components such as Jibillea clearly grew
considerably during Inka rule.
Figure 6.77: Circular house and adjoining corral at Jibillea (YA-093)
The ceramic assemblages at both sites indicate that both Jibillea and CO-061 were
either established or significantly expanded under Inka rule. At Jibillea we recovered 41
Late Horizon sherds but only one Middle Horizon and two LIP (Collagua II) sherds,
while CO-061 produced no LIP ceramics and 40 Late Horizon sherds. Also, a large
297
proportion of the Late Horizon ceramics are fancy Inka wares. Polychromes made up
15% of Late Horizon ceramics from Jibillea and 13% from CO-061only slightly less
than Yanque (YA-041), which had the highest proportion of Late Horizon polychromes
(16%) and comparable to that of Llanka (CO-127) with 15%.99
Both settlements are also situated near large relict reservoirs. Two reservoirs are
located near CO-061one directly north of the residential area of the site, and another
200 m southeast of the site. Another large reservoir is located to the immediate west of
Jibillea. The general expansion of settlement in the puna during the Late Horizon and the
close association between these two sites and large hydraulic features suggest that Jibillea
and CO-061 were established as intensive herding settlements under Inka rule. The high
percentage of elaborate polychrome Inka ceramics at these sites also hints at ties of
reciprocityperhaps these prestige goods were mobilized by the state in exchange for
labor service of these herding populations.
99
Interestingly, both sites also produced probable Chucuito style rimsherds from plates, signaling trade ties
between these herding settlements and those of the northwestern Titicaca Basin.
298
Chapter 7
Spanish Insertions: Early Colonial Settlement and Agro-Pastoral Production
Introduction
In this chapter, I document the archaeological traces of how local and imperial
structures of power were negotiated within the context of early Spanish colonial rule.
The early Spanish colonial presence in the valley is closely associated spatially with
archaeological features that were central to Inka rule of the region, and it is most manifest
in features related to religious indoctrination, notably the remains of early rustic chapels
at the major LIP/LH settlements of Uyu Uyu (YA-050) and San Antonio/Chijra (CO100), and a possible third chapel in Yanque (YA-041). These chapels are prominently
situated adjacent to Inka kallanka structures and their associated plaza spaces at both Uyu
Uyu and San Antonio. I interpret these associations as evidence that early Spanish
evangelical efforts communicated through a syncretic idiom that drew upon explicit
associations between Inkaic and Spanish spaces of public ritual. After reviewing the
documentary evidence for the pre-Toledan colonial history of the valley, I argue more
specifically that Uyu Uyu and San Antonio were two of the doctrinas (parishes for
Christian indoctrination) established by Franciscan friars in the major settlements of the
valley between the 1540s and 1560s.
Colonial political consolidation under the rule of the viceroy Francisco de Toledo
during the 1570s was also constructed on top of the physical and social foundations of the
hybrid Collagua/Inka political hierarchy and settlement pattern. In terms of settlement
patterning, I document how Yanque and other large Inka settlements in the valley became
299
capitals of the colonial repartimientos of the Collaguas province after reduccin. Thus,
while reduccin is often cited as an archetypical example of top-down colonial
impositionresettlement by state decreethe data I present in this chapter illustrate how
the Toledan mandate for reduccin was negotiated in local and provincial contexts.
Nevertheless these negotiations produced unintended consequences for both local
communities and the colonial state, including accelerating demographic decline and the
abandonment of significant areas irrigated terracing.

Colonial ceramics are generally underreported in the Andes, especially in the
highlands, so there is little baseline analysis to aid in chronology building in the Colca
Valley. The work of Rice and colleagues in Moquegua provides the only regional
comparative collection (Rice 1997). Collections from other areas of Spanish America are
dominated by distinct local wares and styles (Deagan 1987; Lister and Lister 1987;
McEwan 1992), and imported fancy wares that are temporally diagnostic are exceedingly
rare in Andean highland contexts. Compounding the problem is the fact that some of the
most common tablewares are stylistically conservative, appearing virtually unchanged in
16th through 20th century contexts (Rice 1997:174-175).
One confounding issue is that the ceramic assemblage during the early Colonial
period likely changed very little from the preceding Late Horizon, and most vessels
during these first decades after the conquest were probably nearly identical to their Late
300
Horizon antecedents. Thus the ceramic data underrepresent the continuity of settlement
between the Late Horizon and early Colonial Period, perhaps to a significant degree.
At this early stage in defining the Colonial Period ceramic chronology in the
Colca Valley, I have divided kiln-fired and glazed wares into three broad temporal
categories: Colonial, Colonial/Republican, and Republican.
Only those in the first
category are relevant for this chapter. These include both glazed domestic wares and
vessels of similar form and decoration as Late Horizon wares, but with vitrified pastes
and poorly executed forms and designs (Figure 7.1). Plates and bowls are by far the most
common vessel forms. They are wheel-thrown and brimmed, with foot rings. I suspect
that the great majority of the colonial tablewares in the valley were manufactured
elsewhere, perhaps in Arequipa. I did document two circular updraft kilns during the
survey, but these appear to have been used for firing ceramic roof tiles, not domestic
ceramics (see below). I present more detailed descriptions of Colonial Period Ceramics
in Appendix A.
301
Figure 7.1: Examples of colonial ceramics. Lot numbers indicated.
Not surprisingly, given the brief four decade span between conquest and
resettlement, the Colonial Period component at non-reduccin sites is represented by very
small ceramic samples. Excluding the reduccin village of Yanque (YA-041), only four
sites produced ten or more sherds that could be unambiguously assigned to the Colonial
Period. While such small collections preclude artifact-based settlement size calculations,
there evidently was considerable continuity in settlement between the Late Horizon and
pre-reduccin colonial period. Virtually all agricultural settlements were abandoned
302
when the reducciones were established in the early 1570s. Most colonial ceramics at the
sites probably date to that early period, although continued cultivation at the sites and
habitation by a few households at the sites could have also contributed to the surface
ceramic scatter. As I discuss below, some settlements such as Uyu Uyu (YA-050), San
Antonio/Chijra (CO-100), Yanque (YA-050), and Coporaque (CO-161) probably grew in
response to early Franciscan evangelization efforts, but the archaeological data do not
permit us to quantify that growth.
The Early Colonial Period Settlement Pattern

In the four decades between conquest and reduccin, most of the local population
continued to live in the same villages they inhabited under Inka rule, although some small
settlements apparently were abandoned. We collected Colonial Period ceramics from
two-thirds (32 of 48) of Late Horizon settlements (including scatters, middens, rubble
mounds and rockshelters). The 16 Late Horizon settlements that lack Colonial Period
components are all small scatters, with the exception of two of the Late Horizon
pastoralist settlements (CO-093 and CO-159) that I argued in the previous chapter were
part of an Inka state strategy to intensify pastoralist production.
abandonment soon after conquest supports this interpretation.
Their apparent
303
Figure 7.2: Colonial Period settlement pattern, including roads and primary trails.
304
Given the overall locational continuity of settlement from the Late Horizon, the
settlement hierarchy probably maintained its overall log-normal rank-size distribution,
with Yanque (YA-041) as the primary political and economic center, followed by the
secondary centers of Uyu Uyu (YA-050), San Antonio/Chijra (CO-100), and
Tunsa/Llactapampa (CO-163/CO-150). Small settlements may have been abandoned as
households resettled to the primary settlements. The primary Late Horizon settlements
probably also grew relative to other settlements after the conquest as a result of the early
evangelical efforts in the valley. There are large colonial structures that I argue were
early chapels at both Uyu Uyu and San Antonio/Chijra, and a possible third chapel is
located on the western edge of Yanque (YA-041). As I discuss below, these chapels
were probably built by Franciscan clerics between the 1540s and 1560s. Tord (1983)
documented a fourth chapel adjacent to the main church in Coporaque (C0-161). It
probably postdates the chapels at Uyu Uyu and San Antonio, having been built around
1565 (Tord 1983:87-89).
The chapels locations at the principal Late Horizon
settlements is consistent with the Franciscan practices in other regions and with the
church doctrine set forth by the first ecclesiastical council of Lima (1550-1552), which
mandated that a doctrina be established in each of the settlements of a cacique principal
(Cook 2002:892).
Here, and throughout Peru, the strategy of early Franciscan
missionaries was to establish their evangelical missions in the seats of Inka-era political
authority (Cook 2002; also see below).
Colonial Period components are absent at the cemeteries of the LIP and Late
Horizon, a fact that further attests to early evangelization efforts. We recovered only one
305
colonial sherd from each of two cemetery sites (YA-008, CO-098).
Clearly, Late
Horizon cemeteries ceased to be used for new interments during the Colonial Period as a
result of prohibitions on indigenous ancestor veneration and mortuary practices (see
below).
But given its cosmological, sociological, and economic importance and its
central role in the negotiation of sociopolitical legitimacy and resource apportionment,

the cult of ancestor veneration almost certainly continued as a clandestine practice for a
considerable period after conversion (Salomon 1995).
The Doctrinas of the Colca Valley, 1540-1595

Colonial documents are unanimous on the point that Franciscan clerics were the
only Spaniards that took up permanent residence in the Colca valley prior to the founding
of the reducciones, and that they established a series of doctrinas in each of the major
settlements of the valley between the 1540s and 1560s (Cook 1992, 2002; Tibesar 1953;
Ulloa Mogolln 1965 [1586]). Fray Juan de Monzn and Fray Juan de Chaves were
probably the first Franciscans to enter the valley, arriving sometime between 1540 and
1545 at the invitation of the valleys encomenderos (Cook 2002:890-891; Tibesar
1953:65).
100
100
Their mission was based in Coporaque, where, according to Mlaga
The exact date of their arrival is not known; however, an ecclesiastical report issued about 1585 states
that they had entered the province about forty years ago (Tibesar 1953:65). Gonzalo Pizarro was
encomendero of Yanquecollaguas, having been granted both parcialidades (Hanan and Urin) by his brother
Francisco in Cuzco on January 22, 1540. Larcollaguas was divided between Marcos Retamoso
(Hanansaya) and Alonso Rodrguez Picado (Urinsaya), granted in the same repartimiento by Francisco
Pizarro during the same repartimiento (for details, see Cook 1982:3-4; Julien 1998; Mlaga Medina
1977:94, 96).
306
(1977:101), they built a chapel dedicated to Saint Ursula.101 Fray Monzn, a member of
the original group of Franciscans in Central America headed by Marcos de Niza, was
remembered for having converted innumerable natives of the province during those early
years. He also undertook an aggressive early campaign to extirpate the idolatrous
practices and objects of the Collaguas, destroying huacas with zeal and erecting crosses
and shrines on the mountaintops. Reviewing an ecclesiastical memorial of the Collaguas
doctrinas from 1585, Tibesar (1953:65) notes:
On one occasion the friars had gathered at Lari such a large quantity of
pagan objects that fifty to sixty Indians were required to carry them to the
spot where the objects were either burned or destroyed and their ashes
strewn in the river. On another occasion while Fray Monzn was
searching for the principal idol of the province, he felt that he was in
danger of his life when his purpose was disclosed to the more than 2000
Indians gathered in the church patio awaiting the start of their doctrinal
instruction.
Despite such initial confrontations, the Franciscans established good rapport with
the Collaguas. In 1547, only shortly after Monzons arrival, the Collaguas provided a
safe haven for the commissary general of the order, Jernimo de Villacarillo, from
Gonzalo Pizarro, then the encomendero of Yanquecollaguas. Villacarillo had fled Lima
after having been threatened by Pizarros notorious co-conspirator, de Carvajal, for
voicing his opposition to the Pizarrist insurgency, and he found refuge amongst the
Collaguas before escaping to Charcas (Tibesar 1953:65).102
101
Mlaga does not cite a source for this information. It is also unclear whether Mlaga was referring to
the village of Coporaque itself, or to the larger community (i.e., the village and surrounding agricultural
fields).
102
As noted by Tibesar (Tibesar 1953:65), the high degree of mutual confidence between the friars and the
Collaguas also hints that already in 1547 the Franciscans were well-established and had been in the
province for some time.
307
With the support of La Gasca and the new encomendero Francisco Noguerol de
Ulloa, the Franciscans redoubled their pastoral efforts amongst the Collaguas after the
defeat of the Pizarrist rebellion in 1548. Villacarillo assigned four friars to the Colca
valley in 1560. Under the leadership of Fray Pedro de los Ros (Cook 2002:891), they
established their headquarters in Coporaque (Cook 2002), building a new chapel
dedicated to San Sebastian, probably in 1565 (Mlaga Medina 1977:101; Tord 1983:8788).103 In keeping with the doctrine of the first ecclesiastical council of Lima (1550-52),
which ordered the construction of a temple in each large village where a principal kuraka
resided (Cook 2002:892; Mlaga Medina 1974:151), the Franciscans also established the
monastery of the Immaculate Conception in Yanque in 1560. They also founded the
doctrinas of Chivay, Coporaque, Callalli, Achoma, Tisco, Tute, Lari, Madrigal, Sibayo,
Yanque, Ichupampa and Maca (Cook 2002:891).104 These doctrinas were apparently
well established by the time of Toledos visita general, given that the reducciones
established in the valley between 1571 and 1574 by the visitador Lope de Suazo bear
these same names.105 Toledo considered reduccin a critical precondition for conversion
103
The structure itself has been identified by Tord (see below). Both Tord and Mlaga state that the date of
construction was 1565, although neither provides a source, and it must post-date Villacarillos orders to
formally establish the doctrina in 1560.
104
The full names of the doctrinas were Asuncin de Chivay, Santiago de Coporaque, San Antonio de
Callalli, Los Reyes de Achoma, San Pedro de Tisco, San Francisco de Tute, Concepcin de Lari-Collaguas,
Santiago de Madrigal, San Juan de Cibayo, Concepcin de Yanqui-Collaguas, San Juan Bautista de
Ichubamba and Santa Ana de Maca (Tibesar 1953:68). Tibesar does not list names for doctrinas in the
Cabanaconde area.
105
Although Mlaga does not cite the source documentation (1977:98-102), he lists the full names of the
reducciones in the repartimiento of Yanquecollaguas as Caylloma, Oropesa de Tisco, Utrera de Cibayo,
Alcntara de Callally, Espinar de Tute, Villanueva de Alcaudete de Coymo, Martn Muoz de Chivay,
Villacastn de Coporaque, Corral de Almoguer de Yanque, and Achoma. Those of Laricollaguas are
Talavera de Lari, Las Brocas, Paradines, Madrigal, El Arrebatacapas, and Miraflores. In the repartimiento
308
(Mlaga Medina 1974), and he advocated the work of the Franciscans amongst the
Collaguas, granting them permanent jurisdiction over the doctrinas of the valley after
reduccin (Tibesar 1953:66).
But within the church, there was debate regarding whether the Franciscan
evangelical mission in the Colca and elsewhere should be considered as a stopgap
measure to cover the acute shortage of secular clergy in the years following the conquest,
or as a more permanent assignment. Within the order, some friars argued that the
Franciscans should eventually relinquish the doctrinas to secular priests under diocesan
authority so that the friars could return to the contemplative monastic life that they
considered more consistent with precepts of the order (Cook 2002:891). At the same
time that the doctrinas in the Colca were expanding in the early 1560s, debate was
stirring amongst Franciscans about when this transfer should take place (Cook 2002:891).
Simultaneously, local communities protested and resisted the churchs attempts to cede
doctrinas to diocesan priests in other areas. In 1569, for example, when Villacarillo
relinquished the Franciscan doctrinas in Cajamarca to secular priests, the natives
protested to the recently-arrived Toledo, and the Church reinstated the friars at his request
(Cook 2002:891).
The doctrinas of the Colca valley remained under Franciscan jurisdiction until
1581, when, shortly after the departure of Toledo for Spain, Villacarillo ordered their
transfer to diocesan clergy under the authority of Gregorio de Montalvo, the Bishop of
Cuzco (Cook 2002:895; Tibesar 1953:66). As in Cajamarca, the natives of the Colca
of Cabanaconde, Mlaga lists the names of the reducciones as Hontiveros, Las Brocas, Oviendo, La Puente
del Arzobispo, Pampamico, Guanca, Lluta, and Yura.
309
valley protested the decision, petitioning the corregidor of the Collaguas Juan de Ulloa
Mogolln on September 15, 1585, and later presenting grievances first to Lima, then to
Spain (Cook 2002:896-897). In a written appeal to an ecclesiastical board in 1586, the
kurakas praised the virtues of the mendicant friars, who the Collaguas admired, because,
in contrast to the encomenderos and secular priests, they lived in poverty (Ulloa
Mogolln 1965 [1586]:332).106 The appeal leveled grave accusations against the secular
priests, alleging that they engaged in commerce and transported wine, activities
prohibited by the third ecclesiastical council of Lima, and that they charged excessively
for the administration of the sacraments (Cook 2002:895-896; Tibesar 1953:66-67).107
The appeals of the native lords, with the aid of the famous Franciscan Friar, Luis
Jernimo de Or (see Cook 1992), also reached Toledos successor, viceroy Fernando
Torres y Portugal, Conde de Villar, who ordered the return of the Franciscans in 1586
(Cook 2002:897). A new group friars led by Or moved to retake the doctrinas that year.
While they were successful in regaining the doctrinas of Yanquecollaguas, which was
then held in encomienda by the Crown, the secular priests resisted their removal in
106
Ulloa Mogolln (1965 [1586]:332, my translation) related their grievance to the board, which included
priests who had recently been appointed to serve in the Collagua doctrinas as follows: The doctrinas of
these villages used to be under the charge of the friars of Saint Francis, where eight friars and one guardian
lived; they founded the churches which exist in this entire province and adorned them with decorations and
other sacred objects of much beauty, made from the silver and gold that they found in the guacas and high
mountaintops and old shrines of ancient times, where they went to destroy them with great charity and
fervor, and they revealed to them the blindness in which they had lived in former times and taught them the
law of God our Lord through their conversion in these doctrinas, until about two years ago, when they left
the doctrinas and returned to their convents at the order of Fray Jernimo de Villacarillo, their commissary.
This saddened the Indians so much that every day they weep for the Franciscan priests and they held such
authority to them that they love and respect them tenderly and they have attempted with all their might that
they return to take charge of these doctrinas. In place of the friars have come clerics of the Order of St.
Peter [the diocesan priests]; these are not as welcome as the friars. They do not have chapels nor is there
anything more than confraternities that do little to order the populace.
107
An ecclesiastical visita by the archdiocese of Cuzco confirmed these abuses (Cook 2002:895-896).
310
Laricollaguas and Cabanaconde, leading to a protracted legal battle between the diocesan
clergy and the Franciscans. On September 15, 1586, acting as canon lawyer for the
order, Fray Or again appealed for the full transfer of the doctrinas to Juan de Ulloa
Mogolln, corregidor of the province (Cook 2002:897). The Collagua kurakas also
protested the continued presence of the secular priests in the lower valley to the audiencia
in Lima (Tibesar 1953:67).108 The corregidor of Arequipa agreed and ordered the return
of the Franciscans, but the secular priests again resisted and remained in Laricollaguas
four more years (Cook 2002:897).
In 1590, the new viceroy, Garca Hurtado de
Mendoza, issued another decree ordering the return of the Franciscans to the valley
(Cook 2002:896).
Twelve friars headed by Or, together with the corregidor of
Laricollaguas, Gaspar Verdugo, were to orchestrate the actual transfer, arriving with
decrees in hand on July 10. The secular priests resisted, especially in Lari, where the
friars forcibly removed the priest Andrs de Arana and took possession of the church.
Over the following days the Franciscans similarly evicted the secular priests of Maca,
Ichupampa, Tuti, and Sibayo (Cook 2002:898-899).
The bishop of Cuzco, Gregorio de Montalvo, vehemently objected to the forced
reinstatement of the Franciscans, and debate regarding the propriety of the friars
continued evangelical work intensified within the order. The provincial, reviewing a
complaint sent by the bishops secretary on November 13, 1590, sought to resolve the
debate and settle the conflict with the diocesan hierarchy. He ostensibly deferred to the
108
The Collagua kurakas who signed the power of attorney were Juan Alanoca, Miguel Nina Taype,
Francisco Inca Pacta, Diego Hacha of Yanqui-Collguas, Miguel and Juan Caqui, Felipe Alpaca, Juan
Coaquira, Juan Suyo of Lari-Collaguas, Luis Ala, and Francisco Anti Ala of Cabana-Collaguas (Tibesar
1953:67, Note 114)
311
bishops authority and conceded diocesan jurisdiction over not only the recently
repossessed doctrinas, but all the doctrinas of the valley. However, the decision was
only partially enforced and diocesan priests were reinstated only to the doctrinas that the
Franciscans had recently retaken (Cook 2002:898). The friars apparently remained in
Yanque and Coporaque, since a year later Or appears in the visita of 1591 to certify the
accounting of births and deaths (Verdugo and Colmenares 1977 [1591]:343).
Only a month after the provincials decision, the canon lawyer of the order, Fray
Mateo de Recalde, summoned all documentation regarding the activities of the
Franciscans in the Colca valley for reconsideration, and he once again argued before the
audiencia that the Franciscans rights to all the doctrinas were legitimate and sanctioned
not only by viceroy Toledo but by the king himself. He noted that some of the secular
priests in the valley had become so strident that they had even fortified their positions,
setting a bad example for the natives (Cook 2002:901). In April, 1591, the authorities in
Lima reinstituted jurisdiction of the Franciscans over those doctrinas in the areas held in
encomienda by the Crown (i.e. Yanquecollaguas). a ruling reiterated two years later by
King Phillip II (Cook 2002:901). Thus, ecclesiastical authority in the region was split
between the Franciscans in Yanquecollaguas and the secular clergy in Laricollaguas and
Cabanaconde. The kurakas continued to denounce the diocesan presence in the lower
valley, and their complaints reached Phillip II himself. His response on January 6, 1594,
ruled in favor of the kurakas, citing Recaldes petition and ecclesiastical visitas that
documented improprieties and onerous levies by the secular priests. The king ordered
that the Franciscans had rightful jurisdiction over not only Yanquecollaguas but also
312
Laricollaguas. Again, the secular priests of Laricollaguas demurred, restating their case
that they gained possession of the doctrinas only after they had been vacated by order of
Villacarillo, and that they did so by the appointment of the bishop. They also pointed out
that the recent accord between the Franciscan prelates and the bishop gave clear
jurisdiction to the diocesan priests in Laricollaguas. In the end, the secular priests
remained in Laricollaguas and Cabanaconde, and the valley remained divided between
the secular priests and the Franciscans until the eighteenth century, when the Franciscans
finally withdrew completely from the region (Cook 2002:901-902).
Archaeological Evidence for the Doctrinas: Early Colonial Chapels

During the survey, I registered two structures that I interpret as two of the early
chapels built by the Franciscans when they initially established doctrinas in the valley.
One is at San Antonio (CO-100) and the other at Uyu Uyu (YA-050). A third possible
chapel is located on the western edge of the reduccin of Yanque, although this structure
is poorly preserved. A number of architectural features set these buildings apart from
local prehispanic architectural traditions and firmly identify them as colonial buildings,
and additional features are consistent specifically with colonial church architectural
canons. These structures are quite similar to the chapel of San Sebastian in Coporaque,
which was built around 1565 as part of the headquarters of the Franciscan evangelical
mission in the valley, as discussed above.
313
Figure 7.3: View of kallanka and probable chapel at San Antonio (CO-100), from the southwest.
The large colonial structures at Uyu Uyu and San Antonio are prominently
situated and closely associated with the Inka kallanka structures that line the plazas of the
two settlements. The chapel at Uyu Uyu is located on the eastern side of its plaza,
directly opposite the kallanka. Its facade, on the short axis of the structure, opened to the
west, facing the plaza and the kallanka. As I discussed in the previous chapter, the
kallanka at Uyu Uyu was also modified to create three internal spaces, and an external
room was constructed against its facade (Figure 7.4).
The secure dating of these
modifications to the kallanka requires excavation, but given their ad-hoc nature and the
fact that such internal divisions are unheard of in kallankas or other prehispanic
structures, they appear to post-date the conquest, and therefore most likely date to the
314
four decade period of continued occupation of the site before reduccin. They may also
have been related to the construction and use of the chapel. At San Antonio, the chapel is
situated on top of a promontory (Sector N) adjacent to the kallanka, which occupies the
saddle between the promontory and the main elite housing sector on the terraced eastern
flanks of Cerro Yuracc Ccacca (Sector M). Its position is both visually impressive from
the surrounding landscape, and it provides views far up and down the valley.
Intriguingly, the top of the hill is also encircled by a double coursed wall that is preserved
on average to a height of around 1.5 m and appears to have served as a fortification for
the structure (Figure 7.3).
The size and proportions of the structures at San Antonio and Uyu Uyu set them
apart from the surrounding prehispanic structures. Only one LIP/LH house in the survey
is larger than the probable chapel at Uyu Uyu, which measures 15.0 x 8.3 m (124.5 m2),
and its 1.8:1 length to width ratio is much more elongated than the vast majority of
LIP/LH houses.109 The long axis of the probable chapel at San Antonio is incomplete and
its facade is no longer preserved, but the wall sections that remain measure 11.5 x 7.2 m
(82.8 m2).
The walls of both structures are very thick and composed of conglomerate
masonry. The walls of the chapel at Uyu Uyu are 80-90 cm thick, and its south and east
walls are also reinforced or buttressed along their bases, forming a 15 cm wide horizontal
facet in the wall profile at 90 cm above the present surface, a feature that does not occur
in any prehispanic structure. The masonry of this chapel is uncoursed rubble (Opus
109
As discussed in the previous chapter, the length to width ratios of 74% of LIP/LH houses falls in the
nearly square range of 1-1.3:1.
315
incertum) with worked tabular corners, door stones, and wall heads similar to Type 2
masonry of LIP/LH houses (see previous chapter). A mud and pebble plaster similar to
that used in the LIP/LH houses was applied to the interior wall surfaces of the chapel, of
which 1 cm thick plaster remnants are present on the northern wall. The walls of the
chapel at San Antonio are also 80-90 cm thick, but there are no signs of buttressing, and
the masonry is of thin, tabular stones similar to elite houses of the Type 5 masonry
common to the LIP/LH houses of this area of the site (Sectors M and N). This similarity
may reflect a continuity in masonry techniques used by local Collagua builders, or the
structure may be constructed partly or entirely from masonry mined from prehispanic
houses at the site. Unfortunately, the San Antonio chapel is badly deteriorated due to
looting and greater exposure to the elements in its promontory setting, and no plaster
remnants are visible in its interior. In both chapels, nearly all cornerstones have been
robbed out, leaving the remaining walls vulnerable to collapse.
316
Figure 7.4: Map of the probable chapel at Uyu Uyu, showing orientation of photos below. Note
the overall size (A), the probable chancel platform on the interior (B), Front elevation with
respect to the plaza (C), the door on short axis and door width (D), and tapering windows (E).
317
Figure 7.5: View of probable chapel at San Antonio, from the northeast.
318
Figure 7.6: Frontal view of probable chapel at San Antonio.
Figure 7.7: Abutting interior join

of vestibule wall in the northeast
corner of the probable chapel at
San Antonio.
The position of the doors and windows of both structures are also inconsistent
with local LIP/LH architectural conventions. Whereas doors of LIP/LH structures are
always situated on the long axis and windows on the short axis (on the gable ends), the
reverse is true of these buildings. Also, the doorway of the chapel at Uyu Uyu is 130 cm
widemuch wider than the extremely narrow doors of Collagua houses (40-80 cm). The
319
two windows of this building are situated opposite one another near the center of the long
axis, rather than on the gable ends, as in local LIP/LH architecture. They are placed 2.5
m above the base of the walls. These windows are also distinct from those of prehispanic
structures in that they taper inward toward the exterior, such that the exterior window
opening is much smaller than the interior opening (Figure 7.4, photo E).110
The facade of the chapel at San Antonio is not preserved, so its doorway is not
observable. The chapels transverse interior walls, however, demarcate a small vestibule
or narthex, creating a transitional entry space before entering the nave, a common
attribute of church architecture.
The transverse walls of the vestibule are poorly
preserved, but their joins with the outer walls are clearly abutting, indicating they were
built later, perhaps during a remodeling episode after the initial construction of the
building.
The structure at Uyu Uyu also has internal features typical of churches. The
remains of a 1 m high, 2 m wide platform spans its eastern end (see Figure 7.4, photo B).
This feature creates a division in the internal space of the church between the nave, where
the congregation was seated, and the elevated chancel, reserved for the clergy, where the
sanctuary and altar would have been. This is the case in all historic church floor plans.
Also, a 60 cm wide patilla or bench feature runs along the sides of the chapel which
could have functioned as additional seating for the congregation.111 The floor of the
110
The height and width of the north window measures 60 x 40 cm on the interior and 45 x 13 cm on the
exterior. The south window measures 70 cm x 40 cm on the interior and 60 x 20 cm on the exterior.
111
This feature runs from the chancel platform westward to just below the windows on both the north and
south walls, after which point it is no longer visible. The interior of the structure is currently used as an
agricultural field, and cultivation has doubtless disturbed the remainder of these and any other floor
features in the nave of the chapel.
320
chapel at San Antonio is extremely pitted and disturbed by intense looting, making
observation of the original state of the floor plan impossible without excavation.
Both structures have high gabled roofs. The chapel at Uyu Uyu reached 7 m at
the top of its gable, and although the gable of the San Antonio chapel is not fully
preserved, it was at least as high. The roof of the chapel at Uyu Uyu was apparently
thatched, but scatters of high-fired ceramic roof tiles next to the chapel at San Antonio
indicate that its roof was tiled, an unambiguous colonial building technique. The tiles are
a surprising find, given that their presence implies the existence of a well-developed
colonial ceramic industry. Although it seems unlikely that ceramic kilns capable of mass
producing roof tiles would have been in place in the pre-reduccin period, when the
Spanish presence in the valley was essentially limited to the Franciscan clerics and
occasional visits by encomenderos and other officials, the weight and the quantity of tiles
required to roof such a large structure would have made their importation from Arequipa
or other centers of tile production impractical. It seems most likely that this chapel was
originally thatched, and the tile roof was added later. In any case, we registered two
Colonial or Republican circular updraft kilns in the survey area. Both are located in or
adjacent to a reduccin village: one (site YA-058) is immediate east of the village of
Ichupampa,112 and the other is in the northwest corner of Yanque.
Interestingly, I
recovered a roof tile waster from the kiln of YA-058, confirming that roof tiles were
112
Specifically, this kiln is situated on the western side of the quebrada Canto Huaycco, which forms the
community boundary between Yanque Urinsaya and Ichupampa and the western edge of the survey
coverage.
321
manufactured in the valley, although it is unclear precisely when these kilns were used or
abandoned.113
The structures at San Antonio and Uyu Uyu also share several features with the
chapel of San Sebastian in Coporaque. This chapel probably slightly postdates the other
two, having been built most likely around 1565, as discussed above. As I discussed
above, ecclesiastical texts document that Coporaque was the headquarters of the
Franciscan mission in the valley, and this chapel reflects this elevated status with its
impressive renaissance style facade (Tord 1983:87-88). It is the earliest known example
of a renaissance-style church in Peru (Tord 1983:88-89). Similar to the chapel at Uyu
Uyu, the chapel of San Sebastian opens to a plaza. On the opposite side of the plaza, the
southern portal of the main church, which was built after reduccin, aligns with the
facade of the chapel (see Figure 7.14). The chapel is of similar size and layout as the two
structures at Uyu Uyu and San Antonio. Its exterior measures 14 x 7.5 m, while the
interior of the nave measures 10.4 x 6 m.
The walls are constructed of coursed
rectangular blocks similar to Type 6 masonry of LIP/LH houses interspersed with

unworked fieldstones.
The front of the chapel has an elaborately decorated facade
enclosed by an open vestibule, likely the way the vestibule of the chapel at San Antonio
originally appeared. The facade is dominated by a portal flanked by doric columns, and
the frieze is adorned with a band of rosettes in relief with the faces of angels in the center
(Figure 7.8). Above the cornice, a pediment encloses an arched niche, now empty. On
113
Rice documented 26 colonial kilns at the wine bodegas of Moquegua. Two primary types of kilns are
present there, calcination kilns and ceramic manufacture kilns, in which large fermentation vessels (tinajas)
and storage/transport bottles (botijas) were made (Rice 1987, 1994; Rice and Van Beck 1993). However,
both types are much larger than those I documented, and Rice (1994:343) could not date their construction
or use because she could not differentiate functional and temporal variability in kiln design.
322
the inside of the doorway, a baroque mural depicting the apical Inka deities, the sun and
moon, appears on the top of the arch, while the image of Saint Sebastian tied to a tree
and impaled by arrows graces the keystone above the door. Together, these images form
a syncretic composition (Figure 7.10). Inside the chapel, a rectangular apse sets the
sanctuary apart from the nave, and a rectangular recess that would have held the crucifix
occupies the rear wall. Like the structure at San Antonio, this structure had a tiled roof,
although this may represent a later modification.
Following the construction of the main church, the chapel of San Sebastian
probably became a secondary chapel. It may have been remodeled in 1889, as that date is
inscribed on a stone found near the chapel (Tord 1983:89), but at some time subsequent
to that date the chapel was abandoned. A restoration project in 1999-2000 by the Spanish
Agency for International Cooperation (AECI) and the Peruvian National Institute of
Culture (INC) reconstructed the roof, murals, and interior decoration (Figure 7.9).
323
Figure 7.8: Facade of the chapel of San Sebastian,

Coporaque. Original from Tord 1983. Reprinted by
permission.
Figure 7.9: Perspective of chapel of San Sebastian after

restoration.
324
Figure 7.10: Interior of doorway, chapel of San Sebastian, Coporaque. Note crescent moon and
sun motif. Image of San Sebastian on keystone is deteriorated but recognizable. Original from
Tord 1983. Reprinted by permission.
A fourth possible chapel is located on the far western edge of the reduccin of
Yanque, along the main road that descends to the north side of the river, before crossing
the colonial bridge across the river gorge en route to the villages of Ichupampa, Lari, and
Madrigal. It is isolated from the houses of the village, 150 m from the nearest house
compound to the east. A modern shrine and crucifix adorns the interior. This structure is
poorly preserved, preventing definitively identifying it as a chapel. It does, however,
share several features with the other three chapels. It is in the same size range, measuring
10.5 x 8.0 m on the exterior, and its principal facade is on the short axis of the structure,
facing the road. Like the chapels of San Antonio and San Sebastian, the front of the
structure was composed of a vestibule and interior doorway, although only one corner of
the transverse wall of the narthex is visible in the northwestern corner, as both the
325
exterior facade of the narthex and the interior portal are not preserved. Unlike the chapel
at San Antonio, the transverse wall forming the vestibule is bonded with the exterior wall,
indicating that it was built at the same time as the exterior walls. The walls are 80-90 cm
thick, and are composed of a mix of uncoursed rubble masonry (Opus incertum) and
irregular courses of lightly dressed stones of varying sizes (Opus quadratum), with belt
courses of thin tabular slabs set at a diagonal angle just below the wall headers on the
interior (Figure 7.11). Similar to the chapel at Uyu Uyu, a low bench feature (patilla)
about 50 cm wide runs the length the west wall and a portion of the east wall. This
features also spans the short axis at the rear of the structure, but it is 10-20 cm wider on
that end, recalling the probable chancel platform at Uyu Uyu.
Figure 7.11: Perspective of possible chapel

in Yanque, from the east.
Figure 7.12: Frontal view of possible chapel

in Yanque. Note corner of vestibule wall on
the right, bench feature and possible chancel
platform.
326
Discussion
The documentary and archaeological evidence all indicates that the colonial
structure at San Antonio was the first temple built in Coporaque by the Franciscans,
preceding the construction of the chapel of San Sebastian. As I mentioned above, Mlaga
(1977:101) makes reference to a church dedicated to Saint Ursula that predated the
chapel of San Sebastian, and this structure may in fact be that chapel. In any case, this
chapel was probably built during the early years of the Franciscan doctrinas between the
early 1540s and the early 1560s. But it also almost certainly continued to be used even
after the construction of the chapel of San Sebastian and the establishment of the
reduccin of Coporaque. The tiled roof and the addition of a vestibule both suggest
remodeling subsequent to the initial construction, although these cannot be dated. The
fortification wall surrounding the chapel, however, resonates with the 1590 testimony of
the Franciscan canon lawyer Recalde that the secular clergy had fortified their positions
in the valley in order to fend off attempts by the friars to retake their doctrinas (Cook
2002:900-901see above). If the fortification at CO-100 corresponds with these efforts,
it would necessarily postdate Villacarillos recall of the Franciscans in 1581that is,
about a decade after the establishment of the reduccionesbut predate their return to
Yanquecollaguas with Or in 1586.
Secular priests might have been particularly
concerned with protecting their position in Coporaque because this had been the
headquarters of the Franciscan mission in the valley. An alternative interpretation is that
the wall functioned as an enclosure for the churchyard. Because clerics did not permit
the unconverted to enter the church, they probably addressed large crowds in front of the
327
church. This kind of open air preaching in front of the church was common throughout
the early viceroyalty (Gisbert and Mesa 1985:185-237). In any case, the evidence points
to the conclusion that the chapel of San Sebastian continued to be used after the
construction of the chapel of San Sebastian in 1565. Given that Coporaque was the
center of Franciscan evangelization efforts, the order might have desired to maintain two
chapels. After they built the San Sebastian chapel, the San Antonio chapel probably
became a secondary chapel, falling into disuse sometime after the main church was built
in the reduccin village.114
Similarly, the colonial structure at Uyu Uyu was probably a chapel built by the
Franciscans between the 1540s and 1560s, which was used until the establishment of the
reduccin of Yanque in the 1570s. The plaza in front of the chapel, which earlier appears
to have been the site of Inka state-sponsored feasts, would have functioned as the
churchyard, and was probably where much of the actual evangelization took place, since,
as I mentioned above, the unconverted were not allowed to enter the church itself. It is
unclear whether Uyu Uyu continued to be partially or intermittently occupied following
reduccin, but there are intriguing parallels between oral historical accounts of Uyu Uyu
and its relationship to the reduccin. The account of the Seor de Exaltacin
115
114
The main church of Coporaque displays an open atrium above the front portal. This was the hallmark of
early open chapels (capillas abiertas) that were designed so that the priests could address large groups in
front of the church (see Gisbert and Mesa 1985:169-234).
115
Gerardo Huaracha Huaracha, a prominent community member of Yanque and our hosts while we lived
in Yanque, recounted the story to me. The story, referred to as San Exaltacin, has also been reported by
Pease (1977:152-156), and Valderrama and Escalante (1987; 1997:78-80) published a shorter version.
328
recounts how the people of Uyu Uyu116 (referred to as Yanque Viejo or Mawka Llacta
YanqueOld Yanque) bore a crucifix from the old village across the river to the
church in the reduccin when they were resettled,117 but the cross escaped at night to
return to its home at Uyu Uyu. The villagers twice returned the crucifix to the new
church, but on the third night it escaped again, this time fleeing to the lands of Cuzco,
where it found a new home.118 Following the exodus of the cross, the water sources dried
up, and the crops and pasturage withered, leaving the valley in its current semi-arid state.
It is not my intent to present an exegesis of the multi-layered syncretic symbolism of the
story ( e.g., the conflation of Christ with the Inka, or Inkarrsee Pease G. Y. 1977:152156), but the salient point here is that the narrative portrays Uyu Uyu as home to an
important Christian icon that was moved to the new temple of the reduccin. Also
consistent with this story, Yanque Urinsaya community members who cultivate fields in
Uyu Uyu universally identify the structure in question as a chapel (capilla). Furthermore,
116
Referred to as the people of Yanque Urinsaya (the lower moiety), among whom the Checcas were
prominent. The reduccin village is referred to as La Brota, home of the upper moiety (Hanansaya),
among whom the Choques (or Choquehuancas) were prominent. To this day, Yanque divides its lands
between Urinsaya on the north side of the river (where Uyu Uyu is located) and Hanansaya on the south
side of the river, surrounding the reduccin. The village itself is also divided by an imaginary line down
the middle, separating Hanansaya to the east from Urinsaya to the west. Also, the Choquehuanca and
Checca families remain prominent families in their respective moieties. For example, the Checcas have
retained the esoteric knowledge and exclusive rights to perform elaborate ritual offerings to the spring
Umajala during the annual cleaning of the Misme canal (the primary feeder canal for Yanque Urinsaya
lands). The Umajala spring is a major source of water for the canal, and is considered the female
counterpart (Mama Umajala) to the male mountain deity of the Misme massif (Tata Misme) (fieldwork
1996, 1999). Valderrama and Escalante (1988) describe the Misme canal cleaning festival and other
hydraulic rites in Yanque in detail.
117
In one version (reported in Pease G. Y. 1977:152-156), the people (Urinsaya/Checcas) of Uyu Uyu
move to Yanque at the invitation of the Choques (or Choquehuancas) of Hanansaya, thus representing the
resettlement as an alliance between the two moieties. In the versions related to me by Mr. Huaracha and
recorded by Valderrama and Escalante (1987:78-80), the people of Uyu Uyu were forcefully resettled by
the Spaniards.
118
In the version recounted by Pease, Yanaoca, in the south of the Department of Cuzco, is specifically
identified as the destination of the spirit or image of the Seor de Exaltacin (Pease G. Y. 1977).
329
in describing the interior of the church of Yanque, Benavides (1994:426-427) notes,
[a]gainst the Evangel wall [i.e., on the left hand side of the church, the side associated
with the Urinsaya moiety] stands a life-size crucifix, related to a legend according to
which the Urinsaya people were forced to move from their old village to the north of
the river to join the Hanansaya people at the present location of Yanque. Although this
crucifix dates to the eighteenth century and thus cannot be the actual artifact moved from
Uyu Uyu, (M. Benavides, pers. comm. 2003), its association with Uyu Uyu and the
narrative just recounted strengthen the interpretation of the colonial structure at Uyu Uyu
as a chapel.
In sum, the large colonial structures at San Antonio and Uyu Uyu almost certainly
represent rustic chapels built by the Franciscans, probably during their early years in the
valley between the 1540s and 1560s. Their locations in two of the largest and most
important prehispanic/early colonial settlements in the central valley; their central
locations within those settlements, adjacent to prehispanic kallankas; and their size,
layout, construction techniques, and interior features are consistent with this
interpretation. The identification of the structure at the western edge of Yanque as a
chapel is more tentative, but the fact that the Franciscans established their monastery in
Yanque during the early 1560s is consistent with the pattern of association between
centers of Inka power and sites of Franciscan evangelization.
The close spatial association between Inka kallankas and the chapels suggests that
the Franciscans capitalized upon established centers of Inka power and ritual for the
locations of their doctrinas. Of course, this kind of association is only a local example of
330
how clergy generally expropriate symbols and loci of Inka power as tools of conversion,
a pattern epitomized by the founding of the monastery of Santo Domingo atop the
Coricancha in Cuzco (MacCormack 1991). Such is probably the case elsewhere in the
valley, although data are scant. In Achoma, Shea (1987:79) describes the ceramics and
architecture at the settlement of Malata as indicative of ...heavy Inca (and possibly
Colonial) influence (emphasis in the original). Systematic survey and excavations being
conducted by Miriam Doutriaux in Lari and Cabanaconde area should shed light on the
Inka to Colonial Period transition in the lower valley.
Toledan Dislocations
Reduccin and Prehispanic Settlements
The Franciscans had consolidated their evangelization mission in the valley by the
1570s, but the proto-reducciones of their doctrinas were eclipsed by the massive
resettlement program initiated between 1570 and 1575 during the visita general of the
viceroy Francisco de Toledo. Long favored by the Crown (Abercrombie 1998:223-229)
and already implemented in New Spain (Gade and Escobar 1982; Mlaga Medina 1974),
the reduccin program became one of the primary policies instituted by Toledo in his
efforts to augment faltering tribute collection, foster the civil and religious indoctrination
of the native population, and secure a steady labor supply for the massive mining
operations of Potos and Huancavelica (Hemming 1983:392-456; Levillier 1935; Mlaga
Medina 1974; Spalding 1984:156-168; Stern 1982a:76-79; Zimmerman 1938).
As
imagined communities of the state, each village was to be a model for a new colonial
331
civitas, with a central plaza surrounded by buildings housing the villages two principal
civic and religious institutions the cabildo, or village hall, and the church for the religious
indoctrination of the native populace (Mlaga Medina 1974). It is commonly noted that
reducciones were designed around a grid of blocks and streets like European cities, but
they actually reified this ideal more perfectly than most European cities of the time. In
this sense, their layout epitomized an urban archetype in miniature.
The resettlement program was an unprecedented experiment in social engineering
that displaced some million and a half native Andeans (Hemming 1983:393), and Toledo
and his predecessors were aware of the problems that could arise from dislocating native
communities. Although at first blush, the reduccin project might seem to be an example
par excellence of top-down state imposition, the decision-making process that determined
how and where to establish reducciones in a given area remains very poorly understood,
in large part because we have yet to recover a protocol detailing the establishment of a
reduccin. Toledos decree, which closely follows the recommendations outlined in the
jurist Juan de Matienzos 1567 treatise on colonial governance in Peru (1910 [1567]),
provides only general guidelines for reduccin settlement planning, and many of its
prescriptive measures must have been adjusted to local circumstances (see Toledo 1924
[1570-1575]).
Only a few features are shared in common between virtually all
reducciones: they are configured as a uniform grid of wide streets around a central plaza,
where spaces were reserved for a church and rectory, an inn, and government buildings
that included a town hall, courthouse, and separate jails for men and women. Each
village was also to have its own infirmary on a separate block from the main plaza. Also
332
a clear feature of state control, Toledo ordered that the enclosed house compounds of the
reducciones have only one door facing the street with no linking doors between
compounds.
Figure 7.13: The reduccin of Yanque, showing community spatial divisions
333
Figure 7.14: The reduccin of Coporaque. Note orientation of church relative to

the plaza. This is the only original 16th century church standing in the valley.
These mandated features aside, many other important aspects of settlement

planning were almost certainly negotiated at the local level as the visitadores met with
334
local clerics, communities, and kurakas. For example, while Toledo ordered (echoing
Matienzo) that the nucleated villages be established on relatively level terrain in healthy
places of good climate near abundant water and cultivable land, he also advocated, in
the interest of efficiency and minimizing disruption, that reducciones be situated in the
locations of previous large prehispanic settlements (Toledo 1924 [1570-1575]:163-164).
At the same time, the viceroy stipulated that the visitadores establish the new settlements
as far as possible from prehispanic cemeteries and huacaspriorities often at odds with
one another (Toledo 1924 [1570-1575]:163-164). Also, while Toledo set the maximum
size of a reduccin at 500 tributaries (i.e., men between 18 and 50 years old), many
exceeded this size considerably (Cook, et al. 1975 [1582]; Gade and Escobar 1982:432).
Even the stringent specifications of the settlement configuration were subject to
considerable variation. For example, it was left ambiguous whether the church should
occupy the same plaza as the buildings of civil authorities or have its own block. But
most importantly, it is unclear how decisions were made regarding whoin terms of
indigenous constructs of communitywas reduced where. Given that ayllu kindreds
often were spread over several settlements and production zones, resettlement had the
potential to fundamentally dislocate and disarticulate native systems of production, and
thus it posed a serious threat to both the communities themselves and to the states ability
to harness tribute revenue.
How were these conflicting priorities negotiated in the Colca valley? Along the
length of the Colca-Majes river drainage, 24 reducciones were established in the early
1570s (Mlaga Medina 1977), and the villages of Yanque and Coporaque illustrate
335
distinct outcomes.
In the case of Yanque, the capital of the colonial province of
Collaguas, colonial authorities placed the reduccin village on top of the primary Inka
political center in the central valley, a site that also was ideally situated according to the
geographic criteria for a reduccin (i.e., on the low, broad pampa above the river gorge).
A large portion of its population therefore would have come from the same locale and
would not have been displaced. The site of Chiquero (YA-032) is the only other large
Late Horizon agriculturalist settlement on the south side of the river, and its population
presumably was moved to Yanque. Resettlement would have been much more disruptive
for the inhabitants of Uyu Uyu (YA-050), Llactarana (YA-054), and YA-045 on the north
side of the river. Although their inhabitants apparently retained their agricultural lands
surrounding these settlements after reduccin, they would have had to cross the river to
reach them, as modern Yanque Urinsaya community members still do today.
The
locational continuity between the capital of the colonial province and its Inkaic
counterpart also appears to fit a general pattern amongst the principal villages of the
valley. Both of the capitals of the other repartimientos of the province, Laricollaguas and
Cabanaconde, are also situated on top of major Inka settlements. In the village of
Cabanaconde, de la Vera Cruz Chvez (1987:107-108; 1988) has documented two houses
of fitted Inka cutstone masonry that are not aligned with the street grid of the reduccin,
as well as Inka-style cutstone blocks used in the masonry of several other buildings. A
similar house of Inka masonry is present in Lari (see Cardona Rosas 2002:123), and
Miriam Doutriaux (pers. comm. 2003) has documented a major Late Horizon settlement
there in her ongoing survey.
336
In the case of Coporaque, its location does not coincide with a major Late
Horizon settlement, and it is situated high on the sloping Qal 4 and 5 surfaces,
surrounded on three sides by Pampa Finaya to the east, Cerro Pumachiri to the north, and
Cerro Yurac Ccacca to the west. Thus it is quite distant from the cluster of LIP/LH
settlements (CO-103, C0-127, CO-150, CO-163, and CO-164) and their fortifications
(CO-165, CO-167, and CO-168) on the valley bottom. The largest LIP/LH settlement in
the vicinity of Coporaque is San Antonio/Chijra (CO-100). Although it served as an early
doctrina, it is situated on steeply sloping hillsides where it would have been impossible to
construct a reduccin. Thus, the reduccin of Coporaque appears to have congregated
the inhabitants of all of these hamlets and villages in a locale where only a small
settlement had been present.119 A similar pattern is found in Achoma, just west of
Yanque, where the reduccin village does not appear to coincide with a major LIP/LH
settlement, but instead derives from the nearby hilltop villages of Pilloniy and
Achomaniy, as well as the settlements of Escalera, Malata, and Potosina (Oquiche
Hernani 1991; Shea 1986b, 1987).
Reduccin and the Agro-Pastoral Landscape

Farmers abandoned many irrigated terrace complexes throughout the valley
during the Colonial Period, and while the timing and causes of their abandonment
119
Mlaga (1977:101) lists the names of the prehispanic villages reduced to Coporaque, but does not
provide an archival source. The list appears to be derived from an ethnographic account, since the protocol
of reduccin by Lope de Suazo has not been identified in the archives. In any case, Mlaga states: Upon
conducting the Visita General of Toledo, Lope de Suazo, visitador of the Collaguas, reduced the villages of
Tunsa, Llanka, Qcuita, Jamallaya y Suripampa, as well as Chiptapampa, Ccanaque, Mosocchacra,
Muraypata, Chucpallu, Marquisahui, Machingaya, Huaynalama, Coporama, Cantupampa, Umaasu and
Ccayra, in the current village of Coporaque.
337
remained poorly understood, previous researchers emphasized depopulation and
reduccin as probable causal factors (Denevan 1987, 2001; Shea 1987; Treacy 1989).
These hypotheses posited that cultivators took land out of production in response to
epidemic-induced demographic decline (a process augmented by the close living quarters
of the reducciones) and as a consequence of moving people away from their fields into
the new nucleated settlements. According to this model, the remoteness of some terrace
complexes made them prohibitively expensive in energetic terms to maintain, especially
in the context of population decline. While the rationality of this kind of cost-benefit
calculus was clearly a motivating factor in terrace abandonment, I argue that the changing
physical and sociological structures of community organization mediated that calculus.
That is, the radically altered relationship between settlement and agro-pastoral
infrastructure in the valley after reduccin produced a dialectic in which the farming and
herding practices of local communities had to either accommodate to or reconfigure the
prehispanic and colonial features of the built landscape. One outcome of this dialectic
was abandonment.
But while most research has focused on terracing, terrace
abandonment was almost certainly epiphenomenal to canal abandonment, since irrigation

requires coordination within and between communities (see previous chapter), and canals
need regular maintenance that can be accomplished only through supra-household labor
organization. Individual terraces, on the other hand, can be maintained at the household
level. Thus, as communities of irrigators were dissembled, displaced, or decimated after
reduccin, some could no longer maintain the canal networks that fed their fields.
338
This explanation can account for the spatial distribution of abandoned canals and
terraces in the survey better than demographic or energetic models.
Clearly, some
abandoned irrigated terraces are located far from the reducciones as would be expected
according to the risk minimization explanations. These include the abandoned terraces
and canals of the undulating hills in extreme eastern Coporaque, as well as the abandoned
canals and terraces and canals on the high slopes above Uyu Uyu and Llactarana in
Yanque Urinsaya (both discussed in the previous chapter). However, there is no simple
distance-from-reduccin correlation for terrace abandonment as demographic/energetic
explanations would predict. In fact, many abandoned terrace complexes in the survey
area are very close to the reducciones.
For example, to the adjacent east of Yanque, two large bench terrace complexes
cover 27.5 ha on the north side of Cerro Pallaclle. The canal that irrigates these two
clusters of terraces descended from the Waranqante canal into a small reservoir on the
mesa-like top of Cerro Pallaclle, and distribution canals radiate from the reservoir to
irrigate each of the terrace groups (Figure 7.15). Survey and excavations in these terraces
in 1997 produced Late Horizon and Colonial Period ceramics, including in situ colonial
sherds from a house floor. Therefore they appear to have been built during the Late
Horizon and abandoned sometime during the Colonial Period, despite the fact that they
are some of the closest canals and terraces to the village.
339
Figure 7.15: Abandoned irrigation features and terraces on Cerro Pallaclle. Note location
adjacent to the southeast corner of the village of Yanque.
In Coporaque, abandoned canals and terraces flank either side of the village. To
the immediate west of the village, a feeder canal network that drew water from the
Chillihuitira and Wasamayo streams flowed into a large reservoir (Chilacotacocha) on the
mesa-like upper reaches of Cerro Yurac Ccacca above the site of Chilacota (CO-154;
Figure 7.16).120 As I discussed in the previous chapter, a secondary canal from one of the
feeder canals fed a small complex of bench terraces on the steep slopes of the
Chunancaya toponym sector to the adjacent west of the village, while the Chilacotacocha
reservoir provided water to the terraces of the Chijra (CO-100) and Alto Ccayra sectors
120
The topographic setting and configuration of the reservoir in relation to the terraces below mirror those
of Cerro Pallaclle on the opposite side of the river in Yanque. Chilacotacocha and its associated canals
were also documented previously by Treacy and Denevan (Treacy 1989:154-155; Treacy and Denevan
1986).
340
on the southern flanks of Cerro Yurac Ccacca. Given that San Antonio/Chijra (CO-100)
continued to be occupied (and probably grew as a doctrina) during early colonial times,
these canals and terraces appear to have been abandoned after reduccin. To the adjacent
east of the village, the abandoned Inca canal (a local name) drew water from the
Sahuara river and irrigated a narrow band (32 ha) of bench terraces on the western and
southern slopes of Pampa Finaya (Figure 7.17). Although we observed no Colonial
Period sherds on these terraces, the abundant LIP/LH sherds on the surface provide a post
quem date for their abandonment, so they were almost certainly abandoned during the
Colonial Period.
341
Figure 7.16: Abandoned irrigation features and terracing on the flanks of Yurac
Ccacca, directly west of Coporaque.
342
Figure 7.17: Abandoned Inca canal and associated terracing on the western slopes
of Pampa Finaya. The canal and terracing continues out of the frame, around the
southwest corner of Pampa Finaya.
343
Chapter Summary
During the first four decades after the conquest, the local Collagua population
continued to live in most of the settlements they inhabited during the Late Horizon.
Franciscan friars began their evangelization efforts in the valley during the 1540s, and
actively associated themselves with local elites and symbols of Inka imperial power by
locating their chapels near Inka kallanka structures at the largest settlements. They
campaigned to extinguish the practices of ancestor veneration, sleuthing out and
destroying mummy bundles, huacas, and regional prehispanic shrines. Yet even as the
clerics undercut the foundational symbols and institutions of community and ethnicity,
colonial administration of the province operated parasitically by shunting tribute through
the remnants of Inka-era sociopolitical structures to the encomenderos.
This faltering hybrid Colonial-Andean system underwent a major overhaul during
viceroy Toledos visita general in the early 1570s. But despite the outward appearance
of comprehensive reorganization, the reduccin villages themselves often grafted onto
the major Late Horizon political centers in the valley. While reduccin clearly displaced
much of the population, the subsequent abandonment of large areas of productive
terracing does not appear to have resulted simply from the movement of people away
from the fields near their old settlements. Rather, I argued that terrace abandonment is
symptomatic of how some communal irrigation systems were disarticulated in the postreduccin landscape.
In the following chapter, I analyze how local communities
negotiated this faulty coupling between old and new imperial systems through analysis of
the land tenure patterns of ayllus and villages using colonial census data.
PART III: ETHNOHISTORICAL RECONSTRUCTION AND

ARCHAEO-HISTORICAL SYNTHESIS
344
Chapter 8
Regional Political-Ecological Syncretism under Inka and Spanish Rule
Introduction
In the previous chapters, I presented the archaeological evidence illustrating how
Inka administration of the Colca valley effectively hierarchized the previous heterarchical
arrangement of supra-settlement communities during autonomous Collagua rule. While
the Inkas established a political center at the site of the reduccin of Yanque, I have
argued that the prominence of Inka public architecture near elite Collagua houses at
several other large settlements suggests that imperial administration was also closely
coordinated through local elites. The early colonial period ecclesiastical presence also
appears to have been closely associated with these centers of Inka power, as Franciscan
friars established a series of chapels at the principle settlements in the survey area, and,
most likely, throughout the rest of the valley. Even the radical reconfiguration of the
local settlement system under the reduccin program of Viceroy Toledo appears to have
been built on the (physical) foundations of Inka rule, as Yanque and the other principle
reducciones of the province were constructed atop the primary centers of Inka
administration in the valley.
In this chapter, I analyze a series of detailed visitas to an area that overlaps with
my survey coverage to provide a complementary view of the sociopolitical organization
of the Inkaic and early colonial province of the Collaguas, and of how those structuring
principles organized regional-scale relations of agro-pastoral production, distribution, and
consumption. I first show how post-Toledan colonial administration siphoned tribute
345
through vestiges of hybrid Collagua/Inka sociopolitical structures. In complement to the
archaeological view of Inka rule from the previous chapters, I then present a detailed
reconstruction of how the Inkas organized the province according to a nested hierarchy of
ayllus that were ranked according to bipartite, tripartite, and decimal categories. Third, I
demonstrate how this ayllu structure mediated households access to variegated and
widely-dispersed agro-pastoral resources and the means of their production. Finally, I
demonstrate how the native lords of the province mobilized staple finance from far-flung
agricultural and pastoral production enclaves to support groups of official state craft
specialists, thereby underwriting a state-sponsored wealth finance economy.
Hybrid Communities: Reduccin, Ayllus and Kurakas

Reduccin, (Re)Visitas, and the Politics of Community
The reduccin program of Toledo effected a decisive break from prehispanic
settlement patterns, and, as I discussed in the previous chapter, represented a sweeping
attempt at imposing an ideal model of social order on native Andean populations.
Ironically, however, these physical reconfigurations belied their status as imagined
communities of the state, since the Toledan reforms were built on uneasy compromises
between native and Spanish institutions of economic extraction and political
administration.
In the case of the Collaguas and throughout the former realm of
Tawantinsuyu, the native institutions that post-Toledan administration relied upon were
themselves hybrid Inka/local constructs whose imagined status became increasingly
apparent after conquest.
Below I briefly contextualize the Toledan reforms, the
346
subsequent crises that produced the flood of post-Toledan censuses such as those that I
analyze and identify those hybrid state/local constructs of community that structured both
Inkaic and post-conquest administration of the province of the Collaguas.
Both sources of state revenue in the Toledan systemtribute levies and corve
labor draftswere to be mediated by ethnic lords with oversight by Spanish magistrates
(corregidores de indios) such that taxes were levied at the provincial or sub-provincial
(repartimiento) level, but their actual collection throughout Peru was left to a hierarchy of
native lords. Both were also modeled on Inka imperial analogues of proportional tribute
and rotational labor, but with the crucial difference that taxes were levied in kind or cash,
not in labor (Murra 1956). Thus, tribute levies were particularly onerous during years of
poor agricultural productivity. Also, population decline from outmigration and from
deaths caused by European-introduced epidemicsa vulnerability aggravated by the
close living quarters of the new reduccionesleft remaining tributaries with higher per
capita tax burdens. Throughout Peru, this scenario of demographic collapse led to a flood
of petitions by kurakas, under pressure by their communities, for population recounts, or
revisitas (Cook 1975; Saignes 1985; Saignes, et al. 1985; Stern 1982a:114-137). As I
discussed in Chapter 3, this was also clearly the case for the Collagua. The kurakas of
the Collaguas were especially adept political actors in this regard, since the visitas to the
Colca valley constitute one of the largest series for any single locale in the New World.
There also were strong pressures from within the colonial administration to raise
tribute levies.
Claiming that the kurakas hid tributaries from previous counts,
corregidores also frequently called for new censuses (Guevara-Gil and Salomon 1994;
347
Stern 1982a:114-137). Special scrutiny would have been warranted in the case of the
Collaguas, given that the province of the Collaguas was the most populous under the
jurisdiction of Arequipa and produced 35% of its tax revenue (Cook 1975; Guillet
1992:23).
The existence of such a large corpus of visitas is in itself suggestive of protracted
demographic, economic, and political crisis in the wake of the Toledan reforms. Many
observers have noted that the contradictions inherent in the new political economy
hammered out by Toledoand its underlying dependence on blunt coercionultimately
led to its disintegration (see, inter alia, Abercrombie 1998; Gledhill 1988; Larson 1998;
Spalding 1982, 1984, 1999; Stern 1982a). However, my primary analytical focus in this
chapter is not on these later effects, but rather on identifying the indigenous structures
and principles of this hybrid colonial/Andean social order in order to illuminate earlier,
Inka-era relations of power and production.
The multi-scalar, flexible concept of ayllu expressed the indigenous construct of
community in the late prehispanic and early colonial Andes (Abercrombie 1986; Cock
Carrasco 1981; Isbell 1997; Platt 1982; Salomon 1991; Spalding 1984). As I discussed in
Chapter 1, the ayllu concept could reference any point along a continuum of social
relationships, from daily interactions among members of an ancestor-focused lineage, to
the large-scale rituals, infrastructural projects, or conflicts involving an entire ethnic
group (Abercrombie 1986; Janusek 1999; Platt 1982; Salomon 1991; Spalding 1984). As
Murra (1980) has illustrated, one important means by which the Inkas were able to so
quickly expand and consolidate their rule was by flexibly incorporating extant ayllus as
348
the primary units of imperial administration. Thus, Murra (1980) and many others (e.g.
Morris and Thompson 1985; Moseley 1992; Stanish 1992) have suggested that, to a large
degree, Inka imperial politics can be seen as an extension of the same strategies
employed by local kurakasthat is, the representation of asymmetrical exchanges
between elites and their ayllu subjects in the same terms as the intimate, symmetrical ties
between consanguines of minimal ayllus. In this way, redistributive relationships
between the state and its subject populations were represented as symmetrical, reciprocal
relationships amongst kin. However, I have argued that the Inkas also significantly
altered intra- and inter-ayllu relations by augmenting and hardening latent inequalities of
rank within and between themin short, by creating hierarchy out of heterarchy.
Perhaps nowhere are the effects of Inka ayllu politics clearer than in the case of
the Collagua (Benavides 1987b; Cock Carrasco 1976-77; Galdos Rodrguez 1984;
Guillet, 1992:18-19; Prssinen 1992:362-371; Pease G. Y. 1977; Rostworowski de Diez
Canseco, 1983:121-123; Treacy 1989; Zuidema, 1964:115-118). Below I present an
overview of the idealized hierarchy of ayllus that created a vertically-integrated,
horizontally-compartmentalized administrative structure in the province of the Collaguas.
Ideal Structures: Collana, Payan, and Cayao

As I discussed in Chapter 3, we know from the description of the provincial
magistrate Juan de Ulloa Mogolln (1965 [1586]) that the colonial and Inkaic Collagua
province was divided spatially and ethnically between the Aymara-speaking Collaguas in
the central and upper valley and the Quechua-speaking Cabanas in the lower valley. The
349
Collaguas were internally divided between the lower ranking Laricollaguas, to the
adjacent east of the Cabanas in the lower-central valley, and the higher-ranking
Yanquecollaguas in the central and upper valley. As is evident in their names, and as I
discussed in the previous chapters, the Late Horizon and colonial political centers of each
of these three provincial subdivisions were Cabanaconde, Lari, and Yanque, respectively.
In turn, the populations of each of these three provincial subdivisionsYanquecollaguas,
Laricollaguas, and Cabanacondewere internally divided between ranked moieties, or
sayas, called parcialidades with colonial administration: Hanansaya (upper moiety)
and Urinsaya (lower moiety).
While such tripartite and dualistic divisions within Inkaic provinces were
common (Prssinen 1992), Ulloa also goes on to explain how the ayllus that composed
each moiety within these subdivisions were divided again according to tripartite
categories of rank and prestige. This tripartite system of rank distinction follows a logic
of high, middle, and low status designations, called respectively, Collana, Payan (or
Pahana), and Cayao. Ulloa explains thus (1965 [1586]:330):
They governed themselves according to that which the Inka had
determined, which was, for their ayllos and parcialidades, he named for
each ayllo a cacique, and they were three ayllus, named Collona,
Pasana [sic, Pahana], Cayao. Each of these ayllos had three hundred
Indians and a headman whom they obeyed, and these three headmen
obeyed the principal cacique, who ruled over all.121
121
Gobernbanse conforme a lo quel inga tena puesto, que era, por sus ayllos e parcialidades nombraba
de cada ayllo un cacique, y eran tres ayllos, llamados Collona, Pasana, Cayao; cada ayllo destos tena
trescientos indios y un principal a quien obedecan, y estos tres principales obedecan al cacique principal,
que era sobre todos.
350
The striking characteristic of this system is its direct parallel with the reckoning of
prestige amongst the royal ayllus (panacas) of Cuzco. In the case of Inka kinship
classification, the prestige and rank of each of these categories was determined by
reference to the calculation of kinship distance from an apical ancestor (Julien 2000;
Kirchoff 1949; Rowe 1985; Zuidema 1977). Thus, members of the Collana group, a
Quechua word meaning of excellent quality, of primary origin (Bertonio 1956
[1612]:50), would be most closely related to the focal ancestor, those of Payan less so,
and those of Cayao would be only distantly related. Zuidema has suggested that Collana
in Inka kinship reckoning referred to groups that claimed both patrilineal and matrilineal
descent from former Inka rulers, those of the intermediate Payan status were related only
patrilineally, and those of the low Cayao status were not related to those of Collana
(Zuidema 1964:101, 115-118; 1973, 1977). These categories were also central to the
mapping of ritual over time and space by the royal ayllus (panacas) of Cuzco, and several
researchers have noted the close parallels between the tripartite hierarchical structure of
Collagua political organization and the ceque system of Cuzco (Bauer 1998:35-37;
Prssinen 1992:362-371; Zuidema 1964:115-118).
These tripartite principles produce an elegant ideal hierarchical sociopolitical
structure for the province (Figure 8.1) (Benavides 1989; Cock Carrasco 1976-77). The
1591 and 1604 visitas of Yanquecollaguas Urinsaya reveal that each of the three ranked
groups (Collana, Payan, and Cayao) noted by Ulloa Mogolln formed macro ayllus of
an ideal size of three hundred households that were each subdivided by the same tripartite
ranking system into three smaller ayllus of one hundred households each (cf. Cock
351
Carrasco 1976-77; Prssinen 1992:362-366).122 Thus, for example, the high ranking
macro ayllu Collana was subdivided into three 100 household ayllus: 1) Collana, (high
ranked), 2) Collana Taypi Pataca (middle ranked)123, and 3) Collana Cayao Pataca (low
ranked) (Fig. 3).124 The pataca designation of these smaller ayllus, meaning 100 in
Aymara, signals that they were considered scalar equivalents of 100 tributary households
within Inka decimal administration (see Julien 1982, 1988). Each of these pataca ayllus
was headed by a kuraka, whose title, pachacuracalord of one hundred households
also reflected his position within the decimal hierarchy.125 The Payan and Cayao macro
ayllus were also divided into three ayllus according to this same logic, thus forming a
total of three macro ayllus and nine micro ayllus each for Hanansaya and Urinsaya, for a
total of 18 pataca-level ayllus (see Figure 8.1).
122
These micro-ayllus are imperfectly reproduced in the visitas, probably due to post-conquest
consolidation and fragmentation. For example, in the 1591 visita of Yanquecollaguas Urinsaya, each of
the three micro ayllus are present for the macro-ayllus Collana and Payan; of the third, Cayao, only the
ayllu Cayao is listed (in the village of Tuti). However, given Ulloas description, and the close
adherence to this stated ideal in the cases of Collana and Payan, there is little reason to doubt that the
overall tripartite structure was present. See also Cock Carrasco (1976-77) and Prssinen (1992:366).
123
Taypi, meaning middle, is a synonym of Payan/Pahana (Bertonio 1956 [1612]:340).
124
On this point, Zuidema has noted that the organization of the nine micro-ayllus of each Collagua moiety
directly parallels the ceque organization of Cuzco. That is, the ceque lines of Cuzco were organized in a
repeating pattern of three groups of three ceques, again named Collana, Payan, and Cayao. This is the
pattern within three of the four quarters (or suyus: Chinchaysuyu, Antisuyu, and Collasuyu ) that divide
Cuzcos ritual space. The fourth, Cuntisuyu, may also follow this pattern, but with certain irregularities in
the naming of the ceques themselves.
125
Pachacuraca is a compound word combining the Quechua term for 100, pachaca (a cognate of the
Aymara pataca) and kuraka [curaca]. The term appears in the preambles for the villages of Tisco and
Sibayo in the 1615-1617 visita, where the visitador called forth quipucamayos and pachacuracas together
with the principal lords and gobernadores (seconds-in-charge) to ensure that all households were present
for accounting (APY Yanquecollaguas Hanansaya 1615-1617:ff. 338v-339r, 380v-381v).
These
declarations not only confirm the existence of pachacuracas, but also document that quipucamayoc
(keepers of quipu knot recordssee Ascher and Ascher 1988, 1997; Quilter and Urton 2002) were still
active and consulted by the pachacuracas and colonial authorities in 1617. Quipucamayoc are also
mentioned in the 1604 Laricollaguas Urinsaya visita. In that case, the quipucamayoc were called forth at
the end of the census as a means of checking for accuracy of the census counts (see Cook 2003).
Figure 8.1: Schematic of the ideal sociopolitical organization of Inkaic Province of Collaguas
352
353
Each of the pataca ayllus may also have been divided in half to form minimal
ayllus of 50 households each.
These minimal ayllus were named simply Taypi
Patacataypi meaning middle, could also be interpreted as meaning half in this

context (cf. Bertonio 1956 [1612]:340; Cock Carrasco 1976-77).
However, only a
handful of Taypi Pataca ayllus are registered in the visitas.126 If the Taypi Pataca ayllus
do represent minimal ayllu units of 50 households, then the vast majority were apparently
aggregated to the pataca-level for the census-taking.127
The corollary of this hierarchical classification system in terms of political
authority is that the kuraka of the highest ranked ayllu (i.e. the pachacuraca of ayllu
Collana) in the upper moiety (Hanansaya) was also the paramount kuraka of the
province as a whole (Figure 8.1see also discussion in Cock Carrasco 1976-77:108110). His structural equivalent in Lari was head of the repartimiento of Laricollaguas
and second in charge of the province (Benavides 1987b, 1988a; Cock Carrasco 1976-77;
Prssinen 1992:366; Pease G. Y. 1981; Treacy 1989). There is some ambiguity regarding
the rank of the head of Cabanaconde in relation to the paramounts of Yanquecollaguas
and Laricollaguas.
Cock Carrasco (1976-77) has suggested that the kuraka of
Cabanaconde could have been ranked in a third position relative to Yanquecollaguas and
Laricollaguas in an analogous fashion to the Collana/Payan/Cayao triadi.e. Yanque
126
In Urinsaya, two Taypi Pataca ayllus were registered in Yanque (Verdugo and Colmenares 1977 [1591]:
217-260 [ff. 16r-42r), two in Achoma (APY Yanquecollaguas Urinsaya 1604: ff. 388v-413v, and one in
Tuti (APY Yanquecollaguas Urinsaya 1604:139v-147r, Verdugo and Colmenares 1977 [1591]:73r-78r).
None were registered in Hanansaya.
127
An alternative interpretation of the name Taypi Pataca is that it was an imprecise reference to one of the
pataca-level ayllus of the middle-ranked Payan/Pahana (i.e. middle-ranked, or Taypi) macro ayllu.
However, this seems unlikely given that all three of the pataca-level ayllus of the Payan/Pahana macro
ayllu appear in the 1591 and 1604 Urinsaya visitas: Pahana Collana Pataca (high-ranked), Pahana Taypi
Pataca (middle-ranked), and Pahana Cayao Pataca (low-ranked).
354
represented Collana, Lari represented Payan, and Cabanaconde represented Cayao.
However, there is no explicit discussion of this in any of the colonial documentation, and
given the fact that the Cabanas were considered a separate ethnic group, it seems
unlikely. I favor an alternative hypothesis, suggested initially by Pease (1977:141) and
later by Prssinen (1992:365-366), which posits that the Cabanas were considered a
discrete group in contrast to the Collaguas as a whole, and the two were administered as
separate ethnic groups within a single province.128
While the logic of these categories would have been familiar locally, I argue that
their application in such a formalized hierarchy indicates that the Inkas substantially
reordered local ayllus according to the same ideal model of sociopolitical organization
that defined hierarchical relationships among the royal ayllus of Cuzco. In this way, the
Inkas amplified and ossified extant, but previously more fluid, inequalities within and
among indigenous ayllus, thereby creating hierarchy from heterarchy.
Articulating Ideal Structures and Collagua Ayllus

There are important differences between the naming conventions of the ayllus
between the two moieties of Yanquecollaguas, suggesting differences in the degree of
state involvement in their constitution. The effects of Inka social engineering are more
evident in the lower-ranked moiety (Urinsaya), while the high-ranked moiety
128
Mlaga Medina (1977:119) suggests a third potential scheme in which Cabanaconde and Laricollaguas
together form a lower-ranking, Urin group in relation to the higher-ranking Yanquecollaguas as the Hanan
group. But, as pointed out by both Prssinen (1992:364) and Pease (Pease G. Y. 1977), this is highly
unlikely given the fact that the colonial documentation is unanimous in describing, on the one hand, the
unity of the Collaguas as a dyadic pairing of the Laricollaguas and Yanquecollaguas groups, and on the
other, the Cabanas, who were a separate ethnic group of distinct origin, language, dress, body modification,
and economic focus.
355
(Hanansaya) appears to have been composed mainly of autochthonous Collagua ayllus
that were subsumed by the tripartite and decimpartite organization established by the
Inkas.
On the one hand, the names of the Urinsaya ayllus conform to the
Collana/Payan/Cayao conventions much more closely than those of Hanansaya,

suggesting they were more thoroughly altered by Inka administration. All but two of the
nine pataca-level ayllu names in Figure 8.1 appear in the Urinsaya visitas (see Table
8.2).129 On the other hand, in the visitas of Hanansaya, only the ayllus Collana and
Pahana Collana Pataca appear in unmodified form.
Others are variations on the
tripartite naming conventions, such as Surcollana (sic, Sur Collana), Pahana Caloca,
Collana Malco, Ilacachibaicayao Taipi Pataca (sic, Ilaca Chivay Cayao Taypi
Pataca)130, and Chapoca Collana Cayao Patacaprobably reflecting how the Inka
fitted these local Hanansaya ayllus to the ideal administrative hierarchy. But the other
Hanansaya ayllus did not follow the ternary and decimal naming conventions at all,
having instead names such as Iumasca [Yumasca], Calloca, Cupi, Checa Malco, and
Icatunga Malco. I argue that this lack of conformity suggests that Hanansaya ayllus
were composed primarily of indigenous Collagua ayllus that the Inkas integrated and
ranked within their ternary and decimal system.131 As I illustrate in further detail in the
129
The only two of the nine pataca-level logical variants of the Collana/Payan/Cayao triad that do not
appear in Urinsaya are Cayao Collana Pataca and Cayao Taypi Pataca. The names of three other ayllus
appear as local variations on these tripartite conventions: Collana Paque, Paragra Pahana Collana Pataca,
and Sibayo Pahana Collana Pataca. See Tables 8.1 and 8.2.
130
131
Ilaca [or hilacata] is the Aymara term for the headman of an ayllu (Bertonio 1956 [1612]:133).
However, the ayllu Yumasca represents an exception to this pattern, since Yumasca is a Quechua, not
Aymara word, suggesting that it was not of indigenous Collagua origin. In any case, yumasca, from the
root yumaysperm or semen (Gonzlez Holgun and Porras Barrenechea 1989 [1608]:371), is the
conjugated participial form, meaning inseminated, suggesting patrilineage. However, the participial
form in Quechua does not orient passive influence interpretation, so it could mean either the one who
356
following chapter, these indigenous ayllus originally may have been organized around
dualistic principles. A full discussion of Hanansaya ayllu onomastics is beyond the
scope of this project, but the names of the Checa and Cupi ayllus are suggestive of
dualism, since cupi (specifically, the compound cupitoque) means right side (Bertonio
1956 [1612]:60), while checa (spelled ccheca in the Bertonio dictionary) means left
side (ibid: 79).132 So these ayllu names can be translated as the right and left ayllus.
My reconstruction of ayllu-level land tenure patterns in the following chapter strongly
indicates that the names of these ayllus also refer to the locations of their agricultural
fields, and by extension, are perhaps indicative of their villages of residence before
reduccin.133
Although it is unclear precisely how these non-Collana/Payan/Cayao
ayllus were ranked within the tripartite and decimpartite sociopolitical structure, they
nonetheless remained subject to the kuraka of the ayllu Collana of Yanque.
inseminated (some entity; for example, its moiety) or the one inseminated by (some entity, perhaps the
Inka) (F. Salomon pers. comm. 2003). Given the seemingly non-local linguistic origin of this ayllu, I
suspect the latter rather than the former, although scrutiny of naming patterns of individuals within the
ayllu and comparison with the other ayllus could potentially clarify whether it was composed of local or
non-local people.
132
Malco (as in the ayllus Checa Malco and Malco), means lord of vassals (seor de vasallos), is the
equivalent of the Quechua term kuraka, meaning lord or chieftain (ibid:212).
133
Also, in the Andes, the right hand side is generally ranked higher than the left hand side and associated
with Hanan; thus the analogy: Cupi (right) : Hanan :: Checa (left) : Urin. So the Cupi ayllus may have
been higher ranked than the Checa [Malco] ayllus, although this must be considered speculative, especially
given the fact that the Checa ayllus are listed before the Cupi ayllus in the visita, and ayllus were generally
listed in descending order of rank.
357
Table 8.1: Villages and ayllus of Yanquecollaguas Hanansaya, 1591/1615-1617*
Village
Ayllo
Surcollana
[Sur
Collana]
Achoma
(incomplete) Unidentified**
Chapoca Collana Cayao Pataca
Chivay
Ilacachibaicayao Taipi Pataca
[Ilaca Chivay Cayao Taypi Pataca]
Collana Malco
Icatunga Malco
Checa Malco
Iumasca [Yumasca]
Coporaque
Calloca
Aipi [Cupi]
Cupi
Oficiales Olleros (Official Potters)
Collana
Guanca
Cupi
Lluta
Collana
Maca
Indios Plateros
Sibayo
Cupi
Anaoca
Collana Malco
Iumasca [Yumasca]
Tisco
Cupi
Cupi Capa Chapi [Cupi Pachapi]
Pahana Collana Pataca
Pahana Caloca
Tuti
Chilpe
Unidentified
Yanque
Yndios Oficiales Cumbicamayos***
(incomplete) Yndios Yanaconas de don Joan Halanoca
(Retainers of don Joan Halanoca)
* Listed in order of appearance in the visitas. From APY Yanquecollaguas Hanansaya 1615-1617, except
Yanque, from APY Yanquecollaguas Hanansaya 1591(Pease G. Y. 1977). Both visitas are incomplete;
only the second half (ff. 303r-611v) of the 1615-1617 visita is preserved, and the 1591 Hanansaya visita is
a small fragment, lacking foliation. The 1591 fragment begins near the end of the accounting of Yanque.
**The 1615-1617 visita fragment begins in the middle of this ayllu. Marginalia indicate that the ayllu is in
Achoma, followed by ayllu Surcollana, also in Achoma. However, there is no summary ledger following
the individual listing of households to identify this ayllu.
***Listed as subjects of don Juan Halanoca (Verdugo and Colmenares 1977 [1591]:415 [no folio number).
Although Halanoca is not included in this visita fragment, Ulloa lists him as principal kuraka of the
provincei.e. kuraka of ayllu Collana of Yanque (Ulloa Mogolln 1965 [1586]:326).
358
Table 8.2: Villages and ayllus of Yanquecollaguas Urinsaya, 1591/1604*
Village
Ayllo
Collana Pataca [1]
Collana Pataca [2]**
Yndios Cunbicamayos deste ayllo (Collana
Achoma
Pataca [2])
(incomplete) Yndios Plateros (Silversmiths) deste ayllo
(Collana Pataca [2])
Taypi Pataca [1]
Taypi Pataca [2]***
Callalli
Cayao
Pahana Taypi Pataca
Canocota
Cayao Pataca
Collana
Pahana Taypi Pataca
Coporaque
Pahana Cayao Pataca
Oficiales Olleros de [Pahana] Cayao Pataca
(Official Potters of [Pahana] Cayao Pataca)
Collana Cayao Pataca
Pahana Taypi Pataca
Guaraoma
Cayao Pataca
(estancia)
Collana Taypi Pataca
Unidentified
Mamaniviri Collana
(estancia)
Taypi Pataca [2]
Collanapaque [Collana Paque]
Paragra Pahana Collana Pataca
Sibayo Pahana Collana Pataca
Sibayo
Pahana Cayao Pataca
Cayao
Cayao Pataca
Collana Cayao Pataca
Tisco
Pahana Taypi Pataca
Pahana Cayao Pataca
Collana
Taypi Pataca
Collana Pataca
Collana Paque
Tuti
Pahana Taypi Pataca
Pahana Cayao Pataca
Collana
Yanque
Taypi Pataca [1]
Taypi Pataca [2]
Cumbicamayos [1]
359
Cumbicamayos [2]
*Listed in order of appearance in the visitas. From APY Yanquecollaguas Urinsaya 1604, except Yanque,
from the 1591 visita (Verdugo and Colmenares 1977 [1591]:191-263 [ff. 1r-44r).
**Collana Pataca is listed as two separate ayllu segments in Achoma, but due to the death of the kuraka of
the second segment during the recording of the visita, they were (re)joined under the kuraka of the first
segment, Miguel apana (APY Yanquecollaguas Urinsaya 1604:ff. 366r, 387v).
***In Achoma Urinsaya the second ayllu Taypi Pataca data are incomplete. The document fragment ends
at folio 413v, in the section recording tributary households from this ayllu.
Structural Grafting: Ayllus and Reducciones

As I discussed in the previous chapter, the reducciones themselves grafted onto
the physical manifestations of Inka rule in the valley, as the Toledan villages (Yanque,
Lari, and Cabanaconde) that served as capitals of the three provincial repartimientos
were each located on the site of a major Inka political center. Similarly, post-Toledan
administration of the province was linked to vestiges of the Collagua-Inka political
structure I outlined above.
The reduccin villages of the central valley (Yanque,
Coporaque, Achoma, and Chivay) were composed exclusively of ayllus from the two
higher-ranking macro-ayllus of Collana and Payan. By contrast, the herding villages
(Callalli, Sibayo, and Tisco) in the upper valley were home to the lower ranking ayllus of
Cayao, as well as outlier segments of the higher-ranking ayllus subject to their respective
kurakas in Yanque and Coporaque.
As I illustrate below, this suggests that the
reducciones and post-Toledan political and economic administration were based on

preceding Inka-era networks of ayllu affiliation and authority.
Supra-Local Networks of Ayllu Authority

Despite reduccin, Kurakas in Yanque and Coporaque continued to hold
authority over a constellation of ayllu satellite populations located in both lower altitude,
360
prime-maize producing valleys to the south, and in the puna grasslands surrounding the
upper reaches of the Colca valley. Settlements in these outlying locales therefore were
composed of disparate ayllu segments that were subject to their respective kurakas in the
agricultural core of the valley. The herding village of Tisco was the primary settlement
in which outlier segments of ayllus based in the central valley were settled.134 This is
why the population of Tisco was divided among so many small ayllu segments: groups
from all six patacas of the high- and middle-ranking macro ayllus Collana and Payan
[Pahana] were settled there, as well as an ayllu from the lowest ranking macro-ayllu
Cayao whose kuraka was resident there (Verdugo and Colmenares 1977 [1591]:264-292
[ff. 44v-64r]) (see Table 8.2 above).135 Listed with most of the outlier ayllus from
Collana and Payan were local mandones, or seconds-in-charge, who would have acted as
proxies for their kurakas in the mobilization of labor and collection of tribute.
The links between kurakas in the central valley and herding populations in the
upper valley and surrounding puna can be reconstructed by tracking the names of ayllus
and kurakas in each of the villages recorded in the visitas. Figure 8.2 illustrates how the
kurakas of ayllus based in Yanque and Coporaque were also listed as the heads of ayllu
segments in pastoralist and maize-production enclaves. Within Urinsaya, the kurakas of
Yanque and Coporaque were responsible for collecting tribute from ayllu segments in
five villages (Canocota, Tuti, Sibayo, Callalli, and Tisco) and three pastoralist hamlets,
(estancias) (Mamaniviri, Villaya/Hivillaya, and Guaraoma) (Figure 8.2).
All of the
134
Similarly, the village of Cailloma was the primary herding settlement for Laricollaguas. An outlier
segment of the ayllu Collana resided there that were subject to its kuraka in Lari (Guillet 1992:22).
135
Of these, three (Collana Cayao Pataca, Pahana Collana Pataca, and Pahana Cayao Pataca) were listed
as being under the direct authority of their respective kurakas in Coporaque.
361
kurakas of Coporaque Urinsaya were listed as the heads of outlier ayllu populations in
three to four herding villages (Figure 8.2). In Yanque Urinsaya, the head of the ayllu
Taypi Pataca (one Diego Chacha) collected tribute from outlier ayllu segments in the
estancias of Mamaniviri and Villaya (APY Yanquecollaguas Urinsaya 1604: ff. 329v330r).136 The paramount kuraka of Urinsaya, (and head of ayllu Collana of Yanque),
was also listed as head of satellite ayllu Collana populations in Tisco and the estancia of
Mamaniviri.137 In addition to the revenue from these herding populations, he also would
have been responsible for collecting the tax levies from each of the pachacuracas of the
moiety (via the apical kurakas of their respective macro-ayllus), and tribute would have
been funneled through him to the paramount kuraka of the province.
A corresponding schematic for Hanansaya cannot be constructed because most of
the Yanque data are missing from the incomplete 1591 and 1616 visitas. However,
included within the fragment that remains of Yanque Hanansaya is an important section
that registers a group of personal retainers (yanacona) of the paramount kuraka of the
province, Joan Halanoca (or Alanoca). As I will discuss in further detail below, these
yanacona were pastoralists that claimed between 10-50 alpacas each, for a total of 510
head. As paramount kuraka of the province, Halanoca was also unique in that he held
authority over production enclaves in prime maize-growing lands of Huanca and Lluta,
136
The estancia of Villaya is also spelled as Hivillaya (APY Yanquecollaguas Urinsaya 1604:f. 336r). If
this second spelling is correct, it may correspond with the large Late Horizon herding settlement of
Jibillea or Jivillea (my phonetic spellings based on local informants) that I recorded as site YA-093 in
the survey (see Chapter 6).
137
Mamaniviri is also spelled Mamanviri (APY Yanquecollaguas Urinsaya 1604:f. 336r). I have not
located this estancia.
362
located 55 aerial kilometers to the south of Yanque near the headwaters of the Siguas
river.
Figure 8.2: Schematic of ayllu outliers in herding villages subject to kurakas in Yanque and Coporaque. Yanquecollaguas Urinsaya.
363
364
The populations of these maize-producing enclaves were even more
heterogeneous than those of the herding settlements. Huanca and Lluta were composed
of ayllu outliers not only from Yaquecollaguas, but also from Cabanaconde and
Laricollaguas (Gelles 2000; Robinson 2003). Additionally, large groups of colonists
from all three repartimientos were settled in the valley of Arequipa, 90 aerial kilometers
to the south (cf. Galdos Rodrguez 1984; Guillet 1992). Although these colonies were
not recorded in the visitas, they were registered in the summary ledger (tasa) from the
Toledan visita general of 1570-1575 (Table 8.3) (Cook, et al. 1975 [1582]). In the tasa,
they were explicitly accounted as part of the Collaguas population and remained subjects
of the kurakas in the Colca valley (ibid: 217-223). Specifically, 481 Collagua and
Cabana tributary households were recorded as residents of the right bank of the Chili
river (La Chimba) in Arequipa141 from Yanquecollaguas, 159 from Laricollaguas,
and 181 from Cabanacondea major presence totaling 2009 individuals (see Table 8.3)
(Cook, et al. 1975 [1582]:217-218, 220-223).138
Table 8.3: Collagua colonists (mitmaqkuna) living in La Chimba of Arequipa, 1582139

Tributaries
(Adult
Elderly/
Females
Repartimiento
Males)
Males (<18 yrs)
Infirm
(All Ages)
Yanquecollaguas
141
14
122
288
Laricollaguas
159
9
159
312
Cabanaconde
181
11
192
421
Totals:
481
34
473
1021
Totals
565
639
805
2009
138
According to Mlaga Medina, they were all reduced to the village of San Juan Bautista de la Chimba
(Mlaga Medina 1981:49).
139
From the Tasa de la visita general de Francisco de Toledo (Cook, et al. 1975 [1582]:217-218, 220-223).
Categories taken from the original.
365
It is unclear whether these outlying colonies in Arequipa, Huanca and Lluta were
established under the Inka aegis or if they had roots in the LIP or earlier times.140 Neither
does the tasa specify if the colonists in Arequipa were exclusively subjects of the
paramount kurakas or if they were from several macro- or pataca-level ayllus. In any
case, the valley of Arequipa under Inka rule was a complex multi-ethnic mosaic of the
sort that Murra predicted would exist in the outlier regions of vertical archipelagos
(Galdos Rodrguez 1987; Murra 1972).141 As I argue below, however, produce from
these outlier enclaves was not used to provision the high altitude population as a whole,
but rather was mobilized as staple finance that underwrote the production of prestige
goods and other activities related to elite prerogatives.
Reduccin, Segmentation, and Succession

Although these supra-village political and economic ties remained (at least
temporarily) intact, the process of reduccin and reorganization had the effect of
segmenting the lower strata of the extant sociopolitical hierarchy into smaller units. The
disarticulation of pataca-level Collagua-Inka structures of authority is evident in the
listing of separate kurakas for the same ayllu in several reducciones in the central valley.
In several cases, the same pataca ayllu appears in more than one village, and each group
140
Archaeological inquiry around Huanca and Lluta has been preliminary in nature (Linares Delgado
1988). Erika Simborth Lozada and Clorinda Orbegoso recently completed a survey of the Huanca/Lluta
area. They have indicated to me that their survey produced abundant Collagua ceramics (pers. comm.
2002); however, detailed analysis and comparison with Colca valley Collagua ceramics has not been
undertaken. As in the case of the Moqugua drainage (Stanish 1985, 1989a; Van Buren 1996), future survey
and excavations in these outlying areas could shed light on this critical question.
141
In addition to the Collaguas and Cabanas, colonists from other ethnic groups resident in the valley at the
time reduccin included the Chumbivilcas, the Canas and Canchis, the Yuminas, among others (Galdos
Rodrguez 1987; Julien 1985).
366
is listed with a separate kuraka, suggesting that they were split into segments that were
administered as separate units in each reduccin. This kind of attenuated disarticulation
probably was negotiated between local elites and the Toledan visitadores during the visita
general. The resulting fragmentation of authority at the base of the political hierarchy
could have been a factor in the breakdown and abandonment of some of the canal and
terrace systems that I discussed in the previous chapter.
The apical ayllu Collana illustrates how pataca-level political leadership was
partitioned between villages while higher-order structures of political authority were
preserved. In the case of ayllu Collana of Hanansaya, most of the data from Yanque are
missing, but the majority of the ayllu probably lived there, given that only a small
segment (registered as ayllu Collana Malco) of 25 tributary households resided in
Coporaque (APY Yanquecollaguas Hanansaya 1615-1617:ff. 480v-493r).142 In any case,
the kuraka of the ayllu Collana segment in Yanque, Joan Halanoca, was also the
paramount kuraka of the entire province.
Of course, colonial authoritiesdoubtless much like their Inkaic counterparts
could and did also intervene in the replacement and succession of specific ayllu leaders.
The visitas register several instances in which the visitador replaced kurakas who hid
tributaries from the census or otherwise neglected tribute collection.143 However, given
142
The headman of the Coporaque segment of ayllu Collana was not exempt from tribute (presumably
because of the small size of the ayllu segment and/or its colonial origin) but nonetheless, as headman of the
highest ranking ayllu segment in the village, he was listed as the mayor (alcalde) of Coporaque (APY
Yanquecollaguas Hanansaya 1615-1617:f. 480v).
143
Such was the case in 1604 for the kuraka of ayllu Collana (Urinsaya) of the herding village of Tisco,
one Felipe Guapucho, who was not present during the visita to that village, but instead was in Yanque
(APY Yanquecollaguas Urinsaya 1604:ff. 106v-107r). When the visitador later arrived in Yanque to
conduct the census, Guapucho, along with a large group of households from his ayllu was still there. Upon
367
the intense pressure from the crown to increase tribute extraction (and thus to minimize
social upheaval), they also had to take indigenous traditions and expectations regarding
succession and legitimacy of rule into consideration.
For example, similar to its
counterpart in the upper moiety, the apical ayllu Collana of Urinsaya also was divided
between Yanque and Coporaque, and its kuraka in Yanque, one Jusepe Guaasuri, was
registered as the paramount kuraka of the moiety (Verdugo and Colmenares 1977
[1591]:f. 1r [192]).144 Consistent with Ulloas description of the ascribed status of the
apical kurakas in the province, Guaasuri, son of the previous paramount, had recently
inherited the office and was only nine years old at the time of the 1591 visita (Verdugo
and Colmenares 1977 [1591]:192 [f. 1r]).
However, Guaasuri was the legitimate
successor only according to the European tradition of primogenitor, and not according to
indigenous rules of succession.
As Ulloa (1965 [1586]:330) explained in detail,
indigenous tradition prescribed priority of succession to the brother of the former

(deceased) principal kuraka over the son, even the son of his primary wife. 145 Therefore,
registering this ayllu while in Yanque, the visitador reprimanded Guapucho and appointed don Francisco
Guaasuri to be the kuraka of the ayllu (ibid: ff. 317v-318r). Visita of Tuesday, October nineteenth of said
year [1604], of the Indians of the ayllo Collana, who are reduced in the village of Tisco, were not
registered there because they were absent and are instead registered here in this village of Yanque because
time was given to the headman to gather them up. [new section header] Ayllo Collana of the village of
Tisco, of whom don Francisco Guasuri is placed as headman because don Felipe Guapucho, who was the
headman, is a fugitive and does not fulfill the collection of taxes (APY Yanquecollaguas Urinsaya 1604:ff.
106v-107r, my translation). "Vissita de Martes diez y nueve de Otubre de el dicho ao de los yndios de el
aillo Collana, que estan reduidos en el pueblo de Tisco, que no se vissitaron alli por estar aussentes y
vissitansse en este pueblo de Ianque porque se le dio tiempo a el prinipal para juntarlos. [new section
header] Ayllo Collana de el pueblo de Tisco, de que se pone por prinipal don Franisco Guasuri, porque
don Felipe Guapucho, que lo era, es imarron y no acude a cobrar la tassa".
144
145
This pattern is true of Laricollaguas as well. See Guillet 1992:21-25.
The principal cacique [i.e. kuraka] had command over all of the other headmen, who were very
[extremely] obedient in everything that he commanded, as in matters of war or justice or the punishment of
offenses. This principal cacique was [had been] placed by the Inka and was succeeded by his sons, and
lacking them [sons] by his brothers, although the caciques legitimate brother was preferred as heir rather
368
according to these rules, Guaasuri, as son of the former kuraka, should have been passed
over in preference to his paternal uncle. However, because of his young age, a proxy was
needed to fulfill Guassuris obligations as moiety paramount, and, as noted earlier by
Cock Carrasco (1976-77:102-103), the choice of this proxy is illustrative. Acting in his
stead, under the title of gobernador (second in charge) was Guaasuris paternal uncle,
don Francisco Yncapacta (Verdugo and Colmenares 1977 [1591]:192, 343 [ff. 1r-2r,
97r]).
So it appears that the legitimate successor from the point of view of the
community (the brother of the former kuraka) remained the de facto authority of
Yanquecollaguas Urinsaya despite the alteration of the rules of succession. The tenure of
Yncapacta as gobernador was temporary, however, and twenty-six years later, in 1617,
this second-in-charge position had reverted to the wealthy lord of the ayllu Pahana
Collana Pataca of Coporaquethat is, the highest ranked pataca of the middle-ranked
macro ayllu Payan [Pahana].146 This is the structurally correct individual to fill this
second-highest post in the moiety, and presumably he was appointed in place of
Yncapacta after Guaasuri reached adulthood and was deemed capable of fulfilling his
role as paramount.
than the son, even if legitimate. (Tena el cacique principal mando y poder sobre todos los dems
principales, los cuales le eran obedientsimos en todo lo que mandaba, as en las cosas de la guerra como en
las cosas de justicia y castigo de delitos. Era este cacique puesto por el inga y subcedan sus hijos y a falta
dellos sus hermanos, aunque era preferidos en la herencia el hermano legtimo del cacique a su hijo, aunque
fuese legtimo.)
146
This kuraka, don Martin Hanco, is listed in the preamble to the visita of Tuti as gobernador of Urinsaya
(APY Yanquecollaguas Hanansaya 1615-1617:f. 391r). He is listed as kuraka of the ayllu Pahana
Collana Pataca of Coporaque in the 1604 Urinsaya visita.
369
A Regional View of Collagua Complemenarity Practices
Village-Scale Agricultural Declarations
The visitas provide detailed evidence for reconstructing how this widelydispersed constellation of ayllus structured complementarity relationships between
agriculturalists and pastoralists. The total field areas and crops declared by each village
in the visitas provide a first index of the organization of agro-pastoral production in
Yanquecollaguas.147 Figure 8.3 represents a composite synchronic view148 of 15911617 field declarations from the visitas. The data presented graphically on the map are
also summarized in Table 8.4.
On the map, pie chart size is proportional to total
agricultural field area declared (in topos). As one would expect, agricultural landholding
declarations per village generally decrease moving up the valley, with the greatest areas
declared by villages in the valleys agricultural coreAchoma, Yanque, and
Coporaquefollowed by the mid-altitude suni zone villages of Chivay, Canocota and
Tuti, and the much smaller areas declared in the high-altitude herding villages of Callalli,
Sibayo, Tisco, Guaraoma, and Mamaniviri. Also, the percentage of maize is greater in
the lower altitude village of Achoma (67%) than in Coporaque (59%) or Yanque
147
The sample is composed of landholding households that with a male heads of household between 18 and
50 years of age (tributary age), including those (e.g. kurakas) exempted from tribute.
148
By composite synchronic, I mean that I have structured my database queries to combine the
declarations of both moieties from the 1591, 1604, and 1615-1617 visitas while excluding redundant
declarations for a given village and moiety. Combining the data from several years is necessary because
there are no complete data for both moieties of all villages from a single census year. Therefore, I use the
1604 visita for the Urinsaya moiety for all villages except Yanque because part of that section is missing
from the original document. In its place, I use its complete counterpart in the 1591 Urinsaya visita, which
begins with the opening section on Yanque missing in the 1604 visita. Likewise, I use the 1615-1617 visita
for the Hanansaya moiety for all villages except Yanque because the Yanque section of this census is
missing. Therefore I use the small fragment of the 1591 Hanansaya visita, which includes only a small
portion of Yanque. Although incomplete, this 1591 visita provides the only census data available for the
Hanansaya moiety of Yanquesee Tables 8.1 and 8.2 above.
370
(42%149), and maize dominated the cultivar regime of the lower-altitude villages of
Huanca and Lluta (96% and 93% respectively).
Clearly, maize was cultivated
preferentially to other crops. By contrast, the proportion of the more frost-tolerant

quinoa, caagua, and potatoes increases with altitude (See Fig. 8.3, Table 8.4).
149
Due to damage to the original document, the crops grown in many fields were illegible in the 1591
Urinsaya visitahence, the high percentage of fields of unknown crop type. So this figure almost
certainly under-represents the percentage of maize in Yanque.
371
Figure 8.3: Crop declarations by village, Yanquecollaguas, 1591/1604/1615-1617. See Table 8.4
below for detailed data summary.
*Due to the fragmentary nature of the original documents, the data for Yanque Hanansaya, Achoma Hanansaya and Achoma Urinsaya are
incomplete. See Tables 8.1 and 8.2 above.
Table 8.4: Crop declarations by village, Yanquecollaguas, 1591/1604/1615-1617.* See Figure 8.3 above for spatial representation.
372
373
The total agricultural landholding area claimed per household also generally
decreases moving up the valley. That is, households amongst villages in the kichwa zone
(Huanca, Lluta, Maca, Achoma, Yanque, Coporaque, and Chivay) tend to have more
agricultural land, with an average of 1.7 topos per household, while households in the
suni zone (Canocota and Tuti) average 1.4 topos, and those in the puna (Callalli, Sibayo,
Tisco, Mamaniviri/Hivillaya, and Guaraoma) average only 0.8 topos per household.150
But there is also considerable inter-village variability within the central valley in terms of
the distribution of landholdings per household. Yanque and Chivay together stand out for
their greater average per household landholdings: 2.4 and 2.3 topos respectively (Figure
8.4). However, these values under-represent the actual wealth of Yanque households as a
group, since they include only the Urinsaya data from 1591 and a small fragment of the
total declarations from the 1591 Hanansaya visita, which, because the original document
is fragmentary, begins with the low-ranking, nearly landless households at end of the
Yanque section (such as the pastoralist yanaconas discussed above). These incomplete
Hanansaya data therefore skew the per capita values downward for Yanque. If they are
excluded from the calculations for Yanque, the per household average amongst Urinsaya
households is 3 topos per householdover double all other villages except Chivay. This
more accurately reflects its high status as the provincial capital.151
150
For these and the following figures regarding land tenure and livestock holdings, I have structured my
database queries such that households includes those headed by men 18 years of age and older, including
elderly households, kurakas, single adult men. These figures also include landless households (as defined
by the criteria just mentioned). They exclude orphans younger than 18 years old (except for one special
case: the paramount kuraka of Yanquecollaguas Urinsaya, Jussepe Guaasuri of ayllu Collana of Yanque),
male heads of household of unspecified or illegible age, widows, and single women.
151
Moreover, this figure almost certainly underrepresents Yanques actual average per-household
landholding area, since it excludes the majority of Yanque Hanansaya households, and includes only what
374
3
Avg. Topos per Household
2.5
2
1.5
1
0.5
Guaraoma
Mamanviri
Tisco
Sibayo
Callalli
Tuti
Canocota
Chivay
Coporaque
Yanque
Achoma
Maca
Lluta
Huanca
Increasing Altitude >>>
Figure 8.4: Average landholding area per household152
were probably the poorest ayllus in the village. The data from Yanque preserved in the 1591 Hanansaya
fragment is limited to a handful of households from an unidentified ayllu (the fragment begins in the
middle of this ayllu) and two ayllus of retainers: cumbicamayos and yanaconas of Joan Halanoca,
paramount kuraka of the province (and kuraka of ayllu Collana of Yanque Hanansaya) (APY
Yanquecollaguas Hanansaya 1591: no foliation).
152
Hanansaya data from Yanquecollaguas Hanansaya 1615-1617, except Yanque, from APY
Yanquecollaguas Hanansaya 1591 (incompletesee Table 8.1 above). Achoma section of 1615-1617
Hanansaya data are incomplete; see Table 8.1. Urinsaya data from APY Yanquecollaguas Urinsaya 1604,
except Yanque, from Yanquecollaguas Urinsaya 1591 (Verdugo and Colmenares 1977 [1591]). Achoma
section of 1604 Urinsaya data incomplete; see Table 8.2.
375
18
16
Topos per Household
14
12
10
8
6
4
2
0
N=
11
23
27
209
208
529
82
71
249
34
109
155
39
15
G
r
ua
ri/
vi
m
ao
i
an
am
M
yo
lli
la
sc
Ti
al
ba
Si
ot
u
aq
y
va
oc
an
hi
e
qu
or
op
t
Tu
a
om
n
Ya
h
Ac
ac
M
a
ut
nc
ua
Ll
H
iv
illa
ya
Figure 8.5: Landholding area per household. See note 38 for data sources.
The boxplot in Figure 8.5 illustrates inter- and intra-village inequalities in land
wealth in further detail. This graph clearly illustrates not only that Yanque households
tended to have more land overall than the other villages, but also that there were greater
inequalities in land wealth within Yanque than in any other village. The wealth of
Yanques households as a group is apparent in the high lower hinge and median values
for Yanque: the median household in Yanque claimed more agricultural field area than
even the wealthiest house in the third quartile (the upper hinge) of the other villages. The
greater stratification of wealth amongst Yanque households is reflected in its broad
midspread (middle 50%) and the great disparity between the first and fourth quartiles (24.25 topos per household).
By contrast, the middle 50% of households in Maca,
376
Achoma, and Coporaque possessed a much smaller range of field areas, illustrating a
large homogeneous middle class of landholders.153
The outlier and extreme cases in the boxplot represent the kurakas and other elite
households that are far wealthier relative to the general population of tributary
households (indios tributarios). As would be expected, those of Yanque were among the
wealthiest, but the kurakas (denoted by the circles and asterisks) of the other villages also
clearly stand apart from their respective tributary populations. Interestingly, in Yanque,
the person with the most land was not the paramount kuraka of Urinsaya (i.e. the kuraka
of Collana ayllu), the nine year old Jussepe Guaasuri, but rather his uncle, Francisco
Yncapacta, who, as I discussed above, was the legitimate heir to the position according
to the indigenous rules of succession, but was serving instead as gobernador. Thus it
appears that although the Spanish interceded in the matter of succession, Yncapacta,
although not the de jure paramount, did not lose access to land-wealth.
Political rank and ayllu affiliation affected not only the total landholdings of a
given household, but also their access to supra-local fields. The proportions of crops
displayed in the map of Figure 8.3 also reveal that herding households in the puna had
access to supra-local fields. This is most obviously evident in maize declarations in the
suni and puna zone villages, because maize agriculture is impossible at these altitudes
(see also Table 8.4 above). Amongst the high altitude herding settlements (Callalli,
153
The seemingly high median and midspread values for Tuti, Sibayo, and Tiscoall in the upper reaches
of the valleycan be explained in part by the high proportion of potato fields claimed in these villages. As
I discussed above, potato fields make up a much higher percentage of the land claimed in these villages
(26% of the field area claimed in Tuti, 68% in Sibayo, and 47% in Tisco) than in the villages in the central
valley, where virtually no potato fields were declared. As a high altitude crop, potatoes are generally
grown in larger fields that require frequent fallowing, so households need more fields overall in order to
manage rotational fallowing.
377
Sibayo, Tisco, and the estancias of Mamaniviri/Hivillaya and Guaraoma) well over a
quarter (29%, 72 of 249 households) claimed maize fields. Access to maize fields was
skewed to wealthier households; as Figure 8.6 illustrates, households in the suni and puna
zone villages with maize fields tend to have more total agricultural landholdings than
those without maize fields. Most of these fields were located in the central valley around
Coporaque and Yanque, and the patterning of their locations illuminates how households
gained access to these fields through their ayllu membership. For example, virtually all
(27.25 of the 28 topos) of the fields located in Coporaque that were claimed by
households from other villages were from Tuti (9.25 topos), Sibayo (8.25 topos), and
Tisco (10.25 topos).154 As I illustrated graphically in Figure 8.2 above, these villages all
have outlier ayllu segments headed by kurakas in Coporaque, so their access to these
fields in Coporaque clearly was mediated by their ayllu affiliation.155
Households in the herding villages with large field holdings also declared fields in
the outlying maize production enclaves of Huanca, Lluta, and Arequipa. Fields in these
three maize production enclaves claimed by households in Callalli, Sibayo, Tisco, and the
estancias total 45.75 topos. Thus, rather than (or at least, in addition to) obtaining kichwa
zone agricultural produce through redistribution by kurakas or direct trade, these
(generally well-off) puna households gained direct access to the means of its production,
probably through their respective kurakas, and cultivated agricultural fields in lower
altitudes themselves. This kind of direct access to supra-local maize fields would have
154
These are conservative figures because they only include those outlier fields that were specifically
located by both toponym and village. Many maize fields claimed in the puna were located by toponym
only, so they cannot be located by village.
155
I discuss how land tenure rights were mediated by kurakas in further detail below.
378
required households to either send family members or proxies to tend their fields, or to
take up part time residence in the villages near their maize fields. An example is found in
the fragmentary section from Yanque in 1604, where the visitador Juan de Rivero noted
that the ayllu segment of Collana of Tisco were not censused in their home village
because they were absent and therefore registered in this village of Yanque because
their leader had time to assemble them (translation by Treacy 1989: 220; APY
Yanquecollaguas Urinsaya 1604:f.317v-318r). This section of the visita was recorded
during the planting month of October, so this ayllu Collana segment from Tisco may
have been in Yanque to cultivate their fields (cf. Treacy 1989:219-220).
Moreover, this kind of household-level access to non-local fields was also quite
common in the villages of the central valley. Agriculturally well-endowed households in
Yanque, Coporaque, Achoma and Chivay claimed far-flung maize fields along the
western bank of the Chili river adjacent to the city of Arequipa and also near the villages
of Huanca and Lluta Figure 8.3).
These fields constitute significant quantities of
Collagua household outlier landholdings, totaling 135 topos (Table 8.5). These fields
almost certainly were associated with large colonies of Collagua households
(mitmaqkuna) settled around the city of Arequipa (see below).
379
4.5
Mean Topos per Household
4
3.5
3
2.5
W ith Maize
No Maize
2
1.5
1
0.5
Guaraoma
Mamanviri
Tisco
Sibayo
Callalli
Tuti
Canocota
Chivay
Coporaque
Yanque
Achoma
Maca
Lluta
Huanca
Figure 8.6: Comparison of average landholdings per household amongst households with maize fields
and households without maize fields.
Table 8.5: Non-local household fields, Yanquecollaguas 1591/1604/1615-1617*
Village of
Residence:
Location of Outlier Fields (topos)

Achoma Yanque Coporaque Canocota
Arequipa
Huanca
Lluta
Achoma**
9.25
n/a
Tuti
0
Callalli
1
Total
10.25
Yanque***
26.75
3.5
0.25
n/a
5.75
38.25
Coporaque
33.25
0.25
2.5
0.5
2.75
n/a
0.75
1.25
41.25
Chivay
3.25
2.25
9.50
Canocota
1.75
0.5
0.25
n/a
0.75
3.25
6.5
7.25
9.25
n/a
26.00
Tuti
Callalli
n/a
12.00
Sibayo
13.5
8.25
25.75
Tisco
19
10.25
11.25
45.50
Mamanviri
1.75
1.75
Guaraoma
4.25
0.5
4.75
Totals:
122.5
2.25
10.5
17.25
9.75
28.00
1.25
24.75
2.00
218.25
*Hanansaya data from Yanquecollaguas Hanansaya 1615-1617, except Yanque, from APY
Hanansaya data are incomplete; see Table 8.1 above. Urinsaya data from APY Yanquecollaguas Urinsaya
1604, except Yanque, from Yanquecollaguas Urinsaya 1591 (Verdugo and Colmenares 1977 [1591]).
Achoma section of 1604 Urinsaya data incomplete; see Table 8.2 above.
**Incomplete data for both Hanansaya and Urinsaya, see note [*] and Tables 8.1 and 8.2.
***Incomplete data for Hanansaya, see note [*] and Table 8.1.
380
Village-Scale Livestock Declarations
The Collagua were renowned for their massive camelid herds, and wool was their
primary source of wealthin fact, none of the colonial tribute of Yanquecollaguas and
Laricollaguas was levied in agricultural goods. Instead, 60% of their tax burden was
levied in camelids and textiles made from alpaca wool, and the remaining 40% was
levied in cash (Cook, et al. 1975 [1582]:217-221). In particular, Collagua textile tribute
was levied in coarse homespun, or ropa de abasca, an important commodity because it
was one of the provisions that Spanish patrons were legally required to furnish for their
native laborers (Guillet 1992:23). But the predominance of pastoralist products in the
Collagua tax assessment almost certainly was built on Inka-era precedents (although, of
course, tribute was levied in labor, not in specie or kind). As I discuss below, the
presence of cumbicamayocsweavers of fine cumbi cloth, the most prestigious
commodity in the Inka empirein Yanque and Achoma also indicates that the
production of fine textiles was an important component of the Collagua provincial
political economy under Inka rule.
Livestock declarations in the visitas fall far short of the actual total herd size of
the Collaguas, because they only record livestock that was the personal property of
individual households. There is no reliable account of the total size of the herds of the
Collagua during early colonial times, but a witness in a judicial inquiry from 1585-1586
estimated that the flocks of Callalli and Tisco alone totaled over 25,000 head (AGI
Justicia 480:f. 1196v, published in Crespo 1977:71-72). The combined declarations of
these two villages in the visitas2354 headis less than 1/10th of this estimate. Also,
381
even in the herding villages of Callalli, Sibayo, and Tisco only 1/3 of households claimed
livestock in the visitas (Figure 8.7). Apparently there were much larger herds held
collectively by ayllus, and many of the remaining 2/3 of households without personal
livestock were probably dedicated to tending these communal flocks.
Percent of Housholds w/ Livestock
40%
35%
30%
25%
20%
15%
10%
5%
Guaraoma
Mamanviri
Tisco
Sibayo
Callalli
Tuti
Canocota
Chivay
Coporaque
Yanque
Achoma
Maca
Lluta
Guanca
0%
Figure 8.7: Proportion of households with livestock, by village.

Hanansaya data from Yanquecollaguas Hanansaya 1615-1617, except Yanque, from APY
Hanansaya data are incomplete; see Table 8.1.
Urinsaya data from APY Yanquecollaguas Urinsaya 1604, except Yanque, from Yanquecollaguas
Urinsaya 1591 (Verdugo and Colmenares 1977 [1591]). Achoma section of 1604 Urinsaya data
incomplete; see Table 8.2.
In contrast to village-level landholdings, the number of head of livestock claimed

per village, and the percentage of households that claim livestock generally increases
with altitude (Figure 8.7). With the exception of Yanque Hanansaya, only 1-3% of
382
households in the kichwa zone villages declared any livestock.
The percentage of
households with livestock increases to between 29% and 37% in Callalli, Sibayo, and
Tisco. Of the 249 households with livestock in these three villages, 69% also claimed
agricultural fields, and, as I discussed above, 29% claim at least one maize field. By
contrast, only 25% of households in these villages claimed fields but no livestock.
Kurakas and other elite households with many agricultural landholdings claimed
especially large numbers of livestock. For example, in Sibayo, the kuraka of ayllu
Paragra Pahana Collana Pataca, don Luis Chacha, claimed 280 head of livestock.
Along with 200 alpacas and llamas (ganado de la tierra), Chacha was one of only a
handful of kurakas that claimed Old World livestock in the visitas40 cattle and 40
goats. He was also wealthy in landholdings, claiming the third largest amount of land in
the visitasa total of 15.5 toposover half of which (8.5 topos) were declared as maize
fields.
In the central valley, the owners of livestock claimed in Yanque and Coporaque
were almost exclusively kurakas and other elite households with large landholdings.
Clearly, livestock ownership was both a marker and motor of wealth and status. For
example, 208 of the 212 head of livestock declared in Coporaque Urinsaya pertained to
Martin Chuquianco, kuraka of the ayllu Pahana Collana Pataca (APY Yanquecollaguas
Urinsaya 1604:f. 236r). All 30 head (alpacas) of Yanque Urinsaya were claimed by the
second-in-charge of ayllu Collana, Garcia Checa (Verdugo and Colmenares 1977
[1591]:193 [f. 2r]). Most significantly, as I mentioned above, the yanaconas of the
paramount lord of the province, Joan Halanoca of Yanque Hanansaya (ayllu Collana)
383
claimed a total of 510 head of livestock. However, these retainers claimed virtually no
agricultural fieldsonly 2 topos amongst all of them. As I discuss below, it appears that
they obtained agricultural produce through the paramount Halanoca in exchange for
wool, which he then used to supply an enclave of weavers of fine cumbi cloth that were
also his personal retainers (a point also raised by Cock Carrasco 1976-77:104-105).
384
Figure 8.8: Livestock declarations by village, Yanquecollaguas 1591/1604/1615-1617.
385
Kurakas, Craft Specialists, and the Provincial Political Economy
Embedded within the ayllu structure of the Collaguas were groups of official Inka
state craft specialists. They manufactured each of the primary classes of prestige goods
that circulated in the imperial political economyfine cumbi cloth, pottery, and metal
(specifically, silver) items. All of these artisans were residents of the central valley in the
villages of Yanque, Coporaque, and Achoma, with an outlier group of silversmiths in
Maca. As mediators of land tenure rights, kurakas almost certainly provisioned these
groups with their agricultural fields.
Cumbi cloth was the most valued prestige good in the Inka imperial economy
(Murra 1962).
As was the case throughout Tawantinsuyu, the four groups of
cumbicamayoc registered in the Yanquecollaguas visitas (three in Yanque and one in

Achoma) were attached specialists held in retainer by their respective kurakas. In the
fragment that remains of the Yanque section of the 1591 Hanansaya visita, a group of
cumbicamayoc who pertained to the paramount kuraka Halanoca precedes the listing of
his yanacona.156 There were 14 tributary households (as well as four elderly households)
listed among these weavers, (APY Yanquecollaguas Hanansaya 1591). Based on their
field declarations, they were clearly self-sufficient in terms of agricultural produce,
claiming a total of 47.25 topos and averaging 2.8 topos of land per householdtwice the
average per household landholdings of Coporaque or Achoma (see Figure 8.4 above).
However, they declared no livestock or other access to wool, and so they would not have
156
Specifically, the oficiales cumbicayos are listed with a headman (one Juan Aromoto), but subject to
don Joan Halanoca. Thus, it appears that Aromoto was the mandn but the group as a whole pertained to
Halanoca. The yanaconas are listed as Yndios Yanaconas de don Joan Halanoca que estan rreduzidos en
este pueblo. While yanacona were normally exempt from tribute, in this case the households headed by
tributary age men were listed as tributaries. There were 10 tributary households, two households headed by
elderly men, and one orphan.
386
been able to provision themselves with the material for their weavings.
The most
parsimonious answer to the question of where they obtained their wool is through their
kuraka, who, as I discussed above, also held authority over pastoralist yanacona and
other ayllu segments of herders. Thus, it appears that the paramount kuraka underwrote
prestige goods production by mobilizing staple finance. That is, he converted maize from
his extensive landholdings and outlier ayllu segments into wool through exchange with
his pastoralist yanacona, and then used that wool to supply his cumbicamoyoc.157
Other kurakas appear to have similarly mobilized their diverse agro-pastoral
resources to provide wool for their cumbicamayoc groups. While they did not have
pastoralist yanacona like the provincial paramount, many had large personal herds and
also held authority over outlier ayllu segments in the puna. There were two separate
cumbicamayoc groups among the ayllus of Urinsaya in Yanque. One (a group of six
tributary households and one elderly household) pertained to the kuraka Joan Guaasuri of
the apical ayllu Collana, who also held authority over ayllu segments in the herding
village of Tisco and the estancia of Mamaniviri (see Figure 8.2 above).158 These weavers
also had plenty of agricultural fields, with an average of 2.5 topos per household (Table
8.6). The other cumbicamyoc group of Yanque Urinsaya were the subjects of the kuraka
157
Undoubtedly, some of the cumbi produced in the Colca was also made from the wool of the vicua
(Vicugna vicugna), a wild Andean camelid abundant in the high puna surrounding the valley. During
Inkaic times, the kurakas or other state administrators probably sponsored festivals to wrangle and sheer
the chest fur of the vicuas that they then supplied to the cumbicamayoc.
158
In the 1591 visita, first ayllu of cumbicamayoc is listed as Yndios Officiales Cumbicamayos del Aillo
Co (damaged) (burned) (Verdugo and Colmenares 1977 [1591]:260 [f.42r]). Fortunately, this section is
included within the fragmentary section of Yanque preserved in the 1604 visita, and it can be confirmed
that this was referring to Collana ayllu: Yndios Cunbicamayos del ayllo Collana de la parialidad de
Urinsaya, de que es prinipal don Jusepe Guasuri, caique prinipal de toda la dicha parialidad (APY
Yanquecollaguas Urinsaya 1604: f. 66r)
387
of the ayllu Taypi Pataca of Yanque, Diego Chacha, who also held authority over herders
in the estancias of Mamaniviri and Hivillaya.159 This group of weavers claimed on
average 1.6 topos per household (Table 8.6). In Achoma Urinsaya, a small group of
cumbicamayoc households (three tributary households, two widows, and one elderly
household) were subjects of the local kuraka of the ayllu Collana Pataca (APY
Yanquecollaguas Urinsaya 1604: ff. 383v-385v). They were also agriculturally selfsufficient, with an average of 1.2 topos per household. However, this kuraka is not listed
as head of any outlying ayllu segments in the puna and did not claim any personal
livestock, so it is unclear how these weavers were supplied; he may have obtained wool
through the paramount kuraka or some other means.
159
In the 1591 visita, the second group of cumbicamayocs is listed as Yndios Cumbicamayos [damaged]
de Don [damaged] de que es principal dicho Martin Taco. In the 1604 visita, this section header is
undamaged: Yndios Cunbicamayos deste ayllo Taypi Pataca, de don Diego Chacha, de que es prinipal
Martin Taca [sic].
388
Table 8.6: Landholdings and livestock of craft specialists, Yanquecollaguas 1591/1604/1615-1617
Village
Moiety
Achoma
Coporaque
Urin
No. of
Households
Year
1604
Landholdings per
Household (topos)
Mean
Sum
Livestock per
Household
Mean
Sum
Cumbicamayocs (of
kuraka of ayllu Collana
Pataca)
1.19
4.75
Silversmiths of ayllu
Collana Pataca
1.75
8.75
Hanan
1616
Official Potters
40
1.02
40.75
Urin
1604
Potters of ayllu Pahana

Collana Pataca
1.13
6.75
Official Potters of ayllu

[Pahana] Cayao Pataca
.88
7.00
Maca
Hanan
1616
Silversmiths (subjects of
paramount of province)
33
.97
32.00
Yanque
Hanan
1591
Cumbicamayocs (of
19
2.49
47.25
Yanaconas (retainers of
20
.10
2.00
26
510
Cumbicamayocs (of
kuraka of ayllu Collana)
2.43
17.00
Cumbicamayocs (of
kuraka of ayllu Taypi
Pataca)
1.56
12.50
Urin
1591
Two ayllu segments of silversmiths were also registered in the visitas, one of
Hanansaya in the village of Maca and one of Urinsaya in Achoma. The silversmiths of
Achoma, like the cumbicamayos there, were part of the ayllu Collana Pataca, and subject
to its local kuraka.160 In addition to his cumbicamayos and yanaconas, the apical kuraka
Halanoca held authority over a large ayllu segment of silversmiths reduced to Maca,
located within the repartimiento of Laricollaguas.161
No such outlier ayllus from
160
There were four tributary households, one widow, and one elderly man recorded under the heading of
yndios plateros of Collana Pataca ayllu in Achoma (APY Yanquecollaguas Urinsaya 1604: ff. 385v-387v).
As I noted in Table 8.2 above, the ayllu Collana Pataca of Achoma had been split in two, but were rejoined
by the visitador at the death of one of the kurakas, and placed under the authority of surviving kuraka, don
Miguel Sapana [apana] (APY Yanquecollaguas Urinsaya 1604:f. 387v).
161
A page (two folios) is missing from the section that records the silversmiths of Maca, but among the
folios present, there were 29 tributary households, five households headed by elderly men, 18 widows and
unmarried women, two orphan boys and one orphan girl.
389
Laricollaguas are found within the territory of Yanquecollaguas, illustrating the one-way
porosity of the boundary, and the power asymmetry between the repartimientos.162
Confirming that the Inka established state pottery production enclaves in the
Colca valley, three groups of potters were listed in the visitas: two in Urinsaya, and one
in Hanansaya. They all resided in Coporaque; two of the groups pertained to Urinsaya
and one to Hanansaya. The two groups of Coporaque Urinsaya potters were registered
as subsections within the ayllus Pahana Collana Pataca and Pahana Cayao Pataca.163
The first of these (those of Pahana Collana Pataca) was a small group of six tributary
households headed by their local kuraka in Coporaque.
The second group, also
composed of six tributary households164, was registered as the official potters of the ayllu
Pahana Cayao Pataca.
Although this second group had also been reduced
(reducidosi.e. resettled from their village original prehispanic village) to Coporaque

they were listed as subject to don Diego Chacha, kuraka of the ayllu Taypi Pataca (and
lord of a group of cumbicamayoc) in Yanque.165 Again, the process of reduccin appears
162
I thank Maria Benavides for this observation (M. Benavides pers. comm. 2003).
163
The second of these two are recorded as the official potters of Cayao Pataca; however, Cayao Pataca
here almost certainly represents a truncated name for the ayllu Pahana Cayao Patacathat is, the lowest
ranked ayllu of the Pahana [Payan] macro ayllu. I arrive at this conclusion because the potters are listed
immediately after an ayllu that was recorded as ayllu Pahana Cayao Pataca in its header (APY
Yanquecollaguas Urinsaya 1604: f. 290r), but similarly truncated as Cayao Pataca in the summary (ibid: f.
309r). Also, these potters are subject to the kuraka of ayllu Taypi Pataca of Yanque (don Diego Chacha),
not to any kuraka of the Cayao macro ayllu (ibid: f. 309vcompare with Verdugo and Colmenares 1977
[1591]:238 [f. 29v]).
164
Two unmarried women and two households in the elderly and infirm category (APY Yanquecollaguas
Urinsaya 1604: ff.309v-311v).
165
In the 1591 visita to Yanquecollaguas Urinsaya, there are two ayllus with the name Taypi Pataca listed
in the Yanque, each with a separate kuraka. The second of these Taypi Pataca ayllus was headed by don
Diego Chacha, who, as I discussed above, also held in retainer a group of cumbicamayoc and a group of
potters in Coporaque. As I discuss below, he also held authority over ayllu segments of herders in the
herding hamlets of Mamaniviri and Villaya (or Hivillaya).
390
to have split this ayllu apart but the original structure of authority was preserved. The
ayllu of Hanansaya potters in Coporaque was larger than either of the Urinsaya groups,
with 19 tributary households, 11 single men of tributary age, 12 households headed by
elderly men and 16 widows and unmarried women. This group was also listed as an
ayllu of official potters, but was not attributed as being a segment of a larger ayllu, but
rather was registered with its own kuraka.166

In this chapter, I first outlined how post-Toledan Spanish colonial administration
of the Collaguas extracted tribute through extant structures of community organization in
order to shed light on earlier, Inkaic administration of the province. Building on previous
research, my reconstruction of the political organization of the Inkaic province of
Collaguas illustrates how those indigenous institutions were themselves hybrid
state/local constructs. The effect of Inka administration was to augment and harden
inequalities within and between indigenous Collagua ayllus by reordering them according
to tripartite, bipartite, and decimal ranking concepts. This arrangement would have
limited imperial administrative burdensboth for the Inkas and the Spanishby
minimizing points of contact with local intermediaries while also providing a means for
166
Because the previous kuraka had been absent for over two years, a provisional ayllu headman was
appointed by the visitador, until the original headman returned. The ayllu header reads: Ayllo of the
official potters of this parcialidad Hanansaya, registered in this village of Santiago de Coporaque, of which
the headman was don Juan Suri, but because he has been absent for more than two years, Francisco Limaya
is placed in his stead until the aforementioned returns. (Ayllo de los ofiiales olleros desta parialidad
Hanansaya, visitados en este pueblo de Santiago de Coporaque, de que estava por prinipal don Juan Suri,
y por estar ausente ms a de dos aos, se pone en su lugar a Franisco Limaia, asta tanto que venga el
susodicho) (APY Yanquecollaguas Hanansaya 1615-1617:f. 603v)
391
gauging productivity and engendering competition between a nested hierarchy of
structurally-equivalent tribute units.
Second, complex, supra-local networks of ayllu affiliation and authority
continued to structure relations of production, exchange, and consumption between
agriculturalists and pastoralists a generation after resettlement. Despite segmentation of
some high-ranking ayllus in the central valley after reduccin, their leaders continued to
hold authority over and collect tribute from members of their ayllus in outlier pastoralist
production enclaves in the puna as well as maize-production enclaves in low-lying
valleys to the south. At the level of domestic economy, commoner agriculturalist and
pastoralist households probably participated in direct exchange with their ayllu cohorts in
other production zones to gain access to each others respective produce. By contrast,
elite households maintained direct access to the means of production in all production
zones: wealthy pastoralist households declared maize fields in the central valley and
beyond, and wealthy agriculturalists in the central valley claimed personal livestock held
in the puna. These distinct complementarity practices therefore developed not only as
adaptations to ecological imperatives, but also as a result of community affiliation and
social status.
In terms of its contribution to the regional political economy, the province of
Collaguas produced the largest single source of revenue collected in Arequipa during
colonial times, and was clearly an important source of regional revenue for the Inkas.
Collagua kurakas not only extracted surplus of staple goods from their variegated ayllu
resources, but they also (re)distributed staple goods, and the means of their production, to
392
finance the production of prestige goods. Some of the kurakas of the high-ranking ayllus
in Yanque and Coporaque also held authority over ayllu segments of official state potters,
silversmiths, and cumbicamayoc. The kurakas almost certainly provided each of these
groups of artisans with access to agricultural fields, and they were generally selfsufficient in terms of subsistence. It also appears that kurakas also underwrote wealth
finance with staple finance by exchanging agricultural produce for wool to provide the
materia prima for the cumbicamayoc.
In sum, while the archaeological evidence suggests that the Inkas consolidated
their rule by establishing an administrative center in Yanque and by coordinating their
administration with local elites at other nearby settlements in the central valley, this
chapter demonstrates the principles and practices that structured that consolidation of
rule. In complement to the regional-scale view of this chapter, the analysis of ayllu-level
land tenure patterns in the next chapter illustrates how household access to agricultural
fields was mediated by community organization at the local scale, and provides a means
for reconstructing where those ayllus lived before reduccin.
393
Chapter 9
Reconstructing Colonial Land Tenure and Late Prehispanic Residence Patterns
Introduction
In the previous chapter I illustrated how hybrid Collagua/Inka community
structures articulated regional scale patterns of agricultural and pastoral production,
exchange, and consumption during Inkaic and early colonial times. In this chapter I
present a complementary, local view of moiety- and ayllu-level land tenure patterning in
Coporaque by mapping out agricultural landholding declarations from the visitas of 1604
and 1615-1617. The two primary goals of the chapter are: 1) to demonstrate how the
dispersed land tenure patterns of Coporaque households were the product not only of
adaptive, risk-minimization strategies, but were also mediated by their moiety and ayllu
affiliation, and 2) to reconstruct the residence patterns of ayllus in the terminal
prehispanic period by comparing early colonial ayllu land tenure patterns to the Late
Horizon settlement pattern.
The analysis in this chapter is based on locating landholding declarations in the
visitas by matching the toponyms used to locate agricultural fields in the visita
declarations with their modern toponymic counterparts. As I discussed in Chapter 4,
household landholding declarations in the visitas included the size of each field, the
predominant crop grown in it, and the location of the field by reference to a toponym.
The 76 modern toponym sectors that also appear in the visitas account for 34% of the
total landholding declarations in the visitas (366.5 out of 1066.25 topos). This means that
I can map out the locations of just over a third of the total landholding declarations in the
394
visitasa sizeable, and probably representative sample, given that they are located in all
physiographic contexts. I use two visitas to provide a complete view of both moieties in
Coporaque: the 1604 visita of Yanquecollaguas Urinsaya, and the 1615-1617 visita of
Yanquecollaguas Hanansaya.167
The Local Crop Mosaic

As I discussed in the previous chapter, overall crop declarations in Coporaque
from 1604 (Urinsaya) and 1615 (Hanansaya) indicate that the local crop mosaic was
composed of 59% maize, 39% quinoa, and the remainder (2%) in other crops (potatoes,
caagua, and unspecified) (cf. Benavides 1986c; Benavides 1986d). Mapping out the
declarations using modern toponyms provides a view of where these crops were grown.
The dot density map of Figure 9.1 displays this data by representing each field that could
be located by toponym as a color-coded dot.168 This map illustrates how in early colonial
times (Crdova Aguilar, et al. 1986; Izaguirre Urbano n.d.; Treacy 1989), maize
cultivation was confined largely to complexes of bench terraces on the lower valley
slopes, while quinoa cultivation was concentrated on broadfield terraces and valley
167
I use these visitas because the Coporaque sections are complete in both, with the exception of a few
missing folios (missing folios in the Coporaque section of the 1604 visita: 245r/v [ayllu Pahana Collana
Pataca]; folios missing from the Coporaque section of the 1615-1617 visita: 595r-596v [ayllu Cupi]).
Coporaque appears in the fragment of the 1591 visita of Yanquecollaguas Urinsaya published in Pease
(Verdugo and Colmenares 1977 [1591]), but the section is incomplete. The Coporaque section of the 1604
Urinsaya visita was recorded between October 5-13, 1604, and the Coporaque section of the 1615-1617
Hanansaya visita was recorded between December 5, 1615 and January 15, 1616 (APY Yanquecollaguas
Hanansaya 1615-1617:ff. 480v-611v). All transcriptions by Laura Gutirrez Arbul. I thank Mara
Benavides for providing me with the photocopies of the original archival documents and the transcriptions.
168
Given that toponym sectors are the minimal unit of provenience in the visitas, the location of the dots
within each toponym is schematicthey are visual representations of the quantity of fields claimed within
each toponym rather than their specific locations within them.
395
bottom fields on the pampas of the Qal 4 and 5 geomorphic surfacesa pattern that
persists even today.
The airphoto detail in Figure 9.1 shows the intensively-terraced slopes of the
Waykiri, Ccayra, and Fayero areas below San Antonio/Chijra (CO-100). These toponym
sectors are situated in a warm microclimate that results from their altitude, slope angle,
orientation, and the terracing itself (Treacy 1989:81-84). Situated just above the valley
floor, these are some of the lowest-altitude terraces in Coporaque. Their southeasterly
orientation promotes early warming in the morning suna beneficial effect accentuated
by the mass and rock facings of the terraces. At night, the terraces radiate heat built up
during the day and further reduce frost risk by shedding cool air, which instead settles
downslope to the pampas and valley bottom (Treacy 1989:81-84). As is evident in the
dot density overlay on the airphoto, visita declarations of crops grown in the fields of the
these areas indicate that they were dedicated almost exclusively to maize production.
Combining the declarations of Ccayra, Fayero, and the Waykiri area, 78% (186 out of
237) of landholdings were declared as maize fields.
Coporaque villagers clearly
preferentially cultivated this culturally-valued crop over others in warm, fertile, low frostrisk areas. Maize, of course, was an esteemed, even sacred, crop and was a central
currency in virtually all social and ceremonial exchanges in the form of corn beer or
chicha (Johannessen and Hastorf 1994).
By contrast, the hardier, more nutritious quinoa was cultivated primarily on the
more frost-prone pampas above the valley floor. For example, the airphoto detail shows
the predominance of quinoa fields declared in the Lama toponym sector on the pampa to
396
the southeast of the Waykiri area (bottom right). Here, 79% (58 out of 73 fields) of
landholdings were declared as quinoa fields. As a key subsistence crop, Coporaque
villagers concentrated quinoa cultivation in the large broadfield terraces and valley
bottom fields.
397
Figure 9.1:Dot density map of crop declarations in Coporaque, 1604/1615-1
398
Moiety-Level Land Tenure Patterning

Benavides (1986e) previously reconstructed the distribution of fields by moiety in
Coporaque, and found that Hanansaya and Urinsaya declarations in the visitas are not
territorially discrete. Rather, households from both moieties share many toponym areas.
Using similar methods, Robinson has also recently documented this kind of interdigitated
moiety arrangement in Laricollaguas (Robinson 2003). My findings are consistent with
theirs. In the map of Figure 9.2, pie chart size is proportional to the combined field area
declared by both moieties, and the pie charts display the percentage of field area declared
by each moiety. Clearly, moieties as collectivities did not cultivate large discrete areas,
and did not control separate irrigation systems.
Although moieties did not control
separate sections of the canal network, the distribution of irrigation water could have
been accomplished through a moiety- or ayllu-based turn-taking system similar to that
employed today in Coporaque (documented in detail by Treacy 1989).
Canal
maintenance was probably also coordinated and executed between moieties.

Given this intermixed moiety land tenure constellation, it is unlikely that a single
moiety made up the entire population of any LIP/Late Horizon settlement, but rather that
settlement populations were composed of both Hanansaya and Urinsaya members.
Below I present a reconstruction that more specifically identifies which ayllus from each
moiety inhabited specific settlements during the Late Horizon.
399
Figure 9.2: Landholdings by moiety, Coporaque 1604/1615-1616
400
Ayllu-Level Land Tenure Patterning
Like the mixed moiety-level pattern, ayllu-level land tenure patterning in
Coporaque was also generally dispersed. Ayllus in Coporaque, and almost certainly in
the rest of the valley as well, did not control large, discrete areas of agricultural fields.
Instead, households in each ayllu declared fields in a constellation of often noncontiguous toponym sectors. As is common in high altitude Andean settings where
drought and frost are constant threats, the household risk-reduction strategy appears to
have been to disperse as many small holdings as possible over different verticallydistributed production zones, rather than maximizing field size in any single area.
Reflecting how households divided their total landholdings between many small fields,
the mean size of fields declared in Coporaque was just 1/3 topo (0.34 topos).169
However, while the dispersion of household landholdings would have minimized
risk, household tenure rights to their scattered fields were also clearly mediated by ayllu
affiliation. There are important, patterned differences in the spatial distribution of the
landholdings of Hanansaya and Urinsaya ayllus. Overall, the fields of Hanansaya ayllus
are more clustered than those of Urinsaya. Specifically, the fieldholdings of seven out of
the eight Hanansaya ayllus are concentrated on either side of the Chillihuitira river, a
major drainage (quebrada) that runs southward through the center of the Coporaque area
(Figure 9.3). The quebrada of the Chillihuitira hydrologically divides Coporaque in two,
as water from the Chillihuitira feeds most of the canals for the fields to the west of the
169
In the Department of Arequipa today, the standardized area of a topo is 3496 m2 (see Chapter 4).
Assuming that this is roughly equivalent to the area of a topo in early colonial times, the average field size
in Coporaque (0.34 topos) measured only about a tenth of a hectare (1189 m2).
401
quebrada, while water from the Sahuara/Cantumayo supplies the canals for most of the
fields to the east (Figure 9.3).
As I discussed in the previous chapter, the names of the ayllus Cupi and Checa
the right side and left side ayllus, respectivelyare suggestive of an underlying
dualistic organization that the Inkas integrated and ranked within their tripartite and
decimpartite organizational structure.
My reconstruction of ayllu-level land tenure
patterns reveals the spatial corollary of this indigenous dualism. On the one hand, the
landholdings of the right side ayllus Cupi and Aipi/Cupi, along with the ayllu Calloca
are located predominantly to the west of the Chillihuitira river. On the other hand, the
lands of the left side ayllu Checca Malco, together with the other two Malco ayllus
from Hanansaya (Collana Malco and Icatunga Malco) and the ayllu Yumasca, are
located predominantly to the east of this quebrada. Facing downstream (towards the
Colca river), west is on the right hand side, and east is on the left (Figure 9.3). In
Hanansaya, only the ayllu of official Inka state pottersclearly an ayllu with close ties
to the Inka statebreaks from this right/left pattern, and is instead more similar to the
dispersed land tenure pattern of lower-ranking ayllus of Urinsaya. As I also discussed in
the previous chapter, the names of the Urinsaya ayllus conform much more closely to the
formal tripartite and decimpartite structures of local Inka governance, suggesting a more
penetrating imperial influence than in the upper moiety. Of the Urinsaya ayllus, only the
high-ranking ayllu Collana displays marked predominance of fieldholding area to one
side (the western, or right side) of the Chillihuitira. The other Urinsaya ayllus, like the
ayllu of official state potters in Hanansaya, are dispersed. Therefore I argue that the
402
ayllus of the higher-ranking moiety (Hanansaya), together with the highest-ranking ayllu
of Urinsaya retained their rights to their traditional lands under Inka rule, while the more
dispersed pattern of the Hanansaya craft specialists and the lower-ranking Urinsaya
ayllus reflect greater Inka state involvement in the distribution of their fields.
Figure 9.3: Panorama of the Coporaque area, highlighting the Chillihuitira river, right and left sides, selected toponym sectors, and principle
LIP/LH settlements.
403
404
Theoretical distributions of fieldholdings unaffected by ayllu affiliation are needed
in order to test for the statistical significance of the observed right (west) and left (east)
ayllu landholding distributions. In this theoretical distribution, the landholdings of an ayllu
would be distributed in direct proportion to the amount of land in the toponym areas on
either side of the Chillihuitira river. This proportion can be derived by first summing the
area of all the toponym areas, and then dividing that sum by the area on either side of the
Chillihuitira dividing line. The total area of the toponyms is 432.7 ha, of which 294.3 ha
are located to the east of the Chillihuitira, and 138.4 are located to the west. In terms of
proportions, 68% of the toponym area is located to the east (i.e. on the left hand side
according to the ayllu names) and 32% to the west (on the right hand side according to
the ayllu names). Therefore, in this theoretical distribution unaffected by ayllu affiliation,
ayllu landholdings would be distributed according to these proportions. Expected values
for the null hypothesis can therefore be generated by multiplying the total mapped area of
the fieldholdings of a given ayllu by these east/west proportions (68% to the east and 32%
to the west). I then use a one-sample chi-square to test for the statistical significance of the
difference between the observed and expected values (Shennan 1997:108-109):
k
2 =
i =1
(Oi Ei ) 2
Ei
where k is the number of categories (in this case, two: east [left] and west [right]), Oi is
the observed cases in category i, and Ei is the expected cases in category i. The results I
report below use the 0.05 level of significance.
405
By comparing these land tenure patterns to the Late Intermediate Period and Late
Horizon settlement patterns I documented in my archaeological survey, I also present a
reconstruction of prehispanic ayllu residence patterns. As I discuss in detail below, this
reconstruction builds upon the right/left dualism I observed in the land tenure patterning. I
suggest that the majority of the ancestral population of the right side ayllus of
Hanansaya, together with ayllu Collana of Urinsaya, resided at the large LIP/Late Horizon
settlement of San Antonio/Chijra (CO-100), which, together with the majority of the field
area of these ayllus, lies to the west of the Chillihuitira.
By contrast, the ancestral
populations of the left side ayllus of Hanansaya appear to have resided primarily in the
LIP/Late Horizon settlements to the east: Kitaplaza (CO-164), Lama (CO-103), Llanka
(CO-127), Tunsa (CO-163), and Llactapampa (CO-150) on the east side of the Chillihuitira.
Finally, given their more dispersed land tenure patterns, the ancestral populations of the
other, lower-ranking ayllus of Urinsaya appear to have been dispersed throughout several
or all of the Late Horizon settlements in Coporaque.
Hanansaya: Right Side Ayllus

As is apparent in Figure 9.4, the lands of the right side ayllu Cupi were clearly
concentrated to the west of the Chillihuitira river. In this and the following maps of ayllu
landholdings, dot sizes are proportional to the amount of land area (in topos) declared in
each toponym by a given ayllu. Of the total area of ayllu Cupi household field declarations
that could be mapped (24.75 topos), 71% (17.5 topos) are located to the west of the
406
Chillihuitira river (Figure 9.4).170
The difference between the expected and observed
distributions is statistically significant171 (x2=17.06), and the null hypothesis is rejected.

Immediately preceding this ayllu in the 1615-1617 visita was an ambiguously named ayllu
that appears to have been another right side ayllu. This ayllu was called Aipi in the
header section introducing the ayllu (f. 565v), but Cupi in the ayllu summary (f. 585r)
probably an error due to miscommunication between the interpreter and the scribe. In any
case, the majority of the field area claimed by households of this ayllu cluster to the west
with those of the other ayllu Cupi (Figure 9.5). The pattern is not as marked as for ayllu
Cupi: 60% of the fields from this ayllu that could be mapped (18.25 out of 30.25 topos),
were located to the west of the Chillihuitira (Figure 9.5).172 Nonetheless, the difference
between the observed and the expected distribution is statistically significant (x2=11.17)
and the null hypothesis is rejected. A third ayllu, Calloca, also has a high percentage (69%,
or 12.75 topos) of its total mapped landholdings (18.5 topos) to the west of the Chillihuitira
(Figure 9.6).
The difference between this observed distribution and the expected
distribution is also statistically significant (x2=11.60).
So the great majority of the
fieldholdings declared by households from the right side ayllu Cupi, together with those
of the probable right side ayllu Aipi/Cupi, and those of ayllu Calloca, are concentrated to
the west of the Chillihuitira river.
Looking at the relationship between the land tenure patterns of these right side
ayllus and the archaeological settlement pattern in Figures 9.4-9.6, the fieldholdings of
170
171
172
These figures exclude one field (0.25 topo) claimed in a toponym that falls within modern Yanque.
Statistical significance (a=0.05) is met when x23.84
These figures exclude two fields (half a topo) mapped in toponyms within modern Yanque.
407
these ayllus are clearly concentrated nearest to the settlement of San Antonio/Chijra (CO100). Therefore, I argue that the majority of the members of these ayllus resided at this site
during the Late Horizon. A smaller proportion may have lived at the much smaller Late
Horizon settlement in the location of the reduccin of Coporaque (CO-161), also located to
the west of the Chillihuitira. As I discussed in Chapter 6, San Antonio/Chijra is the second
largest settlement in the survey, with elaborate elite domestic architecture and rustic Inka
public architecture associated with state-sponsored ritual (kallankas). I argued that the
close spatial association between elite Collagua domestic structures and Inka public
architecture at this and the other principle settlements signaled that Inka administration was
closely coordinated and mediated by local elites. In the previous chapter I presented the
formal political structure through which this mediation took place, and this land tenure
evidence suggests that the Inka established a secondary center at the settlement that was
home to one side of what appears to have been a dualistic indigenous social arrangement.
As I discuss below, Inka public architecture is also present at the settlements near the fields
of the other half of this dualism: the left side ayllus.
408
Figure 9.4: Landholding distribution, ayllu Cupi, Coporaque Hanansaya
409
Figure 9.5: Landholding distribution, ayllu Aipi/Cupi, Coporaque Hanansaya
410
Figure 9.6: Landholding distribution, ayllu Calloca, Coporaque Hanansaya
411
Hanansaya: Left Side Ayllus

The landholdings of the left side ayllu Checa Malco are even more markedly
concentrated on the opposite side of the Chillihuitira from those of the right side ayllus
(Figure 9.7). Virtually all of the fields declared by Checa Malco households are located to
the east of this dividing line. Specifically, 96% (10.75 out of the 11.25 topos) of the total
area of Checa Malco field declarations that could be mapped are located to the east of the
Chillihuitira. The difference between the expected and observed distributions is statistically
significant (x2=3.92).
In the 1615-1617 Hanansaya visita, ayllu Checa Malco is preceded and followed by
two other ayllus that share the honorific title Malco: Collana Malco (the first ayllu listed
in the village) and Icatunga Malco (listed after Checa Malco). Their landholdings are also
concentrated to the east of the Chillihuitira and Cantumayo (Figures 9.8 and 9.9), although
the distributions of these ayllus are not statistically significant (see Table 9.1). In the case
of Collana Malco, 22 out of its 24 fields, and 84% of its field area that could be mapped (8
out of 9.5 topos) were located to the east of the Chillihuitira, but the two fields to the west
were large, totaling 1.5 topos. These two cases in the rather small sample therefore pulled
down the chi square value (x2=1.15), and the null hypothesis cannot be rejected.
Nonetheless, the overall eastern predominance of Collana Malco landholdings is striking.
With only 24 tributaries, Collana Malco of Coporaque almost certainly represents a small,
local segment of the apical ayllu Collana that was cleaved apart by reduccin. While the
headman of this local segment in Coporaque is listed as the mayor (alcalde) of Coporaque,
412
the majority of this ayllu, along with its principal kuraka (Joan Halanoca), who was also the
paramount lord of the entire province, resided in Yanque. Likewise, the great majority of
the landholdings from Icatunga Malco that could be mapped80%, or 15.25 out of 19
toposare located to the east of the Chillihuitira, but this distribution does not differ in a
statistically significant manner from the expected distribution (x2=1.31). 173 Finally, a large
majority of the ayllu Yumasca83%, or 25.25 out of 30.25 toposwere also located to the
east of the Chillihuitira. This distribution is nearly statistically significant (x2=3.32). So
the land tenure pattern of the left side ayllu Checa Malco is heavily concentrated to the east
of the Chillihuitira, in contrast to the right side ayllus Cupi, Aipi/Cupi, and Calloca, whose
lands are found in greater proportions on the western side.
Although the difference
between the observed and expected fieldholding distributions of the other two Malco ayllus
and ayllu Yumasca is not statistically significant, their fields nonetheless are still
predominantly concentrated to the east of the Chillihuitira. Given the evidence from the
ayllu names and their land tenure patterns indicating the emic salience of this left/right
dualism, this general pattern amongst the Yumasca and Malco ayllus as a group may have
been socially meaningful. That is, they may have been considered collectively as the left
side group, in contrast to the right side group of the ayllus Cupi, Aipi/Cupi, and Calloca.
The fieldholdings of these left side ayllus surround the cluster of LIP/LH
settlements on the pampas to the east of the ChillihuitiraKitaplaza (CO-164), Lama (CO103), Llanka (CO-127), Tunsa (CO-163), and Llactapampa (CO-150)so it appears that
most of their populations were distributed amongst these settlements. Given the proximity
173
Icatunga Malco households also claimed 0.75 topos that could be mapped within modern Yanque.
413
of these sites to one another, it is unclear if the populations of these ayllus were each
divided between several settlements or if each ayllu tended to be concentrated in a single
settlement. In the case of the segment of the apical ayllu Collana Malco, all but one (a
small, 1/4 topo field in Canaque) of its mapped fields to the east of the Chillihuitira are
located in toponym sectors immediately surrounding the LIP/Late Horizon settlements of
Tunsa (CO-163) and Llactapampa (CO-150).
This suggests that the majority of its
population resided there during the Late Horizon. As I argued in Chapter 6, these sites,
although registered separately during the survey, probably functioned as distinct housing
sectors of a single large settlement, and, like San Antonio/Chijra, housed elite Collagua
domestic architecture and a prominently-located Inka kallanka in front of a large open
space between the two housing sectors. I interpreted the association between local elite
architecture and Inka imperial architecture associated with public ritual as indicative of how
Inka rule was coordinated rule through local elites at these secondary centers (and, by
extension, of how the authority of local elites became tied to the state). This land tenure
data resonates with this interpretation and provides more specific content regarding the
articulation of local and state institutions and actors. That is, while the paramount kuraka
and the majority of ayllu Collana almost certainly resided at the primary Inka
administrative center in the location of the reduccin of Yanque (YA-041; see Chapter 6), it
appears that Inka rule at Tunsa/Llactapampa was mediated by the elites of this local
segment of the apical ayllu of Hanansaya.
414
Figure 9.7: Landholding distribution, ayllu Checa Malco, Coporaque Hanansaya
415
Figure 9.8: Landholding distribution, ayllu Collana Malco, Coporaque Hanansaya
416
Figure 9.9: Landholding distribution, ayllu Icatunga Malco, Coporaque Hanansaya
417
Figure 9.10: Landholding distribution, ayllu Yumasca, Coporaque Hanansaya
418
To summarize, the names of the right and left ayllus of Hanansaya clearly refer
to the distribution of their landholdings on either side (the west and east sides, respectively)
of the Chillihuitira river. Thus, most of the fields of the Cupi, Aipi/Cupi, and Calloca
would have been irrigated by waters derived from the Chillihuitira (itself derivative of
myriad channelized quebradas in the heightssee Chapter 6), while those of Checca
Malco, Collana Malco, Icatunga Malco, and Yumasca would have been irrigated by waters
from the Cantumayo/Sahuara, as well as the Qachulle canal and the now-abandoned canal
system in the undulating hills of extreme eastern Coporaque (see Chapter 6). The aggregate
pattern of these left and right ayllus are displayed as a dot density map in Figure 9.11,
which shows the overall preponderance of right side ayllu fields to the west and left side
fields to the east. A comparison of the aggregated data using Fishers exact test indicates
that these aggregate distributions are significantly different from one another (P<0.001)
(Table 9.2). As proxy indices for terminal prehispanic residence patterns, these landholding
distributions suggest that the right side ayllus resided primarily at the settlement of San
Antonio, and the left side ayllus lived primarily at the five sites to the east of the
Chillihuitira. The majority of the local segment of the apical ayllu Collana Malco appears
to have been settled at the elite settlement of Tunsa/Llactapampa, which, together with San
Antonio/Chijra (CO-100) and Uyu Uyu (YA-050), occupies the second tier in the Late
Horizon settlement hierarchy below the primary administrative center in the location of the
reduccin of Yanque.
419
Table 9.1: Summary statistics, left/right predominance, Coporaque Hanansaya, 1615-1616
Ayllu
Checa Malco
Collana Malco
Icatunga Mallco
Yumasca
Cupi
Aipi (Cupi)
Calloca
Official Potters
Landholding
Predominance
Left
Left (?)
Left (?)
Left (?)
Right
Right
Right
None
Observed
"Left
Side"
(East)
10.75
8
15.25
25.25
7.25
12
5.75
1.5
Observed
"Right
Side"
(West)
0.5
1.5
3.75
5
17.5
18.25
12.75
2
Expected
Left Side
7.65
6.46
12.92
20.58
16.83
20.58
12.58
2.38
Expected
Right
Side
3.60
3.04
6.08
9.67
7.92
9.67
5.92
1.12
Chi
Square*
3.92
1.15
1.31
3.32
17.06
11.17
11.60
1.02
*Statistical significance (a=0.05) is met when x23.84
Table 9.2: Land tenure summary, left vs. right ayllus, Coporaque Hanansaya, 1615-1616*
Count
Ayllu Class
Total
Left
Right
Location of Fields
Left/East
Right/West
(topos)
(topos)
59
11
25
49
84
60
Total
P<0.001
*Left ayllus: Checa Malco, Collana Malco, Icatunga Malco, and Yumasca
Right ayllus: Cupi, Aipi/Cupi, and Calloca
70
74
144
420
Figure 9.11: Dot density map of left and right ayllus, Coporaque Hanansaya
421
Craft Specialists of Hanansaya: Official State Potters
The only Hanansaya ayllu that breaks from this left/right duality is an ayllu of
official potters (oficiales olleros). This is the largest of the three groups of potters in
Coporaque, and unlike the other two groups in Urinsaya174, it is not listed as a sub-section
of another ayllu. Given their important role in the provincial economy and their direct ties
to the state as officially-sanctioned craft specialists, their land tenure pattern provides an
important contrast with the other ayllus of Hanansaya. The distribution of their 13 mapped
fields (totaling 3.5 topos) does not differ in a statistically significant manner from the
expected distributions (x2=1.02; Figure 9.12). In this sense, their field distribution is
similar to that which predominates among the ayllus of Urinsaya. Nonetheless, half of their
mapped fields (1.75 out of 3.5 topos) were concentrated in a single toponym: the Ccayra
sector, a large complex of linear bench terraces located directly downslope of Chijra/San
Antonio (CO-100). This concentration suggests that a significant portion of this ayllu of
potters may have resided there during the Late Horizon. We did not observe any surficial
evidence for ceramic production (e.g. ceramic wasters) at the site during the survey, nor did
Denevan and colleagues during their test pit excavations in the Chijra sector (Malpass and
de la Vera Cruz Chvez 1990; Neira Avendao 1986; Treacy and Denevan 1986), however,
this remains an intriguing hypothesis that could be tested with more intensive site-specific
survey and larger-scale excavations.
174
As I discussed above, the other two small groups of potters in Coporaque Urinsaya pertained to the ayllus
Pahana Collana Pataca and Pahana Cayao Pataca; however, their mapped landholdings only totaled 3.25
topos (2.25 topos and 1 topo, respectively). I have included these handful of fields with the landholdings of
those ayllus in the Urinsaya discussion below.
422
Figure 9.12: Landholding distribution, ayllu of official state potters, Coporaque Hanansaya
423
Urinsaya Ayllus
Overall, the land tenure patterns of the ayllus of the lower moiety are more
dispersed than those of the upper moiety. Amongst Urinsaya ayllus, only the fieldholdings
of the high-ranking ayllu Collana are concentrated on one side of the Chillihuitira (Figure
9.13). Of the mapped fields of this ayllu, 62% (32.75 out of 52.5 topos), are located on the
western (or right side) side. The difference between the observed and the expected
distributions is statistically significant (x2=22.30). So most of ayllu Collana fields were
interspersed with those of the right side ayllus of Hanansaya. This also indicates that the
lands of the highest-ranked ayllu segments from each moiety were concentrated on opposite
sides of the Chillihuitira: those of Collana of Urinsaya to the west (right), and those of
Collana Malco of Hanansaya to the east (left).
Given the disproportionate concentration of ayllu Collana fields to the west of the
Chillihuitira and nearest to San Antonio/Chijra (CO-100), the terminal prehispanic
population of ayllu Collana appears also to have been concentrated at this settlement. As I
discussed in Chapter 6, some of the largest, most elaborate Collagua houses in the survey
are located at San Antonio/Chijra, and are particularly concentrated near the large,
prominently-situated kallanka structure at the site. This suggests a strong parallel with the
organization of Tunsa/Llactapampa discussed above, which also has elite domestic
architecture associated with a large kallanka. Just as the land tenure data indicate that the
local segment of the apical ayllu of the upper moiety resided primarily at
Tunsa/Llactapampa, the apical ayllu of the lower moiety appears to have resided primarily
424
at San Antonio/Chijra. The populations of both of these and other settlements would have
been politically subordinate to the kuraka of ayllu Collana, who probably resided in the
primary Inka administrative center at the site of the reduccin of Yanque (YA-041). The
archaeological and land tenure data therefore present congruous and complementary views
of how Inka imperial administration was mediated by local communities and their leaders.
The other ayllus of Coporaque Urinsaya are the three pataca-level ayllus that
constitute the middle-ranking macro-ayllu Payan (Pahana): Pahana Collana Pataca (high
ranked), Pahana Taypi Pataca (middle ranked), and Pahana Cayao Pataca (low ranked).
The distribution of landholdings of these three ayllus does not significantly differ from the
expected distributions (Table 9.3). That is, the landholdings of these three ayllus are
distributed roughly in proportion to the amount of land on either side of the Chillihuitira.
As displayed in Figures 9.14 9.16, their fields were dispersed widely over the entire
Coporaque area, and were not concentrated in any particular side or sector. Given this
contrasting pattern with the Hanansaya ayllus, which, as I discussed above, appear to have
been the indigenous ayllus based on their names and land tenure patterns, I interpret this
contrasting dispersed pattern as indicative of greater Inka state involvement in distributing
fields among the lower-ranking ayllus of Urinsaya. Also, these dispersed land tenure
patterns suggest a correspondingly dispersed settlement pattern. The terminal prehispanic
members of these ayllus probably resided at several, if not all, of the settlements in the
Coporaque area.
425
Table 9.3: Summary statistics, left/right predominance, Coporaque Urinsaya, 1604
Ayllu
Collana
Pahana
Collana Pataca
Pahana Taypi
Pataca
Pahana Cayao
Pataca
Official Potters
of [Pahana]
Cayao Pataca
Indian Potters
[of Pahana
Collana
Pataca]
Observed
"Left
Side"
(East)
19.75
Observed
"Right
Side"
(West)
32.75
Expected
"Left
Side"
35.71
Expected
"Right
Side"
16.79
Chi
Square
Value*
22.30
None
42.25
13.5
37.92
17.83
1.55
None
16
10.75
18.19
8.56
0.83
None
18.5
9.25
18.87
8.88
0.02
None
0.5
0.5
Insufficient
Data
Insufficient
Data
Insufficient
Data
None
1.25
Insufficient
Data
Insufficient
Data
Insufficient
Data
Landholding
Predominance
Right
*Statistical significance (a=0.05) is met when x23.84
Sample size too small for statistical testing.
426
Figure 9.13: Landholding distribution, ayllu Collana, Coporaque Urinsaya
427
Figure 9.14: Landholding distribution, ayllu Pahana Collana Pataca, Coporaque Urinsaya
428
Figure 9.15: Landholding distribution, ayllu Pahana Taypi Pataca, Coporaque Urinsaya
429
Figure 9.16: Landholding distribution, ayllu Pahana Cayao Pataca, Coporaque Urinsaya
430
Discussion and Conclusion

This chapter illustrated how specific indigenous community structures continued to
organize and synchronize the land-use practices of local Coporaque households after
conquest and resettlement. I have argued that the early colonial land tenure patterns of the
ayllus of Coporaque also reflect their terminal prehispanic patterns of residence, and, by
extension, provide insights into how Inka imperial administration articulated with local
structures of community organization.
According to the visita declarations, Coporaque farmers continued to cultivate
exclusively native crops three generations after conquest. Some villagers most likely also
experimented with European cultigens, but they were not declared as the predominant crop
grown in any fields in Coporaque during the first two decades of the 17th century.
Coporaque farmers used the most productive, warmest production zones for cultivating
maize, a frost-sensitive crop, while they planted the hardier quinoa in more marginal and
frost-prone areas. This pattern shows the rationality of the local production regime not only
in terms of minimizing risk, but also in the culturally-relative sense of preferentially
cultivating maizea less-nutritious but more culturally-esteemed crop than quinoain
warm, fertile production zones.
Households also maintained access to multiple production zones and minimized
their exposure to localized risks such as pests and frosts by distributing their landholdings
among several small, dispersed fields. However, as reflected in the contrasting patterns of
ayllu landholding distributions, these microverticality (Brush 1977) and risk-
431
minimization strategies were mediated by ayllu affiliation. My land tenure reconstructions
demonstrate that ayllus did not control discrete field areas or territories, but instead ayllu
landholdings consisted of discontinuous and overlapping, but patterned, constellations of
fields. The landholdings of all but one of the ayllus of Hanansaya, as well as the fields of
the apical ayllu of the Urinsaya, were concentrated on either the left or right side of a
hydrological dividing line running through the middle of Coporaque. By contrast, the
landholdings of the other ayllus of Urinsaya, along with the ayllu of official state potters in
Hanansaya, were more broadly distributed across the landscape of Coporaque on both sides
of this dividing line. These distinct Hanansaya/Urinsaya ayllu patterns are also congruent
with a parallel contrast in the naming patterns of the ayllus of each moiety. Hanansaya
ayllu names generally do not follow the tripartite and decimpartite naming conventions of
the ideal Inka sociopolitical organization of the province, and instead appear to be
organized according to a left/right dualism, while those of Urinsaya conform more closely
with the ideal Inka administrative structure (see Chapter 8). Thus, I have argued that the
observed left/right landholding distributions of the Hanansaya ayllus, together with ayllu
Collana of Urinsaya, represent vestiges of indigenous, perhaps pre-Inkaic, land tenure
patterns, while the dispersed patterns of Urinsaya ayllus, together with the ayllu of official
state potters of Hanansaya, reflect greater Inka state intervention in the distribution of their
fields.
Given this evidence for the prehispanic origin of the observed ayllu land tenure
patterns, I argued that they also reflect pre-reduccin site catchment patterns. Thus, by
comparing the ayllu landholding distributions with the Late Horizon settlement pattern, I
432
reconstructed pre-reduccin ayllu residence patterns.
These ayllu residence patterns,
summarized in Table 9.4, provide a more culturally-informed and nuanced view of the
political organization of the local settlement system, both in terms of the supra-settlement
organization of late prehispanic Collagua communities, and in terms of how those
communities articulated with Inka institutions.
Table 9.4: Summary of pre-reduccin ayllu residence patterns
Left Side Ayllus
Right Side Ayllus
Dispersed Ayllus
Majority resided at CO103, CO-127, CO150/164, and CO-163

Hanan
Urin
Majority resided at
CO-100
Population dispersed throughout most/all

Late Horizon settlements in Coporaque
Hanan
Urin
Hanan
Checa Malco
Collana Malco*
Icatunga Malco*
Yumasca*
Cupi
Aipi/Cupi
Calloca
Collana
State Potters
Urin
Pahana Taypi Pataca
Pahana Cayao Pataca
*Difference from expected distribution not statistically significant
Probably settled primarily at Tunsa/Llactapampa (CO-163/150)
As macro-scale imagined communities, moieties were not residentially

coterminous with settlements or groups of settlements, but were distributed over all of the
settlements of Coporaque.
Moiety affiliations cut across settlement and hydrological
boundaries, and it appears that members from both moieties resided at all settlements
(Table 9.4). Clearly , the hydrological interests of moieties as collectivities were widely
dispersed. My analysis of hydrological relationships between feeder canals in Chapter 6
suggested that water distribution was coordinated at a supra-settlement, watershed-scale
during the LIP and Late Horizon. The land tenure patterns I reconstructed in this chapter
indicate that the households of many ayllus cultivated fields that were irrigated by several
feeder canals, suggesting that they would have had an interest in maintaining and
433
coordinating the apportionment of water between them rather than favoring one canal over
another. Coporaque households also declared fields located on the opposite side of the
Colca river near Yanque, as well as fields irrigated by the Misme canal, which draws water
away from the settlements in the Coporaque area to irrigate the terraces surrounding the
LIP/LH settlements of Uyu Uyu (YA-050) and Llactarana (YA-054) to the west (see
Chapter 6). This is especially true of the ayllus whose names and land tenure patterns
appear to be products of direct Inka state intervention: the ayllus of the middle-ranking
Payan macro-ayllu of Urinsaya. Their widely scattered land tenure patterns may reflect an
Inka state policy of dispersing hydrological interests in order to foment this kind of
watershed-scale intensive water management regime, a policy enacted in neighboring
valleys and regions (Trawick 2001; Zimmerer 2000). By contrast, the right/left land
tenure and residence patterns of the indigenous Hanansaya ayllus may reflect older, more
localized hydrological interests that may have factored in the competition and conflict
evident during the LIP.
The Inkas also appear to have attempted to subsume local political interests within a
larger-scale hierarchy of settlements and ayllus. In Chapter 6, I interpreted the association
between elite Collagua houses and prominently-situated kallankas at large LIP/Late
Horizon settlements as indicative of the important intermediary role that local elites played
in Inka imperial administration of the province. My reconstruction of pre-reduccin ayllu
residence patterns reinforces this interpretation because they indicate that the highestranking ayllus of each moiety resided at the two largest LIP/Late Horizon settlements in
Coporaque. Specifically, it appears that a local segment of the highest-ranking ayllu of
434
Hanansaya (Collana Malco) resided primarily at Tunsa/Llactapampa (CO-150/163), while
a local segment of the highest ranking ayllu of Urinsaya (Collana) resided primarily at San
Antonio/Chijra (CO-100). Therefore, Inka administration appears to have been coordinated
through the leaders of these ayllu segments at these secondary centers. However, these
were only local segments of larger ayllus. Other segments of these ayllus also resided in
the reduccin of Yanque (YA-041)previously the primary Inka administrative center for
the central valley (see Chapter 6). The kurakas of the ayllu Collana segments in Yanque
were the paramounts of their respective moieties, and of them, the paramount of the upper
moiety was also lord of the entire province of Collaguas (see previous chapter). This is
consistent with what would be expected, since the Inka-era ancestor of the paramount
lordthe kuraka of Collana of Hanansayawould almost certainly have resided in the
primary administrative center. The Inka-era paramount lord of Urinsaya may also have
resided there, but it seems equally, if not more, likely that he resided at Uyu Uyu (YA-050),
the other secondary administrative center located in present-day Yanque Urinsaya. In any
case, it appears that the segments of the apical ayllus at San Antonio/Chijra and
Tunsa/Llactapampa were subordinate to their counterparts in Yanque (and possibly Uyu
Uyu). This reconstruction therefore reveals how supra-settlement, ranked communities
were distributed over the local settlement system and shows in more culturally-specific
terms how the settlement hierarchy corresponded to the conceptual political hierarchy of the
province.
435
Chapter 10
Summary and Conclusions
Review of the Research Problem

The research agenda I presented in Chapter 1 called for analysis of how community
organization mediates land-use patterning over time in the prehispanic and colonial Andes.
I argued that a community-scale perspective is critical for understanding the articulation of
domestic and political economy, and, by extension, the interface between humans and their
environment. Working from a political ecology framework, I presented a model that posits
a recursive relationship between the structures of the built landscapeitself a synergistic
socio-ecological productand the agency of distinct land-use practices over time.
This project tracked the origins and development of the late prehispanic Collagua
ethnic polity of southern highland Peru, and explored how their later incorporation into the
Inka and Spanish empires related to changes in community organization, land-use
patterning, and the physical configuration of the built landscape of the Colca valley, one of
the most intensively terraced valleys in the Andes. In accord with the compound nature of
this query, the project methodology was multi-scalar, both temporally and spatially. It
combined analysis of long-term diachronic changes in archaeological settlement patterning
and landscape modification with synchronic, regional- and local-scale analysis of land-use
practices during early colonial times using detailed Spanish visitas of an area that overlaps
with the project archaeological survey coverage. In this chapter, I summarize my findings
436
and arguments concerning these long- and short-term processes of interaction between
communities and the household/habitat of the landscapes they inhabit and transform.
Early Agriculturalist Settlement in the Central Colca Valley

The earliest period of sedentary agricultural settlement in the valley was almost
completely obscure prior to this project. No settlements predating the Middle Horizon had
been reported, and ceramic and chronometric data from reconnaissance and excavations
presented widely varying dates for the earliest, unirrigated field systems in the valley,
ranging from as late as the 6th century AD based on ceramics (Treacy 1989), to as early as
the middle of the 3rd millennium BC based on radiocarbon dates (Brooks 1998:378-379).
Previous researchers have suggested that these early, unirrigated field systems represented a
technological adaptation to wetter than average climatological conditions (Brooks 1998;
Treacy 1989). Treacy suggested that the accretionary, low-intensity pattern evident in the
construction and patterning of these fields could have been constructed at the household
scale and did not require coordinated planning (Treacy 1989).
Building on these
observations, I hypothesized that political and economic units at the time of their
construction and use would have been small and minimally hierarchical, with low suprahousehold demands for labor and surplus production.
My conclusions generally concur with those of previous researchers regarding the
form of early agricultural production in the valley; however, my documentation of an
associated Formative Period occupational component places the timing of early local
agriculturalist settlement within a time frame more in line with the general culture-historical
437
trajectory of surrounding locales, and provides a preliminary view of local Formative
Period social organization. The shift from nomadic hunting and foraging to sedentary,
agriculturalist lifeways in the south central Andes was not episodic, but rather occurred
over many hundreds or perhaps over a thousand years during the Terminal Archaic and
Early Formative Period (Aldenderfer 1989, 1998; Stanish 2003:99-109). My survey data in
the Colca hint at the shift from foraging to farming, as open air and rock shelter Archaic
Period sites in the puna continued to be occupied during the Terminal Archaic and Early
Formative. In the kichwa and suni zones within the valleys agricultural core, my survey
revealed a concentration of settlements and agro-mortuary wall complexes on and around
the valley bottom, suggesting that early agricultural production was more concentrated in
these areas than previously suggested. I have suggested that early agricultural production
focused primarily on the broad expanses of the valley bottom, while the valley-side sloping
fields discussed by previous researchers were probably subsequent additions and always
rather marginal or supplementary in terms of their contribution to local subsistence
economy.
In terms of social organization, the Formative Period settlement pattern documented
by the survey was composed of undifferentiated hamlets and villages, and there was no
evidence for a site size hierarchy, suggesting small-scale, egalitarian political organization.
I did not identify any decorated local ceramics that indicate the development of a local
corporate style; rather, local Formative Chiquero ceramics are undecorated utilitarian
wares. The appearance of large settlements such as Chiquero (YA-032) hint at increased
nucleation of settlement, although the coarse nature of the chronology precludes
438
differentiation of what was a very long, and surely socially dynamic time period. At
present there are no indices for trends toward inequality or the growth of supra-local
political units, as in the regional-scale Middle and Late Formative Period cultures of the
Titicaca Basin, groups with whom local peoples clearly interacted, as evidenced by the
distribution of Chivay source obsidian throughout the lacustrine zone and beyond (Brooks
1998; Burger, Asaro, Salas, et al. 1998; Burger, et al. 2000). Future researchers should be
able to refine the Formative ceramic chronology, and with more survey coverage and
specialized investigations, a more nuanced view of this long, important period of initial
sedentary agriculturalist settlement should emerge.
During the subsequent Middle Horizon, there is no evidence for direct imperial
occupation by either Wari or Tiwanaku, although the valley was clearly within the spheres
of influence of both empires. Settlement shifted considerably from the Formative, as 57%
of the 37 Middle Horizon sites represent new occupations. However, there are no indices
of centralization, planned settlement, or architectural evidence in the settlement pattern that
would signal a direct imperial presence. Local ceramics are clearly derivative of regional
Wari variants, and the stamp of Wari imperialism strengthens in the lower reaches of the
Colca/Majes/Caman drainage to the west. On the other hand, the prevalence of Chivay
source obsidian throughout the Tiwanaku heartland to the east evinces continued strong
economic ties to the Titicaca Basin.
Thus, the central Colca valley appears to have
constituted a buffer zone or porous boundary between the political and economic spheres
of, respectively, Wari and Tiwanaku.
Overlying LIP and Late Horizon occupations
complicate further specification of local Middle Horizon social organization, although the
439
expanding diversity of Middle Horizon mortuary featuresranging from simple cyst
tombs, to the first use of chullpas under overhanging rocks and cliffsprovide the first
tentative indices for differentiation of social rank in the local prehispanic sequence.
Community Organization and Production during the LIP and Late Horizon
Following the decline of the Middle Horizon states, all archaeological indices from
the Late Intermediate Period point to a major expansion of settlement and irrigated
agricultural production associated with the development of the Collagua polity. Collagua
ethnogenesis during the LIP is also marked by the appearance of distinctive local domestic
and mortuary architecture, massive complexes of irrigated bench terraces surrounding
settlements, and ceramics that show stylistic and formal continuity from the preceding
Wari-influenced wares.
While these markers suggest an overarching unity of ethnic
identity, I have argued that autonomous Collagua political organization during the LIP was
heterarchical in naturethat is, composed of internally-differentiated communities whose
political rankings were fluid relative to one another. On the one hand, there is strong
evidence for increasing disparities of rank during the LIP, such as intra- and inter-site
disparities in the size and elaboration of domestic architecture, major differences in tomb
elaboration and mortuary treatment, and the development of a site size hierarchy composed
of hamlets, villages, and larger towns. On the other hand, political organization appears to
have been highly decentralized; the LIP settlement rank size distribution is markedly
convex and no settlement dominates the settlement pattern in terms of centrality or
440
elaboration of architecture.
Rather, an elite class resided at several large settlements
dispersed throughout the central Colca valley.

Within this heterarchical political structure, intra-polity community relations appear
to have oscillated between coordination and competition. As I illustrated in Chapter 6,
hydrological relationships between the long feeder canals that carry meltwater from the
surrounding peaks, and the distribution of settlements in relation to those canals, illustrate
that water apportionment was coordinated at a supra-settlement scale during the LIP and
Late Horizon.
This reveals an important aspect of supra-settlement community
organization, since irrigation water is the sine qua non of agricultural production in the
Colca valley. However, frequent violent conflict between Collagua communities, and
perhaps with external polities, during the LIP is also apparent in defensive site locations
and hilltop fortifications (pukaras).
A central thesis of this study posits that the Inkas effectively hierarchized the
formerly heterarchical autonomous political organization of the Collaguas, establishing a
vertically-integrated and horizontally-compartmentalized administrative structure in which
latent distinctions of rank between local ayllus were amplified, ossified, and reordered in
the image of Inka ideals. Settlement in the survey area reached its maximum extent and
became more centralized during the Late Horizon, forming a settlement hierarchy
distributed in a log-normal rank size distribution. I have argued that Inka rule of the central
valley was administered through a large (18 ha) primary political center at the site of
Yanque (YA-041), which alone occupied the top tier of the settlement hierarchy, and was
the only site in the survey with architectural remains made of Inka-style cutstone masonry.
441
The sites that previously occupied the top tier of the settlement hierarchy during the LIP
Uyu Uyu (YA-050), San Antonio/Chijra (CO-100), and Tunsa/Llactapampa (CO163/150)became secondary political centers, where Inka kallanka structures are
prominently situated in close association with elite Collagua domestic structures. Thus, I
suggest that Inka rule was coordinated through local elites at these sites.
My interpretation of the Late Horizon settlement pattern data derives not only from
these corollaries, but also from socio-structural homologies that I reconstructed through
analysis of early colonial visitas from the same area of the valley. As I discussed in
Chapter 8, the ideal political organization of the Inkaic Collagua province was composed of
a nested hierarchy of ayllus that were ranked according to dualistic, ternary, and decimal
categories. Dualism was expressed by moiety divisions: Hanansaya (higher ranking) and
Urinsaya (lower ranking). Within each moiety, macro-ayllus were ranked according to
high, middle, and low status categories (Collana, Payan, and Cayao), each of which was
internally divided into three minimal ayllus of an ideal size of 100 households that were
also ranked according to these same tripartite status categories. The names of the minimal
ayllus of the lower moiety exactly parallel the nomenclature of the ceque system of Cuzco,
suggesting more penetrating imperial reordering of community structures than amongst the
ayllus of the upper moiety, which retained local names.
In contrast to the ternary-
dominated ranking system of the Inkaic administrative structure, the onomastic patterning
amongst the upper moiety ayllus suggest that dualistic principles may have structured
autochthonous political organization and criteria of rank.
442
The covariance of these naming patterns with their contrasting land tenure patterns
further suggest differential articulation of state and local institutions within the province.
Reconstructed ayllu land tenure patterns from the visitas provided a means for retrodicting
their terminal prehispanic residence patterns, and reveal how the imagined, hybrid
local/imperial constructs of community discussed above were mapped onto the local
settlement system. This reconstruction indicates that local segments of the highest ranking
ayllus of each moiety resided at the secondary centers of San Antonio/Chijra (CO-100) and
Tunsa/Llactapampa (CO-163/150). Specifically, a local segment of the ayllu Collana of
Hanansaya (the upper moiety) resided primarily at Tunsa/Llactapampa, while their
Urinsaya counterparts resided primarily at San Antonio/Chijra. Both of these sites were top
tier settlements during the LIP, and subsequently became secondary centers under Inka rule
with prominently-situated kallankas in association with elite Collagua domestic structures.
They are located on opposite sides of a hydrological divide that separated left and right
side ayllus, signaling that underlying dualistic organization was expressed both in terms of
land tenure and settlement patterning. However, just as each of these secondary centers
appear subordinate to the primary administrative center of Yanque by archaeological
indices, the visitas reveal that larger segments of these same apical ayllus, along with their
principal kurakas, resided in Yanque. These kurakas, as leaders of the highest-ranking
ayllus, were also the paramount lords of their respective ayllus, and in the case of the lord
of Collana Hanansaya, of the entire province. Thus, I have argued that the Inka-era
provincial paramount kuraka would also have resided in the administrative center at
443
Yanque, while the paramount of Urinsaya may also have resided there, or at the large
settlement of Uyu Uyu (YA-050), located in the territory of modern-day Yanque Urinsaya.
Collagua Communities and Complementarity during Inkaic

and Early Colonial Times
My analysis of visita landholding and livestock declarations also provided a means
for reconstructing the varied land-use practices of Collagua households and communities
during early colonial times. Previous researchers have presented the regional distribution
of Collagua ethnic colonists in diverse ecological zones as a classic example of a vertical
archipelago (Galdos Rodrguez 1984; Mlaga Medina 1977; Pease G. Y. 1977). However,
as recent re-studies have illustrated in the Lupaqa test case (Van Buren 1993, 1996), outlier
Collagua production enclaves in the low-lying valleys to the south appear not to have
functioned to provision the highland core population as a whole, but rather were attached to
the paramount kurakas, who mobilized their produce to underwrite political activities and a
sizeable prestige goods economy.
Large-scale vertical archipelagoes, in this sense,
primarily provisioned political economy, not subsistence economy, as originally

hypothesized.
While my findings generally support recent critiques that question the
hypothesized adaptive function of vertical archipelagoes (Van Buren 1993, 1996), my

analysis, working from a political ecology framework, suggests that the political and the
ecological need not be counterposed as antithetical. As noted by Goldstein (2000:186),
[i]n regions where resource variability is defined by sharp gradients in altitude, it indeed
appears obvious that access to a variety of altitude-defined ecological zones would be a
444
fundamental goal of human groups. The key point is that these human/environment
interactions are always situated in specific, historically-contingent political contexts and
structures.
The approach to complementarity forwarded in this study highlights this
political mediation as a process rather than as a social formation or structure, and provides a
means for exploring the diverse ways in which households and communities within a single
ethnic group gain access to variegated resources in the vertical Andean environment.
At the local scale, my reconstruction of land tenure patterns from the visitas of
1591, 1604, and 1615-1617 in Coporaque shows how households maintained access to
multiple production zones and reduced their exposure to localized risks by dispersing their
fields widely. However, the contrasting patterns of ayllu landholding distributions illustrate
how this risk-minimizing microverticality (Brush 1977) strategy was mediated by ayllu
affiliation. While ayllus did not collectively own large, contiguous blocks of land, the
fields of ayllu members formed patterned constellations of landholdings. For example, the
landholdings of all but one of the ayllus of Hanansaya, which, based on onomastic
evidence, appear to represent indigenous Collagua ayllus, clustered to either side of the
Chillihuitira river in Coporaque. Conversely, all but one of the ayllus of Urinsaya, whose
names conform exactly to the ideal hierarchical structure of Inka governance, were widely
dispersed throughout Coporaque and Yanque. Thus, household landholding distributions
were not reflective only of adaptive, risk-minimization strategies, but were also patterned
by community (ayllu) affiliation, both in the case of the indigenous communities of
Hanansaya, and the more state-influenced ayllus of Urinsaya.
445
My regional-scale analysis of visita landholding and livestock declarations in
Chapter 8 illustrated how supra-local networks of ayllu affiliation and authority
synchronized
diverse
complementarity
relationships
between
agriculturalists
and
pastoralists under Inka and early colonial rule. At the level of domestic economy,
commoner farming and herding households participated in direct exchange of their
respective produce with their ayllu cohorts. By contrast, elite households declared fields
and livestock in far-flung locales, thereby directly controlling the means of production in
several production zones. Wealthy pastoralist households from the upper reaches of the
valley declared maize fields in the central valley, as well as in maize production enclaves in
Huanca, Lluta, and the valley of Arequipa to the south. Wealthy agriculturalist households
based in the central valley declared personal herds of livestock in the upper valley.
In terms of the regional political economy, the Collaguas provided 35% of the tax
revenue collected in Arequipa in the 1580sthe largest single source under its jurisdiction.
Clearly, the Collaguas were an important regional source of tribute for the Inkas as well.
As extractive agents, colonial Collagua kurakas of Yanque and Coporaque collected tribute
from their far-flung ayllu segments, as their counterparts would have done in Inkaic times,
but they also mobilized staple goods to underwrite a large local Inka imperial prestige
goods industry. Ayllus of official state potters, silversmiths and cumbicamayocs were the
subjects of the paramount kurakas. While these groups of artisans were generally selfsufficient in terms of subsistence, the kurakas supplied them with materia prima for their
crafts by brokering staple goods through tribute and exchange mechanisms. This diversity
of intra-polity mechanisms of exchange and mobilization, ranging from direct trade
446
between households, to mobilization of staple and wealth finance, therefore illustrates how
vertical complementarity amongst the Collaguas was structured not only by ecological
imperatives, but also by political organization and community affiliation.
447
Appendix A:
Ceramic Sequence
Introduction
Delineation of ceramic chronologies remains a critical research problem in many
Pacific slope drainages of southwestern Peru, including the Colca valley. During the
survey, we recovered a total of 3806 ceramic artifacts, ranging from Formative to
Republican times (Table A.1, Figure A.1). Below, I present a preliminary local ceramic
sequence of seven stylistic groups that are divided between five time periods:
1. The Chiquero stylistic group of the Formative Period (1500 BC-AD 500),
2. An unnamed, local Middle Horizon stylistic group (AD 500-1000)
3. Collagua I and Collagua II stylistic groups of the Late Intermediate Period (AD
1000-1450),
4. Collagua III and Collagua Inka stylistic groups of the Late Horizon, (AD 14501532), and
5. An unnamed Colonial Period stylistic group (AD 1532-1821).
The temporal placement of the chronology builds on the sequence developed by
previous research in the Colca valley that began with Neira (Neira Avendao 1961), and
subsequently refined by Malpass and de la Vera Cruz Chvez (de la Vera Cruz Chvez
1988, 1989; Malpass and de la Vera Cruz Chvez 1986, 1990), and Brooks (Brooks 1998).
In the case of previously unidentified Formative Period Chiquero ceramics, I have
conservatively categorized them as broadly diagnostic of the Formative Period, based on
the generally-accepted bracketing of this long time period in the South Central Andes (1500
BC-AD500). Where appropriate, I also make comparisons with other sequences from
neighboring locales such as the Chuquibamba valley (Cardona Rosas 1993; Sciscento
1989), the Titicaca Basin (Julien 1983; Stanish, et al. 1994; Steadman 1995; Tschopik
448
1946), and southern Cuzco (Bauer 1999). I developed the sequence in consultation with
two experts on the ceramic styles of the Department of Arequipa: Willy Ypez Alvarez and
Erika Simborth Lozada, who both worked with me throughout the field component of the
project as well. They also drew the original ceramic diagrams, while I recorded formal,
metric, compositional, and decorative attributes of the ceramics in a data matrix. Bruce
Owen generously provided me with copies of a tabular format coding scheme for recording
metric and qualitative attributes, which I used as a model for building my data matrix and
coding system. I also consulted with Augusto Cardona Rosas on the Formative Period
Chiquero ceramics, and he provided me access to Formative Period collections from his
survey in the Arequipa valley for comparison.
After discussing the methodological approach, recording conventions, and analytical
procedures for devising the sequence, I present the compositional, morphological, and
decorative attributes of each of the stylistic groups. The discussion does not include
analysis of all ceramics collected, but instead focuses on the diagnostic material relevant to
dating occupations of the sites registered during the survey.
Methodology and Laboratory Procedures

The sequence is based on a type-variety methodology, a hierarchical classificatory
approach that proceeds by grouping ceramics into successively more inclusive categories of
similarity (Gifford 1960; Phillips 1958; Smith, et al. 1960; Wheat, et al. 1958). At the most
inclusive level, I defined wares by shared features of manufacturing technology, including
past composition, and/or surface treatment. Next, styles were defined by shared attributes
449
related to vessel form, slip consistency, coverage, and color, and surface treatment. Types
consisted of groups of ceramics within a style that shared similar decorative techniques.
Finally, varieties were the minimal stylistic units of analysis, defined by vessel form details
(e.g. rim profile) and specific decorative colors and motifs within types.
In the laboratory, analysis of the ceramics proceeded in the following fashion: 1)
after washing and labeling, all sherds were separated from their lots and laid out together,
2) ceramics of the same ware were grouped together, 3) ceramics of the same style were
subdivided within each ware group, 4) styles were subdivided by types and varieties,
forming the smallest classificatory units, 5) samples of each variety were drawn and
photographed, 6) quantitative and coded qualitative data on vessel form, dimensions,
composition, surface treatment, firing, slip and paint colors, and decorations of each sherd
was recorded in a data matrix.
Terminology and Coding

I recorded colors using a Munsell Soil Color Chart, and I used a 10X hand lens and
a sand grain size reference card for describing inclusion/temper sizes and shapes. My
inclusion/temper compositional identifications are preliminary, and future petrographic
analysis will provide a more definitive, detailed view of changes in ceramic pastes over
time. At this stage of analysis, differentiation of natural inclusions in the paste matrix from
deliberately added temper is impressionistic. The inclusion/temper size classification terms
I use in my paste descriptions follow the standard Udden-Wentworth grain-size scale:
450
Very Fine Sand:
Fine Sand:
Medium Sand:
Coarse Sand:
Very Coarse Sand:
0.0625-0.125 mm
0.125-0.25 mm
0.25-0.5 mm
0.5-1.0 mm
1.0-2.0 mm
I coded for vessel forms as well, and describe each of these in the discussion below
where relevant. I also recorded coded observations on form details, such as the presence of
lugs, handles (and their cross-section shapes), and rim profile for each sherd. Surface
treatment also proved to be an important index of temporal variation. Building on a scoring
system developed by Stanish for ceramics in the upper Moquegua valley (Stanish 1991:8), I
categorized surface treatment (for both interior and exterior) according to a progressive
scale, as follows:
1
2
3
4
5
6
7
8
9
Polished: highly burnished to a very smooth surface, no burnishing marks evident

Polished, with slight "corrugated" appearance, especially at rim.
Finely burnished: burnished to a polish-like surface, burnishing marks evident.
Burnished: burnished over the entire surface, some smooth pitting or a slightly rough
surface possible
Slightly burnished: occasional burnishing marks present, surface rough and/or
undulating. Possible rough pitting.
Wiped: wiping striations present, possible pitting on a "sandpaper" surface.
Poorly wiped: wiping striations present, surface pitted or rough.
Fiber wiped: deep striations from vegetal fiber instrument (brushed)
Rough: uneven, pitted surface, no burnishing or wiping marks evident.
I also coded observations on firing quality and completeness using the following
normative descriptions as a guide:

IP
IT
IA
IG
CN
CO
VI
Incomplete firing; poor firing throughout, low temp and brittle, brownish
Incomplete, very dark throughout, except outermost layer and slip (where present)
Incomplete firing with dark reduced core
Incomplete firing with light grey, slightly reduced core with blurred edges
Complete but not oxidized red. Buff to brown throughout
Complete and oxidized throughout
Vitrified, clinky (colonial only)
451
Metric observations included rim diameter (to the nearest cm), rim thickness (to
nearest 0.5 mm), thickness of base at thickest point, if available (to nearest 0.5 mm),
thickness of body wall at thickest point (not at carination, join, etc.; to nearest 0.5 mm),
handle width at middle (to nearest 0.5 mm), handle thickness at middle (to nearest 0.5 mm),
handle length (exterior, from center of bottom join curve to center of top join curve; to
nearest 0.5 mm), handle height (maximum distance from body to inside of handle,
perpendicular to handle length; to nearest 0.5 mm), width of slipped area inside rim, if not
fully slipped inside (to nearest 0.5 mm), and width of slipped area outside rim, if not fully
slipped outside (to nearest 0.5 mm).
Summary Counts and Temporal Placement

Table A.1 summarizes the sample size of each of the categories of ceramics from
the survey. Our field collections erred on the side of inclusivity, with the thought in mind
that some sherds that could not be identified in the field may be of chronological value in
the laboratory or in future studies. This, along with a large number of small sherds
(generally < 3 cm2) accounts for the relatively large proportion of unclassified ceramics.
Also, I categorized some small sherds based on superficial appearance, and did not make
detailed paste, surface treatment, or other metric observations on these sherds. For this
reason, n values in some of the histograms and tables (e.g. rim diameter, surface treatment)
below are smaller than these total N values.
As I discuss below, only the Middle Horizon and Late Horizon parts of the sequence
are bolstered by a limited number of carbon dates. The schematic of the sequence in Figure
452
A.1 is meant to provide a preliminary visual reference, and I consider it a provisional
representation, pending revision and refinement through future excavation and radiometric
dates. For the purposes of dating sites, I have combined Collagua I and II as diagnostics for
the LIP, and Collagua III and Collagua Inka as diagnostics for the Late Horizon.
Table A.1: Summary counts and percentages: all ceramics
Chiquero
Middle Horizon
Collagua I
Collagua II
Collagua III
Collagua Inka
Inka
Colonial
Modern
Imports
Utilitarian (prehispanic)
Unclassified/Sm. Frags.
Total
Count
152
217
152
281
218
1120
130
456
58
258
119
645
3806
%
4.0%
5.7%
4.0%
7.4%
5.7%
29.4%
3.4%
12.0%
1.5%
6.8%
3.1%
16.9%
100.0%
453
Figure A.1: Preliminary Colca valley ceramic sequence
454
Figure A.2: Key to ceramic illustrations. The right side of illustration corresponds to the vessel
interior, and the left side corresponds to the exterior. Lot numbers are included to the adjacent right of
each ceramic illustration. Site numbers and proveniences of these lots are listed in Appendix 2
The Formative Period: Chiquero Ceramics

We recovered 152 diagnostic Chiquero sherds, making up 4% of the total ceramic
collections. The majority (64%, 97 sherds) were from a single site, YA-032, Chiquero,
which I use as the type site for this previously-undocumented early ceramic style.
Ware
Chiquero ceramics are made of a distinctive, porous paste with abundant (25-40%)
medium to coarse black sand inclusions, some very coarse, with sparse fine to medium size
white inclusions (probably feldspar and/or quartz) and sparse, fine size reflective, goldcolored inclusions (probably mica or pyrite). A small quantity (n=4) of sherds was also
recovered that appear to be made of a slightly finer paste with fine to medium sand size
temper. Firing is variable; most sherds are buff-to-brown color throughout, and are not
oxidized red, although a few well-oxidized examples were recovered.
Others have a
slightly reduced, grayish core, while some have a dark reduced core. The surface treatment
455
of Chiquero ceramics is also distinctive. It is an unslipped ware, and surface treatment
varies from a wiped, lightly striated surface with pitting to a deeply striated surface from a
brushing technique (Table A.2; Figure A.3). As is evident in Table A.2, scoring of surface
treatment was basically bimodal, with the majority of sherds (62%) displaying deep,
brushed striations (score=8), and most of the remainder (28%) with less pronounced
striations on a rough, sandpaper like surface texture with some pitting (score=6).
Table A.2: Chiquero ceramics surface treatment scores (exterior)
Count
4
5
6
7
8
9
Total
2
3
34
3
75
5
152
%
1.6%
2.5%
27.9%
2.5%
61.5%
4.1%
100.0%
456
Figure A.3: Chiquero body sherds
Vessel Forms
Nearly all Chiquero vessels are variants of the olla form, which I define, following
Steadman (1994:35), as vessels with spheroid to ellipsoid body shapes whose height does
not exceed the maximum vessel diameter. The most common Chiquero vessel form is a
neckless, spheroid olla (olla sin cuello) with a large, constricted orifice (Figures A.5-A.7).
As is evident in Figure A.4, rim diameters of these ollas are in the 26-30 cm range. Rim
profiles of this vessel form are either symmetrical or have reinforcement on the interior.
457
Some have a slight depression or bevel on the exterior (e.g. Figure A.6, G-K). A less
common spheroid olla form has a small, everted collar, such as those in Figure A.8. It
appears that some or all ollas had handles or lugs, although it is unclear where they attached
to the vessel because all were recovered without large sections of the vessel body attached
(Figure A.9). The handles are distinctive in their casual execution, and have a generally
round cross section, unlike later strap-type handles of the Collagua series. The tonguelike lugs illustrated in Figure A.9 are also unique to Chiquero ceramics, and would have
facilitated handling of the vessel. There are also a few examples of jarsthat is, collared
vessels whose orifice diameters are less than their heights (a definition also derived from
Stanish, et al. 1994:35). These are much smaller vessels than the ollas, with rim diameters
of 11-12 cm.
40
Count
30
20
10
Std. Dev = 1.73
Mean = 26
N = 41.00
0
18 - 20
20 - 22
22 - 24
24 - 26
26 - 28
28 - 30
Diameter (cm)
Figure A.4: Rim diameters of Chiquero neckless ollas
458
Dating
Given that no Formative component was identified in the valley prior to this project,
I base the temporal placement of Chiquero ceramics on similarities in form, ware, and
surface treatment with better-documented Formative ceramics in the region. The earliest
ceramics in the Titicaca Basin are fiber-tempered wares, such as Pasiri ceramics in the
southwest Titicaca Basin (Stanish 1997a:40-42), and Early Chiripa on the Taraco Peninsula
(Bandy 2001; Steadman 1999), both of which appear to date to around 1500-1000 BC.
While we recovered no fiber-tempered wares in our survey, Chiquero ceramics do share the
striated, wiped surface treatment of these two early styles, and the spheroid, neckless olla
form is considered a diagnostic form of Early Chiripa ceramics (Bandy 2001:90; Steadman
1999).
The neckless olla is also a hallmark form among the recently-documented
Formative styles in neighboring locales of the Department of Arequipa, such as Ayawala in

the Chuquibamba valley, Soporo in the Andagua valley, and Socabaya in the valley of
Arequipa (Cardona Rosas 2002:40-57; Neira Avendao and Cardona Rosas 2000-2001).
Radiocarbon dates associated with Ayawala ceramics in a well-stratified trench excavated
by Cardona Rosas provide a preliminary chronological placement of these ceramics at
between 1400 and 800 BC (Cardona Rosas 2002:49; Neira Avendao and Cardona Rosas
2000-2001). Socabaya ceramics have been recovered in association with Pukara sherds in
the valley of Arequipa, suggesting they date to between 300 BC and AD 300 (Cardona
Rosas 2002:55). As I discussed in Chapter 5, we recovered a single diagnostic Classic
Pukara sherd at the type site of Chiquero (YA-032), which suggests a similar time frame for
these ceramics. However, in the absence of radiocarbon dates, and given the apparent
459
similarity of vessel shapes with Early Formative ceramics of the Titicaca Basin, I consider
Chiquero ceramics to be broadly diagnostic of the Formative Period, between 1500 BC and
500 AD. The observed variability in the paste and surface treatment might be due to
change through time; future research should refine the chronology, and perhaps reveal
earlier, fiber-tempered wares.
460
461
462
463
Figure A.8: Chiquero collared ollas
464
Figure A.9: Chiquero handles and lugs
465
Figure A.10: Chiquero jars (A-C) and ollas
466
Middle Horizon Ceramics
The survey collections included 217 diagnostic local Middle Horizon sherds,
constituting 5.7% of the total ceramic collections.
Ware
Local Middle Horizon ceramics were made with very different pastes and firing
technique than Chiquero ceramics, marking a major change in ceramic technology. They
are much finer, composed of two paste types that are similar in composition, but with
slightly different texture. About three quarters of the sample (72%) is composed of a very
compact paste with sparse fine to medium white (probably feldspar) inclusions, as well as
very fine- to fine-sized reflective gold inclusions (probably pyrite or mica). Below, I refer
to this paste as Paste 1. The remaining 27% of the sample is composed of a slightly more
porous paste (Paste 2) with more abundant fine to medium white inclusions, also with veryfine to fine-size probable pyrite or mica reflective inclusions.
Ceramics of both pastes were fired to a buff to light-brown surface color, and were
fired in an oxidizing environment. Some of the thicker sherds of the both paste types have
a slightly reduced, grayish core, but thinner sherds are buff throughout. A distinctive red
slip was applied as a band on either side of the rim to the majority of these ceramics. The
slip was generally carefully applied to form a distinctive dark red (10R 4/6 and 5YR 5/4
being the most common values) band with a well-defined border below the vessel rim. As
is evident in Figures A.12-A.14, the band is usually thicker on the exterior than the interior,
although some examples of thick 4-6 cm bands on the interior are also present (A.14 and
467
A.15). Some bowls of these pastes had no slip band, but these were generally well-polished
and even lustrous in appearance.
Vessels of both paste types are well-crafted and
burnished. Virtually all scored a four or lower on my scale of surface treatment, with most
(63%) scoring a four. However, Table A.3 shows how Paste 1 (the more compact, fine
paste) sherds tend to be slightly more polished, with about a third of the vessels scoring 1-3,
while Paste 2 are burnished, but with some slight pitting, scoring homogeneously as 4 on
my scale. Their average scores were 3.18 and 3.98, respectively. Thus, the Paste 1 group
appears to be a slightly finer ware overall.
Table A.3: Middle Horizon ceramic surface treatment scores by paste type
Paste 1
1
2
3
4
5
Total
Count
20
5
43
67
1
136
%
14.7%
3.7%
31.6%
49.3%
.7%
100.0%
Paste 2
Count
Total
%
1
51
1.9%
98.1%
52
100.0%
Count
20
5
44
118
1
188
%
10.6%
2.7%
23.4%
62.8%
.5%
100.0%
Vessel Forms
Nearly all diagnostic Middle Horizon vessels were slightly constricted, cumbrous
bowls and cups. Rim diameters were varied, although the most common size was a
medium-sized bowl in the 12-18 cm range (Figure A.11). A few smaller vessels that
probably functioned as drinking vessels had diameters of 8-12 cm (Figure A.11). The rim
profile is also an important feature for identifying these local Middle Horizon ceramics,
since they are distinct from the subsequent Collagua I and II rim profiles of similar vessel
form. In these Middle Horizon ceramics, rim profiles are generally symmetrical, either
468
unreinforced, or reinforced symmetrically on both interior and exterior, with the lip apex in
the center of the profile axis (see especially Figures A.12 and A.13). On slightly more
open, vertical-walled bowls, exterior rim reinforcement is more common, as in Figure A.14,
A-C. These latter bowls are also distinct because the red slip band is as thick on the interior
as the exterior. We also recovered a few examples of vertical-walled cups, as in Figure
A.17, examples H-J.
Decorations
While the vast majority of the Middle Horizon ceramics in the collection do not
have painted decoration, 13% had decorative motifs, in black, white, and (rarely) red.
Painted motifs were applied all on the exterior, forming a horizontal design band below the
rim. Motifs were non-representational, curvilinear designs, such as X and S figures,
wavy lines, dots, and circles. A few sherds also had pre-fire impressions, such as in Figure
A.15, D-F. Most sherds were quite small and their decorations were therefore fragmentary.
Dating
I have identified these ceramics as pertaining to the Middle Horizon based on their
close similarity in ware, vessel form, and decoration with Middle Horizon ceramics
reported by Malpass and de la Vera (Malpass 1987:59, Figures 7 and 8; Malpass and de la
Vera Cruz Chvez 1986, 1990). They found that variability in material (ware) was a
temporally-sensitive variable for distinguishing Middle Horizon from Late Intermediate
Period ceramics. The Paste 1 and Paste 2 that I described above coincide with their detailed
469
descriptions of Material 1 and Material 2 ceramics, respectively, which, by association
with radiocarbon dates in a well-stratified trench, they considered as diagnostic of the
Middle Horizon, and specifically, of the first half of the period (Malpass and de la Vera
Cruz Chvez 1986:209, 216; 1990:44-46, 57). As I reported in Chapter 3, when calibrated,
these assays return dates of AD 429-694, and AD 436-779 (calibrated 2 sigma). In the
middle Majes valley, recent excavations at the site of Beringa by Tung (2003) have also
produced a series of radiocarbon dates associated with Qosqopa and other Wari-related
ceramics dating to the Middle Horizon. In addition to this identification based on material,
the formal attributes of these bowls suggest a Middle Horizon date, since the slightly
constricted, cumbrous bowl form is typical of Wari-influenced Middle Horizon bowls
throughout the region (Cardona Rosas 1993; de la Vera Cruz Chvez 1987:118, Figure 14,
d and e; Jennings 2002b:282, 289; Neira Avendao 1990:133-134; Sciscento 1989:169;
Ypez, et al. 2001).
Noting similarities in the decorative motifs of these ceramics and local variants of
Wari imperial wares from the Chuquibamba valley (the Qosqopa style), de la Vera Cruz
Chvez referred to similar sherds from the Cabanaconde area as pertaining to the Early
Chuquibamba style (Chuquibama Temprano) (de la Vera Cruz Chvez 1988, 1989).
However, as I discuss below, subsequent Late Intermediate Period ceramics from the
central Colca valley constitute distinctive wares from both these Middle Horizon ceramics
and from contemporaneous LIP Chuquibamba wares, and therefore I prefer not to lump
either style with the Chuquibamba series, as defined by others (Cardona Rosas 1993;
Kroeber 1944; Morante 1939; Sciscento 1989). Also, while the current sample of decorated
470
Middle Horizon ceramics is small, the painted designs also appear similar to Wari-derived
ceramics of southern Cuzco (Paruro Province) reported by Bauer, such as a local variant of
Ocros style ceramics from Muyu Roqo (Bauer 1999:66, Fig. 3-2), as well as the Qotakalli
and Ccoipa styles (Bauer 1999:70-79, Figs. 3-5 and 3-8). The impressed dots and rings on
bowl exteriors in Figures A.15, D-F, and Figure A.16, I-J, are similar to an impressed
blackware style also documented by Bauer in his survey (Bauer 1999:68, Fig. 3-4). Thus,
local Middle Horizon ceramics appear to derive from a broader network of interaction that
includes not only the Chuquibamba/Majes valley area to the southwest, but also the high
altitude zones of southern Cuzco to the northeast. However, given the small sample size
and dispersed distribution of these ceramics, I have not renamed them or attempted a style
definition. What is clear, however, is that local Middle Horizon ceramics are Wari-, not
Tiwanaku-related, and subsequent local LIP styles develop out of this substrate, in contrast
to the Tiwanaku-derived LIP styles of the Titicaca Basin to the east, or the Churajn style
to the south.
Some have questioned whether red-rim slipped ceramics are diagnostic of the
Middle Horizon. For example, Brooks (1998:349-353) and Sciscento (1989: 117-122) both
caution that red rim-slipped ceramics appear to have a wide temporal distribution extending
to the Late Horizon. However, this observation does not take into account the temporallysensitive variables of ware, vessel form, and decoration that differentiate these Middle
Horizon ceramics from the subsequent LIP and Late Horizon wares. As I discuss below,
some local LIP and Late Horizon bowls indeed do have a similar band of red slip below the
rim, and this feature provides an important measure of continuity in the sequence.
471
However, these later LIP and Late Horizon ceramics have distinct paste compositions,
surface treatments, vessel forms, and decorations from local Middle Horizon red rimslipped ceramics.
80
Count
60
40
20
Std. Dev = 3.58
Mean = 16
N = 159.00
0
8 - 10
12 - 14
10 - 12
16 - 18
14 - 16
20 - 22
18 - 20
24 - 26
22 - 24
26 - 28
Diameter (cm)
Figure A.11: Rim diameters of Middle Horizon bowls
472
473
474
475
Figure A.15: Local Middle Horizon red-slipped rim bowls (A-C), and impressed bowls (D-F)
476
Figure A.16: Local Middle Horizon decorated bowls
477
Figure A.17: Local Middle Horizon bowls (A-G) and drinking vessels (H-J)
478
The Late Intermediate Period and Late Horizon: Collagua Ceramics
The results of my analysis of ceramics recovered during the survey provide a
preliminary sequence that begins to address the perennial problem of differentiating local
Late Intermediate Period and Late Horizon wares. My findings build on those of Brooks
(1998), who differentiated the local Collagua style from Chuquibamba and other regional
styles, as well as Malpass and de la Vera Cruz Chvez, who produced a preliminary
ceramic sequence in the 1980s (Malpass and de la Vera Cruz Chvez 1986, 1990). As
opposed to, for example, the Churajn style to the south, or the LIP styles of the Titicaca
Basin to the west, both the Chuquibamba and Collagua styles emerged from a Wari-, rather
than Tiwanaku-influenced substrate. However, Chuquibamba more directly incorporates
some of the common motifs of regional Wari-influenced styles, as well as coastal
influences in the form of specific design elements such as the eight-pointed star.175 Also,
the painted design elements common to both Collagua and Chuquibamba are executed in
distinct ways, with Chuquibamba motifs generally executed with thinner lines in a grayish
paint, as opposed to the thicker, more pronounced execution of Collagua designs in heavy
black mineral paint.
Apart from differentiating Collagua from Chuquibamba, my sequence divides
Collagua ceramics into of four overlapping phases: Collagua I (terminal MH-middle LIP)
Collagua II (general LIP)Collagua III (terminal LIP-LH)Collagua Inka (Figure A.1).
The sequence is bookended at the beginning by radiocarbon dates associated with the
Middle Horizon ceramics described above (Malpass and de la Vera Cruz Chvez 1986,
175
The eight-pointed star is present but rare in Collagua ceramics. We recovered no sherds with the motif,
although Brooks has recovered a handful from the Chivay area (Brooks 1998:322, 334).
479
1990). These assays, ranging from the fifth through eighth centuries AD, provide an ante
quem date for the subsequent Collagua I ceramics. The other end of the sequence is marked
by Inka influence in the decorations of Collagua III and Collagua Inka ceramics, which I
assume to have appeared more or less contemporaneously with Inkaic occupation, most
likely during the mid 15th century. Given the paucity of radiocarbon dates for the LIP, this
part of the sequence is provisional. Accordingly, in the following discussion, I have
lumped Collagua I and II together for the LIP component, and Collagua III and Collagua
Inka together for the Late Horizon component.
Bowl forms constitute most of the diagnostics for Collagua I-III, while Collagua
Inka ceramics includes primarily open plates with lesser quantities of cups, beakers (k'eros),
flaring-rimmed long-necked jars (arbalos), and one-handled pitchers typical of Late
Horizon assemblages throughout the south-central Andes (Brooks 1998; Brgi 1993; de la
Vera Cruz Chvez 1988, 1989; Jennings 2002b; Julien 1983; Lumbreras 1974a, b; Malpass
and de la Vera Cruz Chvez 1986, 1990; Prssinen and Siirriinen 1997; Sciscento 1989;
Stanish 1991; Tschopik 1946; Van Buren 1993). The general trend over the sequence in
terms of form is a change from slightly constricted to vertical-walled bowls with reinforced
rims amongst Collagua I ceramics, to more open, cumbrous and straight-sided bowls of
Collagua II, and finally ending in open, flaring flat-bottomed bowls and shallow plates of
Collagua III and Collagua Inka (Table A.1). Accompanying this change is a shift in the
placement of decoration from the external to internal surface between Collagua II and
Collagua III.
In terms of decorative motifs, designs change from Wari-influenced
geometric and curvilinear decorations arranged in horizontal design fields of Collagua I, to
480
floating thick-lined curvilinear designs on Collagua II bowls, and ending in a more
restricted range of thick wavy lines and simple gestural figures on the interior of Collagua
III bowls. Collagua Inka ceramics show obvious Inka-derived motifs, such as dentate
patterns, finely-executed parallel lines around the rim, concentric diamonds, and other
geometric designs, as well as figurative motifs such as the footplow (chaquitaclla), the
sun, an the maize motifs (see Fernandez Baca 1971, for range of common Inka ceramic
motifs).
The Late Intermediate Period: Collagua I and II

Together, Collagua I and II ceramics made up 11.4% of the ceramic collections
from the survey, with 152 Collagua I sherds (4.0%) and 281 Collagua II sherds (7.4%).
Ware
Both Collagua I and II ceramics are made of a very distinct paste from the Middle
Horizon ware described above. The paste is semi-compact to compact in consistency, with
occasional small voids in the vessel wall. The paste is sand-tempered, varying from sparse
to abundant, in the very fine to fine (and, rarely, medium-sized) size range, with occasional
white (probable feldspar) inclusions.
Vessel firing varies from completely oxidized
throughout to highly reduced, but most sherds show a distinct dark grey reduced core with a
reddish oxidized exterior. The slip is most often cherry red (10R 4/4) although some more
orangish red slips were also used (10R 4/6, 10R 4/8, 10R 5/6 and 10R5/8). Slip coverage
and thickness is variable, and, unlike the red rim-slipped Middle Horizon ceramics, the slip
481
border was not sharply delineated, but uneven and gradually thinned at the edge. In this
sense, and in the variability in firing, Collagua I ceramics appear to have been less carefully
manufactured than the preceding Middle Horizon ware. Burnishing was also variable
amongst Collagua I ceramics, but generally the vessel exterior was well-polished, with 71%
of bowl sherds scoring a 2 for the exterior surface (Table A.4). The average score was
2.21, with a standard deviation of 0.580. Surface treatment of Collagua II ceramics is
slightly more variable than in Collagua I, with treatment scores between 2 and 3 being the
most common, although some lower-quality treatment (scores 4-5) are also present (Table
A.4). The higher average score (2.49) than Collagua I (2.21) suggest that, as a class,
Collagua II ceramics were somewhat less carefully-manufactured as Collagua I ceramics.
Table A.4: Collagua I and II bowl surface treatment scores (exterior surface)
Collagua I
Count
1
2
3
4
5
Total
6
86
27
1
1
121
%
5.0%
71.1%
22.3%
.8%
.8%
100.0%
Collagua II
Count
87
64
5
1
157
Total
%
55.4%
40.8%
3.2%
.6%
100.0%
Count
6
173
91
6
2
278
%
2.2%
62.2%
32.7%
2.2%
.7%
100.0%
Vessel Forms
All Collagua I vessel forms are bowls. There are four principal bowl shapes: 1) a
cumbrous, slightly restricted form (Figure A.20, A-C, E; Figure A.21; Figure A.22, F, G;
Figure A.23), 2) a slightly more open, collared bowl with a flat bottom (Figure A.20, D), 3)
a deeper, vertical-walled bowl (Figure A.22, A-D; Figure A.24; Figure A.25), and 4) a
straight-sided flaring bowl with a flat bottom (Figure A.26). Most Collagua I bowl rim
482
diameters are in the 14-18 cm range (Figure A.18). The shallow, cumbrous form is the
most common. Some of these bowls are quite similar in shape and rim profile as the
Middle Horizon bowls (Figure A.20, A-D, F), but are distinguishable by their distinct paste,
decorative motifs, and less defined slip border on the interior. Also, Collagua I cumbrous
bowls tend to have thicker, more asymmetrical rim profiles than their Middle Horizon
counterparts, with more reinforcement on the exterior side of the rim. The collared bowl
form (Figure A.20, D) is unique to Collagua I bowls, although it is relatively uncommon;
we recovered only six examples. The vertical-walled bowls tend to have thick walls and
thickly reinforced rims; some have small button-like lugs just below the rim on the exterior
(Figure A.21, E; Figure A.25, C), probably to facilitate handling. The more open, straightsided flaring bowls are also thick-walled, and some rimsherds also included small handling
lugs similar to those of the vertical walled bowls (e.g., Figure A.26, C).
Complete
examples of these vessel forms have been recovered from trench excavations around the
foundations of a house at Chijra by Neira (Neira Avendao 1986). Brooks also included
photographs of these bowls in her dissertation (Brooks 1998:364-366).
I separate Collagua II from Collagua I bowls based on the covariance between
formal and decorative attributes. In terms of form, Collagua II bowls include a distinct,
open bowl with a more rounded base than Collagua I bowls (Figure A.28, A and B; Figure
A.29; Figure A.30, A). More straight-sided, flaring open forms are also represented (Figure
A.27; Figure A.28, C-E; Figure A.30, E and F). I have tentatively categorized verticalsided bowls with wavy line motifs as Collagua IIa conservative classification since, as I
discuss below, I consider Collagua II to be broadly diagnostic of the LIP. Bowl sizes are in
483
the same range as for Collagua I, with a slightly higher average rim diameter (16 cm, versus
15 cm for Collagua I)again, reflecting the flaring form of most Collagua II bowls (Figure
A.19). Collagua II forms also include a flared-rim jar with vertical neck walls and spheroid
body (cntaro) (Figure A.31), and a flared-rim cntaro with a more ovaloid body (Figure
A.32, D). While these cntaro forms continue to be used through the Late Horizon, they
can be distinguished on the basis of distinct slip color and decorations (see below).
60
50
Count
40
30
20
10
Std. Dev = 1.63

Mean = 15
N = 111.00
0
8 - 10
10 - 12 12 - 14 14 - 16 16 - 18 18 - 20 20 - 22
Diameter (cm)
Figure A.18: Rim diameters of Collagua I bowls
484
100
80
Count
60
40
20
Std. Dev = 1.38
Mean = 16
N = 134.00
0
10 - 12
12 - 14
14 - 16
16 - 18
18 - 20
20 - 22
Diameter (cm)
Figure A.19: Rim diameters of Collagua II bowls
Decorations
All decorations on Collagua I bowls are on the exterior surface. They are executed
in horizontal design fields on the exterior in black, and, more rarely, white, over the red
slip. Collagua I decorations are varied, ranging from geometric and curvilinear designs to
zoomorphic motifs. Zoomorphorphic motifs include representations of birdsundoubtedly
depictions of the various species of foul abundant around the bofedales of the puna (Figure
A.21, D). Similar avian designs are found on the exterior of Chuquibamba bowls to the
southwest (Sciscento 1989:143) and on the interior of Late Horizon Chucuito Black on Red
and Chucuito Polychrome plates in the Titicaca Basin (Tschopik 1946:28, 29) and in the
outlier colonies of the upper Moquegua (Van Buren 1993:257, 258, 260). Thus, the bird
motif has a wide distribution, beginning during the LIP and extending through the Late
485
Horizon. Curvilinear and geometric motifs of Collagua I bowls include wavy lines that
wrap around the exterior (Figures A23-A.25), pendant triangles (Figure A.20, E), as well as
chained diamonds and rhomboids with parallel and reticulated line fills (Figure A.20, A and
B). Wavy lines often alternate with circles, parallel lines, or dots above and below (Figure
A.23, D-G; Figures A.24-A.25). Circles are often filled with parallel lines Figure A.20, D),
X figures (Figure A.23, G), or a single dot in the center (Figure A.23, D-G). Designs
generally either drape from a single or paired horizontal lines below the rim, or are placed
between horizontal lines around the bowl that delineate the horizontal design field. Many
Collagua I bowls also have black ticks on the rim, generally interpreted as a Wari-derived
trait in the Pacific drainages of Arequipa (Cardona Rosas 1993; de la Vera Cruz Chvez
1988, 1989).
While Collagua I decorations include black and white on red, designs on Collagua II
bowls are all black on red bichromes. In contrast to Collagua I bowls, most Collagua II
decorations are not delineated by horizontal lines to create bordered design fields (see
Figures A.27-A.30). Instead, thick-lined, curvilinear designs typically drape from the rim,
such as single, paired, or triple arc figures (Figure A.27, D-F; Figure A.28, C; Figure A.29;
Figure A.30, A), and large, thick X figures (Figure A.28, B and D). An exception to this
rule is a tight wavy line motif, which appears singly or in pairs between (single or double)
horizontal lines below the rim (Figure A.28, A-C, Figure A.29, A). Most designs are on the
vessel exterior, although some straight-sided flaring bowls are decorated on the interior
(Figure A.30, C-F).
486
Dating
Currently, there are no radiocarbon dates from undisturbed contexts associated with
decorated Collagua I or II ceramics, and continuity in settlement between the LIP and Late
Horizon complicates matters. Previously, Neira (Neira Avendao 1961), de la Vera Cruz
(de la Vera Cruz Chvez 1987, 1988, 1989), and Sciscento (Sciscento 1989), considered the
black on red and black and white on red ceramics that I call Collagua I and II ceramics a
local variant of the Chuquibamba series, originally defined as a Late Intermediate Period
style by Morante (1939) and Kroeber (Kroeber 1944) based on surface collections and
purchased pieces from the Chuquibamba valley (cf. Brooks 1998:353-356; Lumbreras
1974b:212).176
Lumbreras and Amat (1968) noted similarities between Chuquibamba
ceramics and other LIP styles of the south central Andean sierra and altiplano, such as
Churajn of Arequipa (Kroeber 1944), Chucuito and Allita Amaya of Puno (Tschopik
1946), and Mollo in the altiplano (Lumbreras 1974b), which they referred to collectively as
the Southern Tricolor Horizon styles. Subsequently, Malpass and de la Vera Cruz (1986;
1990), noting that previous chronological assessments of Chuquibamba and the Southern
Tricolor Horizon styles lacked supporting chronometric data, suggested instead that black
on red and tricolor ceramics were not isolated to the LIP, but also continued to be used
through the Late Horizon. This more general LIP/Late Horizon assessment was based on
continuity of paste recipe throughout post-Middle Horizon contexts.177 However, they
176
However, Neira (1961; 1990) used the terms Chuquibamba and Collagua interchangeably to refer to the
LIP ceramics of the Colca valley.
177
That is, their Material 4 ceramics supercede their Middle Horizon Materials 1 and 2 (discussed above).
Their description of the sand-tempered, compact Material 4 paste coincides with my description for Collagua
I-Collagua Inka ceramics.
487
considered their interpretation provisional because of turbid stratigraphic contexts in their
terrace excavations (Malpass and de la Vera Cruz Chvez 1986, 1990). Subsequent to these
reports, de la Vera Cruz (1988; 1989) subdivided similar ceramics he collected from
reconnaissance in the Cabanconde area into Early, Middle, and Late Chuquibamba
styles, which approximately correspond to the Middle Horizon, Late Intermediate Period,
and Late Horizon, respectively. More recently, Brooks (1998:317-356) has delineated
stylistic differences between Collagua ceramics from the Colca valley and the
Chuquibamba seriesa distinction that I, as well as others who have extensive experience
working in both areas (Cardona Rosas, pers. comm. 2000; Ypez Alvarez, pers. comm
2000; Simborth Lozada, pers. comm. 2000) find compelling.
However, while Brooks lumps Collagua ceramics into a single LIP/Late Horizon
component, I suggest that several attributes make possible a preliminary distinction
between LIP (Collagua I and II) and Late Horizon (Collagua III and Collagua Inka)
Collagua styles. My sequencing of the Collagua series is based on covariance in distinct
morphological and decorative elements, and in this sense represents a similiary seriation
approach (Rowe 1961:326-327), but it is also undergirded by cross-dating with similar
styles from neighboring locales. Above all, the similarity in vessel form between Middle
Horizon and Collagua I constricted bowl forms suggest that the Collagua I bowls are
derived from that earlier style, but executed with a different paste, slip, and surface
treatment. Decorative motifs on Collagua I bowls, such as the wavy lines, diagonal zigzag
patterns, filled circles, and rim ticking (see especially Figure A.19, D, F; Figure A.21, DG), also appear derivative of the local Wari-influenced Middle Horizon decorated ceramics.
488
Thus, I suggest that Collagua I bowls originate in the early LIP, and perhaps slightly
overlap with the terminal Middle Horizon, and extend at least through the first half of the
LIP (Figure A.1). By contrast, the more open, rounded bowl forms of Collagua II ceramics
appear as an intermediate form between the constricted Middle Horizon and Collagua I
bowls, and the subsequent open plate form typical of Late Horizon assemblages in the
valley and throughout the region. The thick-lined motifs of Collagua II bowls are broadly
similar in execution with other regional LIP styles, such as Pucarani Black on Red (Stanish
1997a:47-48, 153 [230.001-3]), Kelluyo (Stanish 1997a:46-47, Figure 22) and Tanka Tanka
Black on Orange (Hyslop 1976:431-435) in the western Titicaca Basin.
Together, I
consider Collagua I and II to be broadly diagnostic of the Late Intermediate Period, with
Collagua I originating and falling out of use earlier than Collagua II. Future excavations
and chronometric testing should make possible more definitive evaluation and refinement
of the LIP component of the Collagua sequence.
489
Figure A.20: Collagua I cumbrous (A-C, E), and a collared (D) bowls
490
Figure A.21: Collagua I cumbrous bowls
491
Figure A.22: Collagua I vertical-walled (A-D) and cumbrous (E-G) bowls
492
Figure A.23: Collagua I cumbrous (A, F) and slightly open (B-D) bowls
493
Figure A.24: Collagua I vertical walled (A-C, F) and cumbrous (D, E) bowls
494
Figure A.25: Collagua I vertical walled (A-E) and flaring (F) bowls
495
Figure A.26: Collagua I flaring bowls
496
Figure A.27: Collagua II vertical-walled bowls
497
Figure A.28: Collagua II rounded open (A, B), and flaring (C-E) bowls
498
Figure A.29: Collagua II rounded open bowls
499
Figure A.30: Collagua II rounded open (A-C) and flaring (D-F) bowls
500
Figure A.31: Collagua II cntaros
501
The Late Horizon: Collagua III and Collagua Inka
Late Horizon sherds make up the largest class of ceramics in the survey collections,
totaling 1468 sherds, or 38.5% of the collections. These include 218 Collagua III sherds
(5.7% of the total sample), 1120 Collagua Inka sherds (29.4%). I also classified 130 sherds
(3.4%) as Inka (i.e. probable imported imperial Inka wares), owing to their distinct paste
composition and diagnostic Cuzco Polychrome decorative and formal attributes.
Ware
Both Collagua III and Collagua Inka ceramics are made of the same sand-tempered
paste as Collagua I and II ceramics. However, there are significant differences in firing,
slip, and surface treatment between Collagua III and Collagua Inka wares. On the one
hand, firing of local Collagua III ceramics was similar to Collagua I and II ceramics,
varying between highly reduced and oxidized. As is evident in Figure A.33, about 2/3 of
Collagua III ceramics had either a very dark reduced core (34%) or a slightly reduced,
grayish core (31%). These percentages are comparable to Collagua I and II, but with a
slightly greater percentage of sherds with very dark, reduced cores (Figure A.32). On the
other hand, the great majority (82%) of Collagua Inka sherds were high-fired and
thoroughly oxidized throughout (Figure A.32), probably reflecting the states role in quality
control and standardization of production. As I discussed in Chapters 8 and 9, there were
ayllus of official Inka state potters resident in the survey area, and their craftsmanship
appears to be reflected in the firing quality of Collagua Inka ceramics.
502
The slips and surface treatment of Collagua III and Collagua Inka ceramics are also
distinct. The great majority (81%) of Collagua III ceramics were executed in the same dark
red slip as Collagua I and II ceramics, while Collagua Inka ceramics were made with a
bright red and orangish red slip (Figure A.33). Slip coverage of most Collagua III bowls
completely coats the interior, but leaves a thick band below the exterior of the rim, without
a defined border between the slipped and unslipped area. Collagua Inka plates are slipped
over the entire vessel. As is evident in Tables A.5 and A.6, Collagua Inka plates were also
more finely polished and lustrous than any other local style, with the lowest average surface
treatment score in the Collagua sequence (1.76). Surface treatment of Collagua III bowls is
similar to Collagua I and II, but somewhat less carefully executed overall, with an average
score of 2.77 (Table A.5).
503
100
90
80
70
Percent
60
50
Style
40
30
Collagua I
20
Collagua II
10
Collagua III
Collagua Inka
0
IT
IA
IG
CN
CO
Firing Code*
Figure A.32: Comparison of Firing Quality, Collagua I-Collagua Inka

IT
IA
IG
CN
CO
*Key to Firing Codes:

Incomplete, very dark throughout, except outermost layer and slip
Incomplete firing with dark reduced core
Incomplete firing with light grey, slightly reduced core with blurred edges
Complete but not oxidized red. Buff to brown throughout
Complete and oxidized throughout
504
100
90
80
70
Percent
60
50
Style
40
30
Collagua I
20
Collagua II
10
Collagua III
Collagua Inka
0
BRD
BRN
DRD
FBL
ORG
Slip Color*
Figure A.33: Comparison of slip color, Collagua I-Collagua Inka

BRD
BRN
DRD
FBL
ORG
*Key to Color Codes:

Bright red (10R 5/6, 10R 5/6, 10R 5/8)
Brownish red (5YR 5/4, 7.5YR 3/4)
Dark red (10R 4/4, 10R 4/6)
Fire-blackened, obscuring original slip color (7.5 YR 5/3, 7.5 YR 3/1)
Orangish red (2.5 YR 5/6, 2.5 YR 6/8)
Table A.5: Collagua III and Collagua Inka bowl and plate surface treatment scores (interior surface)
Collagua III
Count
1
2
3
4
5
6
7
Total
42
133
1
11
22.3%
70.7%
.5%
5.9%
1
218
.5%
100.0%
Collagua Inka
Count
195
309
119
34
81
28
%
25.5%
40.3%
15.5%
4.4%
10.6%
3.7%
1120
100.0%
Total
Count
195
351
252
35
92
28
1
1338
%
20.4%
36.8%
26.4%
3.7%
9.6%
2.9%
.1%
100.0%
505
Table A.6: Comparison of average surface treatment scores, Collagua I to Collagua Inka*
Style
Mean Score
Std. Dev.
Collagua I
2.21
0.58
Collagua II
2.49
2.21
Collagua III
2.77
0.463
Collagua Inka
1.76
0.681
*For this comparison, I use the area of the vessel which is most visible as the index of
polishing; thus, I use external surface treatment scores for the more closed forms of
Collagua I and II bowls, and interior surface treatment scores for the more open Collagua III
and Collagua Inka bowls and plates.
Vessel Forms
Collagua III is composed primarily of flaring, straight- and curving-sided bowl
forms with flat bottoms that are more open than Collagua I and II bowls, but generally not
as shallow or flat as Inka-style plates. Rim diameters are in the same range as Collagua I
and II bowls, averaging 15 cm (std. dev.=1.60; Figure A.34). The rims of some of the
straight-sided flaring bowls had small protuberances (e.g., Figure A.36, C) that are a
characteristic feature of Cuzco Polychrome and provincial Inka plates (see, e.g., Julien
1983:Plate 30; Rowe 1944:Figure 19; Stanish 1991:45-46).
The diagnostic forms of Collagua Inka ceramics are dominated by shallow open
plates and large, long-necked aryballoid bottles (arbalos), with lesser quantities of cups,
beakers (k'eros), and one-handled pitchers typical of Late Horizon assemblages throughout
the south central Andes (Brooks 1998, Brgi 1993, de la Vera Cruz Chvez 1988, de la
Vera Cruz Chvez 1989, Julien 1983, Lumbreras 1974a, Lumbreras 1974b, Malpass and de
la Vera Cruz Chvez 1988, Malpass and de la Vera Cruz Chvez 1990, Prssinen and
Siirriinen 1997, Stanish 1991, Tschopik 1946, Van Buren 1993, Wernke 2001). The rim
506
diameters of most plates are in the 14-18 cm range, with a mean of 16 cm (std. dev.=1.57;
Figure A.35).
As in the imperial heartland and other provincial settings, the vessel form inventory
of Collagua Inka ceramics was very standardized, and because of this, vessel shapes can
often be reconstructed from small sherds (Bray 2003; Julien 1983; Rowe 1944, 1946a).
For example, some Collagua Inka plates had handles; we recovered a few examples of
handles modeled in the form of a birds head (Figure A.46, G), a typical motif for such Inka
plate handles (Prssinen and Siirriinen 1997:263; Rowe 1944, 1946a; Tschopik 1946:38).
On this plate form, two closely spaced, small protuberances typically appear on the rim
opposite the handle (see, e.g., Julien 1983:Plate 30; Rowe 1944:Figure 19; Stanish 1991:4546; 1997a:47), and this feature was present on several plate rimsherds we collected (e.g.
Figure A.47, J).
Also, locally-produced arbalos follow the standard shape and
construction, with widely flaring necks (Figure A.52), handle joins below the shoulder
(Figure A.53), and large, modeled lugs (Figure A.52). A globular pitcher form is also
common, the handle of which is adorned with a modeled and painted zoormorphic head
(Figures A.50 and A.51). Clearly, the local vessel inventory changed considerably under
Inka state supervision of ceramic production.
507
100
90
80
70
Count
60
50
40
30
20
Std. Dev = 1.60
10
Mean = 15
N = 125.00
10 - 12
14 - 16
12 - 14
18 - 20
16 - 18
22 - 24
20 - 22
Diameter (cm)
Figure A.34: Rim diameters of Collagua III bowls and plates

200
180
160
140
Count
120
100
80
60
40
Std. Dev = 1.57
20
Mean = 16
N = 379.00
10 - 12
12 - 14
14 - 16
16 - 18
18 - 20
20 - 22
Diameter (cm)
Figure A.35: Rim diameters of Collagua Inka and Inka plates
508
Decoration
All Collagua III bowls are decorated on the interior surface, and are all executed in
black on red. By far, the most common motifs were thick, single or paired wavy lines
below the rim, sometimes accompanied by wavy line designs at the base of the interior of
the plate (Figures A.36-A.39). Other designs include M and V shaped motifs, as well
as circles with dots in the center (Figure A.40). Overall, the designs are rather gestural in
manner, and appear not to have been carefully executed, especially in comparison to the
much more rigidly formalistic and standardized decorations of Collagua Inka ceramics.
Decorations on Collagua Inka plates were executed both in bichrome (black on red)
and polychrome, and mimic many of the motifs of Cuzco pottery. Amongst the Collagua
Inka bichrome plates, thin concentric lines inside the rim are the trademark design (Figures
A.41-A.45), to which geometric, curvilinear, or representational motifs were often added on
the plate interior (Figures A.42-A.45). These plates also show stylistic continuities from
external design fields on Collagua style bowls, such as bird motifs, also common to other
southern provincial Inka styles (Figure A.44, A-C, H-K) (Julien 1983:Plates 34, 40;
Prssinen and Siirriinen 1997:263; Tschopik 1946:30). Depictions of camelids in the
center of plates are common (Figure A.45), and are quite similar to those found on
Sillustani and Chucuito ceramics of the western Titicaca Basin (Julien 1983:Plate 30), but
are distinct from the more stylized, repetitive llamita motifs of the Late Horizon Pacajes
(or Pacajes-Inka) style to the south (Albarracin-Jordan and Mathews 1990:171; Rydn
1957:235-238; Van Buren 1993:291, Figure 39, d-f). Camelid depictions are also found on
Late Chuquibamba (or Chuquibamba Inka) ceramics, although, as pointed out by Brooks
509
(1998:357), these are more stick-like in execution than the naturalistic camelid
representations on Collagua ceramics. Camelid head silhouettes often occur in rows below
the rim, both in black and red and polychrome variants (Figure A.45, D and E; Figure A.48,
G). Similar motifs are also found on Late Horizon Chucuito ceramics (Tschopik 1946:2829). Other representational motifs on plate interiors that are typical of the Cuzco Inka style
include the footplow (chakitaclla) (Figure A.47, A-C), fish (Figure A.44, E-G), and
cultigens such as aj (Figure A.48, F-J)(see Fernandez Baca 1971, for comparison with
Cuzco Inka motifs).
Polychrome plates, generally finely executed with thick bands of cream colored
pigment as a background to circumferential designs on plate interiors, display geometric
designs common to the Cuzco Polychrome style, such as dentate (Figure A.48, H and I) and
concentric line (Figure A.48, A) patterns, as well as repetitive parallel lines and cross-hatch
fills (Figure A.48, 48, B-D) (D'Altroy and Bishop 1990; D'Altroy, et al. 1994; Fernandez
Baca 1971; Julien 1983; Prssinen and Siirriinen 1997; Rowe 1944, 1946a; Tschopik
1946).
The most common decoration on large bichrome arbalos is a series of thick stripes
running vertically down the vessel wall and over the tops of the strap handles (Figure
A.53).
On smaller bichrome arbalos, a variety of curvilinear, geometric, and
representational motifs are representedall of which appear in similar form in Cuzco Inka
and other provincial styles (Figure A.51). Polychrome arbalo decorations include vertical
design areas with concentric diamonds, checkerboards, reticulated diamond chains, fineline
crosshatching, and other typically Inkaic geometric designs, as well as stylized
510
representational motifs, such as the branching, maize design (Figure A.55, A-D), and one
example of a feline depiction (Figure A.55, E).
Dating
As discussed above, Collagua Inka ceramics share many morphological and stylistic
attributes with the better-documented Late Horizon styles of the Titicaca Basin, especially
Chucuito Black on Red, Chucuito Polychrome, Sillustani Black on Red, and Sillustani
Polychrome. In fact, some Late Horizon polychrome plates had been identified as altiplano
imports by previous investigators (Linares Delgado 1989; Shea 1987, 1997; Malpass pers.
comm. as cited in Stanish 1991:15). While we did recover a few polychrome plate sherds
of distinct wares that I categorized as probable altiplano/Titicaca Basin imports, my
findings establish that the bichromes and polychromes together share the same paste recipe
as local Collagua I-III wares, and therefore constitute a local provincial Inka style that I call
Collagua Inka. Collagua III appears to represent a local utilitarian ware, which nonetheless
shares certain formal (open shapes, rim protuberances) and decorative (internally-decorated
plates with wavy lines common to regional Late Horizon styles) attributes that allow for a
preliminary terminal LIP to Late Horizon temporal assignment.
At a regional scale, similarities in vessel shapes, the organization of vessel design
spaces, the use of color, and shared motifs all cut across ethnic boundaries and suggest
intensive interaction with the western Titicaca Basin groups. Thus, while early Collagua
(Collagua I and II) and Titicaca Basin LIP styles developed out of contrasting Wari- and
Tiwanaku-derived substrates, there is increasing evidence for stylistic convergence between
511
the two areas through the LIP, a trend bolstered by imperial integration and the introduction
of standardized Inka stylistic attributes. While published reports are lacking for the valley
of Arequipa to the south, the few Late Horizon sites that have been reported show strikingly
similar assemblages (Ypez Alvarez 1991).
There is some debate concerning the timing of the appearance of Inka-influenced
ceramics in the Titicaca Basin relative to political incorporation of the region (Prssinen
and Siirriinen 1997), but most researchers view Inka-influenced ceramics as indicative of
an imperial occupational component that spans the mid-fifteenth century to 1532, in line
with historical accounts which suggest initial Inka incursions by Viracocha Inka and
subsequent consolidation of rule under Pachacuti Inka.
Radiocarbon dates from Inka
settlements on the Island of the Sun bolster this standard chronological view (Bauer and
Stanish 2001:251-255). I expect that the Collagua were incorporated during a similar time
frame, and therefore I bracket the dates of the Late Horizon in the same range as Stanish
and colleagues use for the Titicaca Basin: AD 1450-1532 (Stanish 1997a:14).
512
513
514
515
516
517
Figure A.41: Collagua Inka bichrome bowls and plates
518
519
520
521
522
523
Figure A.47: Collagua Inka bichrome (A-H) and polychrome (I, J) plates
524
Figure A.48: Collagua Inka (A, E-G, I) and Inka (B-D, H) polychrome plates
525
Figure A.49: Collagua Inka bichrome (A, B) and polychrome (C) flat-bottom bowls
526
527
528
Figure A.52: Collagua Inka arbalo rims and lugs
529
Figure A.53: Collagua Inka bichrome arbalo handle and body sherds
530
531
532
Colonial Period Ceramics
The survey collections included 456 sherds categorized as Colonial Period
ceramics12% of the total collections. The classification is necessarily broad at this early
stage in research on colonial ceramics.
Also, as I discussed in Chapter 7, common
tablewares tend to be very stylistically conservative, with many forms virtually unchanged
between the 16th and 20th centuries. Therefore, the large sample size undoubtedly reflects
a palimpsest effect of a long period of deposition.
Ware
There was a great variety of paste recipes in evidence amongst the colonial wares
clearly, a reflection of colonial-era trade networks and commerce. Given the huge diversity
of paste types, I did not record detailed observations for most sherds. In general, however,
Colonial Period ceramics can be grouped into general categories of earthenwares and tinenameled wares.
Colonial sherds are easily recognizable by their high-temperature,
sometimes vitrified firing, usually in a oxidizing environment.
In contrast to coiled
prehispanic ceramics, Colonial Period ceramics were wheel-thrown.
Vessel Forms
The most common tableware form were various open, flat-bottomed bowl forms,
some brimmed and thick-walled, and others with symmetrical, unreinforced rims. Some
bases had foot rings. We recovered some high-fired rimsherds of rounded open bowls and
straight-sided flaring bowls that probably date to the early colonial period, showing
533
continuity in vessel form and decoration from the Collagua Inka and Collagua III bowls
(see Figure A.56, C-E; Figures A.57-A.60). Rim profiles of the rounded bowls are slightly
reinforced on the interior, forming a small bevel just below the lip on the interior (Figure
A.57). The more straight-sided flaring bowls have tapering rim profiles (Figures A.58A.60).
Decoration
Decorations on Colonial Period tablewares were executed in a variety of paint and
glaze colors, the most common being black, red, maroon, brown, and green. Decorations
were very gestural and generally casually executed.
Designs range from wavy lines,
curvilinear motifs, and reticulated polygons, to naturalistic representations of birds. A

small sample of bowls with sloppily executed linear designs similar to those on Collagua III
and Collagua Inka bowls may represent transitional early colonial designs (e.g. Figures
A.57-A.60).
Dating
Given the preliminary state of knowledge concerning colonial ceramics in the south
central highlands, I have conservatively grouped colonial ceramics within the broad range
of AD 1532-1821.
The only collections reported in the region that can be used for
comparison are those of Rice and Van Buren from the Moquegua valley (Rice 1997; Van
Buren 1993:309-316). However, Rice (1997:174-175) also reports the difficulty of dating
local tin-enameled tablewares and industrial (wine tinajas and botijas) ceramics because of
534
they have changed so little over the last four centuries. More detailed, future research holds
promise for identifying subtle changes in vessel form and decoration through time.
535
Figure A.56: Colonial Period bowls
536
Figure A.57: Probable early colonial rimsherds. Left: YA-046, lot 206; right: CO-167, lot 566
Figure A.58: Probable early colonial rimsherd from YA-054 (Llactarana), lot 220
537
Figure A.59: Probable early colonial rimsherd from YA-041 (Yanque), lot 179
Figure A.60: Probable early colonial rimsherd from YA-050 (Uyu Uyu), lot 259
538
Appendix B:
Projectile Point Illustrations
Lot numbers are indicated on right side of illustrations. See Appendix D for lot
number proveniences.
All type identifications in the figure captions are based on descriptions and
illustrations in Klink and Aldenderfer (in press). Corresponding date ranges (approximate)
for each of the types in the regional chronology are as follows:
Table B.1: Projectile point types and date ranges (uncalibrated bp),
based on Klink and Aldenderfer (in press)
Type
1A/1B
2A
2B
2C
Period
Early Archaic
Early-Middle Archaic
Middle Archaic
Middle Archaic
3A
Early/Middle Archaic
3B
3C/3D
3E
3F
4A/4B
4C
4D
4E
Early/Middle Archaic
Archaic (General)
Middle Archaic
Late Archaic
Early Archaic
Terminal Archaic
Late Archaic
Middle Horizon
Late/Terminal
Archaic
Terminal Archaic
Terminal Archaic Middle Horizon
4F
5A
5B
Years bp
10,000-8000
9000-7000
7000-6000
8000-6000
10,00006000
8000-6000
10,000-3200
8000-6000
6000-4400
10,000-8000
3600-3000
6000-4400
1450-850
5000-3600
4400-3600
4400-850
5C
Terminal Archaic Late Formative
4400-1450
5D
Terminal Archaic Late Horizon
3600-contact
539
Figure B.1: Projectile points. Type 1A: A-C; Type 1B: D; Type 2A: E, F; Type 2C: G
540
Figure B.2: Projectile points. Type 3F: A, B; Type 4D: C; Type 4F: D; Type 5C: E, F; Type 5D: G-I
541
Appendix C:
Archaic Period Sites
Table C.1: Projectile point counts, grouped by time period*
Site No.
21
32
33
34
40
66
67
73
78
82
83
84
86
87
88
89
93
94
102
104
105
106
110
111
145
152
156
157
158
159
EA
E-MA
0
0
0
7
0
1
2
0
0
0
0
0
2
1
2
0
0
0
0
0
0
2
0
0
1
0
0
0
0
0
MA
0
0
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
LA
0
0
0
2
0
0
0
0
0
1
0
0
0
0
0
0
1
0
0
0
0
1
0
0
0
0
0
0
0
0
TA
0
0
1
16
0
1
0
0
0
0
1
0
0
1
0
0
0
0
0
0
0
6
1
0
0
0
0
0
0
0
GA
0
0
0
2
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
1
0
0
0
0
0
0
0
1
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
TA-FOR
0
0
0
1
0
0
0
0
0
0
1
0
0
1
0
0
0
2
0
1
0
2
0
0
0
0
0
0
0
1
TA-LH^
1
5
1
2
1
1
0
2
3
0
1
1
0
0
1
2
0
3
1
0
0
9
0
1
0
4
1
1
0
1
*Based on projectile point chronology for the south central highlands (Klink and Aldenderfer in press).
^Included for comparison, but Type 5D points are broadly diagnostic of Terminal Archaic to Late Horizon
Periods.
Codes for Table C.1
Code
EA
E-MA
MA
LA
TA
GA
TA-FOR
TA-LH
Period
Early Archaic
Early-Middle Archaic
Middle Archaic
Late Archaic
Terminal Archaic
General Archaic
Terminal Archaic-Formative
Terminal Archaic-Late Horizon
Point Types
1A/1B/4A
3A
2A/2C
3F/4D
4F/5A
3D
5C
5D
542
Figure C.1: Archaic Period sites. See Table C.1 for component-specific projectile point counts.
543
Lot District
1 YA
Site no. Sector

1
2 YA
3 YA
1
1
4 YA
5 YA
6 YA
7 YA
8 YA
9 YA
10 YA
2
2
11
12
13
14
15
16
17
3
4
4
4
4
4
YA
YA
YA
YA
YA
YA
YA
A
B
18 YA
19
20
21
22
YA
YA
YA
YA
5
6
6
7
23 YA
24 YA
25
26
27
28
29
30
31
32
33
34
8
8
8
8
8
9
9
9
9
9
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
B
B
C3
C2
C1
C3
C1
A
B
C
C
D
Ceramics Lithics Metals

Provenience
General surface collection, exterior
1
of structure
Rubble mound surface
5
Surface collection, terrace surfaces
14
to east of structure
General surface collection
3
Special collection 3, terrace
3
surfaces
13
surfaces
1
surfaces
Rubble mound
24
Rubble mound
4
6
surfaces
Isolated tomb
2
Mound A
8
Mound F
4
Wall Fill (Wall G)
4
Mound D
1
Mound E
6
15
surfaces
13
surfaces
3
Superior part of rockshelter
3
Superior part of rockshelter
1
Surface of mounded mortuary
44
feature
6
feature
1
feature
Superior part of boulder
47
12
4
1
1
Wall fill
13
Surface collection, terraces
19
28
4
45
Other
544
Lot District
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
Site no. Sector

9
D
10
A
10
B
12
B
12
B
12
A
14
E
14
E
14
E
14
E
14
D
14
D1
14
D2
14
D
14
B
14
C
14
C
14
E3
14
E
14
E4
14
E5
14
A
14
A
14
A
14
A
15
15
15
17
18
21
21
21
21
25
25
26
27
28
29
30

Provenience
4
Rubble mound
13
Rubble mound
2
11
1
41
Wall surfaces
29
Wall surfaces
6
Wall surfaces
5
Wall surfaces
1
Wall surfaces
19
Wall surfaces
2
Surface collection
3
Wall surfaces
1
Wall surfaces
3
Wall surfaces
5
Wall surfaces
2
Wall surfaces
8
Wall surfaces
18
Wall surfaces
1
Wall surfaces
2
Wall surfaces
34
Wall surfaces
2
Wall surfaces
1
Wall surfaces
2
Rubble mound
3
Rubble mound
1
Rubble mound (painted tablets)
Wall surfaces
1
Interior of rockshelter
1
Special collection, terrace surfaces
1
5
Hearth (?)
1
Exterior of rockshelter
9
1
Special collection 9, hillside context
1
3
Interior of chullpa
1
Wall surfaces
2
Wall surfaces
3
Wall surfaces
1
Wall surfaces
4
Wall surfaces
4
Other
545
Lot District
Site no. Sector

30
30
30
30
32
A
32
A
32
A
32
A
32
B
32
B
32
B
32
B
78
79
80
81
82
83
84
85
86
87
88
89
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
90
91
92
93
94
95
96
97
98
99
100
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
32
32
32
32
32
32
32
32
32
32
32
B
C
C
C
C
C
D
D
D
D
E
101 YA
32
102 YA
32
103 YA
32
104 YA
32
105 YA
32
106 YA
32
107 YA
32
108 YA
33
109 YA
33
110 YA
33

Provenience
Wall surfaces
1
Wall surfaces
4
Wall fill
1
Wall surfaces
1
General surface collection, terraces
25
9
3
21
18
11
3
(malacological sample)
8
44
25
1
2
12
24
12
1
1
25
General surface collection,
mounded, flat area
29
mounded, flat area
2
mounded, flat area
12
mounded, flat area
24
mounded, flat area
31
and slope
26
and slope
6
and slope
General surface collection, top of
2
hill
General surface collection, top of
21
hill
3
mounded, flat area
Other
546
Lot District
111 YA
Site no. Sector

33
B
112 YA
33
113 YA
32
114 YA
32
115 YA
32
116 YA
32
117
118
119
120
121
122
123
124
125
126
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
34
34
34
34
34
34
34
34
34
34
A
A
A
B
B
B
C1
C1
C1
C2
127 YA
34
C2
128 YA
34
C2
129
130
131
132
133
134
135
136
YA
YA
YA
YA
YA
YA
YA
YA
34
34
34
40
40
40
C2A
C2A
C2A
32
137 YA
32
138 YA
32
139 YA
32
140 YA
141 YA
142 YA
41
41
41
X
X
W

Provenience
5
mounded, flat area
52
mounded, flat area
8
and slope
1
and slope
11
and slope
12
and slope
General surface collection, corral
3
18
1
Rockshelter and exterior scatter
3
20
1
5
89
15
2
mounded (?) area
23
mounded (?) area
20
mounded (?) area
Ash lens
8
Ash lens
40
Ash lens
38
7
5
1
Special collection, hillside
1
General surface collection, plaza of
10
modern chapel
8
modern chapel
1
modern chapel
General surface collection, around
43
modern cross
Wall fill
13
Street surface
3
Wall fill
13
Other
547
Lot District
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
Site no. Sector

41
W
41
W
41
V
41
V
41
U
41
U
41
U
41
T
41
S
41
S
41
S
R
41
R
41
R
41
R
41
Q
41
Q
41
41
P
41
P
41
P
41
O
41
O
41
N
41
N
41
N
41
N
41
M
41
M
41
L
K
41
41
J
41
J
41
I
41
I
41
H
41
H
41
G
41
G
41
G
41
F
41
E
41
E
41
E
Provenience
Street surface
Street surface
Wall fill
Street surface
Wall fill
Street surface
Wall fill
Wall fill
Wall fill
Street surface
Wall fill
Wall fill
Street surface
Wall fill
Wall fill
Wall fill
Street surface
Wall fill
Street surface
Wall fill
Wall fill
Street surface
Wall fill
Street surface
Wall fill
Street surface
Wall fill
Street surface
Wall fill
Wall fill
Wall fill
Street surface
Wall fill
Street surface
Wall fill
Street surface
Wall fill
Street surface
Wall fill
Wall fill
Wall fill
Street surface
Street surface
Ceramics
Lithics
10
2
6
8
24
9
1
20
22
6
1
15
3
2
1
15
11
15
1
1
12
1
13
25
20
8
15
22
8
14
16
10
9
4
24
9
22
1
1
17
17
3
1
Metals
Other
548
Lot District
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
Site no. Sector

41
E
41
D
41
D
41
C
41
C
41
C
41
C
41
B
41
B
41
A
41
A
42
42
43
44
45
A
202 YA
203 YA
45
45
A
A
204
205
206
207
208
209
210
211
YA
YA
YA
YA
YA
YA
YA
YA
45
45
46
46
46
47
48
49
B
B
212 YA
213 YA
214 YA
52
53
54
215 YA
54
216 YA
54
217 YA
54
218 YA
54
219 YA
54
220 YA
54
A1
A1

Provenience
Street surface
5
Wall fill
20
Street surface
2
Wall fill
20
Street surface
11
Wall fill
1
Wall fill
1
Wall fill
4
Street surface
8
Wall fill
16
Street surface
22
Rubble mound
10
Rubble mound
4
Rubble mound
16
Rockshelter/burials
7
33
around structures 6 and 7
Wall fill, structure 6
2
4
around structures 8
Rubble mound
7
Rubble mound
4
7
8
1
2
Collapsed wall fill, structure 1
8
General surface collection around
1
looted tombs
11
Corrals
5
General surface collection around
9
looted tombs
General surface collection on top
29
of hill
General surface collection on top
3
of hill
General surface collection, west
22
slope, terrace
4
slope, terrace
24
slope, terrace
96
slope
Other
549
Lot District
221 YA
Site no. Sector

54
C
222 YA
54
223 YA
54
224 YA
54
225 YA
54
226 YA
54
227 YA
54
228 YA
54
229 YA
54
230
231
232
233
234
235
236
237
238
239
240
241
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
54
54
54
54
55
55
56
56
57
58
58
58
C
C
D
D
242 YA
58
243
244
245
246
AR
AR
AR
YA
50
247
248
249
250
251
252
253
YA
YA
YA
YA
YA
YA
YA
50
50
50
50
50
50
50

Provenience
3
slope
5
slope
11
around structures 2 and 3
2
around structure 4
2
around structure 4
4
instrument station 3
21
11
1
Interior of structure 6
12
Interior of structure 6
5
27
3
8
8
2
Exterior (east side) of rockshelter
31
Terrace fill
4
Terrace fill
1
General surface collection (clay)
Kiln wall (fire-reddened clay
plaster)
Kiln wall (fire-reddened clay
plaster)
Special collection (clay)
Surface collection, exterior of
1
structure 2
11
8
Terrace surface, structure 3
1
18
2
4
1
Other
1
4
1
1
1
1
550
Lot District
254 YA
255 YA
Site no. Sector

50
50
256 YA
50
257 YA
50
258 YA
50
259 YA
260 YA
50
50
261 YA
262 YA
50
50
263
264
265
266
267
268
269
YA
YA
YA
YA
YA
YA
YA
50
50
50
50
50
50
50
270 YA
50
271
272
273
274
YA
YA
YA
YA
50
50
50
50
275 YA
50
276 YA
50
277 YA
278 YA
50
50
279 YA
280 YA
50
50
281 YA
282 YA
50
50
283 YA
50
Provenience
Surface collection, interior of
structure 48
structure 48
structure 53
structure 59
structure 60
structure 67
Wall fill, structure 67B/C
structure 60
structure 77
structure 97
Surface collection, surface of
terrace to east of structure 102
Surface collection, surface of
terrace of structure 104
structure 112
structure 117
structure 119
structure 125
Ceramics
Lithics
15
8
6
20
12
1
1
2
3
13
2
2
1
2
1
1
2
1
2
2
1
2
4
4
1
1
3
7
1
1
Metals
Other
551
Lot District
284 YA
285 YA
Site no. Sector

50
50
286 YA
50
287 YA
50
288 YA
50
289
290
291
292
YA
YA
YA
YA
50
50
50
50
293 YA
50
294 YA
50
295
296
297
298
299
300
301
302
303
YA
YA
YA
YA
YA
YA
YA
YA
YA
50
50
50
50
50
50
50
50
50
304
305
306
307
308
309
310
311
312
313
314
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
50
50
50
50
50
50
50
50
50
50
50
315 YA
316 YA
50
50
317 YA
318 YA
50
50
Provenience
structure 125A
structure 130
structure 144
structure 151
structure 157
structure 158
structure 158
Structure 93, terrace fill
Interior, structure 146 (recovered
from wall niche)
Terrace fill
Terrace fill
Wall fill
Terrace fill
General surface collection, terrace
structure 64
Subterranean collared tomb
(interior)
Ceramics
Lithics
2
2
40
4
6
2
3
2
2
2
5
1
4
6
3
2
6
4
6
2
5
7
7
15
5
15
4
4
4
14
1
15
3
1
1
Metals
Other
552
Lot District
Site no. Sector

50
50
50
50
50
50
319
320
321
322
323
324
325
YA
YA
YA
YA
YA
YA
YA
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
64
66
66
66
67
67
67
73
73
73
73
74
76
76
76
78
78
78
78
78
78
78
80
81
82
351 YA
82
352 YA
82
353 YA
82
354
355
356
357
83
83
83
83
YA
YA
YA
YA

Provenience
1
7
1
1
1
1
7
surfaces
Surface collection, interior of corral
8
Exterior of rockshelter, north slope
5
10
49
7
30
1
3
42
4
19
4
3
1
3
13
1
1
7
17
14
6
7
8
2
of rockshelter
1
of rockshelter
12
of rockshelter
1
rockshelter
Erosive slope on exterior of cave
1
14
7
Surface collection, interior of cave
1
Other
553
Lot District
358 YA
Site no. Sector

84
359 YA
84
360 YA
84
361 YA
84
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
YA
85
85
86
86
87
87
87
87
88
88
88
88
88
88
89
A
A
B
B
A
A
A
B
B
B
A
377 YA
89
378 YA
89
379 YA
89
380
381
382
383
384
385
386
387
388
389
390
391
392
89
34
34
59
61
90
B
D
D
93
93
93
93
93
93
A
B
B
B
A
B
YA
YA
YA
YA
CO
YA
YA
YA
YA
YA
YA
YA
YA

Provenience
General surface collection, nucleus
3
A
17
A
4
B
1
B
15
6
55
13
15
5
7
4
5
8
3
58
22
1
4
rockshelter
1
rockshelter
Erosive slope on exterior of
30
rockshelter
Erosive slope on exterior of
3
rockshelter
Surface collection, corrals
9
40
10
Terrace fill
28
55
22
Special collection, obsidian sample
4
16
6
3
Interior, tomb 1
9
Interior, tomb 2
1
Interior, tomb 3
15
Other
554
Lot District
393 YA
Site no. Sector

93
C
394 YA
94
395 YA
94
396 YA
94
397 CO
100
398 CO
399 CO
100
100
C
E
400
401
402
403
404
405
406
CO
CO
CO
CO
CO
CO
CO
100
100
100
100
100
100
100
F
F
F
G
G
G
G
407 CO
100
408
409
410
411
CO
CO
CO
CO
100
100
100
100
G
H
I
J
412
413
414
415
416
417
418
419
CO
CO
CO
CO
CO
CO
CO
CO
100
100
100
100
100
100
100
100
J
J
K
K
K
K
K
K
420 CO
421 CO
422 CO
100
100
100
G
L
M
423 CO
424 CO
425 CO
100
100
100
M
M
M

Provenience
General surface collection, east
10
slope of site
9
rockshelter
11
rockshelter
39
rockshelter
1
disturbed tomb
Interior of disturbed tomb
5
12
disturbed tomb
Interior of tomb 1
2
Top of tomb 1
2
Top of tomb 1
1
Wall fill, chullpa 1
3
1
1
Interior of chullpa 1, sector G
(vegetal fiber cord)
Interior of chullpa 1, sector G
(human hair braid)
2
Interior of tomb 2
2
Exterior of tomb2
6
17
terraces surfaces
1
1
8
2
Exterior of tomb 1
1
Top of tombs 3 and 4
7
17
Exterior of tomb 1 (vegetal fiber
cord)
Interior of chullpa 1 (textile)
28
General surface collection (vegetal
fiber cord)
Exterior of tomb
1
1
50
Other
3
2
1
1
1
555
Lot District
426 CO
Site no. Sector

100
M
427 CO
100
428 CO
100
429 CO
430 CO
100
100
N
N
431
432
433
434
435
436
CO
CO
CO
CO
CO
CO
100
100
101
102
102
102
N
N
437 CO
102
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
103
104
104
105
105
105
106
106
106
106
106
106
106
106
106
109
110
110
111
111
111
114
114
118
118
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
A
A
B
A
A
B
B
B
B
B
B
C
B
B

Provenience
Exterior of tomb (vegetal fiber
cord)
Surface collection around probable
7
chapel
Surface collection around probable
5
chapel
Terrace fill, southern slope of hill
35
Surface collection, southern slope
of hill
Terrace fill, eastern slope of hill
44
Terrace fill, north slope of hill
34
2
Erosive slope outside of rockshelter
1
3
1
rockshelter
22
rockshelter
24
3
3
6
9
37
2
7
Profile
1
Profile
17
Profile
17
Profile
2
Profile
11
Profile
2
25
11
5
47
4
1
8
Surface collection, exterior of tomb
2
4
Interior of tomb 1
1
Interior of tomb 1 (vegetal fiber
cord)
Other
556
Lot District
463 CO
Site no. Sector

119
464 CO
465 CO
120
120
466 CO
120
467 CO
468 CO
121
121
A
A
469 CO
121
470 CO
471 CO
472 CO
114
114
98
473 CO
474 CO
98
98
A
A
475
476
477
478
CO
CO
CO
CO
98
98
98
98
A
A
A
A
479 CO
98
480 CO
98
481
482
483
484
485
486
487
488
CO
CO
CO
CO
CO
CO
CO
CO
98
98
98
98
98
98
98
98
A
A
A
A
A
A
A
A
489 CO
98
490 CO
98
491 CO
492 CO
493 CO
98
98
98
A
A
A

Provenience
1
1
Interior of tomb 2
1
1
2
cord)
Rubble mound fill
42
Rubble mound fill (painted tablets-ceramic)
9
structure 1
1
Interior of tomb (textile)
1
5
Interior of tomb 10 (textile)
wrapping)
Top of tomb 4
1
Interior of tomb 6
2
11 (textile)
cord)
12 (textile)
wrapping)
wrapping)
wrapping)
Other
1
1
1
1
1
1
2
2
1
2
1
2
1
1
1
1
1
1
1
557
Lot District
494 CO
Site no. Sector

98
A
495 CO
98
496
497
498
499
500
501
502
503
504
505
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
98
98
98
98
98
98
98
98
98
98
A
A
A
A
A
A
A
A
A
A
506
507
508
509
510
511
512
513
514
515
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
98
118
122
123
169
169
169
169
169
169
A
B
D
516 CO
517 CO
169
169
518 CO
169
519 C0
169
520
521
522
523
524
CO
CO
CO
CO
CO
169
169
169
128
134
525 CO
134
526 CO
134
527 CO
131

Provenience
13 (textile)
12
13
Interior of tomb 15
1
Interior of tomb 19
2
Interior of tomb 24
1
Wall fill, tomb 27
1
Top of tomb 28
1
Top of tomb 29
5
2
38
1
Wall fill, chullpa
1
2
Terrace fill
2
Wall surfaces
77
Rubble mound fill
25
Wall fill, chullpa1
1
Fill around offering box 2
12
3
chullpa (painted tablets)
Offering box 1 (painted tablets)
Surface collection around box 2
(painted tablets)
Wall fill around box 2 (painted
tablets)
Surface collection around box 3
(painted tablets)
Interior, box 4 (painted tablets)
Wall fill, chullpa 2 (painted tablet)
Wall fill
4
Exterior of collared tomb
4
11
structure 10
3
structure 10
1
structure 10
2
chullpa
Other
1
1
1
3
30
7
8
18
1
558
Lot District
Site no. Sector

98
B
127
127
140
528
529
530
531
CO
CO
CO
CO
532
533
534
535
536
537
538
539
540
CO
CO
CO
CO
CO
CO
CO
CO
CO
145
147
147
147
147
147
147
147
147
541
542
543
544
545
546
547
548
549
550
551
552
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
147
147
98
148
148
148
148
148
149
149
149
163
553 CO
163
554
555
556
557
558
559
560
561
562
563
564
565
566
567
163
163
163
163
165
165
165
166
167
167
167
167
167
168
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
A
B
C
C
C
C
C
C
D
D
B
A
A
C
D
E
A
B
B

Provenience
Wall fill
21
68
3
structure 3
Erosive slope outside of cave
1
Exterior of corral
2
Wall surfaces
26
Wall surfaces
4
Exterior of tomb 1
3
Exterior of tomb 2
1
Exterior of tomb 2 (textile)
Exterior of tomb 2 (textile)
Exterior of tomb 2 (vegetal fiber
cord)
Terrace surface
4
Wall surfaces
5
Interior of tomb 6
2
Wall fill (painted tablets)
Wall surfaces
15
Wall surfaces
2
Wall surfaces
7
Wall surfaces
1
Wall surfaces
26
Wall surfaces
40
12
9
structure 57
14
structure 56
5
Wall surfaces
10
4
11
Wall surfaces
15
Interior of pukara, central structure
1
Exterior of chullpa
1
Wall surfaces
62
Exterior of tomb 3
2
Wall surfaces
7
Wall fill (painted tablet)
Wall surfaces
57
Wall surfaces
13
Wall surfaces
38
Other
1
1
1
559
Lot District
568
569
570
571
572
573
574
575
CO
CO
CO
CO
CO
CO
CO
CO
Site no. Sector

168
168
168
168
168
168
164
164
576 CO
164
577
578
579
580
CO
CO
CO
CO
164
164
164
163
581
582
583
584
585
586
587
588
589
590
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
164
164
150
150
150
150
150
150
150
151
591 CO
151
592 CO
151
593 CO
594 CO
595 CO
152
152
152
596
597
598
599
600
601
602
603
604
153
154
154
154
154
158
158
158
156
CO
CO
CO
CO
CO
CO
CO
CO
CO
A
A
B
B
C
C
D
B
B
C
C
D

Provenience
Interior of tomb 3
1
Interior of pukara, cemetery
7
Exterior of tomb 4
2
Interior of tomb 1
5
Wall surfaces of fortification walls
5
Interior of tomb 2
2
Wall fill (painted tablets--ceramic)
9
residential area
7
cemetery
Wall fill
24
Wall surfaces
7
Wall surfaces
6
26
cemetery
Wall fill
20
33
9
7
Rubble mound fill
3
Wall fill
1
Eastern slope of sector, profile
12
Eastern slope of sector, profile
15
15
32
abandoned terraces
1
abandoned terraces
2
abandoned terraces
Exterior of tomb, slope
6
Exterior of tomb, slope
22
Interior of tomb and associated
11
scatter
Peak of Yuracc Ccacca
5
Exterior of chullpas 8-11
2
Exterior of chullpa 12
3
Exterior of chullpa 13 (textile)
Exterior of tomb 16
2
Eastern slope of Pumachiri
3
Peak of Pumachiri
3
Peak of Pumachiri
1
20
Other
560
Lot District
605
606
607
608
609
610
611
612
613
614
615
616
617
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
YA
Site no. Sector

156
156
157
A
157
A
157
A
159
159
159
161
A
161
B
161
D
161
F
34
C
Provenience
Street surface
Street surface
Street surface
Street surface
Ash lens
Totals:*
Ceramics
Lithics
Metals
Other
136
1
7
5
15
3
17
18
3
11
7
3
3
1
3847
1673
*Totals are from original field counts. After reconstructing some vessels in the laboratory, the ceramic count
lowered to 3806, as reflected in the ceramic analysis statistics (Appendix 1).
561
Appendix E: Ceramic Distributions by Site
Table E.1 lists the distribution of ceramics by style. Table E.2 aggregates the ceramic
data by phase definition (i.e., Collagua I and II = LIP; Collagua Inka and Inka=LH; and
diagnostic non-local sherds [imports] are assigned to phases). Site number M.E. indicates
special collections (not declared as sites, but collections taken)
Codes for Table E.1

CHI:
MH:
CO1:
CO2:
CO3:
COI:
INK:
IMP:
COL:
MOD:
UKN:
UTI:
Chiquero (Formative)
Middle Horizon
Collagua I
Collagua II
Collagua III
Collagua Inka
Inka
Import/Exotic (various time periods)
Colonial Period
Republican Period
Unclassified/Small Fragments
Utilitarian (prehispanic, non-diagnostic)
562
Site No.
1
1
1
1
1 Total
2
2
2 Total
3
3 Total
4
4
4
4
4
4 Total
6
6 Total
7
7
7 Total
8
8
8
8 Total
9
9
9
9
9 Total
10
10 Total
12
12
12
12 Total
14
14
14
14
14
14
14
14
14
14
14
14
Sector Lot CHI MH CO1 CO2 CO3 COI INK IMP COL MOD UKN UTI Total
1
1
1
2
1
1
1
1
2
6
3
1
2
1
5
4
3
16
4
1
1
1
3
0
1
2
3
0
2
2
0
6
4
5
1
26
A
8
2
2
1
3
15
23
B
9
5
5
2
0
0
2
1
3
0
0
0
0
20
0
28
11
1
1
2
0
0
0
0
0
1
0
0
0
0
1
0
2
12
1
1
4
2
8
13
1
1
1
1
4
14
2
1
1
4
15
1
1
16
1
1
1
2
5
0
1
1
1
2
8
1
0
3
0
4
1
22
B
20
3
1
4
0
3
0
0
0
1
0
0
0
0
0
0
4
22
2
7
4
1
1
4
1
25
45
24
1
1
0
2
0
7
4
1
0
1
4
1
26
0
46
C1
27
1
1
1
1
4
C2
26
12
12
C3
25
5
1
1
38
1
46
1
5
1
0
1
1
0
0
1
0
51
1
62
A
30
1
4
1
1
6
13
B
31
2
2
5
1
7
1
18
C
32
25
2
1
1
29
D
34
1
1
2
25
5
4
3
3
44
2
2
1
2
2
59
7
0
7
0
17
5
104
A
36
5
1
7
13
0
0
0
0
0
5
0
0
1
0
7
0
13
A
40
18
2
2
14
36
B
38
6
1
1
3
11
B
39
1
1
0 24
0
0
0
4
0
0
3
0
17
0
48
A
56
1
5
2
8
14
1
31
A
57
2
2
B
49
1
1
1
3
C
50
2
1
4
7
C
51
1
1
D
45
1
1
2
1
2
1
10
18
D1
46
1
1
2
D2
47
2
2
4
E
41
1
1
3
3
3
2
1
16
30
E
42
1
2
4
7
E
43
2
1
2
5
E
53
3
1
1
1
8
2
16
563
Site No.
14 Total
15 Total
17 Total
18 Total
21 Total
25 Total
26 Total
27 Total
28 Total
29 Total
30 Total
32 Total
33 Total
14 E3
52
1
1
5
7
14 E4
54
1
1
2
14 E5
55
3
3
4
3
3
14
9
21
1
0
7
3
66
7
138
15
60
1
2
3
0
0
0
0
0
1
0
0
0
0
2
0
3
17
63
1
1
0
0
0
0
0
0
0
0
0
0
1
0
1
18
64
1
1
0
0
0
0
0
0
0
0
0
0
1
0
1
21
66
1
4
5
21
67
1
1
0
0
0
0
0
0
0
0
1
0
5
0
6
25
71
1
2
3
25
72
1
1
0
0
0
0
0
1
0
0
0
2
1
0
4
26
73
1
1
2
0
0
0
0
0
1
0
0
0
0
1
0
2
27
74
1
2
3
0
0
0
0
0
0
0
0
1
0
2
0
3
28
75
1
1
2
0
0
0
0
0
1
0
0
0
0
1
0
2
29
76
1
1
2
4
0
0
0
1
0
1
0
0
0
0
2
0
4
30
77
1
3
4
30
78
1
1
30
79
1
1
2
4
30
80
1
1
30
81
1
1
1
1
1
1
0
1
0
0
0
0
6
0
11
32 A
82 14
1
1
10
26
32 B
86
4
1
8
1
1
3
18
32 B
87
1
1
2
32 C
91 23
1
1
8
1
1
8
43
32 C
94
1
1
2
32 D
96 13
1
1
6
2
23
32 D
99
1
1
2
32 E
100
7
1
1
1
10
2
3
25
32 E
136
3
4
1
1
1
10
32 E
138
2
2
32 E
139 10
1
1
15
4
1
2
6
1
41
32 F
105 16
6
1
8
31
32 G
113
2
1
2
3
8
96
3
4
3
2
59
9
1
11
1
43
1
233
33 A
108
1
1
2
33 B
110
1
2
3
0
0
0
0
0
0
0
0
0
2
3
0
5
34 A
117
1
1
1
3
564
Site No.
34
34
34
34
34
34
Sector
B
C1
C2
C2
C2A
C2A
34 Total
40
40 Total
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
A
A
B
B
C
C
C
D
D
E
E
E
F
G
G
G
H
H
I
I
J
J
K
L
M
M
N
N
N
N
O
O
P
P
P
Q
Q
R
R
Lot CHI MH CO1 CO2 CO3 COI INK IMP COL MOD UKN UTI Total
120
2
1
3
123
2
3
5
126
1
1
2
127
1
1
129
1
1
1
1
2
6
131
1
1
3
1
0
1
0
2
0
0
3
2
9
0
21
132
2
3
2
7
0
0
0
0
0
2
0
0
3
0
2
0
7
195
1
1
7
3
3
1
16
196
1
1
8
2
5
3
20
193
1
3
4
194
1
4
1
1
7
189
2
10
5
2
19
190
1
6
1
1
3
12
192
1
1
187
2
5
1
7
1
5
21
188
1
1
2
183
1
7
5
3
1
17
184
4
4
186
3
1
1
5
182
1
7
2
3
4
17
179
1
1
1
8
2
4
4
1
22
180
1
1
181
1
1
177
1
13
2
2
1
4
1
24
178
1
1
2
1
1
3
9
175
1
2
1
2
1
2
9
176
1
3
4
173
8
5
1
1
15
174
1
1
1
4
1
1
9
172
1
2
2
4
5
14
171
1
2
3
1
7
169
2
4
2
2
1
2
1
14
170
7
2
6
4
19
165
1
1
11
13
166
1
1
6
1
10
3
4
1
27
167
9
3
4
2
1
19
168
3
1
4
8
163
4
7
1
12
164
1
1
160
1
5
7
2
15
161
1
1
162
1
1
158
2
12
1
15
159
1
2
7
1
11
154
3
1
4
2
6
16
155
1
1
2
565
Site No.
41
41
41
41
41
41
41
41
41
41
41
41
41
41
Sector
R
S
S
S
T
U
U
V
V
W
W
W
X
X
41 Total
42
42 Total
43
43 Total
44
44 Total
45
45
45
45
A
A
A
B
45 Total
46
46 A1
46 A1
46 Total
47
47 Total
48
48 Total
49
49 Total
50
50
50
50
50
50
50
50
50
50
50
50
156
2
2
151
6
1
9
7
23
152
1
1
2
1
2
7
153
1
1
150
3
3
9
3
1
19
147
4
1
11
1
3
2
22
148
3
5
1
1
10
145
1
4
1
6
146
2
5
2
9
142
2
6
1
2
1
12
143
1
1
7
2
11
144
2
2
140
1
10
1
1
13
141
2
2
1
8
10
16
6 168 35
1 204
20
92 12
573
197
1
1
1
7
10
0
1
0
0
1
1
0
0
0
0
7
0
10
199
3
12
15
3
0
0
0
0
0
0
0
0
0
12
0
15
200
1
1
4
6
0
0
1
0
0
0
0
0
1
0
4
0
6
201
4
2
2
17
1
3
3
2
34
202
1
1
1
3
203
3
1
4
204
2
1
2
2
7
2
0
4
3
4
22
1
0
3
0
7
2
48
206
1
1
1
1
3
7
207
1
7
8
208
1
1
2
1
0
0
0
1
0
0
1
0
11
0
16
209
2
2
0
0
0
2
0
0
0
0
0
0
0
0
2
210
1
7
8
0
0
0
1
0
7
0
0
0
0
0
0
8
211
1
1
0
0
0
0
0
1
0
0
0
0
0
0
1
247
1
5
1
2
2
11
248
1
5
2
8
249
1
1
250
2
2
251
2
2
252
4
4
253
1
1
254
1
9
1
3
14
255
2
6
8
256
1
3
1
1
6
257
3
11
2
1
2
19
258
3
4
1
1
1
1
11
566
Site No.
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
259
1
1
2
260
1
1
261
2
1
3
262
2
2
263
1
8
1
2
12
264
1
1
2
265
1
1
266
1
1
267
2
2
269
1
1
2
270
3
3
271
1
1
272
1
1
2
273
1
1
274
1
1
275
2
2
276
3
1
4
277
5
1
6
278
1
1
279
1
1
280
2
1
1
4
281
2
2
2
6
282
1
1
283
1
1
284
1
1
2
285
2
2
286
6
1
23
10
40
287
3
1
4
288
1
1
1
1
1
1
6
289
1
1
2
290
2
2
291
1
1
2
292
2
2
293
1
1
2
294
1
2
1
1
5
296
1
3
1
5
297
1
1
2
1
5
298
2
1
3
299
2
2
300
6
6
301
1
2
1
4
302
2
2
1
1
6
303
1
1
2
304
5
5
305
2
3
2
7
306
3
1
3
7
307
2
1
5
1
4
2
15
308
5
5
567
Site No.
50
50
50
50
50
50
50
50
50
50
50
50
50
50 Total
52
52 Total
53
53 Total
54
54
54
54
54
54
54
54
54
54
54
54
54
54 Total
55
55 Total
56
56
56 Total
57
57 Total
58
58 Total
59
59 Total
61
61 Total
64
64 Total
66
309
2
1
5
1
2
3
14
311
3
1
4
312
3
1
1
5
313
4
1
9
1
15
315
2
1
1
6
2
2
1
15
317
1
1
318
1
1
319
1
1
320
4
1
1
1
1
8
321
2
2
322
1
1
323
1
1
2
324
1
1
0
1
39
12
9 183 15
0
18
1
63 19
360
212
1
1
3
6
11
0
0
0
1
1
3
0
0
0
0
6
0
11
213
1
2
3
6
0
0
0
1
0
2
0
0
3
0
0
0
6
A
214
1
1
3
3
1
9
A
215
3
4
11
1
2
4
3
28
A
217
1
2
3
1
10
4
3
24
B
219
2
1
2
9
1
6
2
23
C
220
1
1
8
15
10
38
2
9
9
93
C
222
1
1
C
223
2
2
6
1
11
C
224
1
1
2
C
226
2
1
2
5
C
227
1
4
3
9
1
1
19
C
229
1
1
C
230
2
2
2
3
1
1
1
12
D
232
1
3
13
1
3
1
4
26
4
4
19
35
23 101
5
0
17
1
33 12
254
234
1
1
1
5
8
1
0
0
0
0
1
0
0
1
0
5
0
8
236
2
2
237
4
1
1
13
19
0
0
4
1
1
0
0
0
0
0
15
0
21
B
238
1
1
2
4
0
1
0
0
0
0
0
0
1
0
2
0
4
239
1
1
0
0
0
0
0
0
0
0
1
0
0
0
1
383
1 12
1
1
2
2
8
27
1 12
0
1
1
2
0
0
2
0
8
0
27
384
33
7
4
7
1
3
1
56
0
0
0
0
0
33
7
4
7
1
3
1
56
326
6
1
7
0
0
0
0
0
0
0
0
6
0
1
0
7
327
1
2
1
1
5
568
Site No.
66 Total
67
67 Total
74
74 Total
76
76 Total
78
78
78 Total
81
81 Total
82
82 Total
83
83 Total
88
88
88 Total
89
89 Total
90
90 Total
93
93
93
93
93
93
93 Total
94
94 Total
98
98
98
98
98
98
98
98
98
98
98 Total
100
100
100
100
100
0
0
1
0
0
2
0
0
1
0
1
0
5
332
1
1
0
0
0
0
0
0
0
0
1
0
0
0
1
337
1
2
3
0
0
0
0
0
0
0
0
1
2
0
0
3
338
3
3
0
0
0
0
0
0
0
0
3
0
0
0
3
342
1
1
343
5
3
8
0
0
0
0
0
1
0
0
5
0
3
0
9
349
8
8
0
0
0
0
0
0
0
0
8
0
0
0
8
350
1
1
1
3
1
0
0
1
0
0
0
0
0
0
1
0
3
354
1
1
0
0
0
0
0
0
0
0
0
1
0
0
1
A
370
4
4
A
371
1
1
0
0
0
1
0
0
0
0
4
0
0
0
5
B
380
1
8
9
0
0
0
0
0
0
1
0
8
0
0
0
9
385
1
1
17
2
2
23
0
1
0
0
0
1
0
0
17
2
2
0
23
A
387
8
3
1
2
14
A
391
1
1
B
388
1
2
2
1
6
B
390
8
1
9
B
392
1
1
9
2
2
15
C
393
2
3
4
9
0
1
0
2
2
31
8
1
0
0
9
0
54
394
1
1
7
9
0
1
0
0
0
1
0
0
0
0
7
0
9
A
472
1
1
A
475
1
1
A
476
1
1
2
A
495
1
1
1
2
6
11
A
497
1
1
2
A
499
1
1
A
503
1
1
A
504
1
1
1
1
4
A
505
1
1
B
543
1
1
0
1
3
3
3
4
0
0
1
2
8
0
25
B
397
1
1
C
398
3
1
1
5
E
399
10
2
12
F
400
1
1
1
3
6
F
401
3
3
569
Site No.
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100 Total
101
101 Total
102
102 Total
103
103 Total
105
105 Total
106
106
106
106 Total
109
109 Total
111
111 Total
114
114
114 Total
118
118
118 Total
119
119 Total
120
120 Total
Sector
G
G
G
G
H
I
J
J
J
K
K
K
K
L
M
M
N
N
N
N
N
A
B
B
B
B
403
1
2
3
404
1
1
408
1
1
2
420
1
1
409
2
2
410
4
2
6
411
5
1
2
1
8
17
412
1
1
413
1
1
414
1
1
2
3
7
416
1
1
417
2
2
2
1
7
418
2
3
8
2
2
17
421
1
3
2
15
2
1
1
1
1
27
423
1
1
425
1
6
6
25
3
5
2
48
427
7
7
428
1
3
1
5
429
1
4
4
2
8
4
1
1
2
7
34
431
2
2
3
7
3
7
2
1
11
8
46
432
4
7
3
10
1
1
1
4
2
33
11 10
14
32
28
97 11
3
15
2
55 16
294
433
2
2
0
0
0
0
0
0
0
0
2
0
0
0
2
436
1
1
0
0
0
0
1
0
0
0
0
0
0
0
1
438
2
1
1
13
1
1
4
23
0
0
2
1
1
13
1
0
1
0
0
4
23
441
2
1
1
2
6
2
0
0
1
0
1
0
0
0
0
2
0
6
444
1
2
3
446
1
1
449
1
1
1
0
0
0
1
0
0
0
0
0
3
0
5
453
4
6
1
11
0
0
0
0
0
4
0
0
6
0
1
0
11
456
1
1
1
1
4
0
0
0
0
0
1
1
0
1
0
1
0
4
459
1
1
2
460
1
2
1
4
0
0
1
0
0
3
0
0
0
0
2
0
6
461
1
1
507
1
1
0
0
0
2
0
0
0
0
0
0
0
0
2
463
1
1
0
0
1
0
0
0
0
0
0
0
0
0
1
464
1
1
0
0
0
1
0
0
0
0
0
0
0
0
1
570
Site No.
121
121
121 Total
122
122 Total
123
123 Total
127
127
127 Total
128
128 Total
131
131 Total
134
134
134
134 Total
140
140 Total
147
147
147
147
147
147
147
147
147
147 Total
148
148
148
148
148 Total
149
149
149
149 Total
150
150
150
150
150
150
150 Total
151
151 Total
A
467
2
2
4
9
6
11
1
1
4
2
42
B
469
2
1
3
1
1
8
2
4
4
10
6
14
1
1
0
0
5
3
50
D
508
1
1
0
0
0
0
1
0
0
0
0
0
0
0
1
509
1
1
1
3
0
0
0
1
0
1
0
0
0
0
1
0
3
529
2
3
9
2
3
1
1
21
530
3
1
3
35
9
11
3
4
69
0
0
3
3
6
44 11
0
14
0
4
5
90
523
2
1
1
4
0
0
0
2
1
1
0
0
0
0
0
0
4
527
2
2
0
0
0
0
0
0
0
0
0
0
2
0
2
524
2
4
2
3
11
525
2
1
1
4
526
1
1
0
0
0
0
0
5
0
0
4
3
4
0
16
531
1
1
1
3
0
0
0
0
0
1
1
0
0
0
1
0
3
A
533
1
1
B
534
3
2
3
5
6
5
24
C
535
1
1
2
4
C
536
2
1
3
C
537
1
1
C
539
1
1
2
C
540
1
1
D
541
3
1
1
5
D
542
3
1
1
5
0
9
3
4
7
11
1
0
1
0
10
0
46
A
545
3
1
1
3
3
4
15
C
546
1
1
2
D
547
1
1
2
3
7
E
548
1
1
0
4
2
1
4
5
0
0
2
0
7
0
25
A
549
1
1
5
3
3
1
2
9
25
B
550
1
9
4
12
3
7
4
40
B
551
6
2
2
2
12
0
1
2
14
7
21
1
0
7
0
18
6
77
A
583
3
4
1
8
A
584
1
4
1
1
7
B
585
1
1
1
3
B
586
2
2
C
587
4
1
5
1
11
D
589
1
2
2
1
1
1
7
15
6
4
2
8
3
8
0
0
0
0
12
3
46
590
14
2
1
5
1
9
32
0 14
2
1
0
5
0
1
0
0
9
0
32
571
Site No.
152
152
152 Total
153
153 Total
154
154
154
154
154 Total
156
156
156
156 Total
158
158
158 Total
159
159 Total
161
161
161
161
161 Total
163
163
163
163
163
163
163
163 Total
164
164
164
164
164
164
164
164 Total
165
165
165
165 Total
166
166 Total
167
167
593
1
1
595
1
7
8
1
0
0
0
0
0
0
0
0
0
8
0
9
B
596
1
1
3
5
0
0
0
0
0
1
0
0
1
0
3
0
5
B
597
1
2
3
C
598
1
1
1
3
C
599
1
1
2
D
600
2
2
0
2
0
4
0
2
0
0
0
1
1
0
10
604
1
1
605
1
1
606
1
1
1
9
12
0
0
0
1
0
2
1
0
1
0
9
0
14
601
1
1
1
3
602
1
2
3
0
0
0
0
2
3
0
0
1
0
0
0
6
610
14
1
2
17
0
0
0
0
0
14
1
0
0
0
2
0
17
A
613
1
2
1
2
5
11
B
614
6
1
7
D
615
1
2
1
4
F
616
1
1
2
0
0
1
1
0
11
2
0
2
0
2
5
24
552
1
1
1
3
3
9
553
1
1
1
8
2
1
14
554
3
1
1
5
555
1
3
3
1
1
1
10
556
2
1
1
4
557
1
2
5
3
11
580
1
2
1
9
3
3
1
5
2
27
1
3
2
13
10
27
1
0
1
1
15
6
80
575
1
1
5
1
1
1
10
576
1
1
1
3
1
7
577
2
6
4
7
1
1
3
24
578
2
4
1
7
579
3
2
5
581
1
6
4
2
1
7
21
582
5
4
4
11
1
1
2
4
32
0
4
6
17
19
32
3
0
2
2
16
5
106
558
1
1
2
3
3
5
15
559
1
1
560
1
1
0
1
1
4
0
3
0
0
3
0
5
0
17
561
1
3
16
18
7
1
1
13
1
61
1
0
3
16
18
7
1
0
1
0
13
1
61
562
1
1
1
3
563
4
1
3
8
572
Site No.
167
565
9
3
10
9
9
4
15
59
167
566
2
2
3
5
1
13
167 Total
0 13
4
10
11
11
0
0
9
0
24
1
83
168
567
1
2
2
8
9
3
12
37
168
568
1
1
168
569
1
1
5
7
168
570
1
1
168
571
1
2
1
4
168
572
1
1
3
5
168
573
1
1
2
168 Total
1
3
3
10
13
4
0
0
1
0
22
0
57
169
510
2
2
169
511
1 49
5
2
4
1
1
10
73
169
512
7
2
3
3
1
9
25
169
513
1
1
169
514
1
2
1
1
1
5
11
169
515
1
1
2
169
517
1
1
169 Total
2 60
0
7
6
9
0
2
2
1
26
0
115
M.E. 3
5
3
2
5
M.E. 3 Total
0
0
0
0
0
3
0
0
0
0
2
0
5
M.E. 4
6
2
1
4
6
1
14
M.E. 4 Total
0
0
0
0
0
2
1
0
4
0
6
1
14
M.E. 5
7
1
1
M.E. 5 Total
0
0
0
0
0
1
0
0
0
0
0
0
1
M.E. 6
10
1
1
1
3
6
M.E. 6 Total
0
1
0
0
0
0
0
0
1
1
3
0
6
M.E. 7
17
4
7
2
1
1
15
M.E. 7 Total
0
0
0
0
0
4
0
0
7
2
1
1
15
M.E. 8
18
1
2
1
7
2
1
14
M.E. 8 Total
0
1
0
2
1
7
0
0
2
0
1
0
14
M.E. 9
70
1
1
M.E. 9 Total
0
1
0
0
0
0
0
0
0
0
0
0
1
M.E. 10
325
2
2
2
1
7
M.E. 10 Total
0
2
2
0
0
2
0
0
0
0
1
0
7
Grand Total
152 215 152 281 218 1120 130 15 456
58 890 119 3806
573
Codes for Table E.2
FOR:
MH:
LIP:
LH:
COL:
MOD:
UKN:
Chiquero (Formative)
Middle Horizon
Late Intermediate Period
Late Horizon
Colonial Period
Republican Period
Unknown
574
Site No.
1
1
1
1
1 Total
2
2
2 Total
3
3 Total
4
4
4
4
4
4 Total
6
6 Total
7
7
7 Total
8
8
8
8 Total
9
9
9
9
9 Total
10
10 Total
12
12
12
12 Total
14
14
14
14
14
14
14
14
14
14
14
14
Sector Lot FOR MH LIP LH COL MOD UKN Total

1
1
1
2
2
2
2
6
3
3
1
5
4
3
16
4
1
1
1
3
0
1
5
4
6
4
6
26
A
8
2
2
4
15
23
B
9
5
5
2
0
2
4
0
0
20
28
11
1
1
2
0
0
0
1
0
0
1
2
12
1
5
2
8
13
1
2
1
4
14
2
1
1
4
15
1
1
16
1
2
2
5
0
1
2
11
3
0
5
22
B
20
3
1
4
0
3
0
1
0
0
0
4
22
2
7
6
4
1
25
45
24
1
1
0
2
7
6
4
1
26
46
C1
27
1
1
1
1
4
C2
26
12
12
C3
25
5
2
39
46
1
5
1
2
1
0
52
62
A
30
1
5
1
6
13
B
31
2
2
6
8
18
C
32
25
2
2
29
D
34
1
1
32
4
6
44
2
2
3
68
7
0
22 104
A
36
5
1
7
13
0
0
0
5
1
0
7
13
A
40
18
2
2
14
36
B
38
6
1
1
3
11
B
39
1
1
0 24
0
4
3
0
17
48
A
56
6
10
15
31
A
57
2
2
B
49
2
1
3
C
50
2
1
4
7
C
51
1
1
D
45
1
4
2
1
10
18
D1
46
1
1
2
D2
47
2
2
4
E
41
1
4
6
2
1
16
30
E
42
1
2
4
7
E
43
2
1
2
5
E
53
3
1
1
1
10
16
575
Site No.
14 Total
15 Total
17 Total
18 Total
21 Total
25 Total
26 Total
27 Total
28 Total
29 Total
30 Total
32 Total
33 Total

14 E3
52
2
5
7
14 E4
54
1
1
2
14 E5
55
3
3
4
3 17
31
7
3
73 138
15
60
1
2
3
0
0
0
1
0
0
2
3
17
63
1
1
0
0
0
0
0
0
1
1
18
64
1
1
0
0
0
0
0
0
1
1
21
66
1
4
5
21
67
1
1
0
0
0
0
1
0
5
6
25
71
1
2
3
25
72
1
1
0
0
0
1
0
2
1
4
26
73
1
1
2
0
0
0
1
0
0
1
2
27
74
1
2
3
0
0
0
0
1
0
2
3
28
75
1
1
2
0
0
0
1
0
0
1
2
29
76
1
1
2
4
0
0
1
1
0
0
2
4
30
77
1
3
4
30
78
1
1
30
79
1
1
2
4
30
80
1
1
30
81
1
1
1
1
2
1
0
0
6
11
32 A
82
14
1
1
10
26
32 B
86
4
1
9
1
3
18
32 B
87
1
1
2
32 C
91
23
1
1
9
1
8
43
32 C
94
1
1
2
32 D
96
13
1
7
2
23
32 D
99
1
1
2
32 E
100
7
2
13
3
25
32 E
136
3
5
1
1
10
32 E
138
2
2
32 E
139
11
1
1
19
2
7
41
32 F
105
16
6
1
8
31
32 G
113
2
1
2
3
8
97
3
7
70
11
1
44 233
33 A
108
1
1
2
33 B
110
1
2
3
0
0
0
0
0
2
3
5
34 A
117
1
1
1
3
576
Site No.
34
34
34
34
34
34
Sector
B
C1
C2
C2
C2A
C2A
34 Total
40
40 Total
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
A
A
B
B
C
C
C
D
D
E
E
E
F
G
G
G
H
H
I
I
J
J
K
L
M
M
N
N
N
N
O
O
P
P
P
Q
Q
R
R
Lot FOR MH LIP LH COL MOD UKN Total

120
2
1
3
123
2
3
5
126
1
1
2
127
1
1
129
1
1
1
1
2
6
131
1
1
3
1
1
2
3
2
9
21
132
2
3
2
7
0
0
0
2
3
0
2
7
195
1
11
3
1
16
196
2
10
5
3
20
193
1
3
4
194
1
4
1
1
7
189
2
10
5
2
19
190
1
6
1
1
3
12
192
1
1
187
2
6
7
1
5
21
188
1
1
2
183
1
7
5
4
17
184
4
4
186
3
1
1
5
182
1
9
3
4
17
179
1
1
11
4
5
22
180
1
1
181
1
1
177
1
15
2
1
5
24
178
2
3
1
3
9
175
1
3
2
1
2
9
176
4
4
173
8
5
2
15
174
2
6
1
9
172
5
4
5
14
171
1
2
3
1
7
169
2
6
2
1
3
14
170
9
6
4
19
165
1
1
11
13
166
1
1
7
10
3
5
27
167
12
4
2
1
19
168
4
4
8
163
4
7
1
12
164
1
1
160
6
7
2
15
161
1
1
162
1
1
158
2
12
1
15
159
1
2
7
1
11
154
4
4
2
6
16
155
1
1
2
577
Site No.
41
41
41
41
41
41
41
41
41
41
41
41
41
41
Sector
R
S
S
S
T
U
U
V
V
W
W
W
X
X
41 Total
42
42 Total
43
43 Total
44
44 Total
45
45
45
45
A
A
A
B
45 Total
46
46 A1
46 A1
46 Total
47
47 Total
48
48 Total
49
49 Total
50
50
50
50
50
50
50
50
50
50
50
50

156
2
2
151
7
9
7
23
152
2
2
3
7
153
1
1
150
6
9
4
19
147
5
11
1
5
22
148
3
5
1
1
10
145
1
4
1
6
146
2
5
2
9
142
2
6
1
3
12
143
2
7
2
11
144
2
2
140
1
10
1
1
13
141
2
2
1
8 26 210 204
20 104 573
197
1
2
7
10
0
1
0
2
0
0
7
10
199
3
12
15
3
0
0
0
0
0
12
15
200
1
1
4
6
0
0
1
0
1
0
4
6
201
6
20
3
5
34
202
1
1
1
3
203
3
1
4
204
2
3
2
7
2
0
7
27
3
0
9
48
206
1
1
1
1
3
7
207
1
7
8
208
1
1
2
1
0
1
1
0
11
16
209
2
2
0
0
2
0
0
0
0
2
210
1
7
8
0
0
1
7
0
0
0
8
211
1
1
0
0
0
1
0
0
0
1
247
1
6
2
2
11
248
1
5
2
8
249
1
1
250
2
2
251
2
2
252
4
4
253
1
1
254
1
10
3
14
255
2
6
8
256
5
1
6
257
3
13
1
2
19
258
3
5
1
2
11
578
Site No.
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50

259
1
1
2
260
1
1
261
2
1
3
262
2
2
263
9
1
2
12
264
1
1
2
265
1
1
266
1
1
267
2
2
269
1
1
2
270
3
3
271
1
1
272
2
2
273
1
1
274
1
1
275
2
2
276
3
1
4
277
5
1
6
278
1
1
279
1
1
280
2
1
1
4
281
2
2
2
6
282
1
1
283
1
1
284
1
1
2
285
2
2
286
7
23
10
40
287
3
1
4
288
1
1
3
1
6
289
1
1
2
290
2
2
291
1
1
2
292
2
2
293
2
2
294
1
2
1
1
5
296
1
3
1
5
297
1
3
1
5
298
2
1
3
299
2
2
300
6
6
301
1
2
1
4
302
2
2
1
1
6
303
1
1
2
304
5
5
305
2
3
2
7
306
4
3
7
307
2
6
1
6
15
308
5
5
579
Site No.
50
50
50
50
50
50
50
50
50
50
50
50
50
50 Total
52
52 Total
53
53 Total
54
54
54
54
54
54
54
54
54
54
54
54
54
54 Total
55
55 Total
56
56
56 Total
57
57 Total
58
58 Total
59
59 Total
61
61 Total
64
64 Total
66

309
2
7
2
3
14
311
3
1
4
312
4
1
5
313
4
10
1
15
315
3
9
3
15
317
1
1
318
1
1
319
1
1
320
5
1
1
1
8
321
2
2
322
1
1
323
1
1
2
324
1
1
0
1 51 207
18
1
82 360
212
1
4
6
11
0
0
1
4
0
0
6
11
213
1
2
3
6
0
0
1
2
3
0
0
6
A
214
2
3
4
9
A
215
3
16
2
7
28
A
217
1
5
11
7
24
B
219
2
1
2
9
1
8
23
C
220
1
1 23
50
9
9
93
C
222
1
1
C
223
2
8
1
11
C
224
1
1
2
C
226
2
1
2
5
C
227
5
12
1
1
19
C
229
1
1
C
230
2
4
4
2
12
D
232
4
14
3
1
4
26
4
4 54 129
17
1
45 254
234
1
1
1
5
8
1
0
0
1
1
0
5
8
236
2
2
237
5
1
13
19
0
0
5
1
0
0
15
21
B
238
1
1
2
4
0
1
0
0
1
0
2
4
239
1
1
0
0
0
0
1
0
0
1
383
1 12
1
3
2
8
27
1 12
1
3
2
0
8
27
384
1
43
7
1
4
56
0
1
0
43
7
1
4
56
326
6
1
7
0
0
0
0
6
0
1
7
327
1
2
1
1
5
580
Site No.
66 Total
67
67 Total
74
74 Total
76
76 Total
78
78
78 Total
81
81 Total
82
82 Total
83
83 Total
88
88
88 Total
89
89 Total
90
90 Total
93
93
93
93
93
93
93 Total
94
94 Total
98
98
98
98
98
98
98
98
98
98
98 Total
100
100
100
100
100

0
0
1
2
1
0
1
5
332
1
1
0
0
0
0
1
0
0
1
337
1
2
3
0
0
0
0
1
2
0
3
338
3
3
0
0
0
0
3
0
0
3
342
1
1
343
5
3
8
0
0
0
1
5
0
3
9
349
8
8
0
0
0
0
8
0
0
8
350
1
1
1
3
1
0
1
0
0
0
1
3
354
1
1
0
0
0
0
0
1
0
1
A
370
4
4
A
371
1
1
0
0
1
0
4
0
0
5
B
380
1
8
9
0
0
0
1
8
0
0
9
385
1
1
17
2
2
23
0
1
0
1
17
2
2
23
A
387
12
2
14
A
391
1
1
B
388
5
1
6
B
390
9
9
B
392
1
12
2
15
C
393
2
3
4
9
0
1
2
42
0
0
9
54
394
1
1
7
9
0
1
0
1
0
0
7
9
A
472
1
1
A
475
1
1
A
476
1
1
2
A
495
1
2
2
6
11
A
497
1
1
2
A
499
1
1
A
503
1
1
A
504
1
2
1
4
A
505
1
1
B
543
1
1
0
1
6
7
1
2
8
25
B
397
1
1
C
398
4
1
5
E
399
10
2
12
F
400
1
1
1
3
6
F
401
3
3
581
Site No.
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100 Total
101
101 Total
102
102 Total
103
103 Total
105
105 Total
106
106
106
106 Total
109
109 Total
111
111 Total
114
114
114 Total
118
118
118 Total
119
119 Total
120
120 Total
Sector
G
G
G
G
H
I
J
J
J
K
K
K
K
L
M
M
N
N
N
N
N
A
B
B
B
B

403
1
2
3
404
1
1
408
1
1
2
420
1
1
409
2
2
410
4
2
6
411
5
3
1
8
17
412
1
1
413
1
1
414
1
1
2
3
7
416
1
1
417
2
2
3
7
418
2
13
2
17
421
1
4
19
1
2
27
423
1
1
425
7
34
7
48
427
7
7
428
1
4
5
429
1
5
6
13
2
7
34
431
2
2 10
10
2
1
19
46
432
11
14
1
7
33
11 12 46 136
15
2
72 294
433
2
2
0
0
0
0
2
0
0
2
436
1
1
0
0
0
1
0
0
0
1
438
3
15
1
4
23
0
0
3
15
1
0
4
23
441
2
1
1
2
6
2
0
1
1
0
0
2
6
444
1
2
3
446
1
1
449
1
1
1
0
0
1
0
0
3
5
453
4
6
1
11
0
0
0
4
6
0
1
11
456
2
1
1
4
0
0
0
2
1
0
1
4
459
1
1
2
460
1
2
1
4
0
0
1
3
0
0
2
6
461
1
1
507
1
1
0
0
2
0
0
0
0
2
463
1
1
0
0
1
0
0
0
0
1
464
1
1
0
0
1
0
0
0
0
1
582
Site No.
121
121
121 Total
122
122 Total
123
123 Total
127
127
127 Total
128
128 Total
131
131 Total
134
134
134
134 Total
140
140 Total
147
147
147
147
147
147
147
147
147
147 Total
148
148
148
148
148 Total
149
149
149
149 Total
150
150
150
150
150
150
150 Total
151
151 Total

A
467
2
3 13
18
6
42
B
469
2
1
3
2
8
2
5 14
21
0
0
8
50
D
508
1
1
0
0
0
1
0
0
0
1
509
1
1
1
3
0
0
1
1
0
0
1
3
529
2
14
3
2
21
530
4
47
11
7
69
0
0
6
61
14
0
9
90
523
2
2
4
0
0
2
2
0
0
0
4
527
2
2
0
0
0
0
0
0
2
2
524
2
4
2
3
11
525
2
1
1
4
526
1
1
0
0
0
5
4
3
4
16
531
2
1
3
0
0
0
2
0
0
1
3
A
533
1
1
B
534
3
5
11
5
24
C
535
1
1
2
4
C
536
2
1
3
C
537
1
1
C
539
1
1
2
C
540
1
1
D
541
4
1
5
D
542
3
1
1
5
0
9
7
19
1
0
10
46
A
545
3
2
6
4
15
C
546
1
1
2
D
547
2
2
3
7
E
548
1
1
0
4
3
9
2
0
7
25
A
549
1
6
7
2
9
25
B
550
10
16
3
11
40
B
551
6
2
4
12
0
1 16
29
7
0
24
77
A
583
3
4
1
8
A
584
1
4
1
1
7
B
585
1
2
3
B
586
2
2
C
587
4
1
6
11
D
589
1
2
3
2
7
15
6
4 10
11
0
0
15
46
590
15
3
5
9
32
0 15
3
5
0
0
9
32
583
Site No.
152
152
152 Total
153
153 Total
154
154
154
154
154 Total
156
156
156
156 Total
158
158
158 Total
159
159 Total
161
161
161
161
161 Total
163
163
163
163
163
163
163
163 Total
164
164
164
164
164
164
164
164 Total
165
165
165
165 Total
166
166 Total
167
167

593
1
1
595
1
7
8
1
0
0
0
0
0
8
9
B
596
1
1
3
5
0
0
0
1
1
0
3
5
B
597
1
2
3
C
598
1
1
1
3
C
599
1
1
2
D
600
2
2
0
2
4
2
0
1
1
10
604
1
1
605
1
1
606
1
1
1
9
12
0
0
1
3
1
0
9
14
601
2
1
3
602
3
3
0
0
0
5
1
0
0
6
610
15
2
17
0
0
0
15
0
0
2
17
A
613
1
3
7
11
B
614
6
1
7
D
615
1
2
1
4
F
616
2
2
0
0
2
13
2
0
7
24
552
2
4
3
9
553
1
1
9
3
14
554
4
1
5
555
1
6
1
2
10
556
2
2
4
557
1
7
3
11
580
1
2 10
6
1
7
27
1
3 15
38
1
1
21
80
575
1
7
2
10
576
2
5
7
577
2
6
11
1
1
3
24
578
6
1
7
579
3
2
5
581
1
6
6
1
7
21
582
9
16
1
6
32
0
4 23
54
2
2
21 106
558
1
3
3
3
5
15
559
1
1
560
1
1
0
1
5
3
3
0
5
17
561
1
19
26
1
14
61
1
0 19
26
1
0
14
61
562
1
1
1
3
563
4
1
3
8
584
Site No.
167
565
9 13
18
4
15
59
167
566
4
3
6
13
167 Total
0 13 14
22
9
0
25
83
168
567
1
2 10
12
12
37
168
568
1
1
168
569
1
1
5
7
168
570
1
1
168
571
1
2
1
4
168
572
1
1
3
5
168
573
1
1
2
168 Total
1
3 13
17
1
0
22
57
169
510
2
2
169
511
1 50
5
6
1
10
73
169
512
8
2
6
9
25
169
513
1
1
169
514
1
2
1
1
1
5
11
169
515
1
1
2
169
517
1
1
169 Total
2 62
7
15
2
1
26 115
M.E. 3
5
3
2
5
M.E. 3 Total
0
0
0
3
0
0
2
5
M.E. 4
6
3
4
7
14
M.E. 4 Total
0
0
0
3
4
0
7
14
M.E. 5
7
1
1
M.E. 5 Total
0
0
0
1
0
0
0
1
M.E. 6
10
1
1
1
3
6
M.E. 6 Total
0
1
0
0
1
1
3
6
M.E. 7
17
4
7
2
2
15
M.E. 7 Total
0
0
0
4
7
2
2
15
M.E. 8
18
1
2
8
2
1
14
M.E. 8 Total
0
1
2
8
2
0
1
14
M.E. 9
70
1
1
M.E. 9 Total
0
1
0
0
0
0
0
1
M.E. 10
325
2
2
2
1
7
M.E. 10 Total
0
2
2
2
0
0
1
7
Grand Total
153 222 433 1474 456
58 1010 3806
585
Appendix F: Site and Sector Registry
Notes for Table F.1:
UTM coordinates correspond to UTM Zone 19S, WGS 1984 Datum.
All areas in hectares (ha). In cases where residential area is different from total site
area, parentheses are around the total site area.
Habitational site size classes in the table correspond are included as a heuristic aid,
not as analytical categories. They correspond to the following areal measures:
Class
Isolate
Hamlet
Small Village
Large Village
Town
Center
Area (ha)
0.01-0.25
0.25-0.5
0.5-1.75
1.75-3
3-9
>9
586
Figure F.1: Overview map showing locations of Maps F.2-F.5
587
Figure F.2: All sites registered, northwest block
588
Figure F.3: All sites registered, northeast block
589
Figure F.4: All sites registered, west-central block
590
Figure F.5: All sites registered, south block
District Site No.

YA
1
YA
1
YA
1
YA
1
YA
2
YA
2
YA
2
YA
3
YA
4
YA
4
YA
4
YA
4
YA
5
YA
6
YA
6
YA
7
YA
7
YA
7
YA
7
YA
8
YA
8
YA
8
YA
8
YA
8
YA
9
YA
9
YA
9
YA
9
YA
9
YA
10
YA
10
YA
11
YA
12
A/D
B
B-D
B-D
B-D
C
C
C
C
B
B
Sector(s)
A
A
B/C
B/C
A/B
A/B
A/B
A/B
UTM N
8268291
8268291
8268291
8268291
8268489
8268489
8268489
8268979
8269015
8269015
8269015
8269015
8268910
8268989
8268989
8268241
8268241
8268241
8268241
8268977
8268977
8268977
8268977
8268977
8268680
8268680
8268680
8268680
8268680
8268533
8268533
8267732
8267557
UTM E Altitude Ecozone

216265
3444 Kichwa
216265
3444 Kichwa
216265
3444 Kichwa
216265
3444 Kichwa
215683
3495 Kichwa
215683
3495 Kichwa
215683
3495 Kichwa
215955
3414 Kichwa
215826
3413 Kichwa
215826
3413 Kichwa
215826
3413 Kichwa
215826
3413 Kichwa
215948
3413 Kichwa
215356
3471 Kichwa
215356
3471 Kichwa
215513
3489 Kichwa
215513
3489 Kichwa
215513
3489 Kichwa
215513
3489 Kichwa
214855
3403 Kichwa
214855
3403 Kichwa
214855
3403 Kichwa
214855
3403 Kichwa
214855
3403 Kichwa
214900
3478 Kichwa
214900
3478 Kichwa
214900
3478 Kichwa
214900
3478 Kichwa
214900
3478 Kichwa
215022
3482 Kichwa
215022
3482 Kichwa
214507
3465 Kichwa
214079
3460 Kichwa
Time Pd.
LH
COL
MH
LIP
FOR
LIP
LH
LH
MH
LIP
LH
COL
UKN
MH
LH
MH
LIP
LH
COL
FOR
MH
LIP
LH
COL
MH
FOR
LIP
LH
COL
LH
COL
UKN
MH
Site Class
Habitational: Isolate
Midden/Rubble Mound
Midden/Rubble Mound
Midden/Rubble Mound
Midden/Rubble Mound
Midden/Rubble Mound
Small Cemetery
Midden/Rubble Mound
Midden/Rubble Mound
Midden/Rubble Mound
Midden/Rubble Mound
Rock Art Panel
Small Scatter
Midden/Rubble Mound
Midden/Rubble Mound
Midden/Rubble Mound
Midden/Rubble Mound
Small Cemetery
Small Cemetery
Small Cemetery
Small Cemetery
Small Cemetery
Small Scatter
Small Scatter
Midden/Rubble Mound
Ag. Feature/Scatter
Midden/Rubble Mound
Midden/Rubble Mound
Tomb Feature
Small Scatter
Table E.1: Site and sector registry

Area
0.51
0.51
0.51
0.51
0.06
0.06
0.06
0.01
0.61
0.61
0.61
0.61
0.01
0.04
0.04
0.01
0.01
0.01
0.01
0.04
0.04
0.04
0.04
0.04
0.73
UKN
0.73
0.73
0.73
0.03
0.03
0.01
0.08
591
District Site No. Sector(s)

YA
12
YA
12
YA
13
YA
14 A-E
YA
14 A-E
YA
14 A-E
YA
14
YA
14
YA
15
YA
16
YA
17
YA
18
YA
19
YA
20
YA
21
YA
22
YA
23
YA
24
YA
25
YA
26
YA
27
YA
28
YA
29
YA
29
YA
30
YA
30
YA
30
YA
30
YA
31
YA
32
YA
32
YA
32
YA
32
UTM N
8267557
8267557
8267005
8266390
8266390
8266390
8266390
8266390
8267242
8267564
8265301
8265179
8265726
8265900
8265951
8264776
8263964
8264654
8264813
8265071
8265268
8265335
8265333
8265333
8265689
8265689
8265689
8265689
8264934
8265587
8265587
8265587
8265587

214079
3460 Kichwa
214079
3460 Kichwa
214372
3497 Kichwa
214201
3533 Kichwa
214201
3533 Kichwa
214201
3533 Kichwa
214201
3533 Kichwa
214201
3533 Kichwa
215395
3509 Kichwa
215631
3564 Kichwa
215614
3720 Suni
215468
3722 Suni
214776
3643 Suni
214771
3597 Kichwa
214659
3587 Kichwa
214674
3700 Suni
214845
3777 Suni
214492
3702 Suni
213472
3666 Suni
214448
3658 Suni
214237
3658 Suni
214134
3621 Suni
214017
3615 Suni
214017
3615 Suni
213605
3575 Kichwa
213605
3575 Kichwa
213605
3575 Kichwa
213605
3575 Kichwa
213607
3622 Suni
214584
3665 Suni
214584
3665 Suni
214584
3665 Suni
214584
3665 Suni
Time Pd.
LH
COL
UKN
LIP
LH
COL
FOR
MH
LH
UKN
UKN
UKN
UKN
UKN
TA-LH
UKN
UKN
UKN
LH
LH
COL
LH
LIP
LH
FOR
MH
LIP
LH
UKN
FOR
MH
LIP
LH
Site Class
Scatter
Habitational: Isolate (prob. modern)
Tomb Feature
Rock Art Panel
Rock Art Panel
Rock Art Panel
Small Rockshelter
Rock Art Panel
Rock Art Panel
Rock Art Panel
Large Scatter
Small Scatter
Habitational: Large Village

Area
0.08
0.08
0.04
91.75
91.75
91.75
UKN
UKN
0.19
0.01
0.02
0.01
0.01
0.02
0.01
0.01
0.01
0.04
0.03
0.33
0.33
0.01
0.02
0.02
UKN
UKN
23.00
23.00
0.01
3.88
UKN
1.86
1.86
592

YA
32
YA
32
YA
33
YA
33
YA
33
YA
34
YA
34
YA
34
YA
34
YA
34
YA
34
YA
34
YA
34
YA
34
YA
34
YA
34
YA
34
YA
34
YA
35
YA
36
YA
37
YA
38
YA
39
YA
40
YA
40
YA
40
YA
41 A/C/E/N
YA
41 A-X
YA
41
YA
41
YA
41
YA
42
YA
42
UTM N
8265587
8265587
8263696
8263696
8263696
8263465
8263465
8263465
8263465
8263465
8263465
8263465
8263465
8263465
8263465
8263465
8263465
8263465
8265304
8265259
8264955
8263664
8263834
8263896
8263896
8263896
8268034
8268034
8268034
8268034
8268034
8268471
8268471

214584
3665 Suni
214584
3665 Suni
216513
4160 Puna
216513
4160 Puna
216513
4160 Puna
216578
4187 Puna
216578
4187 Puna
216578
4187 Puna
216578
4187 Puna
216578
4187 Puna
216578
4187 Puna
216578
4187 Puna
216578
4187 Puna
216578
4187 Puna
216578
4187 Puna
216578
4187 Puna
216578
4187 Puna
216578
4187 Puna
215018
3754 Suni
215222
3727 Suni
215434
3763 Suni
217129
4297 Puna
217103
4246 Puna
216654
4149 Puna
216654
4149 Puna
216654
4149 Puna
214962
3484 Kichwa
214962
3484 Kichwa
214962
3484 Kichwa
214962
3484 Kichwa
214962
3484 Kichwa
214135
3448 Kichwa
214135
3448 Kichwa
Time Pd.
TA-LH
COL
E-MA
LA
TA-LH
EA
E-MA
MA
LA
TA
TA-FOR
FOR
GA
MH
LIP
LH
TA-LH
COL
UKN
UKN
UKN
UKN
UKN
LH
TA-LH
COL
MH
LIP
FOR
LH
COL
MH
LH
Site Class
Scatter
Small Scatter
Small Scatter
Small Scatter
Large Scatter
Large Scatter
Large Scatter
Large Scatter
Large Scatter
Large Scatter
Large Scatter
Large Scatter
Large Scatter
Large Scatter
Rock Art Panel
Rock Art Panel
Rock Art Panel
Rock Art Panel
Rock Art Panel
Small Scatter
Small Scatter
Small Scatter
Habitational:Center
Habitational:Center
Midden/Rubble Mound
Midden/Rubble Mound

Area
UKN
1.86
0.19
0.19
0.19
2.19
2.19
2.19
2.19
2.19
2.19
2.19
2.19
2.19
2.19
2.19
2.19
2.19
0.01
0.01
0.01
0.01
0.01
0.15
0.15
0.15
UKN
UKN
UKN
17.96
17.96
0.01
0.01
593
District Site No. Sector(s) UTM N UTM E Altitude Ecozone

YA
43
8268597 214363
3417 Kichwa
YA
44
8269129 214611
3510 Kichwa
YA
44
8269129 214611
3510 Kichwa
YA
45
8269379 214546
3556 Kichwa
YA
45
8269379 214546
3556 Kichwa
YA
45
8269379 214546
3556 Kichwa
YA
45
8269379 214546
3556 Kichwa
YA
46
8269045 214467
3494 Kichwa
YA
46
8269045 214467
3494 Kichwa
YA
46
8269045 214467
3494 Kichwa
YA
46
8269045 214467
3494 Kichwa
YA
47
8268991 214278
3497 Kichwa
YA
48
8269166 213784
3554 Kichwa
YA
48
8269166 213784
3554 Kichwa
YA
49
8268827 213794
3480 Kichwa
YA
50
8269021 213730
3527 Kichwa
YA
50
8269021 213730
3527 Kichwa
YA
50
8269021 213730
3527 Kichwa
YA
50
8269021 213730
3527 Kichwa
YA
51
8269367 213558
3596 Kichwa
YA
52
8269209 213617
3559 Kichwa
YA
52
8269209 213617
3559 Kichwa
YA
53
8268836 212937
3674 Suni
YA
53
8268836 212937
3674 Suni
YA
53
8268836 212937
3674 Suni
YA
54
8268580 213272
3572 Kichwa
YA
54
8268580 213272
3572 Kichwa
YA
54
8268580 213272
3572 Kichwa
YA
54
8268580 213272
3572 Kichwa
YA
54
8268580 213272
3572 Kichwa
YA
55
8268638 213583
3485 Kichwa
YA
55
8268638 213583
3485 Kichwa
YA
55
8268638 213583
3485 Kichwa
Time Pd.
FOR
LIP
COL
FOR
LIP
LH
COL
FOR
MH
LH
COL
LIP
LIP
LH
LH
MH
LIP
LH
COL
UKN
LIP
LH
LIP
LH
COL
FOR
MH
LIP
LH
COL
FOR
LH
COL
Site Class
Midden/Rubble Mound
Small Cemetery
Small Rockshelter
Small Scatter
Habitational: Hamlet
Small Scatter
Large Scatter
Scatter
Scatter
Small Cemetery
Small Group Below Ground
Small Scatter
Habitational:Town
Habitational:Town
Habitational:Town
Rock Art Panel
Scatter
Scatter
Small Scatter
Small Scatter
Habitational: Small Village
Small Scatter
Scatter
Scatter

Area
0.01
0.01
0.01
1.51
1.51
1.51
1.51
0.34
0.34
0.34
0.34
0.01
0.04
0.04
0.04
UKN
4.26
4.26
4.26
0.01
0.03
0.03
0.25
0.25
0.25
1.75
UKN
1.75
1.75
1.75
0.03
0.03
0.03
594

YA
56
8268148 213513
3521 Kichwa
YA
56
8268148 213513
3521 Kichwa
YA
57
8267579 213723
3431 Kichwa
YA
57
8267579 213723
3431 Kichwa
YA
58
8268153 212385
3497 Kichwa
YA
59
8270617 214512
3808 Suni
YA
59
8270617 214512
3808 Suni
YA
59
8270617 214512
3808 Suni
YA
59
8270617 214512
3808 Suni
YA
59
8270617 214512
3808 Suni
YA
60
8270746 214360
3905 Puna
CO
61
8272761 213249
4360 Puna
CO
61
8272761 213249
4360 Puna
CO
61
8272761 213249
4360 Puna
YA
62
8267696 216201
3780 Suni
YA
63
8266839 216519
3858 Puna
YA
64
8265903 216923
4074 Puna
YA
65
8266375 216705
3928 Puna
YA
66
8266268 216288
3864 Puna
YA
66
8266268 216288
3864 Puna
YA
66
8266268 216288
3864 Puna
YA
66
8266268 216288
3864 Puna
YA
66
8266268 216288
3864 Puna
YA
66
8266268 216288
3864 Puna
YA
67
8265507 216971
4154 Puna
YA
68
8266867 216657
3909 Puna
YA
69
8266531 216804
3945 Puna
YA
70
8265748 217355
4137 Puna
YA
71
8265631 217132
4162 Puna
YA
72
8265716 216764
4114 Puna
YA
73
8265245 216516
3962 Puna
YA
74
8265286 216758
4035 Puna
YA
75
8265215 217412
4215 Puna
Time Pd.
LIP
LH
MH
COL
COL
FOR
MH
LIP
LH
COL
UKN
MH
LH
COL
UKN
UKN
COL
UKN
EA
LA
LIP
LH
TA-LH
COL
EA
UKN
UKN
UKN
UKN
UKN
TA-LH
COL
UKN
Site Class
Scatter
Scatter
Small Scatter
Small Scatter
Kiln
Small Scatter
Small Scatter
Scatter
Ag. Feature/Scatter
Ag. Feature/Scatter
Rock Art Panel
Small Scatter
Rock Art Panel
Rock Art Panel
Large Rockshelter
Large Rockshelter
Rockshelter
Large Rockshelter
Large Rockshelter
Large Rockshelter
Small Scatter
Rock Art Panel
Rock Art Panel
Rock Art Panel
Rock Art Panel
Rock Art Panel
Small Scatter
Rock Art Panel

Area
0.02
0.02
0.14
0.14
0.01
0.25
0.25
0.25
0.25
0.25
0.01
UKN
2.83
2.83
2.41
0.01
0.05
0.01
0.06
0.06
0.06
0.06
0.06
0.06
0.04
0.01
0.01
0.01
0.01
0.01
0.25
1.00
0.01
595

YA
76
YA
77
YA
78
YA
78
YA
78
YA
79
YA
80
YA
81
YA
82
YA
82
YA
82
YA
83
YA
83
YA
83
YA
84
YA
85
YA
86
YA
87
YA
87
YA
87
YA
87
YA
88
YA
88
YA
88
YA
88
YA
88
YA
89 A
YA
89 B
YA
89 B
YA
90
YA
90
YA
90
YA
91
UTM N
8265009
8265160
8264939
8264939
8264939
8264594
8259370
8261620
8262286
8262286
8262286
8262361
8262361
8262361
8262592
8263261
8262895
8263074
8263074
8263074
8263074
8262997
8262997
8262997
8262997
8262997
8262793
8262793
8262793
8271951
8271951
8271951
8270973

217354
4204 Puna
217682
4246 Puna
217557
4259 Puna
217557
4259 Puna
217557
4259 Puna
216817
3997 Puna
218443
4436 Puna
217531
4544 Puna
218401
4623 Puna
218401
4623 Puna
218401
4623 Puna
217591
4467 Puna
217591
4467 Puna
217591
4467 Puna
216785
4271 Puna
216490
4223 Puna
217095
4331 Puna
217208
4337 Puna
217208
4337 Puna
217208
4337 Puna
217208
4337 Puna
216918
4286 Puna
216918
4286 Puna
216918
4286 Puna
216918
4286 Puna
216918
4286 Puna
216581
4233 Puna
216581
4233 Puna
216581
4233 Puna
213417
4300 Puna
213417
4300 Puna
213417
4300 Puna
213638
4177 Puna
Time Pd.
UKN
UKN
LH
TA-LH
COL
UKN
UKN
COL
MA
FOR
LIP
LA
TA-FOR
FOR
TA-LH
UKN
EA
EA
LA
TA-FOR
FOR
EA
GA
LIP
TA-LH
COL
TA-LH
LH
COL
MH
LH
COL
UKN
Site Class
Rock Art Panel
Rock Art Panel
Small Scatter
Small Scatter
Small Scatter
Rock Art Panel
Small Scatter
Small Rockshelter
Small Rockshelter
Rockshelter
Small Rockshelter
Small Rockshelter
Small Rockshelter
Small Scatter
Small Scatter
Small Scatter
Small Scatter
Small Scatter
Small Scatter
Small Scatter
Small Scatter
Small Scatter
Small Scatter
Small Rockshelter
Small Scatter
Rock Art Panel

Area
0.01
0.01
0.35
0.35
0.35
0.01
0.04
1.00
0.01
0.01
0.01
0.01
0.01
0.01
0.05
0.06
0.04
0.09
0.09
0.09
0.09
0.70
0.70
0.70
0.70
0.70
0.01
0.26
0.26
UKN
1.00
1.00
0.01
596

YA
92
YA
93 B
YA
93
YA
93
YA
93
YA
94 A
YA
94 A
YA
94 B/C
YA
94 B/C
YA
94 B/C
YA
94 B/C
YA
95
YA
96
YA
97
CO
98
CO
98
CO
98
CO
98
CO
99
CO
100 G/J/K/M/N
CO
100
CO
100
CO
100
CO
100
CO
101
CO
102
CO
102
CO
103
CO
103
CO
103
CO
104
CO
104
CO
104
UTM N
8270697
8272359
8272359
8272359
8272359
8271998
8271998
8271998
8271998
8271998
8271998
8270549
8269179
8268836
8270500
8270500
8270500
8270500
8269625
8269972
8269972
8269972
8269972
8269972
8269599
8272637
8272637
8269577
8269577
8269577
8276550
8276550
8276550

213644
4088 Puna
212448
4292 Puna
212448
4292 Puna
212448
4292 Puna
212448
4292 Puna
211952
4254 Puna
211952
4254 Puna
211952
4254 Puna
211952
4254 Puna
211952
4254 Puna
211952
4254 Puna
212361
4065 Puna
213834
3554 Kichwa
214676
3427 Kichwa
215317
3763 Suni
215317
3763 Suni
215317
3763 Suni
215317
3763 Suni
215089
3606 Suni
215348
3666 Suni
215348
3666 Suni
215348
3666 Suni
215348
3666 Suni
215348
3666 Suni
216407
3541 Kichwa
217359
4068 Puna
217359
4068 Puna
216622
3558 Kichwa
216622
3558 Kichwa
216622
3558 Kichwa
217530
4362 Puna
217530
4362 Puna
217530
4362 Puna
Time Pd.
UKN
MH
MA
LIP
LH
MH
LH
TA
TA-FOR
FOR
TA-LH
UKN
UKN
UKN
MH
LIP
LH
COL
UKN
MH
FOR
LIP
LH
COL
COL
LH
TA-LH
LIP
LH
COL
TA
TA-FOR
FOR
Site Class
Tomb Feature
Small Scatter
Small Scatter
Small Scatter
Small Rockshelter
Small Rockshelter
Small Rockshelter
Small Rockshelter
Rock Art Panel
Small Cemetery
Large Cemetery
Large Cemetery
Large Cemetery
Large Cemetery
Large Cemetery
Rock Art Panel
Small Scatter
Small Scatter
Habitational:Town
Habitational:Town
Habitational:Town
Small Rockshelter
Small Rockshelter
Small Scatter
Small Scatter
Small Scatter

Area
0.01
UKN
0.75
0.75
0.75
0.04
0.04
0.01
0.01
0.04
0.01
0.01
0.25
0.10
0.05
0.05
0.05
0.05
0.01
UKN
UKN
8.65 (12.35)
8.65 (12.35)
8.65 (12.35)
0.10
0.01
0.01
0.25
0.25
0.25
0.01
0.01
0.01
597

CO
105
8276151 217418
4348 Puna
CO
105
8276151 217418
4348 Puna
CO
105
8276151 217418
4348 Puna
CO
105
8276151 217418
4348 Puna
CO
106
8276717 217007
4397 Puna
CO
106
8276717 217007
4397 Puna
CO
106
8276717 217007
4397 Puna
CO
106
8276717 217007
4397 Puna
CO
106
8276717 217007
4397 Puna
CO
106
8276717 217007
4397 Puna
CO
106
8276717 217007
4397 Puna
CO
107
8276970 216995
4426 Puna
CO
108
8276508 216768
4390 Puna
CO
109
8275673 216403
4340 Puna
CO
109
8275673 216403
4340 Puna
CO
110
8274175 215597
4392 Puna
CO
110
8274175 215597
4392 Puna
CO
111
8274320 215685
4355 Puna
CO
111
8274320 215685
4355 Puna
CO
111
8274320 215685
4355 Puna
CO
112
8274692 215958
4328 Puna
CO
113
8272502 216979
3843 Puna
CO
114
8271034 216953
3665 Suni
CO
114
8271034 216953
3665 Suni
CO
115
8271307 217203
3781 Suni
CO
116
8271240 217093
3724 Suni
CO
117
8271158 217117
3763 Suni
CO
118
8270976 217006
3697 Suni
CO
119
8270660 217088
3759 Suni
CO
120
8270321 217049
3706 Suni
CO
121
8269211 216548
3537 Kichwa
CO
121
8269211 216548
3537 Kichwa
CO
121
8269211 216548
3537 Kichwa
Time Pd.
TA
FOR
LIP
LH
EA
MA
LA
TA-FOR
FOR
LH
TA-LH
UKN
UKN
LH
COL
LA
TA-LH
LH
TA-LH
COL
UKN
UKN
LIP
LH
UKN
UKN
UKN
LIP
LIP
LIP
FOR
MH
LIP
Site Class
Small Scatter
Small Scatter
Large Rockshelter
Large Rockshelter
Large Rockshelter
Large Rockshelter
Large Rockshelter
Large Rockshelter
Large Rockshelter
Rock Art Panel
Small Rockshelter/Rock Art Panel
Small Scatter
Small Scatter
Rock Art Panel
Rock Art Panel
Small Cemetery
Small Cemetery
Tomb Feature
Rock Art Panel
Tomb Feature
Small Cemetery
Small Cemetery
Small Cemetery
Midden/Rubble Mound
Small Scatter
Midden/Rubble Mound

Area
0.06
0.06
0.06
0.06
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.01
0.01
1.00
1.00
0.20
0.20
0.07
0.07
0.07
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.04
0.01
0.01
0.01
0.01
598

CO
121
CO
122 A
CO
123
CO
123
CO
124
CO
125
CO
126
CO
127
CO
127
CO
127
CO
128
CO
128
CO
129
CO
130
CO
131
CO
132
CO
133
CO
134
CO
135
CO
136
CO
137
CO
138
CO
139
CO
140
CO
141
CO
142
CO
143
CO
144
CO
145
CO
146
CO
147 A-D
CO
147 B/C/D
CO
147
UTM N
8269211
8270191
8270556
8270556
8271198
8269972
8269714
8268928
8268928
8268928
8269491
8269491
8269351
8270456
8270641
8270837
8270923
8271161
8270653
8270243
8269851
8268479
8268279
8268184
8271308
8271563
8271086
8271482
8271498
8271115
8268162
8268162
8268162

216548
3537 Kichwa
217427
3769 Suni
217059
3757 Suni
217059
3757 Suni
217379
3811 Puna
217776
3790 Suni
217302
3638 Suni
217416
3563 Kichwa
217416
3563 Kichwa
217416
3563 Kichwa
218098
3606 Suni
218098
3606 Suni
218497
3627 Suni
220026
3724 Suni
219951
3697 Suni
220249
3697 Suni
220166
3702 Suni
220534
4144 Puna
219676
3679 Suni
219656
3680 Suni
219556
3659 Suni
219193
3619 Suni
219340
3595 Kichwa
218899
3587 Kichwa
219766
3826 Puna
219768
3908 Puna
219596
3745 Suni
219339
3967 Puna
219221
4038 Puna
219749
3741 Suni
218359
3579 Kichwa
218359
3579 Kichwa
218359
3579 Kichwa
Time Pd.
LH
LH
LIP
LH
UKN
UKN
UKN
LIP
LH
COL
LIP
LH
UKN
UKN
UKN
UKN
UKN
UKN
UKN
UKN
UKN
UKN
UKN
UKN
UKN
UKN
UKN
UKN
EA
UKN
LIP
MH
LH
Site Class
Midden/Rubble Mound
Small Cemetery
Small Cemetery
Small Cemetery
Rock Art Panel
Habitational:Town
Habitational:Town
Habitational:Town
Rock Art Panel
Rock Art Panel
Habitational: Hamlet (prob. modern)
Rock Art Panel
Small Rockshelter

Area
0.01
0.63
0.01
0.01
0.01
0.01
0.04
4.23
4.23
4.23
0.01
0.01
0.01
0.01
0.02
0.16
0.01
0.25
0.25
0.09
0.01
0.01
0.01
0.16
0.04
0.01
0.02
0.01
0.01
0.01
27.94
25.72
27.94
599

CO
147
CO
148 A/C
CO
148 A-E
CO
148
CO
148
CO
149 A
CO
149 A/B
CO
149
CO
149
CO
150 A
CO
150 A
CO
150 B/C
CO
150 B/C/D
CO
150
CO
151
CO
151
CO
151
CO
152
CO
152
CO
153
CO
153
CO
154
CO
154
CO
154
CO
155
CO
156
CO
156
CO
156
CO
156
CO
157
CO
158
CO
158
CO
159
UTM N
8268162
8268930
8268930
8268930
8268930
8268285
8268285
8268285
8268285
8268378
8268378
8268378
8268378
8268378
8270357
8270357
8270357
8270788
8270788
8270674
8270674
8270144
8270144
8270144
8275979
8275213
8275213
8275213
8275213
8275105
8273611
8273611
8273897

218359
3579 Kichwa
218303
3573 Kichwa
218303
3573 Kichwa
218303
3573 Kichwa
218303
3573 Kichwa
217586
3552 Kichwa
217586
3552 Kichwa
217586
3552 Kichwa
217586
3552 Kichwa
216916
3513 Kichwa
216916
3513 Kichwa
216916
3513 Kichwa
216916
3513 Kichwa
216916
3513 Kichwa
215101
3822 Puna
215101
3822 Puna
215101
3822 Puna
215012
3884 Puna
215012
3884 Puna
215270
3858 Puna
215270
3858 Puna
215247
3767 Suni
215247
3767 Suni
215247
3767 Suni
214300
4600 Puna
214893
4448 Puna
214893
4448 Puna
214893
4448 Puna
214893
4448 Puna
214937
4430 Puna
215554
4564 Puna
215554
4564 Puna
215155
4419 Puna
Time Pd.
COL
MH
LIP
LH
COL
MH
LIP
LH
COL
LIP
LH
FOR
MH
COL
MH
LIP
LH
FOR
TA-LH
LH
COL
MH
LIP
LH
UKN
LIP
LH
TA-LH
TA-LH
TA-LH
TA
LH
TA-FOR
Site Class
Scatter
Midden/Rubble Mound
Midden/Rubble Mound
Small Scatter
Small Scatter
Small Scatter
Scatter
Scatter
Small Cemetery
Small Cemetery
Large Cemetery
Rock Art Panel
Small Rockshelter
Small Rockshelter
Small Rockshelter
Small Rockshelter
Small Rockshelter
Scatter
Large Fortification
Small Scatter

Area
27.94
7.44
35.46
35.46
UKN
UKN
23.24
23.24
UKN
1.40
1.40
0.15
0.19
1.40
0.40
0.40
0.40
0.18
0.18
0.01
0.01
0.02
0.02
0.02
0.01
0.01
0.01
0.01
0.01
0.01
UKN
1.90
0.36
600

CO
159
CO
159
CO
159
CO
160
CO
161 A/B/D/F
CO
161
CO
161
CO
162
CO
163
CO
163
CO
163
CO
163
CO
163
CO
164
CO
164
CO
164
CO
164
CO
165
CO
165
CO
165
CO
165
CO
166
CO
166
CO
166
CO
166
CO
167
CO
167
CO
167
CO
167
CO
168
CO
168
CO
168
CO
168
UTM N
8273897
8273897
8273897
8273086
8270521
8270521
8270521
8266652
8268280
8268280
8268280
8268280
8268280
8269331
8269331
8269331
8269331
8268518
8268518
8268518
8268518
8268603
8268603
8268603
8268603
8268741
8268741
8268741
8268741
8269008
8269008
8269008
8269008

215155
4419 Puna
215155
4419 Puna
215155
4419 Puna
214353
4233 Puna
216353
3626 Suni
216353
3626 Suni
216353
3626 Suni
216499
3839 Puna
217264
3528 Kichwa
217264
3528 Kichwa
217264
3528 Kichwa
217264
3528 Kichwa
217264
3528 Kichwa
217199
3556 Kichwa
217199
3556 Kichwa
217199
3556 Kichwa
217199
3556 Kichwa
217384
3543 Kichwa
217384
3543 Kichwa
217384
3543 Kichwa
217384
3543 Kichwa
217273
3545 Kichwa
217273
3545 Kichwa
217273
3545 Kichwa
217273
3545 Kichwa
217100
3537 Kichwa
217100
3537 Kichwa
217100
3537 Kichwa
217100
3537 Kichwa
217079
3547 Kichwa
217079
3547 Kichwa
217079
3547 Kichwa
217079
3547 Kichwa
Time Pd.
FOR
LH
TA-LH
UKN
LIP
LH
COL
UKN
FOR
MH
LIP
LH
COL
MH
LIP
LH
COL
MH
LIP
LH
COL
FOR
LIP
LH
COL
MH
LIP
LH
COL
FOR
MH
LIP
LH
Site Class
Small Scatter
Small Scatter
Rock Art Panel
Small Scatter
Habitational:Town
Habitational:Town
Maqueta Cluster
Small Scatter
Small Scatter
Habitational:Town
Habitational:Town
Habitational:Town
Small Scatter
Habitational:Town
Habitational:Town
Habitational:Town
Small Fortification
Small Fortification
Scatter
Scatter
Small Fortification
Small Fortification
Scatter
Small Fortification
Small Fortification

Area
0.36
0.36
0.36
0.01
UKN
4.85
4.85
0.65
UKN
UKN
5.75
5.75
5.75
UKN
3.76 (5.70)
3.76 (5.70)
3.76 (5.70)
1.71
1.71
1.71
1.71
UKN
5.29
5.29
5.29
UKN
7.28
7.28
UKN
UKN
9.46
9.67
9.67
601

CO
168
8269008 217079
3547 Kichwa
CO
169
8269450 217378
3568 Kichwa
CO
169
8269450 217378
3568 Kichwa
CO
169
8269450 217378
3568 Kichwa
CO
169
8269450 217378
3568 Kichwa
CO
169
8269450 217378
3568 Kichwa
Time Pd.
COL
FOR
MH
LIP
LH
COL
Site Class
Scatter

Area
9.67
UKN
1.86
1.86
1.86
1.86
602
603
References Cited
Abbott, M. B., M. W. Binford, M. Brenner and K. R. Kelts
1997 A 3500 14C yr high-resolution record of water-level changes in Lake
Titicaca, Bolivia/Peru. Quaternary research 47(169-180).
Abbott, M. B., G. O. Seltzer, K. R. Kelts and J. Southon
1997 Holocene paleohydrology of the tropical Andes from lake records.
Quaternary research 47:70-80.
Abercrombie, T. A.
1986 The politics of sacrifice: an Aymara cosmology in action. Ph.D. dissertation,
University of Chicago.
1998 Pathways of memory and power: ethnography and history among an Andean
people. University of Wisconsin Press, Madison.
Adams, R. M.
1966 The evolution of urban society. Aldine, Chicago.
Albarracin-Jordan, J. and J. E. Mathews
1990 Asentamientos prehispnicos del valle de Tiwanaku I. CIMA, La Paz.
Aldenderfer, M. S.
1989 Archaic period settlement patterns in the high sierra of the Osmore basin. In
Ecology, settlement, and history in the Osmore drainage, Peru, edited by D. S. Rice,
C. Stanish and P. R. Scarr, pp. 129-166. British Archaeological Reports
International Series, Oxford.
1998 Montane foragers: Asana and the south-central Andean Archaic. University
of Iowa Press, Iowa City.
Anderson, B.
1991 Imagined communities: reflections on the origins and spread of nationalism.
Verso, London.
Ascher, M. and R. Ascher
1988 Code of the quipu: a study in media, mathematics, and culture. University of
Michigan Press, Ann Arbor.
1997 Mathematics of the Incas: code of the quipu. Dover Publications, Mineola,
New York.
604
Bale, W. L.
1998 (editor) Advances in historical ecology. Columbia University Press, New
York.
Bandy, M. S.
2001 Population and history in the ancient Titicaca Basin. Ph.D. dissertation,
University of California at Berkeley.
Barth, F.
1969 Introduction. In Ethnic groups and boundaries. The social organization of
culture difference, edited by F. Barth, pp. 9-38. Allen & Unwin, London.
Bauer, B. S.
1998 The sacred landscape of the Inca: the Cuzco ceque system. University of
Texas Press, Austin.
1999 The early ceramics of the Inca heartland. Fieldiana, Anthropology New
Series no. 31. Field Museum of Natural History, Chicago.
Bauer, B. S. and C. Stanish
2001 Ritual and pilgrimage in the ancient Andes: the islands of the sun and the
moon. 1st ed. University of Texas Press, Austin.
Benavides, M. A.
1986a Cambios en la tenencia y el uso de tierras desde el siglo XVI hasta el
presente en el valle del Colca (Caylloma, Arequipa). In The cultural ecology,
archaeology, and history of terracing in the Colca valley of southern Peru.
Technical report to the National Science Foundation and the National Geographic
Society, edited by W. M. Denavan, pp. 509-524. vol. 1. Department of Geography,
University of Wisconsin, Madison.
1986b Coporaque in the 1591 visita of Yanquecollaguas Urinsaya. In The cultural
ecology, archaeology, and history of terracing and terrace abandonment in the
Colca valley of Southern Peru. Technical report to the National Science
Foundation and the National Geographic Society, edited by W. M. Denevan, pp.
390-405. vol. 1. Department of Geography, University of Wisconsin-Madison,
Madison.
1986c Coporaque in the 1604 visita to Yanquecollaguas Urinsaya. In The cultural
605
Madison.
1986d Coporaque in the 1615-1617 visita of Yanquecollaguas Hanansaya. In The
cultural ecology, archaeology, and history of terracing and terrace abandonment in
the Colca valley of Southern Peru. Technical report to the National Science
Madison.
1986e Coporaque toponyms: comparison and analysis of place names in the 16th,
17th, and 20th centuries. In The cultural ecology, archaeology, and history of
terracing and terrace abandonment in the Colca valley of Southern Peru. Technical
report to the National Science Foundation and the National Geographic Society,
edited by W. M. Denevan, pp. 450-492. vol. 1. Department of Geography,
University of Wisconsin-Madison, Madison.
1986f Introduction: ethnohistorical research for the terrace abandonment project,
Colca valley. In The cultural ecology, archaeology, and history of terracing and
terrace abandonment in the Colca valley of Southern Peru. Technical report to the
National Science Foundation and the National Geographic Society, edited by W. M.
Denevan, pp. 386-393. vol. 1. Department of Geography, University of WisconsinMadison, Madison.
1987a Anlisis del uso de tierras registrado en las visitas de los siglos XVI y XVII
a los Yanquecollaguas, Arequipa, Per. In Pre-hispanic agricultural fields in the
Andean region, edited by W. M. Denevan, K. Mathewson and G. Knapp, pp. 129145. International Series 359(1). British Archaeological Reports, Oxford.
1987b Apuntes histricos y etnogrficos del valle del ro Colca (Arequipa, Per)
1575-1980. Boletn de Lima 50:7-20.
1988a La divisin social y geogrfica hanansaya/urinsaya en el valle del Colca y la
provincia de Caylloma (Arequipa, Per). Boletn de Lima 60:49-53.
1988b The Yanque Parish Archive. In The Cultural ecology, archaeology, and
history of terracing and terrace abandonment in the Colca valley of southern Peru.
Society, edited by W. M. Denevan, pp. 25-37. vol. 2. Department of Geography,
606
1989 Las visitas a Yanquecollaguas de 1604 y 1615-1617: estructuras de mando y
tenencia y herencia de tierras. Bulletin del Instituto Francs de Estudio Andinos
18(2):241-267.
1990a El archivo parroquial de Yanque. Boletn del Instituto Riva Agero 17:47-70.
1990b Tenencia de tierras agrcolas en el valle del Colca (Caylloma, Arequipa).
Revista peruana de ciencias sociales 2(1):49-76.
1991 Libro de fbrica de la iglesia de Yanque. In Actas del primer congreso de
investigacin histrica, edited by H. Rodrguez Pastor, pp. 133-139. vol. 3.
CONCYTEC, Lima.
1992
The Yanque Parish Archive. The Americas April, XLVIII(4):519-535.
1994 The Franciscan church of Yanque (Arequipa) and Andean culture. The
Americas 50(3):419-436.
1995 Cambios en el paisaje agroecolgico de la provincia de Collaguas: un
anlisis de documentos en los archivos de Arequipa, Per. Revista del Archivo
Arzobispal de Arequipa 2:15-46.
1997 Las batallas de Chachayllo: la lucha por el agua de riego en el valle del
Colca (Arequipa, Per). Boletn de la Sociedad Geogrfica de Lima 110:127-148.
Bender, B.
1992 Theorizing landscape and the prehistoric landscapes of Stonehenge. Man
27:735-755.
Bernal, I.
1949 Exploraciones en Coixtlahuaca, Oaxaca. Revista mexicana de estudios
antropolgicos 10:5-76.
1958 Exploraciones en Cuilapan de Guerrero, 1902-1954. In Informes 7,
Direccin de monumentos prehispnicos. Instituto Nacional de Antropologa e
Historia, Mexico City.
1966 The Mixtecs in the archaeology of the valley of Oaxaca. In Ancient Oaxaca:
Discoveries in Mexican archaeology and history, edited by J. Paddock, pp. 345-366.
Stanford University Press, Stanford.
Bertonio, P. L.
1956 [1612]
Vocabulario de la lengua Aymara. Ediciones CERES, Cochabamba.
607
Billman, B. R.
1996 The evolution of prehistoric political organizations in the Moche Valley,
Peru. Ph.D. dissertation, University of California, Santa Barbara.
Binford, L., R.
1967 Smudge pits and hide smoking: the use of analogy in archaeological
reasoning. American antiquity 32(1):1-12.
Binford, M. W. and A. L. Kolata
1996 Then natural and human setting. In Tiwanaku and its hinterlands:
archaeology and paleoecology of an Andean civilization, edited by A. L. Kolata, pp.
23-56. vol. 1, Agroecology. Smithsonian Institution Press, Washington, DC.
Binford, M. W., A. L. Kolata, M. Brenner, M. T. Seddon, M. B. Abbott and J. Curtis
1997 Climate variation and the rise and fall of an Andean civilization. Quaternary
research 47:235-248.
Blaikie, P. M. and H. Brookfield
1987 Land degredation and society. Methuen, New York.
Blanton, R. E.
1978 Monte Albn: settlement patterns at the ancient Zapotec capital. Academic
Press, New York.
Blanton, R. E., S. Kowalewski, G. Feinman and J. Appel
1981 Ancient Mesoamerica: a comparison of change in three regions. Cambridge
University Press, Cambridge.
1982 Monte Alban's hinterland, part I: the prehispanic settlement patterns of the
central and southern parts of the valley of Oaxaca, Mexico. Memoirs of the
museum of anthropology, number 15. University of Michigan, Ann Arbor.
Blom, D. E., B. Hallgrimsson, L. Keng, M. C. Lozada C and J. E. Buikstra
1998 Tiwanaku 'colonization': bioarchaeological implications for migration in the
Moquega Valley, Peru. World archaeology 30(2):238-261.
Botkin, D. B.
1990 Discordant harmonies: a new ecology for the twenty first century. Oxford
University Press, New York.
Bourdieu, P.
608
1977 Outline of a theory of practice. Translated by R. Nice. Cambridge University
Press, Cambridge.
Braudel, F.
1972 The Mediterranean and the Mediterranean world in the age of Philip II.
Harper & Row, New York,.
Bray, T.
2003 Inca pottery as culinary equipment: food, feasting, and gender in imperial
state design. Latin American antiquity 14(1):1-22.
Brooks, S. O.
1998 Prehistoric agricultural terraces in the ro Japo basin, Colca valley, Peru.
Ph.D. dissertation, University of Wisconsin-Madison.
Brooks, S. O., M. D. Glascock and M. Giesso
1997 Source of volcanic glass for ancient Andean tools. Nature (386):449-450.
Brooks, S. O. and M. Olivares Ayala
1998 The pre-Inca Huarancante canal, Colca Valley, Peru. Paper presented at the
26th Annual Midwest Conference on Andean and Amazonian Archaeology,
Anthropology, and Ethnohistory, University of Illinois, Champaign-Urbana.
Brush, S. B.
1977 Mountain, field, and family: the economy of human ecology of an Andean
village. University of Philadelphia Press, Philadelphia.
Bryant, R.
1992 Political ecology: an emerging research agenda in third world studies.
Political geography 11:12-36.
Burger, R. L.
1989 Overview of Peruvian archaeology (1976-1986). Annual review of
anthropology 18:37-69.
Burger, R. L., F. Asaro, G. Salas and F. Stross
1998 The Chivay obsidian source and the geological origin of Titicaca Basin-type
obsidian artifacts. Andean past 5:203-223.
Burger, R. L., F. Asaro, P. Trawick and F. Stross
1998 The Alca obsidian source: the origin of raw material for Cuzco type obsidian
artifacts. Andean past 5:185-202.
609
Burger, R. L., K. L. M. Chvez and S. J. Chvez
2000 Through the glass darkly: Prehispanic obsidian procurement and exchange in
southern Peru and northern Bolivia. Journal of world prehistory. 14(o. 3, 2000.):.
Brgi, P. T.
1993 The Inka empire's expansion into the coastal sierra region west of Lake
Titicaca. Ph.D. dissertation, University of Chicago.
Cardona Rosas, A.
1993 Caractersticas geogrficas del patrn de asentamiento para el valle de
Chuquibamba - Arequipa, durante el perodo comprendido entre el Horizonte
Medio, y el Horizonte Tardo. Licenciado thesis, Universidad Catlica Santa Mara.
2002
Arqueologa de Arequipa: de sus albores a los Incas. CIARQ, Arequipa.
Caviedes, C. N.
1975 El Nino 1972: its climatic, ecological, human and economic implications.
Geographical review. 65(4):493-509.
1984
El Nio 1982-83. Geographical review 74:267-290.
Cock Carrasco, G.
1976-77
Los kurakas de los Collaguas: poder poltico y poder econmico.
Historia y cultura (10):95-118.
1978 Ayllu, territorio y frontera en los Collaguas. Etnohistria y antropologa
andina:29-32.
1981 El ayllu en la sociedad andina: alcances y perspectivas. Ethnohistria y
Antropologa andina:231-253.
Cook, A. P. and N. D. Cook
1991 Good faith and truthful ignorance: a case of transatlantic bigamy. Duke
University Press, Durham, North Carolina.
Cook, N. D.
1975 Introduccin. In Tasa de la visita general de Francisco de Toledo, edited by
N. D. Cook, A. Mlaga Medina and T. Bouysse-Cassagne. Universidad Nacional
Mayor de San Marcos, Lima.
1982 The people of the Colca valley: a population study. Westview Press,
Boulder.
610
1992 Luis Jernimo de Or: una aproximacin. In Symbolo catholico indiano, pp.
35-61, Australis, Lima.
2002 'Tomando posesin' Luis Gernimo de Or y el retorno de los franciscanos a
las doctrinas del valle del Colca. In El hombre y los Andes: homenaje a Franklin
Pease G.Y., edited by J. Flores Espinoza and R. Varn Gabai, pp. 889-903. vol. 2.
Pontificia Universidad Catlica del Per, Lima.
2003 Introduccin. In Collaguas II, Lari Collaguas: economa, sociedad y
poblacin, 1604-1605, edited by D. J. Robinson, pp. xv-xxxv. Pontifcia
Universidad Catlica del Per, Lima.
Cook, N. D., A. Mlaga Medina and T. Bouysse-Cassagne (editors)
1975 [1582] Tasa de la visita general de Francisco de Toledo. Universidad
Nacional Mayor de San Marcos, Lima.
Crdova Aguilar, H., L. Gonzalez Ilizarbe and C. Guevara Tello
1986 Agriculture in Coporaque. In The cultural ecology, archaeology, and history
of terracing and terrace abandonment in the Colca valley of Southern Peru.
Society, edited by W. M. Denevan, pp. 60-87. vol. 1. Department of Geography,
Costin, C. L. and T. K. Earle
1989 Status distinction and legitimation of power as reflected in changing patterns
of consumption in late prehispanic Peru. American antiquity (54):691-714.
Crespo, J. C.
1977 Los Collaguas en la visita de Alonso Fernndez de Bonilla. In Collaguas I,
edited by F. Pease G. Y., pp. 53-91. Pontfica Universidad Catlica del Per, Lima.
Crumley, C. L.
1975 Toward a locational definition of state systems of settlement. American
anthropologist 78:59-73.
1979 Three locational models: an epistemological assessment for anthropology
and archaeology. In Advances in archaeological method and theory, edited by M. B.
Schiffer, pp. 141-173. vol. 2. Academic Press, New York.
1987 A dialectical critique of hierarchy. In Power relations and state formation,
edited by T. C. Patterson and C. W. Gailey, pp. 155-169. Archeology Section,
American Anthropological Association, Washington, DC.
611
1994 Historical ecology: a multidimensional ecological orientation. In Historical
ecology: cultural knowledge and changing landscapes, edited by C. L. Crumley, pp.
1-16. School of American Research Press, Santa Fe.
1999 Sacred landscapes: Constructed and conceptualized. In Archaeologies of
landscape, edited by W. Ashmore and A. B. Knapp, pp. 269-276. Blackwell,
Oxford.
D'Altroy, T. N.
1987 Transitions in power: Centralization of Wanka political organization under
Inka rule. Ethnohistory 34(1):78-102.
1992 Provincial power in the Inka empire. Smithsonian Institution Press,
Washington.
2002
The Incas. The Peoples of America. Blackwell, Malden, Mass.
D'Altroy, T. N. and R. L. Bishop

1990 The provincial organization of Inka ceramic production. American antiquity
55(1):120-138.
D'Altroy, T. N., A. M. Lorandi and V. Williams
1994 Produccin y uso de cermica en la economa poltica Inka. In Tecnologa y
organizacin de la produccin cermica prehispnica en los Andes, edited by I.
Shimada, pp. 395-441. Pontifica Universidad Catlica del Per Fondo Editorial
1994, Lima.
Dvila, J. and J. C. Thouret
1999 Volcn Huaynaputina (sur del Per): erupcin pliniana e ignimbrtica de
gran magnitud en 1600 D.C. In Volumen Jubilar No 5, 75 Aniversario Sociedad
Geolgica del Per, edited by J. Machar, V. Benavides and S. Rosas, pp. 141-168.
Sociedad Geolgica del Per, Lima.
de la Vera Cruz Chvez, P.
1987 Cambios en los patrones de asentamiento y el uso y abandono de los andenes
en Cabanaconde, Valle del Colca, Per. In Pre-Hispanic agricultural fields in the
Andean region, edited by W. M. Denevan, K. Mathewson and G. Knapp, pp. 89128. B.A.R. International Series, Oxford.
1988 Estudio arqueolgico en el valle de Cabanaconde, Arequipa. BA Thesis,
Universidad Catlica Santa Mara.
612
1989 Cronologa y corologa de la cuenca del ro Caman - Majes - Colca Arequipa. Licentiature Thesis, Universidad Catlica Santa Mara.
1996 El papel de la sub region norte de los valles occidentales en la articulacion
entre los Andes centrales y los Andes centro sur. Integracion Surandina: Cinco
Siglos Despues:135-157.
Deagan, K. A.
1987 Artifacts of the Spanish colonies of Florida and the Caribbean, 1500-1800.
Smithsonian Institution Press, Washington, D.C.
Denevan, W. M.
1980 Tipologa de configuraciones agrcolas prehispnicas. Amrica Indgena
40:610-652.
1986a The cultural ecology, archaeology, and history of terracing and terrace
abandonment in the Colca valley of Southern Peru. Technical report to the
National Science Foundation and the National Geographic Society. 1. Department
of Geography, University of Wisconsin-Madison, Madison.
1986b Introduction: the Ro Colca Abandoned Terrace Project. In The cultural
Foundation and the National Geographic Society, edited by W. M. Denevan, pp. 846. vol. 1. Department of Geography, University of Wisconsin-Madison, Madison.
1987 Terrace abandonment in the Colca Valley. In Prehispanic agricultural fields
in the Andean region, edited by W. M. Denevan, K. Mathewson and G. Knapp, pp.
1-43. British Archaeological Reports. vol. 359. B.A.R. International Series, Oxford.
1988a The cultural ecology, archaeology, and history of terracing and terrace
abandonment in the Colca valley of southern Peru. Technical report to the National
Science Foundation and the National Geographic Society. 2. 2 vols. Department of
Geography, University of Wisconsin-Madison, Madison.
1988b Measurement of abandoned terracing from air photos, Colca valley, Peru.
Yearbook, Conference of Latin Americanist Geographers 14:20-30.
2001 Cultivated landscapes of Native Amazonia and the Andes: triumph over the
soil. Oxford University Press, Oxford.
613
Denevan, W. M. and L. Hartwig
1986 Measurement of terrace abandonment in the Colca valley. In The cultural
Madison.
Denevan, W. M., K. Mathewson and G. W. Knapp (editors)
1987 Pre-Hispanic agricultural fields in the Andean region. B.A.R. International
Series, Oxford.
Denevan, W. M., J. M. Treacy and J. A. Sandor
1986 Physical geography of the Coporaque region. In The cultural ecology,
archaeology, and history of terracing and terrace abandonment in the Colca valley
of Southern Peru. Technical report to the National Science Foundation and the
National Geographic Society, edited by W. M. Denevan, pp. 47-59. vol. 1.
Department of Geography, Unversity of Wisconsin-Madison, Madison.
Dick, R. P., J. A. Sandor and N. S. Eash
1994 Soil enzyme activities after 1500 years of terrace agriculture in the Colca
valley, Peru. Agriculture, ecosystems, and environment 50:123-131.
Dillehay, T. D. (editor)
1995 Tombs for the living: Andean mortuary practices. Dumbarton Oaks,
Washington, D.C.
Dillehay, T. D. and L. Nez
1988 Camelids, caravans, and complex societies in the south-central Andes. In
Recent studies in pre-Columbian archaeology, edited by N. J. Saunders and O. de
Montmollin, pp. 603-634. BAR International Series. vol. 421, Oxford.
Donkin, R. A.
1979 Agricultural terracing in the aboriginal new world. Viking Fund
Publications in Anthropology. University of Arizona Press, Tucson, Arizona.
Duviols, P.
1973 Huari y llacuaz. Agricultores y pastores. Un dualismo prehispanico de
oposicion y complementaridad. Revista. tomo 39:153-191.
Earle, T. K.
614
1987 Specialization and the production of wealth: Hawaiian chiefdoms and Inka
power. In Specialization, exchange, and complex societies, edited by E. M. Brumfiel
and T. K. Earle, pp. 64-75. Cambridge University Press, Cambridge.
Eash, N. S. and J. A. Sandor
1995 Soil chronosequence and geomorphology in a semiarid valley in the Andes
of southern Peru. Geoderma 65(59-79).
Ehrenreich, R. M., C. L. Crumley and J. E. Levy (editors)
1995 Heterarchy and the analysis of complex societies. Archeological papers of
the American Anthropological Association ; no. 6. American Anthropological
Association, Arlington.
Erickson, C. L.
1988 An archaeological investigation of raised field agriculture in the Lake
Titicaca Basin of Peru. Ph.D. Dissertation, University of Illinois at UrbanaChampaign.
1999 Neo-environmental determinism and agrarian 'collapse' in Andean
prehistory. Antiquity 73(281).
2000 The Lake Titicaca Basin: a precolumbian built landscape. In Imperfect
balance: Landscape transformations in the precolumbian Americas, edited by D. L.
Lentz, pp. 311-356. Columbia University Press, New York.
Falconer, S. E. and S. H. Savage
1995 Heartlands and hinterlands: alternative trajectories of early urbanization in
Mesopotamia and the southern Levant. American antiquity 60:37-58.
Femenias, B. B.
1997 Ambiguous emblems: gender, clothing, and representation in contemporary
Peru. Ph.D. dissertation, University of Wisconsin-Madison.
1998 Ethnic artists and the appropriation of fashion: embroidery and identity in
Caylloma, Peru. Chungara 30(2).
Fernandez Baca, C.
1971 Motivos de ornamentacin de la cermica Inca-Cuzco. Libreria Studium,
S.A., Lima.
Gade, D. W. and M. Escobar
1982 Village settlement and the colonial legacy in southern Peru. Geographical
review 72:430-449.
615
Galdos Rodrguez, G.
1984 Expansin de los Collaguas hacia el valle de Arequipa. El Derecho 296:81152.
1987 Comunidades prehispnicas de Arequipa. Fundacin Manuel J. Bustamante
De la Fuente, Arequipa.
Garca Mrquez, M. and R. Bustamante Montoro
1990 Arqueologa del valle de Majes. Gaceta arqueolgica andina V(18/19):2540.
Gelles, P. H.
1990 Channels of power, fields of contention: the politics and ideology of
irrigation in an Andean peasant community (Peru). Ph.D. dissertation, Harvard
University.
1993 Cabaneo ethnohydrology: the cosmological referents and historical roots of
an Andean irrigation system. Culture and environment: a fragile coexistence:24.
1995 Equilibrium and extraction: dual organization in the Andes. American
Ethnologist. 22(o. 4, 1995.):.
2000 Water and power in highland Peru. Rutgers University Press, New
Brunswick.
Giddens, A.
1979
Central problems of social theory. University of California Press, Berkeley.
Gifford, J. C.
1960 The type-variety method of ceramic classification as an indicator of cultural
phenomena. American antiquity 25(3):341-347.
Gisbert, T. and J. d. Mesa
1985 Arquitectura andina, 1530-1830: historia y anlisis. Embajada de Espaa en
Bolivia, La Paz.
Glave Testino, L. M. and M. I. Remy
1983 Estructura agraria y vida rural en una regin andina : Ollantaytambo entre
los siglos XVI-XIX. Archivos de historia andina 3. Centro de Estudios Rurales
Andinos "Bartolom de las Casas", Cusco.
Gledhill, J.
616
1988 Legacies of empire: political centralization and class formation in the
Hispanic-American world. In State and society: the emergence and development of
social hierarchy and political centralization, edited by J. Gledhill, B. Bender and T.
L. Mogens, pp. 302-319. Unwin Hyman, London.
Goldstein, P. S.
1989 Omo, a Tiwanaku provincial center in Moquegua, Peru. Ph.D. dissertation,
University of Chicago.
1993 House, community, and the state in the earliest Tiwanaku colony: Domestic
patterns and state integration at Omo M12, Moquegua. In Domestic architecture,
ethnicity, and complementarity in the south-central Andes, edited by M. S.
Aldenderfer, pp. 25-41. University of Iowa Press, Iowa City.
2000 Communities without borders: The vertical archipelago and diaspora
communities in the southern Andes. In The archaeology of communities, edited by
M. A. Canuto and J. Yaeger, pp. 182-209. Routledge, New York.
Gonzlez, F. L.
1995 Fences, fields and fodder: Enclosures in Lari, Valle del Colca, Southern
Peru. M.S., University of Wisconsin-Madison.
Gonzlez Holgun, D. and R. Porras Barrenechea
1989 [1608] Vocabulario de la lengua general de todo el Per, llamada lengua
qquichua o del Inca. 3a ed. Universidad Nacional Mayor de San Marcos, Lima.
Graffam, G.
1990 Raised fields without bureaucracy: an archaeological examination of
intensive wetland cultivation in the Pampa Koani zone, Lake Titicaca, Bolivia.
Ph.D. Dissertation, University of Toronto.
1992 Beyond state collapse: rural history, raised fields, and pastoralism in the
south Andes. American anthropologist 94(4):882-904.
Grosboll, S.
1987 Ethnic boundaries within the Inca Empire: evidence from Huanuco, Peru.
Ethnicity and culture:18.
Guaman Poma de Ayala, F., J. V. Murra, R. Adorno and J. Urioste
1987 Nueva cronica y buen gobierno. 1a ed. Cronicas de America ; 29a-29c.
Historia 16, Madrid.
617
Guerra Santander, E. M. and P. A. Aquize Cceres
1996 Patrn arquitectnico y patrn de asentamiento del sitio prehispnico de
Uskallacta Chivay (Valle del Colca) - Arequipa. Licenciado Thesis, Universidad
Catlica Santa Mara.
Guevara-Gil, A. and F. Salomon
1994 A 'personal visit': colonial political ritual and the making of Indians in the
Andes. Colonial Latin American review 3(1-2):3-36.
Guillet, D. W.
1978 The supra-household sphere of production in the Andean peasant economy.
Paper presented at the Actes du XXLe Congrs International des Amricanistes,
Paris.
1981 Land tenure, ecological zone, and agricultural regime in the central Andes.
American ethnologist 8(1):139-156.
1987 Terracing and irrigation in the Peruvian highlands. Current anthropology.
28(o. 4, 1987,):.409-430,.
1992 Covering ground: communal water management and the state in the
Peruvian highlands. University of Michigan Press, Ann Arbor.
Guillet, D. W., L. Furbee, J. A. Sandor and R. A. Benfer
1995 On a methodology for combining cognitive and behavioral research: the Lari
soils project in Peru. In The cultural dimension of development: indigenous
knowledge systems, edited by D. M. Warren, L. J. Slikkerveer and D. Brokensha,
pp. 71-81. Intermediate Technology Publications, London.
Haggett, P.
1965
Locational analysis in human geography. Edward Arnold, London.
Hastorf, C. A. (editor)
1999 Early settlement at Chiripa, Bolivia: research of the Taraco archaeological
project. University of California, Berkeley.
Hemming, J.
1983 The conquest of the Incas. Rev. ed. Penguin Books, Harmondsworth,
Middlesex, England ; New York, N.Y., U.S.A.
Hyslop, J.
1976 An archaeological investigation of the Lupaqa kingdom and its origins.
Ph.D. dissertation, Columbia University.
618
1984
The Inka road system. Studies in archaeology. Academic Press, Orlando.
1990
Inka settlement planning. 1st ed. University of Texas Press, Austin.
Isbell, W. H.
1997 Mummies and mortuary monuments: a postprocessual prehistory of central
Andean social organization. 1st ed. University of Texas Press, Austin.
2000 What we should be studying: the "imagined community" and the "natural
community". In The archaeology of communities, edited by M. A. Canuto and J.
Yaeger, pp. 243-266. Routledge, New York.
Izaguirre Urbano, L. E.
n.d.
Estudio del subsistema agrcola del Distrito de Coporaque. Master's Thesis,
Universidad Nacional Agraria.
Janusek, J. W.
1999 Craft and local power: Embedded specialization in Tiwanaku cities. Latin
American antiquity 10(2):107-131.
Jennings, J. M.
2002a Description and method of exploitation of the Alca obsidian source, Peru.
Latin American antiquity 13(1):107-118.
2002b Prehistoric imperialism and cultural development in the Cotahuasi valley,
Peru. Department of Anthropology, University of California, Santa Barbara.
Submitted to Ph.D. dissertation.
Johannessen, S. and C. A. Hastorf (editors)
1994 Corn and culture in the prehistoric New World. Westview Press, Boulder.
Johnson, A. M.
1976 The climate of Peru, Bolivia, and Ecuador. In Climates of Central and South
America, edited by W. Schwerdtfeger, pp. 147-200. World survey of climatology.
vol. 12. Elsevier Scientific Publishing Company, New York.
Johnson, G. A.
1987 The changing organization of Uruk administration on the Susiana plain. In
Archaeological perspectives on western Iran: from prehistory to the Islamic
conquest, edited by F. Hole, pp. 107-140. Smithsonian Institution Press,
Washington, DC.
619
Joyce, R. A. and J. A. Hendon
2000 Heterarchy, history, and material reality: "communities" in Late Classic
Honduras. In The archaeology of communities, edited by M. A. Canuto and J.
Yaeger, pp. 143-160. Routledge, New York.
Julien, C. J.
1982 Inca decimal administration in the Lake Titicaca region. Inca and Aztec
states, 1400-1800: anthropology and history, edited by G. A. Collier, R. Rosaldo
and J. D. Wirth, pp. 119-151. Academic Press, New York.
1983 Hatunqolla, a view of Inca rule from the Lake Titicaca region. University of
California publications in anthropology ; v. 15. University of California Press,
Berkeley.
1985 Guano and resource control in sixteenth-century Arequipa. In Andean
ecology and civilization: an interdisciplinary perspective on Andean ecological
complementarity, edited by S. Masuda, I. Shimada and C. Morris, pp. 185-231.
University of Tokyo Press, Tokyo.
1988
How Inca decimal administration worked. Ethnohistory 35:257-279.
1993 Finding a fit: archaeology and ethnohistory of the Incas. In Provincial Inca,
edited by M. A. Malpass, pp. 177-233. University of Iowa Press, Iowa City.
1998 La encomienda del Inca. In Actas del IV congreso internacional de
etnohistoria, pp. 489-516. vol. II. Pontfica Universidad Catlica Fondo Editorial,
Lima.
2000
Reading Inca history. University of Iowa Press, Iowa City.
Julien, D. G.
1993 Late pre-Inkaic ethnic groups in Highland Peru: an archaeologicalethnohistorical model of the political geography of the Cajamarca region. Latin
Kidder, A. V.
1915 Pottery of the pajarito plateau and of some adjacent regions in New Mexico.
American Anthropological Association, Memoir 2, No. 6. American
Anthropological Association.
1916 Archaeological explorations at Pecos, New Mexico. Paper presented at the
Proceedings of the National Academy of Science.
620
Kirch, P. V. and T. L. Hunt
1997 Historical ecology in the Pacific Islands: prehistoric environmental and
landscape change. Yale University Press, New Haven.
Kirchoff, P.
1949 The social and political organization of the Andean peoples. In Handbook of
South American Indians, edited by J. Steward, pp. 293-311. Bureau of American
Ethnology Bulletin 143. vol. 5. U.S. Government Printing Office, Washington, DC.
Klink, C. and M. S. Aldenderfer
in press
A projectile point chronology for the south-central Andean
Highlands. In Titicaca Basin prehistory, edited by C. Stanish, A. Cohen and M. S.
Aldenderfer. Cotsen Institute of Archaeology, Los Angeles.
Knapp, A. B. and W. Ashmore
1999 Archaeological landscapes: constructed, conceptualized, ideational. In
Archaeologies of landscape, edited by W. Ashmore and A. B. Knapp, pp. 1-32.
Blackwell, Oxford.
Kolata, A. L.
1986 Agricultural foundations of the Tiwanaku state: a view from the heartland.
1993
The Tiwanaku. Basil Blackwell, Oxford.
1996a Tiwanaku and its hinterland: archaeology and paleoecology of an Andean

civilization. Smithsonian Institution Press, Washington, DC.
1996b Tiwanaku and its hinterlands: archaeology and paleoecology of an Andean
civilization. 1, Agroecology. Smithsonian Institution Press, Washington, DC.
Kolb, M. J. and J. E. Snead
1997 It's a small world after all: Comparative analyses of community organization
in archaeology. American antiquity 62(4):609-628.
Kroeber, A. L.
1916 Zuni potsherds. Anthropological papers, American Museum of Natural
History 18.
1944 Peruvian archeology in 1942. Viking Fund Publications in Anthropology,
Number 4. Viking Fund, Inc., New York.
621
La Capra, D.
1985 History and criticism. Cornell University Press, Ithaca.
Lansing, J. S.
1991 Priests and programmers: technologies of power in the engineered
landscape of Bali. Princeton University Press, Princeton, N.J.
Larson, B.
1998 Cochabamba, 1550-1900: colonialism and agrarian transformation in
Bolivia. Duke University Press, Durham.
Levillier, R.
1935 Don Francisco de Toledo, supremo organizador del Per. Coleccin de
publicaciones histricas de la Biblioteca del Congreso argentino. 3 vols, Buenos
Aires.
LeVine, T. (editor)
1992 Inka storage systems. University of Oklahoma Press, Norman.
Linares Delgado, L. R.
1988 Estudio arqueolgico en el valle de Huanca - Arequipa. B.A. thesis,
1989 Cronologa y relaciones culturales pre-hispnicas del Valle del Chili Arequipa. Licenciado thesis, Universidad Catlica Santa Mara.
Linares Mlaga, E.
1993 Arequipa tierra ma. Correo, Arequipa.
Lister, F. C. and R. H. Lister
1987 Andalusian ceramics in Spain and New Spain : a cultural register from the
third century B.C. to 1700. University of Arizona Press, Tucson.
Lumbreras, L. G.
1974a Los reinos post-Tiwanaku en el rea altiplnica. Revista del Museo Nacional
40:55-85.
1974b The peoples and cultures of ancient Peru. Smithsonian Institution Press,
Washington, DC.
Lumbreras, L. G. and H. Amat Olazaval
622
1968 Secuencia arqueolgica del altiplano occidental del Titicaca. In Actas y
memorias, pp. 75-106. vol. 2. 37th International Congress of Americanists, Buenos
Aires.
MacCormack, S.
1991 Religion in the Andes: vision and imagination in early colonial Peru.
Princeton University Press, Princeton, N.J.
Mlaga Medina, A.
1974 Las reducciones en el Per (1532-1600). Historia y cultura (8):141-172.
1977 Los Collagua en la historia de Arequipa en el siglo XVI. In Collaguas I,
edited by F. Pease G. Y., pp. 93-130. Pontifica Universidad Catolica del Peru, Lima.
1981
Arequipa, estudios histricos. 1a ed. Biblioteca Arequipa, Arequipa, Peru.
Mlaga Nez-Zeballos, A.
2002 El enojo de los dioses: terremotos y erupciones en Arequipa del siglo XVI.
In El hombre y los Andes: homenaje a Franklin Pease G.Y., edited by J. Flores
Espinoza and R. Varn Gabai, pp. 905-913. vol. 2. Pontificia Universidad Catlica
del Per, Lima.
Malpass, M. A.
1986 Late prehistoric terracing at Chirja in the Colca valley, Peru: preliminary
report. In Perspectives on Andean prehistory and protohistory, edited by D.
Sandweiss and P. Kvietok, pp. 19-34. Cornell University Latin American Studies
Program, Ithaca.
1987 Late prehistoric terracing at Chijra in the Colca valley, Peru: Preliminary
report II. In Prehistoric agricultural fields in the Central Andes, edited by W. M.
Denevan and K. Mathewson. British Archaeological Reports, Oxford.
1988 Irrigated versus non-irrigated terracing in the Andes: Environmental
considerations. Paper presented at the 7th Northeast Conference of Andean
Archeology and Ethnohistory, Amherst, MA November 5-6, 1998, Amherst, Ma.
Malpass, M. A. and P. de la Vera Cruz Chvez
1986 Ceramic Sequence from Chijra, Colca Valley, Peru. In The cultural ecology,
archaeology, and history of terracing and terrace abandonment in the Colca valley
of Southern Peru. Technical report to the National Science Foundation and the
Department of Geography, University of Wisconsin-Madison, Madison.
623
1990 Cronologa y secuencia de la cermica de Chijra, valle del Colca. Gaceta
Arqueolgica Andina 18/19:41-57.
Marcus, J.
2000 Toward an archaeology of communities. In The archaeology of communities,
edited by M. A. Canuto and J. Yaeger, pp. 231-242. Routledge, New York.
Martin, D.
1986 Archaeology of terraces and settlement at Chilacota, Coporaque. In The
cultural ecology, archaeology, and history of terracing and terrace abandonment in
the Colca valley of southern Peru. Technical report to the National Science
Madison.
Masuda, S., I. Shimada and C. Morris (editors)
1985 Andean ecology and civilization: an interdisciplinary perspective on Andean
ecological complementarity. University of Tokyo Press, Tokyo.
Matienzo, J. d.
1910 [1567] Gobierno del Per. Compaa sud-americana de billetes de banco,
Buenos Aires,.
Mayer, E.
1985 Production zones. In Andean ecology and civilization: An interdisciplinary
perspective on Andean ecological complementarity, edited by S. Masuda, I.
Shimada and C. Morris, pp. 45-84. University of Tokyo Press, Tokyo.
McEwan, B. G.
1992 The role of ceramics in Spain and Spanish America during the 16th century.
Historical Archaeology 26(1):92-108.
McGlade, J.
1995 Archaeology and the ecodynamics of human-modified landscapes. Antiquity
69:113-132.
Mitchell, W. P.
1976 Irrigation and community in the central Peruvian highlands. American
anthropologist 78(1):25-44.
Moore, J. D.
1996 Archaeology of plazas and the proxemics of ritual: three Andean traditions.
American anthropologist 98(4):789-802.
624
Morante, J. M.
1939 Arqueologa del Departamento de Arequipa; Condesuyos y Caman
precolumbinos. Ph.D. dissertation, Universidad Nacional de San Agustn.
Morris, C.
1966
El tampu real de Tunsucancha. Cuadernos de investigacion (1):95-116.
1972 State settlements in Tawantinsuyu. A strategy of compulsory urbanism. In

Contemporary archaeology: A guide to theory and contributions, edited by M. P.
Leone, pp. 393-401. Southern Illinois University Press, Carbondale, Illinois.
Morris, C. and D. E. Thompson
1985 Hunuco Pampa: An Inca city and its hinterland. Thames and Hudson,
London.
Moseley, M. E.
1992 The Incas and their ancestors. Thames and Hudson, New York.
Moseley, M. E., R. A. Feldman, P. S. Goldstein and L. Watanabe
1991 Colonies and conquest: Tiahuanaco and Huari in Moquegua. In Huari
administrative structure: prehistoric monumental architecture and state
government, edited by W. H. Isbell and G. F. McEwan, pp. 121-140. Dumbarton
Oaks, Washington, DC.
Moseley, M. E., P. S. Goldstein, L. Watanabe and R. A. Feldman
1991 Colonies and conquest: Tiahuanaco and Huari in Moquegua. Huari
administrative structure: prehistoric monumental architecture and state
government:121-140.
Murdock, G. P.
1949 Social structure. Macmillan Co., New York,.
Murra, J. V.
1956 The economic organization of the Inca state. Ph.D. dissertation, University
of Chicago.
1962
The function of cloth in the Inca state. American anthropologist 64:710-728.
1964 Una apreciacin etnolgica de la visita. In Visita hecha a la provincia de

Chucuito por Garci Diez de San Miguel en el ao 1567, edited by J. V. Murra, pp.
419-444. Casa de la Cultura del Per, Lima.
625
1968
An Aymara kingdom in 1567. Ethnohistory 15:115-151.
1972 El control vertical de un mximo de pisos ecolgicos en la economia de las

sociedades andinas. In Visita de la Provincia de Len de Hunuco II, edited by J. V.
Murra, pp. 429-476. Universidad Hermilio Valdizn, Hunuco, Peru.
1980
The economic organization of the Inka State. JAI Press, Greenwich, Conn.
Murra, J. V. and C. Morris

1976 Dynastic oral traditon, administrative records and archaeology in the Andes.
World archaeology. 7(3):269-279.
Neira Avendao, M.
1961 Los Collaguas. MA thesis, Universidad Nacional de San Agustin.
1968 Un nuevo complejo ltico y pinturas rupestres en la gruta Su-3 de Sumbay.
Universidad Nacional de San Agustn revista de la facultad de letras 5:43-75.
1986 Excavaciones arqueolgicas en las ruinas de Chishra (Chijra), Coporaque. In
The cultural ecology, archaeology, and history of terracing and terrace
abandonment in the Colca valley of Southern Peru. Technical report to the
National Science Foundation and the National Geographic Society, edited by W. M.
Denevan, pp. 167-197. vol. 1. Department of Geography, University of WisconsinMadison, Madison.
1990 Arequipa prehispnica. In Historia general de Arequipa, edited by M. Neira
Avendao, G. Galdos Rodrguez, A. Mlaga Medina, E. Q. Paz Soldn and J. G.
Carpio Muoz, pp. 5-213. Fundacin M.J. Bustamante de la Fuente, Arequipa.
1998 Arqueologa de Arequipa. Cronos 1:9-50.
Neira Avendao, M. and A. Cardona Rosas
2000-01 El Periodo Formativo en el rea de Arequipa. In Proyecto arqueolgico
condesuyus, edited by M. Zilkowski and A. Beln Franco, pp. 27-60. Andes:
Boletn de la Misin Arqueolgica Andina. vol. 1, no. 3. Warsaw University,
Warsaw.
Nelson, N. C.
1916 The chronology of the Tano ruins, New Mexico. American anthropologist
18:159-180.
626
Niles, S. A.
1984 Architectural form and social function in Inca towns near Cuzco. BAR
international series 210:205-223.
1987 Callachaca: style and status in an Inca community. University of Iowa
Press, Iowa City.
1993 The provinces in the heartland: stylistic variation and architectural
innovation near Inca Cuzco. In Provincial Inca: archaeological and ethnohistorical
assessment of the impact of the Inca state, edited by M. A. Malpass, pp. 145-176.
University of Iowa Press, Iowa City.
ONERN
1973 Inventario, evaluacin y uso racional de los recursos naturales de la costa:
Cuenca de Ro Caman-Majes 2. 2 vols. Oficina Nacional de Evaluacin de
Recursos Naturales, Lima.
Oquiche Hernani, A. A.
1991 La ocupacin Collagua Inca en el sector de Achoma Valle del Colca. B.A.
thesis, Universidad Catlica Santa Mara.
Or, L. J.
1992 [1598]
Symbolo catholico indiano. Australis, Lima.
Orlove, B. S.
1994 Sticks and stones: ritual battles and play in the southern Peruvian Andes. In
Unruly order: violence, power, and cultural identity in the high provinces of
southern Peru, edited by D. Poole, pp. 133-164. Westview Press, Boulder, CO.
Orlove, B. S., C. H. Chiang and M. A. Cane
2000 Forecasting Andean rainfall and crop yield from the influence of El Nio on
Pleiades visibility. Nature 403:68-71.
Ortloff, C. R. and A. L. Kolata
1993 Climate and collapse: agro-ecological perspectives on the decline of the
Tiwanaku state. Journal of archaeological science.
Owen, B.
1993 A model of multiethnicity: state collapse, competition, and social complexity
from Tiwanaku to Chiribaya in the Osmore valley, Peru. Ph.D. dissertation,
University of California, Los Angeles.
627
Palacios, O., J. de la Cruz, N. de la Cruz, B. A. Klinck, R. A. Allison and M. P. Hawkins
1993 Geologa de la cordillera Occidental y altiplano al oeste de Lago Titicaca
sur del Per. Boletn No. 42. INGEMMET, Repblica del Per, Sector Energa y
Metalrgico, Lima.
Parker, A. C.
1916 Excavations in an Erie Indian village and burial site at Ripley, Chautauqua
County, New York. New York State Museum, Bulletin 117.
Parker, A. C. and M. R. Harrington
1922 Archaeological history of New York. New York State Museum, Bulletins
235-236.
Parsons, J. R., C. M. Hastings and R. Matos M
1997 Rebuilding the state in highland Peru: herder-cultivator interaction during
the late Intermediate period in the Tarama-Chinchaycocha region. Latin American
antiquity. 8(4):317-341
Parsons, J. R., C. M. Hastings and R. Matos Mendieta
2001 Prehispanic settlement patterns in the upper Mantaro and Tarma drainages,
Junin, Peru. Memoirs of the Museum of Anthropology, University of Michigan.
University of Michigan Museum, Ann Arbor.
Prssinen, M.
1992 Tawantinsuyu: the Inca state and its political organization. SHS, Helsinki.
Prssinen, M. and A. Siirriinen
1997 Inka-style ceramics and their chronological relationship to the Inka
expansion in the southern Lake Titicaca area (Bolivia). Latin American antiquity
8(3):255-271.
Patterson, T. C.
1994 Toward a properly historical ecology. In Historical ecology: cultural
knowledge and changing landscapes, edited by C. L. Crumley, pp. 223-237. School
of American Research Press, Santa Fe.
Patterson, T. C. and C. W. Gailey (editors)
1987 Power relations and state formation. American Anthropological
Association, Washington, D.C.
Paynter, R. W.
628
1983 Expanding the scope of settlement analysis. In Archaeological hammers and
theories, edited by J. A. Moore and A. S. Keene, pp. 233-275. Academic Press, New
York.
Pease G. Y., F.
1976-77 Etnohistoria andina: un estado de la cuestin. Historia y cultura 10:206228.
1977 Collaguas: una etna del siglo XVI. In Collaguas I, edited by F. Pease, pp.
131-168. Pontifica Universidad Catolica del Peru, Lima.
1981 Ayllu y parcialidad, reflexiones sobre el caso de Collaguas. In Etnohistoria y
antropologa andina, segunda jornada del Museo Nacional de Historia, edited by
M. K. d. Paredes and M. M. d. Pease, pp. 19-34. Museo Nacional de Historia y la
Comisin para Intercambio Educativo entre los Estados Unidos y el Per, Lima.
Phillips, P.
1958 Application of the Wheat-Gifford-Wasley taxonomy to eastern ceramics.
Platt, T.
1982 Estado boliviano y ayllu andino: tierra y tributo en el norte de Potos.
Instituto de Estudios Peruanos, Lima.
1986 Mirrors and maize: the concept of yanantin among the Macha of Bolivia. In
Anthropological history of Andean polities, edited by J. V. Murra, N. Wachtel and J.
Revel, pp. 228-259. Cambridge University Press, New York.
Polo de Ondegardo, J.
1917 [1571] Relacin de los fundamentos acerca del notable dao que resulta de
no guardar a los indios sus fueros..." In Coleccin de libros y documentos referentes
a la historia del Per, pp. 45-188. vol. 1, no. 3, Lima.
Protzen, J.-P. and R. Batson
1993 Inca architecture and construction at Ollantaytambo. Oxford University
Press, New York.
Quilter, J. and G. Urton
2002 Narrative threads: accounting and recounting in Andean Khipu. University
of Texas Press, Austin.
629
Ralegh, W.
1997 The discoverie of the large, rich, and bewtiful Empyre of Guiana, edited by
N.L. Whitehead. University of Oklahoma Press, Norman.
Ratti de Luchi Lomellini, M. and A. Zegarra Arenas
1987 Reconocimiento del yacimiento arqueolgico de Beringa. BA thesis,
Redfield, R.
1955 The little community; viewpoints for the study of a human whole. University
of Chicago Press, Chicago.
1956 Peasant society and culture; an anthropological approach to civilization.
University of Chicago Press, Chicago.
Reinhard, J.
1998 Discovering the Inca ice maiden: my adventures on Ampato. National
Geographic, Washington, DC.
Rice, D. S., C. Stanish and P. R. Scarr (editors)
1989 Ecology, settlement and history in the Osmore drainage, Peru. B.A.R.
International Series, 545, Oxford, England.
Rice, P. M.
1987
138.
The Moquegua Bodegas Survey. National Geographic research 3(2):136-
1994 Kilns of Moquegua, Peru: technology, excavations, and functions. Journal of

field archaeology 21:325-344.
1997 Tin-enameled wares of Moquegua, Peru. In Approaches to the historical
archaeology of Mexico, Central, and South America, edited by J. Gasco, G. C.
Smith and P. Fournier-Garca, pp. 173-180. The Institute of Archaeology,
University of California, Los Angeles, Los Angeles.
Rice, P. M. and S. L. Van Beck
1993 Spanish colonial kiln tradition of Moquegua, Peru. Historical archaeology
27:65-81.
Ritchie, W. A.
1932 The Algonkin sequence in New York. American anthropologist 34:406-414.
1938 A perspective of northeastern archaeology. American antiquity 4(2):94-112.
630
Robinson, D. J.
2003 Estudio. In Collaguas II; Lari Collaguas, edited by D. J. Robinson, pp.
xxxvii-cxii. Pontificia Universidad Catlica del Per, Lima.
Romaa, M. d., J. Blassi and J. Blassi
1987 Descubriendo el valle del Colca. Francis O. Patthey and Sons, Madrid.
Rossignol, J. and L. Wandsnider (editors)
1992 Space, time, and archaeological landscapes. Plenum Press, New York.
Rostworowski de Diez Canseco, M.
1983 Estructuras andinas del poder: ideologia religiosa y politica. Instituto de
Estudios Peruanos, Lima.
Rowe, J.
1961
Stratigraphy and seriation. American antiquity 26(3):324-330.
Rowe, J. H.
1944 An introduction to the archaeology of Cuzco. The Museum, Cambridge,
Mass.,.
1946a Inca culture at the time of the Spanish conquest. In Handbook of South
American Indians, edited by J. H. Steward, pp. 183-330. vol. 2. Smithsonian
Institution Bureau of American Ethnology, Washington, D.C.
1962 Stages and periods in archaeological interpretation. Southwestern journal of
anthropology 18(1):40-54.
1985
La constitucin Inca del Cuzco. Histrica 9(1):35-73.
Rydn, S.
1957 Andean excavations I. Publication no. 4. The Ethnographical Museum of
Sweden, Stockholm.
Saignes, T.
1985 Los Andes orientales : historia de un olvido. Instituto Frances de Estudios
Andinos, Lima, Peru.
1995 Indian migration and social change in seventeenth-century Charcas. In
Ethnicity, markets, and migration in the Andes: at the crossroads of history and
anthropology, edited by B. Larson and O. Harris, pp. 167-195. Duke University
Press, Durham.
631
Saignes, T., P. Garner and T. Platt
1985 Caciques, tribute and migration in the Southern Andes : Indian society and
the 17th century colonial order: (Audencia de Charcas). Occasional papers
(University of London. Institute of Latin American Studies), 15. University of
London Institute of Latin American Studies, London.
Salomon, F.
1986 Native lords of Quito in the age of the Incas: The political economy of northAndean chiefdoms. Cambridge University Press, Cambridge.
1991 Introductory essay: The Huarochir manuscript. In The Huarochir
manuscript, pp. 1-38. Translated by F. S. a. G. L. Urioste. University of Texas
Press, Austin.
1995 'The beautiful grandparents': Andean ancestor shrines and mortuary ritual as
seen through colonial records. In Tombs for the living: Andean mortuary practices,
edited by T. D. Dillehay, pp. 315-353. Dumbarton Oaks, Washington, DC.
Sanders, W. T., J. R. Parsons and R. S. Santley
1979 The Basin of Mexico: ecological processes in the evolution of a civilization.
Academic Press, New York.
Sandor, J. A.
1986 Report on soils in agricultural terraces in the Colca valley. In The cultural
Madison.
1987a Initial investigation of soils in agricultural fields in the Colca valley, Peru. In
Pre-Hispanic agricultural fields in the Andean region, edited by W. M. Denevan, K.
Mathewson and G. Knapp, pp. 163-192. vol. 1. British Archaeological Reports,
International Series 359, Oxford, England.
1987b Soil conservation and redevelopment of agricultural terraces in the Colca
valley, Peru. Journal of the Washington, D.C. academy of sciences 77:149-154.
1988 Investigation of agricultural soils at Lari, Colca valley, Peru. In Cognitive,
behavioral, and agronomic studies of soil management in Lari, Colca valley, Peru.
Report to the National Science Foundation. Department of Anthropology, Catholic
University, Washington, DC.
632
1992 Long-term effects of prehistoric agriculture on soil: examples from New
Mexico and Peru. In Soils in archaeology: landscape evolution and human
occupation, edited by V. T. Holliday, pp. 217-245. Smithsonian Institution Press,
Washington, D.C.
Sandor, J. A. and N. S. Eash
1991 Significance of ancient agricultural soils for long-term agronomic studies
and sustainable agricultural research. Agronomy journal 83:29-37.
1995 Ancient agricultural soils in the Andes of southern Peru. Soil science society
of America journal 59:170-179.
Sandor, J. A. and L. Furbee
1996 Indigenous knowledge and classification of soils in the Andes of southern
Peru. Soil science society of America journal 60:1502-1512.
Schreiber, K. J.
1992 Wari imperialism in Middle Horizon Peru. Anthropological papers /
Museum of Anthropology, University of Michigan ; no. 87. Museum of
Anthropology University of Michigan, Ann Arbor.
Sciscento, M. M.
1989 Imperialism in the high Andes: Inka and Wari involvement in the
Chuquibamba valley, Peru. Ph.D. dissertation, University of California, Santa
Barbara.
Seddon, M. T.
1998 Ritual, power, and the development of a complex society: The Island of the
Sun and the Tiwanaku State. Ph.D. dissertation, University of Chicago.
Shea, D. E.
1986a The Achoma archaeology pilot study. In The cultural ecology, archaeology,
and history of terracing and terrace abandonment in the Colca valley of Southern
Peru. Technical report to the National Science Foundation and the National
Geographic Society, edited by W. M. Denevan, pp. 313-362. vol. 1. Department of
Geography, University of Wisconsin-Madison, Madison.
1986b Preliminary discussion of prehistoric Settlement at Achoma. In The cultural
ecology, archaeology, and history o terracing and terrace abandonment in the
Colca valley of southern Peru. Technical report to the National Science
633
313-362. vol. 1. 2 vols. Department of Geography, University of WisconsinMadison, Madison.
1987 Preliminary discussion of prehistoric settlement and terracing at Achoma,
Colca Valley, Peru. In Pre-Hispanic agricultural fields in the Andean region, edited
by W. M. Denevan, K. Mathewson and G. Knapp, pp. 67-88. B.A.R. International
Series, vol. 359 (i). B.A.R., Oxford.
1997 Discussion of prehistoric settlement at Achoma. In Achoma archaeology: A
study of terrace irrigation in Peru, edited by D. E. Shea, pp. 49-58. Logan Museum
of Anthropology, Beloit, Wisconsin.
Shennan, S.
1997
Quantifying archaeology. 2nd ed. Edinburgh University Press, Edinburgh.
Sherbondy, J. E.
1982 The canal systems of hanan Cuzco. Ph.D. dissertation, University of Illinois
at Urbana-Champaign.
Smith, M.
1992a Braudel's temporal rhythms and chronology theory in archaeology. In
Archaeology, annales, and ethnohistory, pp. 23-34. Cambridge University Press,
Cambridge.
1992b Rhythms of change in postclassic Mexico: archaeology, ethnohistory, and
the Braudelian model. In Archaeology, annales, and ethnohistory, edited by A. B.
Knapp, pp. 51-74. Cambridge University Press, Cambridge.
Smith, R. E., G. R. Willey and J. C. Gifford
1960 The type-variety concept as a basis for the analysis of Maya pottery.
Spalding, K.
1982 Exploitation as an economic system: the state and the extraction of surplus in
colonial Peru. In The Inca and Aztec states, 1400-1800: anthropology and history,
edited by G. A. Collier, R. Rosaldo and J. D. Wirth, pp. 321-342. Academic Press,
New York.
1984 Huarochir, an Andean society under Inca and Spanish rule. Stanford
University Press, Stanford.
1999 The crises and transformations of invaded societies: Andean area (15001580). In The Cambridge history of the native peoples of the Americas: Volume III,
634
South America, Part I, edited by F. Salomon and S. B. Schwartz, pp. 904-964.
Cambridge University Press, New York.
Spencer, J. E. and G. A. Hale
1961 The origin, nature, and distribution of agricultural terracing. Pacific
viewpoint 2:1-40.
Staff, S. S.
1975
Soil Taxonomy. Agriculture Handbook 436. U.S. Department of Agriculture.
1990 Keys to soil taxonomy. 4th ed. SMSS Technical Monograph 19, Blacksburg,
Virginia.
Stanish, C.
1985 Post-Tiwanaku regional economies in the Otora valley, Southern Peru.
Ph.D. dissertation, University of Chicago.
1989a An archaeological evaluation of an ethnohistorical model. In Ecology,
settlement, and history in the Osmore drainage, edited by D. S. Rice, C. Stanish and
P. R. Scarr, pp. 303-320. vol. 545. British Archaeological Reports, International
Series, Oxford.
1989b Household archaeology: testing models of zonal complementarity in the
south central Andes. American anthropologist (91):7-24.
1991 A late pre-hispanic ceramic chronology for the upper Moquegua valley,
Peru. Fieldiana 16 (New Series). Field Museum of Natural History, Chicago.
1992
Ancient Andean political economy. University of Texas Press, Austin.
1994 The hydraulic hypothesis revisited: Lake Titicaca basin raised fields in
theoretical perspective. Latin American antiquity 5(4):312-332.
1997a Archaeological survey in the Juli-Desaguadero region of Lake Titicaca
Basin, Southern Peru. Fieldiana. Anthropology, new series, no. 29. Field Museum
of Natural History, Chicago.
1997b Nonmarket imperialism in the prehispanic Americas: The Inka occupation of
the Titicaca Basin. Latin American antiquity 8(3):195-216.
2003 Ancient Titicaca: the evolution of complex society in southern Peru and
northern Bolivia. University of California Press, Berkeley.
635
Stanish, C., R. L. Burger, L. M. Cipolla, M. D. Glascock and E. Quelima
2002 Evidence for early long-distance obsidian exchange and watercraft use from
the southern Lake Titicaca Basin of Bolivia and Peru. Latin American antiquity
13:444-454.
Stanish, C., L. Steadman and M. T. Seddon
1994 Archaeological research at Tumatumani, Juli, Peru. Field Museum of
Natural History, Chicago.
Steadman, L.
1995 Excavations at Camata: an early ceramic chronology for the western
Titicaca Basin, Peru. Ph.D. Dissertation, University of California, Berkeley.
1999 The ceramics. In Early settlement at Chiripa, Bolivia, edited by C. A.
Hastorf, pp. 61-72. Contributions of the University of California Archaeological
Research Facility. University of California Archaeological Research Facility,
Berkeley.
Stern, S. J.
1982a Peru's Indian peoples and the challenge of Spanish conquest. University of
Wisconsin Press, Madison.
1982b The social significance of judicial institutions in an exploitive society:
Huamanga, Peru, 1570-1640. In Inca and Aztec states 1800: : anthropology and
history, edited by G. A. Collier, R. Rosaldo and J. D. Wirth, 289-320. Academic
Press, New York.
Steward, J. H.
1942 The direct historical approach to archaeology. American antiquity 7(4):337343.
Strong
1933 The plains culture area in the light of archaeology. American anthropologist
35(2):271-287.
Strong, W. D.
1933 An introduction to Nebraska archaeology. Smithsonian miscellaneous
collections, vol. 93, No. 10. Smithsonian Institution, Washington, DC.
1940 From history to prehistory in the northern great plains. In Essays in
historical anthropology of North America. Smithsonian Miscellaneous Collection
100.
636
Stuiver, M. and P. J. Reimer
1993 Extended 14C database and revised CALIB radiocarbon calibration
program. Radiocarbon 35:215-230.
Stuiver, M., P. J. Reimer and T. F. Braziunas
1998 High-precision radiocarbon age calibration for terrestrial and marine
samples. Radiocarbon 40:1127-1151.
Sutter, R. C.
2000 Prehistoric genetic and culture change: A bioarchaeological search for preInka altiplano colonies in the coastal valleys of Moquegua, Peru, and Azapa, Chile.
Latin American antiquity 11(1):43-70.
Thompson, L. G.
1992 Ice core evidence from Peru and China. In Climate since A.D. 1500, edited
by R. S. Bradley and P. D. Jones, pp. 517-548. Routledge, New York.
1993 Reconstructing the paleo ENSO records from tropical and subtropical ice
cores. Bulletin, Institut francais d'etudes andines. 22(o. 1, 1993.):.
1995 Late holocene ice core records of climate and environment from the tropical
Andes, Peru. Bulletin de l'Institut Francais d'tudes Andines 24:619-629.
Thompson, L. G., M. E. Davis, E. Mosley-Thompson and K.-b. Liu
1988 Pre-Incan agricultural activity recorded in dust layers in two tropical ice
cores. Nature 336:763-765.
Thompson, L. G. and E. Mosley-Thompson
1989 One-half millenium of tropical climate variability as recorded in the
stratigraphy of the Quelccaya ice cap, Peru. In Climate change in the eastern Pacific
and western Americas, edited by D. Peterson, pp. 15-31. American Geophysical
Union Monograph 55. American Geophysical Union, Washington, DC.
Thompson, L. G., E. Mosley-Thompson, J. F. Bolzan and B. R. Koci
1985 A 1500-yr record of tropical precipitation in ice cores from the Quelccaya
ice cap, Peru. Nature 336:763-765.
Thompson, L. G., E. Mosley-Thompson, W. Dansgaard and P. M. Grootes
1986 The Little Ice Age as recorded in the stratigraphy of the tropical Quelccaya
ice cap. Science 234:361-364.
637
Thompson, L. G., E. Mosley-Thompson and M. E. Davis
1994 Glacial records of global climate: a 1500-year tropical ice core record of
climate. Human ecology. 22(o. 1, 1994.):.
Tibesar, A.
1953 Franciscan beginnings in early colonial Peru. Academy of American
Franciscan History, Washington, DC.
Toledo, F. d.
1924 [1570-1575]
Libro de la visita general del virrey Toledo (1570-75). Revista
Histrica, Organo del Instituto Histrico del Per 7(2):114-216.
Tord, L. E.
1983
Templos coloniales del Colca - Arequipa. Atlas, Lima.
Trawick, P.
2001 The moral economy of water: equity and antiquity in the Andean commons.
American anthropologist 103(2):361-379.
Treacy, J. M.
1989 The fields of Coporaque: agricultural terracing and water management in
the Colca valley, Arequipa, Peru. Ph.D. dissertation, University of WisconsinMadison.
1994 Las chacras de Coporaque: Andeneria y riego en el Valle del Colca. Serie:
Estudios de la sociedad rural. Instituto de Estudios Peruanos, Lima, [Peru].
Treacy, J. M. and W. M. Denevan
1986 Survey of abandoned terraces, canals, and houses at Chijra, Coporaque. In
The cultural ecology, archaeology, and history o terracing and terrace
abandonment in the Colca valley of southern Peru. Technical report to the National
Science Foundation and the National Geographic Society, edited by W. M.
Denevan, pp. 198-220, vol. 1. Department of Geography, University of WisconsinMadison, Madison.
Tschopik, M. H.
1946 Some notes on the archaeology of the department of Puno, Peru. Papers of
the Peabody Museum of American Archaeology and Ethnology XXVII, no. 3.
Peabody Museum of American Archaeology and Ethnology, Cambridge, MA.
638
Tung, T. A.
2003 A bioarchaeological perspective on Wari imperialism in the Andes of Peru:
a view from heartland and hinterland skeletal populations. Ph.D. dissertation,
University of North Carolina.
Ulloa Mogolln, J. d.
1965 [1586] Relacin de la provincia de los Collaguas para la discrepcin de las
Indias que su magestad manda hacer. In Relaciones geogrficas de Indias, edited by
M. Jimenez de la Espada, pp. 326-333. vol. I. 3 vols. Ediciones Atlas, Madrid.
Valderrama, R. and C. Escalante
1987 Tres mitos de origen, valle del Colca. Revista Veintiuno 4.
1988 Del tata mallku a la mama pacha: riego, sociedad y ritos en los Andes
peruanos. Centro de Estudios y Promocion del Desarrollo, Lima.
1997 La doncella sacrificada: mitos del Valle del Colca. Universidad Nacional de
San Agustn; Instituto Francs de Estudios Andinos, Arequipa.
Van Buren, M.
1993 Community and empire in southern Peru: the site of Torata Alta under
Spanish rule. Ph.D. dissertation, University of Arizona.
1996 Rethinking the vertical archipelago: Ethnicity, exchange, and history in the
south central Andes. American anthropologist 98(2):338-351.
1997 Continuity or change?: Vertical archipelagos in southern Peru during the
early colonial period. In Approaches to the historical archaeology of Mexico,
Central and South America, edited by J. Gasco, G. C. Smith and P. Fournier-Garca,
pp. 155-164. The Institute of Archaeology, University of California, Los Angeles,
Los Angeles.
van de Guchte, M.
1990 Carving the world: Inca monumental sculpture and landscape. Ph.D.
dissertation, University of Illinois, Urbana-Champaign.
1999 The Inca cognition of landscape: archaeology, ethnohistory, and the
aesthetic of alterity. In Archaeologies of landscape: contemporary perspectives,
edited by W. Ashmore and A. B. Knapp, pp. 149-168. Blackwell Publishers,
Oxford.
639
Verdugo, G. and G. d. Colmenares
1977 [1591] Visita de Yanquecollaguas Urinsaya. In Collaguas I, edited by F.
Pease G. Y, pp. 191-406. Pontifcia Universidad Catlica del Per, Lima.
Wachtel, N.
1977 The vision of the vanquished : the Spanish conquest of Peru through Indian
eyes, 1530-1570. Barnes & Noble, New York.
Wagstaff, J. M.
1987 Landscape and culture: geographical and archaeological perspectives. Basil
Blackwell, London.
Waugh, R. A. and J. M. Treacy
1986 Hydrology of the Coporaque irrigation system. In The cultural ecology,
archaeology, and history o terracing and terrace abandonment in the Colca valley
of southern Peru. Technical report to the National Science Foundation and the
Department of Geography, University of Wisconsin-Madison, Madison.
Wedel, W. R.
1938 The direct-historical approach in Pawnee archaeology. Smithsonian
miscellaneous collections 97. Smithsonian Institution, Washington, DC.
1940 Culture sequence in the central great plains. In Essays in historical
anthropology of North America. Smithsonian Miscellaneous Collection 100.
Wernke, S. A.
1997 Excavaciones preliminares en el sitio de Pallaclle, Distrito de Yanque,
Provincia de Caylloma, Departamento de Arequipa. Report to the Instituto
Nacional de Cultura, Lima.
2001a El Proyecto Prospeccin Regional Valle del Colca. Technical report
submitted to the Instituto Nacional de Cultura del Peru, Lima, Peru.
2001b A reassessment of Collagua and provincial Inka ceramic styles of Arequipa,
Peru. Paper presented at the 66th Annual Meeting for the Society for American
Archaeology, New Orleans, LA.
Wheat, J. B., J. C. Gifford and W. W. Wasley
1958 Ceramic variety, type cluster, and ceramic system in southwestern pottery
analysis. American antiquity 24(1):34-47.
640
Whitehead, N. L.
1995 The historical anthropology of text. Current anthropology 36(1):53-74.
1998 Ecological history and historical ecology: diachronic modeling versus
historical explanation. In Advances in historical ecology, edited by W. Bale, pp.
30-41. Columbia University Press, New York.
Williams, P. R.
1997 The role of disaster in the development of agriculture and the evolution of
social complexity in the south-central Andean sierra. Ph.D. dissertation, University
of Florida.
Wilson, D. J.
1988 Prehispanic settlement patterns in the lower Santa Valley, Peru: a regional
perspective on the origins and development of complex north coast society.
Smithsonian Institution Press, Washington, D.C.
Winterhalder, B.
1993 Rainfall and temperature patterning and predictability with respect to water
management in the central Andes of southern Peru. In Irrigation at high altitudes:
the social organization of water control systems in the Andes, edited by W. P.
Mitchell, D. Guillet and I. Bolin. Society for Latin American Anthropology
publication series, v. 12. Society for Latin American Anthropology, American
Anthropological Association, Arlington, VA.
Wolf, E. R.
1956 Aspects of group relations in complex society: Mexico. American
anthropologist 58:1065-1078.
Yaeger, J.
2000 The social construction of communities in the Classic Maya countryside:
Strategies of affiliation in western Belize. In The archaeology of communities,
edited by M. A. Canuto and J. Yaeger, pp. 123-142. Routledge, New York.
Yaeger, J. and M. A. Canuto
2000 Introducing an archaeology of communities. In The archaeology of
communities, edited by M. A. Canuto and J. Yaeger, pp. 1-15. Routledge, New
York.
Ypez Alvarez, W.
1991 Estudio de la cermica de las fases Churajn Tardo y Chuquibamba Tardo
en el sitio arqueolgico de Umacollo, valle del Chili Arequipa. Universidad
Catlica de Santa Mara. Submitted to Technical report.
641
Ypez, W., J. Jennings, C. Quequeza L., M. Lpez H. and A. Cardona Rosas
2001 Rethinking the Qosqopa style: The rise and fall of the Wari empire in the
valleys of Arequipa, Peru. Paper presented at the 66th Annual Meeting of the
Society for American Archaeology, New Orleans, LA.
Zimmerer, K. S.
1999 The overlapping patchworks of mountain agriculture in Peru and Bolivia:
toward a regional-global landscape model. Human Ecology 27(1):135-165.
2000 Rescaling irrigation in Latin America: the cultural images and political
ecology of water resources. Ecumene 7(2):150-175.
Zimmerman, A. F.
1938 Francisco de Toledo, fifth viceroy of Peru, 1569-1581. The Caxton Printers
ltd., Caldwell, Id.,.
Zuidema, R. T.
1964 The ceque system of Cuzco: The social organization of the capital of the
Inca. E.J. Brill, Leiden.
1973 Kinship and ancestor cult in three Peruvian communities: Hernandez
Principe's account of 1622. Bulletin del IFEA. vol 2(10):16-33.
1977 The Inca kinship system: a new theoretical view. In Andean kinship and
marriage, edited by R. Bolton and E. Mayer, pp. 240-292. Special Publication no. 7.
American Anthropological Association, Washington, DC.

Arqueologia Del Colca

Uploaded by

Copyright:

Available Formats

Arqueologia Del Colca

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Arqueologia Del Colca

Uploaded by

Copyright:

Available Formats

AN ARCHAEO-HISTORY OF ANDEAN COMMUNITY AND LANDSCAPE:

THE LATE PREHISPANIC AND EARLY COLONIAL COLCA VALLEY, PERU

A dissertation submitted in partial fulfillment of

Chapter 1: Community and Landscape in the Andes

Chapter 2: Physiographic, Climatic, and Anthropogenic Characteristics of the

Chapter 3: The Collagua in Regional and Historical Context: Previous Research

Chapter 4: Methodology and Research Operations

PART II: ARCHAEOLOGICAL FINDINGS

Chapter 6: The Late Intermediate Period and Late Horizon Occupations

Chapter 7: Spanish Insertions: Early Colonial Settlement and Agro-Pastoral

PART III: ETHNOHISTORICAL RECONSTRUCTION AND

Chapter 9: Reconstructing Colonial Land Tenure and Late Prehispanic

Chapter 10: Summary and Conclusions

Appendix B: Projectile Point Illustrations

Appendix C: Archaic Period Sites

Appendix D: Survey Artifact Registry

Appendix E: Ceramic Distributions by Site

Appendix F: Site and Sector Registry

funding of a Dissertation Research Grant from the Wenner-Gren Foundation for

In completing this project, I am reminded of what a privilege it has been to be

I warmly thank the Mayors of Yanque and Coporaque, Ramn

I thank Karen Wise and Augusto Cardona Rosas, Co-Directors of

CIARQ, for all manner of logistical and bureaucratic support.

ethnohistorical portion reconstructs the areas local- and regional-scale economic

During this period, inter-settlement political relations appear to have

oscillated between competition and cooperation. Defensible settlement locations and

PART I: INTRODUCTION AND BACKGROUND

Although they identify relevant variables, these

Yet in the absence of systematic

This study tracks the origins and

The project is multi-scalar, both spatially and

A Political-Ecological Approach to Community and Landscape

Recent approaches are working in this

Ecological anthropologists and

Here, rather than documenting the formal attributes of a vertical

complementarity system as a social formation or structure, the processual variability of

Communities, Natural and Imagined: The Case of the Andean Ayllu

Ayllus of different scales of

Also, the protohistoric Andean

Ayllu members reaffirmed and reified their community affiliations by

Examples of complex, orderly, heterarchical systems abound in nature,

For example, the

Alternative Hypotheses and Their Correlates

Early Agriculture and Settlement in the Colca Valley

The Middle Horizon: Expansionist States and Their Local Influence

However, economic links with Tiwanaku are evident in the continued

First, if Wari political influence was

The Late Intermediate Period: Community Organization and Agro-Pastoral Production

In short, I hypothesize that Collagua political

Inter-community relations probably shifted between coordination and

Differences in the size and elaboration of domestic

Previous researchers have noted the distinctive features that

Portable artifact media suggested considerable Inka

In the censuses, each of the fields declared by

Restructuration and Reduccin: Colonial Period Adjustments

These researchers have suggested that colonial

Location, Geology, and Geomorphology

Figure 2.3: Volcanic features and obsidian sources

massive eruptions of the stratovolcano Huaynaputina, on February 19 and 20, 1600,

Figure 2.4: Geomorphic surfaces in the central Colca valley