Prehistoric Cultural Development and Interregional
Interaction in the Tropical Montane Forests of Peru
Volume 1
A Dissertation
Presented to the Faculty of the Graduate School
of
Yale University
in Candidacy for the Degree of
Doctor of Philosophy
by
Warren Brooks Church
Dissertation Director: Richard L. Burger
December, 1996
© 1996 by Warren Brooks Church
ALL RIGHTS RESERVED
Dedicated to my mother and father,
Sylvina W. and Thomas T. Church
ABSTRACT
PREHISTORIC CULTURAL DEVELOPMENT-AND INTERREGIONAL
INTERACTION IN THE TROPICAL MONTANE FORESTS OF PERU
Warren Brooks Church
Yale University
1996
The forested eastern slopes of the Central Andes have
been characterized alternately as the impenetrable eastern
frontier of Andean Civilization, as an empty migratory
corridor, a sparsely populated buffer zone separating Andean
and Amazonian populations, and as a remote ecological zone
servicing highland political economies.
This dissertation
presents a new cultural sequence excavated from the
stratified site of Manachaqui Cave at the upper edge of
Peru's northeastern Andean montane forest.
Analyses of the
archaeological materials including ceramic, lithic, faunal
and botanical remains are featured.
These data undermine
popular population movement and colonization hypotheses,
instead providing evidence for autochthonous montane forest
cultural development dating from the Preceramic Period
(prior to 2000 B.C.).
Manachaqui Cave's location beside a
pre-Hispanic road, coupled with ethnographic analogies and
ethnohistorical and archaeological evidence indicate that
the rockshelter functioned primarily as a wayside camp for
persons engaged in long-distance travel and exchange between
the Central Andean, Amazonian and Northern Andean regions.
The interpretation of Manachaqui Cave as a wayside station
allows a unique perspective on the development of
interregional interaction before, during and after the
emergence of Central Andean civilization.
Evidence from
Manachaqui Cave and from other localities on the eastern
slopes suggests that, rather than a remote frontier, the
montane forest was the locus of intense boundary
interaction.
The mediation of long-distance exchange by
autonomous montane forest populations strategically situated
along the Andean slopes was vital to the development and
maintenance of political economies and exchange networks in
adjacent highland and lowland regions throughout prehistory.
ACKNOWLEDGEMENTS
Many kind individuals and generous institutions
provided crucial support getting to Manachaqui Cave,
performing the fieldwork there, processing the
archaeological materials in Trujillo (Peru), and finally
writing the dissertation back in New Haven (USA) .
I would
first like to thank Dr. Thomas J. Lennon for offering me the
opportunity to join the Rio Abiseo National Park Research
Project in 1984, and for his ongoing support and
encouragement.
The 1990 excavations at Manachaqui were
funded by a generous donation to the University of ColoradoBoulder's Rio Abiseo Project by Lee and Betsy Turner.
Permission to undertake research in and around the Rio
Abiseo National Park was granted by the Ministry of
Agriculture's Oficina del Director General de Forestal y
Fauna in Lima.
The 1990 archaeological fieldwork was
performed under the auspices of the University of ColoradoBoulder, and authorized by the Institute Nacional de Cultura
(INC), Lima, and Resoluci6n Ministerial No. 1514-90-ED.
I
am grateful to Dr. Rogger Ravines, then of the INC, and to
Rolando Paredes for providing valuable guidance.
Thanks to
Sr. Enrique Goicochea Hoyle and his family for making me
feel at home in Lima.
In Trujillo, Lie. Ana Maria Hoyle and
Dr. Santiago Uceda of the Institute Regional de CulturaRegi6n La Libertad kindly facilitated the proposed
fieldwork.
iii
iv
I am grateful to the people of Pataz and Los Alisos for
receiving us with such kindness.
The 1990 field season
benefitted from a team of enthusiastic and dedicated
archaeologists from Trujillo that worked long days and weeks
under difficult conditions.
Edgardo Silva V. served as INC
field supervisor, and Victor Pimentel S. assisted direction
of the fieldwork.
The archaeologists were Cesar A. Cornelio
Lecca, Lucia Medina de la Cruz, Monica Panaifo Texiera,
Victor Hugo Rios Cisneros and Laura Nacarino Flores.
From
Pataz, Tito Hurtado B. served as cook and managed the camp,
while Macedonia Gonzales, Marcos Salirrosas, Jose Escobedo,
Celestino Ramirez, Julio Armas Villalobos, Jorge Alguilar
and Mariano Alguilar assisted with the day to day work.
Carlos Sillones provided short-wave radio support from
Trujillo.
The subsequent laboratory work in Trujillo was funded
by Yale University's Joseph Albers and Augusta Hazzard
Funds, and National Science Foundation Dissertation Research
Grant No. DBC-9200799.
Laboratory processing of the
archaeological materials was overseen with the assistance of
Daisy Barreto Cedamanos, while Cesar Cornelio Lecca, Elena
Goicochea Diaz and Violeta Chamorro Castillo participated.
Lie. Segundo Vasquez S. generously gave me access to his
preliminary description and illustrations of ceramics
recovered from Manachaqui Cave in 1988.
Edwin Blas Carranza
drew most of the artifact illustrations with help by
v
Marisabel Paredes de Pimentel and Manuel Tam Chang.
I am
most grateful to my parents-in-law, the late Dr. Otto Cedr6n
Alva and Sra. Estela Viuda de Cedr6n and their family for
allowing me to utilize a large portion of their back yard
for the laboratory work.
The processing of Manachaqui Cave's faunal and
botanical materials, and their analyses by Drs. Jonathan
Kent and Deborah Pearsall was funded by the Wenner-Gren
Foundation for Anthropological Research Grant No. 5425.
Exportation of archaeological materials for analysis was
authorized by Resoluci6n Suprema Nos. 007-93-ED, 033-93-ED
and 035-93-ED.
Ms. Patricia Moore provided invaluable and
varied kinds of assistance from the States during this stage
of the project.
I should also point out that Dr. Kent was
in Peru when the chronological analysis of Manachaqui Cave's
stratigraphy was completed, and he did not have the
opportunity to check the recalculations of phase totals
featured in this dissertation.
Back in New Haven, support for the writing of my
dissertation was provided by a Yale University Mellon
Dissertation Writing Fellowship.
Many specialists
graciously supplied useful suggestions and access to
unpublished data at the interpretive stages of my work,
especially Drs. Kenneth Young, Donald Rodbell, Ruth Shady
S., Warren DeBoer, Theresa Topic, Karen Stothert, MaryElizabeth Reeve, Sabina Hyland, Michael Coe and Frank Hole.
vi
Thanks also to archaeologists Arthur Rostoker, Ismael Perez
C. and Alfredo Melly for offering unpublished information
regarding their work.
I am certain that I neglect to
mention many other individuals whose generosity made this a
better thesis.
My wife Elke assisted my work on occasion, and showed
considerable patience during the dissertation process.
The
success of this project is in large part owed to the
unwavering support and friendship offered by my mentor Dr.
Richard Burger and by his wife Lucy Burger Salazar.
That I
was able to complete the project is due to the loving and
unconditional support of my father and mother to whom I have
dedicated this dissertation.
of course, my responsibility.
Any of its short-comings are,
TABLE OF CONTENTS
ACKNOWLEDGEMENTS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii
LIST OF ILLUSTRATIONS ................................ xviii
LIST OF PLATES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxix
LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
XXX
Chapter
1. THE PROBLEM OF MONTANE FOREST PREHISTORY ...........
1
Andean Environments and Culture Areas ............. 11
The Central Andes. . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
The Central Andean Tropical
Montane Forests ............................ 17
The Eastern Montane and
Premontane Forests ......................... 19
Interpretive Frameworks for
Andean Migrations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Unitary Origins and Population Movements ..... 26
Andean Political Economy, Verticality
and Colonization. . . . . . . . . . . . . . . . . . . . . . . . . . . 29
The Problem of Montane Forest Prehistory:
Migration or Interaction? ...................... 34
Concepts and Strategies for Evaluating
Migration Hypotheses ..................... 39
Contexts for Inquiry: Manachaqui Cave and
the Greater Eastern Montane Forest ......... 51
The Alternative Hypothesis: Local
Development and interaction ................ 58
2. CIVILIZING MIGRATIONS, VERTICALITY COLONIZATION
AND THE EASTERN MONTANE FORESTS .................. 64
Civilizing Migrations and the Eastern
Montane Forest. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Primordial Migrations and the Eastern
Montane Forest. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
vii
viii
Civilizing Chachapoyas ....................... 74
Population Movements and "Cuelap
Civilization" ......................... 75
Highlandizing the Montane Forests ....... 78
Neolithic Migrations and the Eastern
Montane Forest. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
Quechuas, Maize and the Eastern
Montane Forest. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Verticality, Colonization and the Eastern
Montane Pores t . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 9
The Southeastern Montane Forest .............. 105
The Central Montane Forest ................... 106
Colonization as Huari
Political Economy ..................... 108
Verticality Colonies in the
Tarma Canyon .......................... 113
The Northeastern Montane Forest .............. 120
Alternative Viewpoints ............................ 125
3. THE PATAZ-ABISEO STUDY AREA: PRESENT AND PAST ...... 129
Life Zones and Land-use in the
Pataz-Abiseo Area ............................... 13 6
The Dry Forest Zone ......................... 139
The Moist Montane Zone ...................... 139
The Tropical Alpine Zone .................... 142
The Tropical Montane Rain Forest Zone ....... 143
The Huallabamba and Huallaga Valley
Premontane Forests ........................ 146
Ethnohistory of the Pataz-Abiseo Area ............. 148
Southern Chachapoyas ......................... 149
Chachapoyas Culture ..................... 150
ix
Chachapoyas Under the Incas ............. 154
Spanish Conquest and Administrative
Units in Southern Chachapoyas ......... l55
Demographic Collapse .................... 157
The Central Huallaga Valley .................. 159
Early Missionary Contacts ............... 160
Ethnohistoric Evidence for HighlandLowland Boundaries ......................... 168
Ethnohistoric Evidence for HighlandLowland Interaction ........................ 172
Archaeology of Southern Chachapoyas and
the Central Huallaga ............................ 175
Archaeology of the Pataz-Abiseo Area ......... 180
Pataz District .......................... 181
The Abiseo Drainage and Central
Huallaga Lowlands ..................... 185
Summary ........................................... 18 8
4. MANACHAQUI CAVE AND ITS ENVIRONMENTAL CONTEXT ...... 195
The Changing Manachaqui Valley Environment ........ 198
Geology and Pleistocene Glaciation ........... 199
Holocene Environmental Change ................ 201
Flora and Fauna .............................. 203
The Manachaqui Valley and Manachaqui Cave ......... 205
Manachaqui Cave .............................. 2 0 6
The Pre-Hispanic Road System ................. 209
Hypothetical Functions of Manachaqui Cave ......... 212
Modern Uses of Highland Rockshelters ......... 213
Ancient Uses of Highland Rockshelters ........ 215
Manachaqui Cave as a Wayside Station ......... 216
X
Surnrnary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 2 5
5. EXCAVATIONS AT MANACHAQUI CAVE ..................... 228
The 1988 Excavations .............................. 23 0
The 1990 Excavations .............................. 234
Site Formation and Stratigraphy ................... 238
Site formation ............................... 239
Stratigraphy ................................. 245
Stratum 4 ............................... 247
Stratum 3 ............................... 247
Stratum 2 ............................... 248
Stratum 1 ............................... 249
Archaeological Features, Floors
and Hearths ................................ 2 50
Sector A Features and Floors ............ 251
Sector B Features and Floors ............ 261
Changing Hearth Morphology and Function ........... 264
Radiocarbon Evidence for Chronology ............... 266
Sector A Radiocarbon Evidence ................ 266
Sector B Radiocarbon Evidence and
Stratigraphic Correlations with
Sector A ................................... 269
Materials Analysis ................................ 272
Ceramic Analysis ............................. 273
Lithic Analysis .............................. 277
Analysis of Organic Remains .................. 277
Botanical Analysis ...................... 278
Faunal Analysis ......................... 279
xi
6 . THE MANACHAQUI PHASE . . . • . . • • • . • . . • . • . . . . . . . . . . . . . . . 2 8 2
Manachaqui Phase Antecedents . . . . . . . . . . . . . . . . . . . . . . 282
Manachaqui Phase Ceramics . . . . . . . . . . . . . . . . . . . . . . . . . 285
Ceramic Paste A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
Morphology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
Decoration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 0 9
Ceramic Paste Group B . . . . . . . . . . . . . . . . . . . . . . . . 321
Paste B1 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 322
Paste B2 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 323
Paste B3 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 324
Lithic Remains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 2 4
Rocks and Minerals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326
Botanical Remains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
Faunal Remains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
Manachaqui Phase Chronology . . . . . . . . . . . . . . . . . . . . . . . 328
Manachaqui Phase Ceramic Relationships ............ 329
Paste A Relationships:
The Central Andes . . . . . . . . . . . . . . . . . . . . . . . . . . 331
Paste A Relationships:
The Amazonian Lowlands . . . . . . . . . . . . . . . . . . . . . 341
Paste A Relationships:
The Northern Andes . . . . . . . . . . . . . . . . . . . . . . . . . 346
Paste B Origins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
7. The Sui tacocha Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 55
Suitacocha Phase Ceramics . . . . . . . . . . . . . . . . . . . . . . . . . 355
Ceramic Paste Group A . . . . . . . . . . . . . . . . . . . . . . . . 357
Morphology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 359
Decoration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 7 2
xii
Ceramic Paste Group B . . . . . . . . . . . . . . . . . . . . . . . . 3 8 4
Paste B4 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 384
Paste B5 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 3 85
Lithic Remains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 8 6
Rocks and Minerals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 87
Botanical Remains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 87
Faunal Remains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 88
Suitacocha Phase Chronology . . . . . . . . . . . . . . . . . . . . . . . 388
Paste A Relationships: The Central Andes .......... 390
Paste A Relationships: The Amazonian
Lowlands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 398
Paste A Relationships: The Northern Andes ......... 404
Paste B Origins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 411
Paste B4 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 412
Paste B5 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 413
8. THE COLPAR PHASE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414
Colpar Phase Ceramics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415
Ceramic Paste Group A . . . . . . . . . . . . . . . . . . . . . . . . 417
Morphology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 417
Decoration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 421
Ceramic Paste Group B . . . . . . . . . . . . . . . . . . . . . . . . 423
Paste B6 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 424
Paste B7 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 424
Paste B8 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 425
Paste B9 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 425
Paste B10 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 42 6
Ceramic Paste C1 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 42 6
xiii
Lithic Remains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427
Rocks and Minerals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428
Botanical Remains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 8
Faunal Remains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429
Radiocarbon Evidence for Colpar
Phase Chronology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 9
Paste A Relationships: The Central Andes .......... 430
Paste A Relationships: The Amazonian
Lowlands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443
Paste A Relationships: The Northern Andes ......... 445
Paste B Origins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 448
Paste B6 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 449
Paste B7 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 449
Paste B8 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 450
Paste B9 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 451
Paste B10 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 451
Paste C1 Origins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 451
Ceramic Evidence for Colpar
Phase Chronology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 452
9. THE EMPEDRADA PHASE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 454
Empedrada Phase Ceramics . . . . . . . . . . . . . . . . . . . . . . . . . . 455
Ceramic Paste Group A . . . . . . . . . . . . . . . . . . . . . . . . 457
Morphology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457
Decoration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 464
Ceramic Paste Group B . . . . . . . . . . . . . . . . . . . . . . . . 465
Paste B11 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 465
Paste B12 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 466
Paste B13 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 466
xiv
Paste B14 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 466
Paste B15 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 467
Paste B16 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 467
Paste B17 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 468
Paste B18 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 468
Ceramic Paste Group C . . . . . . . . . . . . . . . . . . . . . . . . 469
Paste C2 a-h . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 7 0
Lithic Remains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 7 4
Rocks and Minerals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 476
Botanical Remains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 77
Faunal Remains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 77
Radiocarbon Evidence for Ernpedrada
Phase Chronology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 77
Paste A Relationships: The Central Andes .......... 478
Paste A Relationships: The Amazonian
Lowlands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 484
Paste A Relationships: The Northern Andes ......... 486
Paste B Origins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 487
Paste B11 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 487
Paste B12 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 488
Paste B13 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 488
Paste B14 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 489
Paste B15 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 490
Paste B16 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 491
Paste B17 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 491
Paste B18 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 492
Paste C2 Origins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 494
XV
Paste C2 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 494
Paste C2 a-h . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 495
Lithic Evidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 497
Ceramic Evidence for Empedrada
Phase Chronology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 499
10. MANACHAQUI CAVE AND MONTANE
FOREST MIGRATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 501
Evidence for Migrations from
Historical Linguistics . . . . . . . . . . . . . . . . . . . . . . . . . . 502
Chachapoyas Languages . . . . . . . . . . . . . . . . . . . . . . . . 503
Chol6n and Hivito Languages . . . . . . . . . . . . . . . . . . 505
Evidence for Migrations from
Physical Anthropology . . . . . . . . . . . . . . . . . . . . . . . . . . . 508
Evidence for Migrations from Archaeology:
Manachaqui Cave and the Pataz-Abiseo Area ....... 509
Pre-Lavasen Occupations and the
Lavas en Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 513
The Manachaqui Phase Occupation . . . . . . . . . . . . . . 518
Evidence from Botanical and
Faunal Remains . . . . . . . . . . . . . . . . . . . 519
Evidence from Artifacts . . . . . . . . . . . . . . . . . 521
Manachaqui Phase Evidence for
Migrations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 527
The Suitacocha Phase Occupation . . . . . . . . . . . . . . 533
Evidence from Botanical and
Faunal Remains . . . . . . . . . . . . . . . . . . . . . . . . 533
Evidence from Artifacts . . . . . . . . . . . . . . . . . 534
Suitacocha Phase Evidence for
Migrations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 536
The Colpar Phase Occupation . . . . . . . . . . . . . . . . . . 541
Colpar Phase Evidence for
Migrations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 542
xvi
The Empedrada Phase Occupation . . . . . . . . . . . . . . . 543
Evidence from Botanical and
Faunal Remains . . . . . . . . . . . . . . . . . . . . . . 544
Evidence from Artifacts . . . . . . . . . . . . . . . 545
Empedrada Phase Evidence for
Migrations . . . . . . . . . . . . . . . . . . . . . . . . . . 546
Summary of Evidence for Migrations
at Manachaqui Cave . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 549
Evidence for Colonization in the Eastern
Montane Forest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 552
The Southeastern Montane Forest . . . . . . . . . . . . . . 552
The Central Montane Forest . . . . . . . . . . . . . . . . . . . 553
The Northeastern Montane Forest . . . . . . . . . . . . . . 556
11. MONTANE FOREST CULTURAL DEVELOPMENT AND
INTERREGIONAL INTERACTION . . . . . . . . . . . . . . . . . . . . . . . . 560
Pre-Lavasen and Lavasen Phase Evidence
for Interaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 563
Manachaqui Phase Evidence for
Interaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 565
Suitacocha Phase Evidence for
Interaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 570
The Chavin Horizon Hiatus . . . . . . . . . . . . . . . . . . . . 575
Colpar Phase Evidence for Interaction ........ 577
Empedrada Phase Evidence for
Interaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 580
Boundary Interaction and the Central Andean
Tropical Montane Forests . . . . . . . . . . . . . . . . . . . . . . . . 583
Pre-Chavin Horizon Boundary
Interaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589
Post-Chavin Horizon Boundary
Interaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 599
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 608
xvii
REFERENCES CITED ........................................ 619
APPENDIX A: ILLUSTRATIONS AND ARTIFACT
PROVENIENCE ........................................... 67 6
APPENDIX B: PLATES ...................................... 803
APPENDIX C: TABLES ...................................... 812
APPENDIX D: MACROCHRONOLOGY ............................. 826
APPENDIX E: VESSEL SHAPES AND RIM FORMS BY PHASE ........ 836
APPENDIX F: REPORT ON BOTANICAL REMAINS
BY DEBORAH M. PEARSALL ................................ 8 6 8
LIST OF ILLUSTRATIONS
Fig. 1. Map of the Central Andes, western
Amazonia and the Northern Andes (after
Institute Geografico Nacional 1984) . . . . . . . . . . . . . 677
Fig. 2. Map of the Central Andean Montane Forests
in Peru (after Young and Leon 1993: Fig. 1).
The highland montane forest is not shown ........ 678
Fig. 3. Map of some sites mentioned in the text ......... 679
Fig. 4. Map of the northeastern Peruvian Andes.
Shaded area is above 3,000 m (after Institute
Geografico Nacional 1984 and NASA 1976) ......... 680
Fig. 5. Map of the Pataz-Abiseo study area and
macro-ecological zones described in the text
(after Young 1993: Fig. 3 and Young et al.
1994: Fig. 1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 681
Fig. 6. Profile of the Maraffon-Huallaga divide
showing distribution of macro-ecological
zones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 682
Fig. 7. Map of the Manachaqui and Montecristo
River valleys and the pre-Hispanic roads ........ 683
Fig. 8. Plan map of the Manachaqui Cave site
complex (Site M-1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 684
Fig. 9. Plan map with detail of Manachaqui Cave
(Site M-1A) interior. The principal interior
space is shaded . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 8 5
Fig. 10. Map of Manachaqui Cave excavation
units 1988-1990 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 686
·Fig. 11. West profile of Units 1, 2 and 3 . . . . . . . . . . . . . . . 688
Fig. 12. West profile of Units 4, 5, 6 and 35.
Features R-2 and R-3 are shown . . . . . . . . . . . . . . . . . 689
Fig. 13. West profile of Units 38' 39, 7 and 8 .......... 690
Fig. 14. East profile of Units 3' 2 and 1 . . . . . . . . . . . . . . . 691
xviii
xix
Fig. 15. East profile of Units 35, 6, 5 and 4.
Feature R-4 is shown ........................... 692
Fig. 16. East profile of Units 7, 8, 39 and 38 .......... 693
Fig. 17. East profile of Units 37, 32 and 28 ............ 694
Fig. 18. South profile of Units 11, 1, 13 and 12.
Feature R-5 is shown ........................... 695
Fig. 19. North profile of Units 12 and 13. Feature
R-5 is shown and Sector A floors are
indicated in right margin ...................... 696
Fig. 20. South profile of Units 21, 20, 19 and 18.
Feature R-3 and Floor Z's Hearths 1 and
2 are shown .................................... 697
Fig. 21. North profile of Units 29, 30, 35, 31
and 3 2 ......................................... 69 8
Fig. 22. Schematic diagram illustrating
relationships between 1988 excavation
levels and strata as seen in Unit 5
and Unit 6 east profiles ....................... 699
Fig. 23. Schematic diagram illustrating
relationships between 1990 excavation
levels and strata as seen in Unit 31 and
Unit 32 north profiles ......................... 700
Fig. 24. Plan view of Feature R-5, Sector A............. 701
Fig. 25. Sector A floors in Units 14-17. a: Floor
FF and associated rock-filled hearth; b:
Floor EE and associated rock-filled hearth ..... 702
Fig. 26. Sector A floors in Units 14-17. a: Floor
CC and associated rock-filled hearth; b:
Floor BB and associated hearth ................. 703
Fig. 27. Feature R-7 in rear of Sector A, Units 9
and 10. a: north face of stone wall in
Unit 10. b: east profile of Feature R-7.
Cave wall protrusion is shaded ................. 704
Fig. 28. Plan view of Manachaqui Cave interior
showing rock wall features R-6, R-7 and
R-8 ............................................ 705
XX
Fig. 29. Sector A floors in Units 14-17. a: Floor
AA and associated hearth; b: Floor Z and
associated Hearths 1 and 2 . . . . . . . . . . . . . . . . . . . . . 706
Fig. 30. Sector A Floor P and associated hearth
in Units 14-17 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 707
Fig. 31. Sector A floors containing hearths with
embedded stones. a: Floor S in Unit 17;
b: Floor M? in Unit 16 . . . . . . . . . . . . . . . . . . . . . . . . . 708
Fig. 32. Plan views of Sector Brock wall features.
a: Feature R-6 in Units 23 and 28
(Rockshelter exterior is shaded). b:
Feature R-8 in Units 18 and 19 . . . . . . . . . . . . . . . . . 709
Fig. 33. Sector B Feature R-4 in Unit 31, Level 16.
a: Plan view of hearth. b: Profile of
hearth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 710
Fig. 34. Manachaqui Phase Paste A body profiles.
a-h: globular, semi-carinated and
carinated profiles. i-p: same, with
applique medial ribs. q-t: rare "stepped"
shoulder profiles with shoulder ribs ........... 718
Fig. 35. Manachaqui Phase Paste A, Shape A rim
p7ofil~s. a~h: Ri~ 1. i: hypothetical
R1m 1 Jar. J-p: R1m 2 . . . . . . . . . . . . . . . . . . . . . . . . . . 719
Fig. 36. Manachaqui Phase Paste A, Shape A rim
profiles. a, b, g: Rim 2 decorated rims.
c-f: Rim 2 vessels with large orifices.
h: hypothetical Rim 2 jar . . . . . . . . . . . . . . . . . . . . . . 720
Fig. 37. Manachaqui Phase Paste A, Shape A rim
profiles. a-c: Rim 3. d-f: Rim 4. g-i:
Rim 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 21
.Fig. 38. Manachaqui Phase Paste A, Shape A rim
profiles. a-e: Rim 6. f-i: Rim 7. j:
hypothetical Rim 7 jar. k-m: Rim 8.
n -p: Rim 9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 2 2
Fig. 39. Manachaqui Phase Paste A and B rim
profiles. a-e: Shape A Rim 10. f: Shape
A Rim 11. g: Shape A Rim 12. h: Shape A
Rim 13. i-p: Shape BRim la. q-r: Shape B
Rim lb. y-aa: Shape BRim lc . . . . . . . . . . . . . . . . . . . 723
xxi
Fig. 40. Manachaqui Phase Paste A, Shape B rim
profiles. a: Rim la with handle. b:
hypothetical Rim lb jar. c-p: Rim 2a ........... 724
Fig. 41. Manachaqui Phase Paste A, Shape B rim
profiles. a, b: Rim 2a profile and top view.
c-i: Rim 2b. j-p: Rim 3a. q-w: Rim 3b .......... 725
Fig. 42. Manachaqui Phase Paste A, Shape Brim
profiles. a-h: Rim 4. i-p: Rim 5 ............... 726
Fig. 43. Manachaqui Phase Paste A, Shape Brim
profiles. a, b: Rim 5 with applique
decoration. c: hypothetical Rim 5 jar .......... 727
Fig. 44. Manachaqui Phase Paste A, Shape B rim
profiles. a: Rim 5 with rare notched
decoration. b-d: Rim 5 with incision and
applique decoration ............................ 728
Fig. 45. Manachaqui Phase Paste A, Shape B rim
profiles. a-g: Rim 6. h-1: Rim 7. m-o:
Rim 8 .......................................... 7 2 9
Fig. 46. Manachaqui Phase Paste A, Shape B rim
profiles. a: Rim 8 partially reconstructed
j ar . b, c : Rim 9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 3 0
Fig. 47. Manachaqui Phase Paste A, Shape B
Rim 9 profiles ................................. 731
Fig. 48. Manachaqui Phase Paste A, Shape B rim
profiles. a, b: Rim 9. c: Rim 9 partially
reconstructed vessel. d, e: Rim 10 ............. 732
Fig. 49. Manachaqui Phase Paste A, Shape B rim
profiles. a: Rim 10 partially reconstructed
jar. b, c: Rim 10 profiles with top views ...... 733
.Fig. 50. Manachaqui Phase Paste A, Shape C rim
profiles. a-k: Rim la. 1-p: Rim lb. q:
hypothetical Rim la bowl ....................... 734
Fig. 51. Manachaqui Phase Paste A, Shape C rim
profiles. a-i: Rim 2a. j-n: Rim 2b. o:
hypothetical Rim 2a bowl ....................... 735
Fig. 52. Manachaqui Phase Paste A, Shape C rim
profiles. a, b: Rim 3. c-e: Rim 4. f:
Rim 5 . g : Rim 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 3 6
xxii
Fig. 53. Manachaqui Phase Paste A, ShapeD rim
profiles. a-h: Rim 1. i-1: Rim 2. m:
hypothetical Rim 3 bowl ........................ 737
Fig. 54. Manachaqui Phase Paste A, Shape D rim
profiles. a-d: Rim 3. e-g: Rim 4. h-j:
Rim 5 . k, 1 : Rim 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 3 8
Fig. 55. Manachaqui Phase Paste A, Shape E rim
profiles. a, b: Rim 1. c, d: Rim 2. e-h:
Rim 3. i, j: Rim 4. k: Rim 5 ................... 739
Fig. 56. Manachaqui Phase Paste A, Shape E rim
profiles. a: Rim 5. b: Rim 6. c: Rim 7 ......... 740
Fig. 57. Manachaqui Phase Paste A miscellaneous
shapes. a: spoon fragment. b, c: fragments
of unidentified artifacts ...................... 741
Fig. 58. Manachaqui Phase Paste A applique rib
decorations. a-d: unembellished medial
ribs. e-g: unembellished shoulder ribs.
h-m: Notched A medial ribs ..................... 742
Fig. 59. Manachaqui Phase Paste A applique rib
decorations. a, b: Notched A medial ribs.
c: Notched B shoulder rib. d: Notched B
medial rib. e, g: Incised A medial ribs.
f: Incised A shoulder rib ...................... 743
Fig. 60. Manachaqui Phase Paste A applique rib
and band decorations. a: Incised A medial
rib. b-e: Incised B medial and shoulder
flanges. f-k: notched bands .................... 744
Fig. 61. Manachaqui Phase Paste A applique band
decorations. a, b: notched bands with
Incised A ribs. c, d: bands with circular
punctations. e, f: bands with ovoid
incisions. g: applique serpent head ............ 745
Fig. 62. Manachaqui Phase Paste A assorted
applique decorations. a: Incised A
rib/band. b: lug with notched medial rib
ana notched band. c: incised button on
unembellished rib. d: incised button on
Incised A medial rib. e, f: snake adornos
on unembellished ribs .......................... 746
Fig. 63. Manachaqui Phase Paste A adornos. a: head
of amphibean or fish. b: bird wing or
fish fin. c: zoomorphic heads .................. 747
xxiii
Fig. 64. Manachaqui Phase Paste A incised
decoration. a-c: Incised Vessel 1 with
zoned punctation. d, e: Incised Vessel 2
with zoned punctation. f: incised sherd ........ 748
Fig. 65. Manachaqui Phase Paste Bl vessels. a:
Vessel 1, b: Vessel 2, c: Vessel 3 ............. 749
Fig. 66. Manachaqui Phase Paste Group B. a:
Paste Bl' Vessel 4. b: Paste Bl Vessel 5.
c: Paste B2 vessel ............................. 750
Fig. 67. Manachaqui and Suitacocha Phase artifacts.
a: Manachaqui Phase Paste B3 vessel. b:
Manachaqui Phase incised slate disk. c:
Suitacocha Phase sherd with net impressions
on interior surface ............................ 751
Fig. 68. Suitacocha Phase Paste A, Shape A and B
rim profiles. a-c: Shape A Rim 14. d-f:
Shape A rim 15. g, h: Shape A Rim 16. i:
Shape A Rim 17. j-s: Shape BRim lla ........... 752
Fig. 69. Suitacocha Phase Paste A, Shape Brim
profiles. a-c: Rim lla. d-k: Rim llb.
1-r: Rim llc ................................... 753
Fig. 70. Suitacocha Phase Paste A, Shape Brim
profiles. a-h: Rim lld. i-1: Rim lle.
m-s: Rim llf. t-w: Rim llg ..................... 754
Fig. 71. Suitacocha Phase Paste A, Shapes B and
C rim profiles. a-f: Shape B Rim 12. g-1:
Shape B Rim 13. m-u: Shape C Rim 7 ............. 755
Fig. 72. Suitacocha Phase Paste A, Shape C rim
profiles. a: Rim 7. b-d: Rim 8 ................. 756
Fig. 73. Suitacocha Phase Paste A, Shape C and
E rim profiles. a-c: Shape C Rim 9. d:
Shape C Rim 10. e-h: Shape E Rim 8. i:
Shape E Rim 9. j: Shape E Rim 10. k, 1:
Shape E Rim 11 ................................. 757
Fig. 74. Suitacocha Phase Paste A partially
reconstructed Shape F Rim 1 vessel ............. 758
Fig. 75. Suitacocha Phase Paste A, Shape F rim
profiles. a-n: Rim la. o-v: Rim lb ............. 759
Fig. 76. Suitacocha Phase Paste A, Shape F rim
profiles. a-k: Rim 2. 1-n: Rim 3 ............... 760
xxiv
Fig. 77. Suitacocha Phase Paste A, Shape F Rim 3
profiles ....................................... 7 61
Fig. 78. Suitacocha Phase Paste A, Shape F rim
profiles. a: Rim 3. b-d: Rim 4a. e, f:
Rim 4b. g, h: Rim 5. i, j: Rim 6. k:
Rim 7 .......................................... 7 6 2
Fig. 79. Suitacocha Phase Paste A, Shape F and
X rim profiles. a-c: Shape F Rim 8. d-j:
Shape X Rim 1. k-n: Shape X Rim 2. o-r:
Shape X Rim 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 6 3
Fig. 80. Suitacocha Phase Paste A rim profiles
and basal sherds. a: Shape X Rim 4. b:
Shape X Rim 5. c: Shape X Rim 6. d: jar or
bowl base. e-h: basal angles from carinated
bowls .......................................... 764
Fig. 81. Suitacocha Phase Paste A miscellaneous
shapes. a-c: fragments of mammiform vessel
legs. d-h: handle fragments? i:
unidentified fragment .......................... 7 65
Fig. 82. Suitacocha Phase Paste A sherds with
notched and punctate rib or band decoration.
a-d: notched. e: parallel notched. f, g:
high relief notched. h, i: ovoid
punctation. j: round punctation ................ 766
Fig. 83. Suitacocha Phase Paste A sherds with
applique and incised decoration. a: round
punctation. b, c: unembellished applique.
d-f: flanges. g-i: incised lines. j:
divergent arrays of parallel lines. k, 1:
cross-hatching ................................. 767
Fig. 84. Suitacocha Phase Paste A decorated sherds
with stamped circles. a: in row. b, h:
with paint. c-f, i-1: with incision. j, k:
with punctation ................................ 768
Fig. 85. Suitacocha Phase Paste A decorated sherds.
a-f: stamped circles and applique with
round punctations. g, h: punctation and
incision ....................................... 7 69
Fig. 86. Suitacocha Phase Paste A decorated sherds
with punctation and incision ................... 770
XXV
Fig. 87. Suitacocha Phase Paste A sherds with
punctation. a-h: with punctation and
incision. i: punctation and applique with
round punctation. j: punctation, incision,
notched applique and red paint . . . . . . . . . . . . . . . . . 771
Fig. 88. Suitacocha Phase Paste A assorted
decorated sherds. a, b: punctation,
incision and notched applique. c, d:
combing. e: brushing. f: rouletting. g:
incised boss. h: fine-line scratched ........... 772
Fig. 89. Suitacocha Phase Paste A assorted sherds.
a, b: mat and/or fabric impressed. c:
zoomorphic adorno depicting parrot head.
d: zoomorphic adorno {opposum head?}. e:
zoomorphic adorno (probably a bat's head
and face} . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 773
Fig. 90. Suitacocha Phase Paste A adorno and
Group B sherds. a: anthropomorphic head
adorno. b: Paste Be. c, d: Paste B4 jar
rims. e-g: Paste B4 sherds with punctation
and incision. h: Paste B5 sherd from bowl ...... 774
Fig. 91. Colpar Phase Paste A, Shape B and C rim
profiles. a-e: Shape BRim 11h. f-h:
Shape BRim 11i. i, j: Shape BRim 11j.
k-o: Shape C Rim 11. p-s: Shape C Rim 12.
t: Shape C Rim 13. u: Shape C Rim 14. v:
Shape C Rim 15 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 775
Fig. 92. Colpar Phase Paste A, Shape E rim
profiles and decorated sherds. a: Rim 3,
b, c: Rim 9. d-j: Rim 12. k, 1: Rim 13.
m: Rim 14. n-p: notched applique ............... 776
Fig. 93. Colpar Phase Paste Group B. a, b:
Paste B6 rim and basal angle profiles.
c-e: Paste ~ rim profiles. f: Paste ~
rim with iridescent red paint. g: Paste B8
rim with white-on-red decoration ............... 777
Fig. 94. Colpar Phase Paste Groups B and C.
a, b: Paste B8 rim profiles. c-e: Paste
B9 rims with red paint. f: Paste B10 rim
with orange-red paint. g: Paste C1 rim
with red paint . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 8
Fig. 95. Ernpedrada Phase Paste A Shape E rim
profiles. a-c: Rim 3. d-k: Rim 9. 1-s:
Rim 15. t-w: Rim 16 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 779
xxvi
Fig. 96: Empedrada Phase Paste A Shape E rim
profiles. a-d: Rim 17. e-h: Rim 18. i-k:
Rim 19. 1-n: Rim 20. o: Rim 21. p, q:
Rim 22. r: Rim 23. s: Rim 24 ................... 780
Fig. 97. Ernpedrada Phase Paste A Shape GRim
profiles. a-m: Rim 1. . . . . . . . . . . . . . . . . . . . . . . . . . . 781
Fig. 98. Ernpedrada
profiles.
Rim 4. o:
s, t: Rim
Phase Paste A Shape G rim
a-h: Rim 2. i-k: Rim 3. 1-n:
Rim 5. p, q: Rim 6. r: Rim 7.
8. u: Rim 9. v: Rim 11 ............... 782
Fig. 99. Ernpedrada Phase Paste A Shape H rim
profiles. a-d: Rim 1. e: Rim 2. f: Rim 3.
g : Rim 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 8 3
Fig. 100. Ernpedrada Phase artifacts. a, b: Paste A
spoon fragments. c: unidentified artifact.
d-e: notched applique. f: Paste Bu jar
rim profile. g: Paste Bu rim profile ......... 784
Fig. 101: Empedrada Phase Paste Group B. a-c:
Paste Btl bowl with red paint. d-f: Paste
B14 sherds from bowls with red and brown
paint ......................................... 7 85
Fig. 102. Empedrada Phase Paste Group B. a, b:
Paste B15 bowl with dark red paint, pale
orange slip on exterior. c: Paste B16 :
bowl with iridescent red paint. d, e: Paste
B17 sherds with orange and white paint on
brown slip. f: Paste B18 bowl with orange
slip and negative-resist decoration ........... 786
Fig. 103. Empedrada Phase Paste Group C2 • a-p:
unpainted. q-w: traces of red paint. x:
Paste C2a with red paint ...................... 787
Fig. 104. Empedrada Phase Paste Group C2 • a-e:
Paste C2a with red paint ...................... 788
Fig. 105. Empedrada Phase Paste Group C2 • a-e:
Paste C2 a with red paint ...................... 789
Fig. 106. Empedrada Phase Paste Group C2 • a-e:
Paste C2a with red paint. f-n: Paste C2a
with orange slip .............................. 790
Fig. 107. Empedrada Phase Paste C2 • a-e: Paste C2b
with black paint .............................. 791
xxvii
Fig. 108. Empedrada Phase Paste C2 • a-d: Paste C2b
with black paint. e-g: Paste C2c with
red and black paint ........................... 792
Fig. 109. Empedrada Phase Paste C2 • a-c: Paste C2c
with red and black paint. d-f: Paste C7d
with tan paint. g, h: Paste C1 e with
black paint on orange slip .................... 793
Fig. 110. Empedrada Phase Pastes C2 • a: Paste C2 f
with brown slip (bottom), negative-resist
smudging (top) and white painted circles
on natural tan paste (middle) . b: Paste
C2g with red paint on light brown paste.
c: Paste C2h unpainted jar rim. d: notched
ground slate disk from Unit 10 Level 2 ........ 794
Fig. 111. Pre-Lavasen Phase animal size
distribution (NISP=Number of Individual
Specimens) .................................... 7 9 5
Fig. 112. Lavasen Phase animal size distribution
(NISP=Number of Individual Specimens) ......... 795
Fig. 113. Manachaqui Phase animal size distribution
(NISP=Number of Individual Specimens) ......... 796
Fig. 114. Suitacocha Phase animal size distribution
(NISP=Number of Individual Specimens) ......... 796
Fig. 115. Colpar Phase animal size distribution
(NISP=Number of Individual Specimens) ......... 797
Fig. 116. Empedrada Phase animal size distribution
(NISP=Number of Individual Specimens) ......... 797
Fig. 117. Guaman Perna's rendering of a camelid with
a cloth or net alforja containing jars for
transport (from Guaman Poma 1936:524) ......... 798
Fig. 118. Map showing relative degrees of
stylistic similarity to the Manachaqui
Phase Paste A style, and the probable
sources of Paste Group B sherds. Darkest
shading indicates closest similarity to
Manachaqui Paste A............................ 799
xxviii
Fig. 119. Map showing relative degrees of
stylistic similarity to the Suitacocha
Phase Paste A style, and the probable
sources of Paste Group B sherds. Darkest
shading indicates closest similarity to
Suitacocha Paste A ............................ 800
Fig. 120. Map showing relative degrees of
stylistic similarity to the Colpar Phase
Paste A style, and the probable sources of
Paste Group B and C sherds. Darkest
shading indicates closest similarity
to Colpar Paste A............................. 801
Fig. 121. Map showing relative degrees of
stylistic similarity to the Empedrada
Phase Paste A style, and the probable
sources of Paste Group B and C sherds.
Darkest shading indicates closest
similarity to Empedrada Paste A. Q=
source of obsidian ............................ 802
LIST OF PLATES
Plate I.
View of the Manachaqui Cave site complex
from the north side of Manachaqui Valley
facing south . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 804
Plate II.
Manachaqui Cave and 1988 trench looking
east . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 0 5
Plate III.
Manachaqui Cave 1990 work in progress ....... 805
Plate IV.
Manachaqui Cave Unit 6 north
profile 1988 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 806
Plate V.
Manachaqui Cave Unit 13 north
profile 1990 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 807
Plate VI.
Manachaqui Cave Units 14-17
Floor CC, 1990 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 807
Plate VII.
Manachaqui Phase Shape B jar
(see Fig. 48c) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 808
Plate VIII. Manachaqui Phase Paste B2
(see Fig. 66c) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 808
Plate IX.
Ground slate points. m: Manachaqui
Phase. K, 1: Suitacocha/Colpar Phase.
f, g: Colpar Phase. a-e, h-j:
Empedrada Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 809
Plate X.
Suitacocha Phase Shape F jar
(see Fig. 74) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 810
Plate XI.
Colpar Phase. Paste B9
(see Fig. 94c-e) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 810
Plate XII.
Empedrada Phase Paste C2 a kaolin sherds ..... 811
Plate XIII. Ernpedrada Phase projectile points . . . . . . . . . . . 811
xxix
LIST OF TABLES
Table 1.
Soil sediments from Unit 15, Sector A ......... 813
Table 2 .
Soil sediments from Unit 36, Sector B ......... 813
Table 3 .
Hearth classification, measurements
and chronology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 814
Table 4.
Radiocarbon dates from Manachaqui Cave ........ 815
Table 5.
Manachaqui Phase Shape A rims ................. 816
Table 6.
Manachaqui Phase Shape B rims ................. 816
Table 7.
Manachaqui Phase Shape
Table 8 .
Manachaqui Phase Shape D rims ................. 817
Table 9.
Manachaqui Phase Shape E rims ................. 817
c
rims ................. 817
Table 10. Totals pertaining to rib and flange
embellishment techniques ...................... 817
Table 11. Suitacocha Phase Shape A rims ................. 818
Table 12. Suitacocha Phase Shape B rims ................. 818
Table 13. Suitacocha Phase Shape
c
rims ................. 818
Table 14. Suitacocha Phase Shape E rims ................. 818
Table 15. Suitacocha Phase Shape F rims ................. 819
Table 16. Empedrada Phase Shape E rims .................. 819
Table 17. Empedrada Phase Shape G rims .................. 819
Table 18. Botanical remains by phase .................... 820
Table 19. Fauna of the Lavas en Phase .................... 821
Table 20. Fauna in Lavasen/Manachaqui Strata ............ 821
Table 21. Fauna of the Manachaqui Phase ................. 822
Table 22. Fauna in Manachaqui/Suitacocha Strata ......... 822
Table 23. Fauna of the Suitacocha Phase ................. 822
Table 24. Fauna in Suitacocha/Colpar Strata ............. 823
XXX
xxxi
Table 25. Fauna of the Colpar Phase ..................... 823
Table 26. Fauna in Colpar/Ernpedrada Strata .............. 823
Table 27. Fauna of the Empedrada Phase .................. 824
Table 28. Animal size by phase. Sums are NISP
(Number of Individual Specimens) .............. 825
CHAPTER 1
THE PROBLEM OF TROPICAL MONTANE
FOREST PREHISTORY
It is ... a zone of tropical conditions wherein Nature takes
on terrifying proportions--terrifying luxuriance of
vegetation, terrifying onrush of rivers, terrifying animal
life. Never, at any rate so far as we now know, have either
the lower portions of the ceja de la montana or the montana
itself been the seat of any stable and advanced community;
on the contrary, they are probably the habitat of arboreal
man in an archaic stage of culture beyond which progress is
not possible save under very definite suasion from the
outside world.
{Philip A. Means 1931:22-23)
Pitt downed the last of his brandy.
"It doesn't seem
possible an advanced civilization existed in such a remote
region without some kind of outside influence."
{Dirk Pitt, action hero in Clive
Cussler's novel Inca Gold 1994)
This thesis presents a description and interpretation
of new archaeological data from the edge of the eastern
tropical montane forest in northern Peru.
These high-
altitude forests are referred to variously in published
literature as "cloud forest," "ceja de selva" and "ceja de
montana."
The eastern tropical montane forest covers the
slopes of the Central Andean cordillera between
approximately 3,500 and 1,500 m, and marks the environmental
and cultural transition from highland Andes to lowland
Amazon jungles.
In modern-day Peru, economic and cultural
linkage between Andean and Amazonian regions is tenuously
1
2
maintained by air traffic, and by half a dozen dangerous,
winding roads that are frequently rendered impassible by
stormy weather and landslides.
Western society conditions
us to presume that without airplanes, chainsaws, bulldozers,
dynamite and powerful economic incentives, substantial
interchange between the Andes and eastern lowlands would be
virtually impossible to sustain.
This perception of the
montane forest as a remote and inhospitable frontier has a
long history, and it lingers still, as the quote from
Cussler's best-selling tall tale of treasure and treachery
in the montane forests of Chachapoyas attests.
Traditional views of an empty eastern frontier have
been based on assumptions like those listed by J. Scott
Raymond (1976:205-206):
1) That the rugged terrain and dense vegetation of the ceja
inhibited or prevented travel between the highlands and the
Montana.
2) That disease, heat and pests made colonization of the tropical
forest difficult, if not impossible.
3) That the poor soils, heavy rainfall and steep slopes of the
eastern highlands severely restricted the agricultural potential.
4) That historic and present demographic patterns in the montana
are a good indication of the pre-Columbian demography.
As Raymond points out, these assumptions have no basis in
archaeological evidence.
The final assumption regarding
demography has been particularly pervasive.
Both popular
and scholarly perceptions have had to overcome ambivalent
and negative characterizations of the "largely unpopulated"
montane forest (Steward 1948:508).
Through pioneering
studies of the 1960s and 70s that documented surprisingly
3
plentiful evidence of pre-Hispanic montane forest
occupation, mainstream scholarly thought regarding the
forested eastern slopes has been revolutionized such that
Lyon (1981:8) has remarked,
"Wherever a reasonable amount of
research has been carried out we find continuous occupation
and utilization of the land from the highlands into the
montana."
Despite the new perspective provided by a half-dozen
projects over the past three decades, scholars still tend to
regard the tropical montane forest alternately as: 1) a
barrier containing the eastward spread of Andean
civilizations, 2) a threshold (or doormat) for early
prehistoric upslope and downslope population movements and
3) the intractable recipient of late prehistoric highland
agricultural colonies.
Each of these positions perpetuates
stereotypes of sparse and inconsequential montane forest
populations passively witnessing repeated incursions by
"culturally superior" migrants bound for elsewhere (RenardCasevitz et al. 1988:5), or colonists from expansionist
highland polities.
The montane forest's potential for human settlement and
utilization was first scientifically evaluated by Tosi
(1960:148) who concluded that the montane wet forest (bosque
muy humedo montane) constituted a nearly useless, and even
dangerous,
formation for farmers and herders.
He attributed
repeated and failed attempts to colonize the upper forest to
4
the excessively humid climate and unusually steep terrain
(Ibid.:l54).
The region's low economic potential was later
reiterated by the Peruvian government's Oficina Nacional de
Evaluaci6n de Recursos Naturales (ONERN 1976).
Archaeologists have repeated the same themes.
(1967:11), for example, wrote that:
Lanning
"Because of the
combination of intense rains and steep mountain sides, the
high cloud forest is a dangerous area which may not have
been much exploited either for agriculture or as pasture."
In reference to the eastern limits of Andean civilization,
Lanning remarked that "the Incas failed to import their own
civilization to the montana, the Spanish failed in their
turn, and the modern Peruvians have failed in spite of
repeated attempts"
(Ibid.:197).
Inca "failure" to conquer
the forest lands to the east has been an oft-repeated theme
(Means 1931:22-23; Meggers 1954:808; Gade 1972, 1979).
Traditional views of historical demography fostered by the
lack of documentary evidence for habitation of the upper
forests have been further reified by the inability of modern
Peruvians (excepting the recent coca boom) to intensively
colonize the region.
Archaeologists habitually cite this
repeated failure as a measure of the montane forest's
inhospitable disposition with regard to human utilization
(e.g. Bonavia 1968:76; Kauffmann 1986:6-7).
Despite slowly shifting perspectives, indigenous
montane forest prehistory is still subordinated in the
5
interest of furthering "macro-scale analyses"
(Lightfoot and
Martinez 1995) of highland Andean cultural evolution and
political economy.
Scholars working within evolutionist
paradigms have postulated Andean-Amazonian population
movements to
su~port
universalist, ecologically-based
explanations of cultural innovation and emerging sociopolitical complexity in the Andes, and in Amazonia as well.
Andeanists effecting macro-scale analyses of highland
political economy view montane forest colonization as an
adjunct feature of imperial expansion in the adjacent
highlands.
The population movement hypotheses
characteristically ignore the possibility of montane forest
inhabitants, while the colonization models accord them
parenthetical treatment at best.
Although the aforementioned paradigms continue to shape
the predominant perceptions of montane forest prehistory, a
small group of archaeologists, mostly Peruvians, has worked
outside of these macro-scale frameworks, documenting longterm montane forest occupations and crediting montane forest
populations with a dynamic capacity in prehistoric
interregional interaction. Ironically, some of these
scholars have emphasized the unique trajectory of montane
forest cultural development (Morales 1993).
Lumbreras
(1974:149), in his synthesis of Peruvian prehistory, even
suggests that the eastern montane forest represents "a new
culture area, distinct from the Central and Northern ones."
6
Such disparity in opinion has few parallels in world
archaeology, and it exemplifies the considerable power that
theoretical frameworks exert over data interpretation.
Despite a recent modest resurgence, migration theories have
fallen out of mainstream archaeological discourse.
Nonetheless, hypothesized Andean-Amazonian population
movements retain a measure of credibility because they
highlight anomalous instances of similar material patterning
in these highly dissimilar, yet juxtaposed regions (e.g.
Tello 1942; Meggers 1971:146-149; Lathrap 1970, 1971, 1974;
Rivera 1975, 1991).
Documentary evidence supports models
postulating montane forest colonial enclaves (Murra 1964,
1967, 1975) that are still formulated and re-formulated
today.
In effect, these proposed migratory episodes have
come to represent the salient developmental events in
eastern montane forest prehistory.
Therefore, any
archaeological treatment of this region must address issues
of population intrusion.
As archaeology is developing
increasingly sophisticated frameworks for examining
archaeological evidence from frontiers and boundaries (e.g.
Schortman and Urban 1992; Lightfoot and Martinez 1995) a
reevaluation of these montane forest models is appropriate
and timely.
With this thesis, I intend to weigh both new and extant
montane forest evidence for hypothesized long-distance
population movements and highland colonization against the
7
alternative hypothesis that the tropical montane forest
harbored substantial indigenous populations with long
histories of in situ cultural development.
A re-analysis of
evidence once utilized to infer prehistoric migrations can
now be shown to demonstrate that montane forest societies
participated in complex modes of economic interaction,
acting alternately as primary purveyors and intermediary
conveyors in long-distance exchange connecting distant
Central Andean, Amazonian and Northern Andean highland and
lowland regions.
An analysis of continuity and change in cultural
remains recently excavated from Manachaqui Cave constitutes
the core of my thesis.
Today, this small rockshelter at the
upper edge of the montane forest infrequently lodges hunters
and cattle herders from nearby villages.
Excavations reveal
that Manachaqui Cave was more intensively utilized during
prehistory.
Because the remains of a road that connected
prehistoric highland population centers and tropical forest
destinations lies only a few meters to the north, it is
reasonable to suppose that Manachaqui Cave once sheltered
travelers routinely crossing the ecotone.
Manachaqui Cave's stratified deposits contain a record
of human utilization beginning approximately 8,000 B.C. and
the early Preceramic Period, and ending with the arrival of
the Spanish conquistadors in 1532.
Period, Initial Period,
The Late Preceramic
the first half of the Early Horizon,
8
the Early Intermediate Period and the Late Intermediate
Period and/or Late Horizon are particularly wellrepresented.
Conspicuous gaps in the sequence coincide with
the Chavin horizon (500- 200 B.C.), the Middle Horizon
(A.D. 700 - 900) and possibly the Late Intermediate Period.
Study of the cultural sequence at Manachaqui Cave offers a
window on the dynamics of Andean-Amazonian interaction, and
related cultural processes before and after the emergence of
Chavin civilization.
The rockshelter's proximity to modern
Ecuador renders a similar opportunity to examine the Central
Andes' changing relationship to the Northern Andes.
My interpretation of the Manachaqui evidence suggests
that popular "verticality" or "zonal complementarity" models
are inappropriate frameworks for determining modes and
motivations involved in most of the interactions documented
at Manachaqui.
I call attention to the need for the
development and application of alternative models of Andean
interaction and exchange that avoid the a priori assumptions
that interaction on the forested slopes was always vertical,
or even ecologically complementary.
It would seem more
appropriate to view the so-called montane forest "frontier"
as a dynamic "cultural interface" where populations and
population segments consciously manipulated styles and
symbols to express changing intra-group and inter-group
boundary relations, and to facilitate certain kinds of
interaction (Hodder 1982; Lightfoot and Martinez 1995:485).
9
Rather than remote peripheries, Central Andean montane
forests were loci of creativity where interaction generated
innovations with potential developmental ramifications in
adjoining regions.
The conclusion that autochthonous populations did
develop independently in the eastern montane forests is
significant for several reasons.
First, the archaeological
sequence invalidates assumptions that montane forest
populations are always intrusive.
Second, while theories of
early population movements probably cannot be "disproved,"
they are rendered highly suspect as the burden of proof must
now be borne by migration advocates.
Third, the long record
of interregional interaction suggests that analyses of
cultural evolution emphasizing only internally generated
processes neglect a significant and dynamic component of
Central Andean cultural development.
A new awareness of
montane forest prehistory is especially crucial to the
development of theories that accord interregional
interaction a powerful role in stimulating the growth of
cultural complexity in the Andes and Amazonia (e.g. Shady
and Rosas 1979; Burger 1992).
In the following section, I will describe the Central
Andes,
the Central Andean tropical montane forests, and the
eastern tropical montane forest in particular, in terms of
geography and environment.
I summarize the predominant
interpretive frameworks utilized by Andeanists that have
10
shaped our understanding of montane forest prehistory.
Emphasized are the roles of evolutionary paradigms with
unilinear assumptions of unitary origins, and "verticality"
or "complementarity" models of Andean political economy
commonly invoked to account for eastern montane forest
settlement.
Then, by specifying explicit sets of
archaeological correlates for hypothetical population
movement and colonization scenarios distilled from the
extant theoretical literature, I outline a methodology for
evaluating the explanatory effectiveness of both models.
I
point out some difficulties involved in correctly inferring
migration from the kinds of evidence available, and I
suggest a more fruitful manner of conceptualizing boundary
interaction that partially reconciles the deep division
between competing migration and interaction explanations.
Chapter 2 will describe the proposed population
movement and colonization scenarios postulated by
Andeanists, as well as the larger theoretical agendas that
their hypotheses served.
Chapter 3 introduces the Pataz-
Abiseo study area, presents a scheme of ecological zonation
and describes modern economic activities.
An analysis of
Pataz-Abiseo culture history offers my interpretations of
existing ethnohistoric and archaeological evidence for
changing historic and prehistoric demography, settlement
patterns, cultural boundaries and interaction networks.
Here I present new documentary evidence for politically
11
autonomous montane forest societies, and show how postConquest events transformed the Pataz-Abiseo area from a
node of interregional interaction into the "imaginary
frontier" described by Lyon (1981).
This background sets the stage for description of the
Manachaqui site environment and delineation of the working
hypothesis that Manachaqui Cave functioned as a wayside
station servicing prehistoric interregional travel and
commodity transport (Chapter 4) .
Chapters 5 through 9
present the archaeological data recovered during the 1988
and 1990 field seasons.
Chapter 10 will first present an
evaluation of Manachaqui Cave's postulated wayside station
function, and then show how the new data support an
interpretation of long-term autochthonous development rather
than the population movement or colonization hypotheses.
In
Chapter 11, I implement a theoretical framework emphasizing
interregional exchange and boundary interaction to interpret
Manachaqui Cave's archaeological record.
Utilization of
this alternative framework reveals that the tropical montane
forests were loci of early Central Andean population
nucleation, technological innovations and complex social
formations.
While migration explanations of culture change
tend to stifle further inquiry, an interregional interaction
framework provides a starting point for future research.
Andean Environments and Culture Areas
This thesis deals with an archaeological problem of
12
ample spatial and temporal dimensions, and so must include
definition of pertinent geographic regions and cultural
sequences for heuristic purposes.
The regions are: the
Central Andes, western Amazonia and the Northern Andes, but
the following paragraphs will focus on environmental
distributions in the Central Andes.
Geographic, as well as
cultural distinctions between the Central Andean, Amazonian
and Northern Andean regions (Fig. 1) have been consistently
recognized by scholars during the mid-twentieth century
(e.g. Steward 1949; Steward and Faron 1959; Willey 1971)
although their terminology has varied.
The term "Amazonia"
refers to the Amazon River Basin below 1,500 mas delineated
by Meggers and Evans (1983).
corresponds to Lumbreras'
The term "Northern Andes"
(1981)
"Andes Septentrionales" and
especially modern Ecuador (of course the prehistoric
significance of modern political boundaries should not be
assumed).
Throughout this thesis, Rowe and Menzel's (1967)
Central Andean prehistoric master sequence is employed as
the least encumbered by evolutionary assumptions.
The
sequence is as follows: Preceramic Period (10,000-1,800
B.C.), Initial Period (1,800-900 B.C.), Early Horizon (900200 B.C.), Early Intermediate Period (200 B.C.- A.D. 700),
Middle Horizon (A.D. 700-900), Late Intermediate Period
(A.D. 900-1350), and Late Horizon (A.D. 1450-1532).
Within
the Early Horizon, Burger (1988, 1992, 1993) defines the
13
Chavin horizon (500-200 B.C.) as the apex of the Chavin
cult's influence and a pivotal event marking the emergence
of ranked socio-economic hierarchies and state-level
political organization in the Central Andes.
The Central Andes
The term "Central Andes" as it is utilized in this
study refers to the coastal and highland environments within
the borders of modern Peru.
The southern limits of the
Central Andes area are not directly pertinent to this study,
but Bennett's (1948) and Burger's (1984a) identification of
a northern boundary or buffer zone separating the Peruvian
Co-tradition from the Northern Andes justifies their
separation as distinct units of analysis in the present
study.
The antiquity of this boundary and its permeability
at different prehistoric moments are issues of current
debate (Hocquenghem et al. 1993).
The Sechura Desert in
extreme north-coastal Peru, the Catamayo Valley in interior
Ecuador and the low (2100 m) Andean pass at Porculla are
landmarks within the inter-Andean area.
Bennett regarded the tropical forest edges as the
eastern limit of the Peruvian Co-tradition (Bennett and Bird
(1949:95-96), but explorers (e.g. Savoy 1970) and
archaeologists (e.g. Raymond 1985, Hastings 1985;
Schjellerup 1992) have recently documented evidence of preHispanic Andean settlement deep within Amazonian tropical
forests.
The degree to which the pre-Hispanic Central
14
Andean montane forests covering the eastern slopes can be
conceived of culturally as "Central Andean,"
"Amazonian" or
distinct from either (cf. Lumbreras 1974:149, 1981:32;
Kauffmann et al. 1989:6; Morales 1993:642-653) remains to be
demonstrated.
Problems regarding the eastern boundary's
locations, antiquity and permeability are central to this
thesis.
The contrasting environments of the arid coast and
seasonally-moist highlands are usually distinguished in
published descriptions of Peru and the Central Andes (e.g.
Moseley 1992; Burger 1992) because they afforded differing
subsistence opportunities to human inhabitants from earliest
times.
The cold Humboldt Current which sweeps north against
the Peruvian coast before veering out to sea near the
Ecuadorian border conditions the coastal climate.
The
combination of the cold water current, easterly winds and
tropical latitudes creates a temperature inversion which
results in extreme aridity relieved only by winter fog and
drizzle from May to November, especially between 300 and 800
m.
The Humboldt Current also supports unusually high
quantities of marine
fa~,a
that comprised a resource of
primary dietary importance to coastal dwellers.
Cyclic
changes in the ocean currents every seven to 10 years
generates heavy rains and occasional catastrophic flooding
called El Nino.
The dry coastal plain separating the steep western
15
flanks of the Central Andean cordillera and the Pacific
Ocean narrows southward from 100 kms to less than 20 kms.
Aridity correspondingly increases moving south from the
humid mangrove swamps at the Ecuadorian border.
This
coastal desert stretches from Peru's northern border well
into Chile, but is broken at regular intervals by river
valleys oriented east-west and draining wet-season highland
precipitation between October and May.
Low, barren and
rocky divides separate coastal valley oases where human
populations have always clustered, especially since plant
cultivation became a subsistence priority.
A belt of steep, semi-arid slopes, canyons and thorn
forests constitutes the transition from coastal to highland
environments.
The Andean highlands become higher, wider and
drier from north to south.
The geographer Troll (1958)
distinguishes between the high "puna Andes" of central Peru
and Bolivia, and the lower "paramo Andes" of Colombia,
Ecuador and northern Peru.
Salomon summarizes the
distinction between the two Andean landscapes:
In the former, the upper slopes are characterized by scant
rainfall, strong insolation, and sharp diurnal temperature
variation regularly producing nighttime freezes.
In the
latter, the upper slopes receive considerable rainfall and
drizzle but little direct insolation. They do not freeze as
regularly.
The true "paramo Andes," that is, montane areas
in which no puna occurs, extend northward from a line
running approximately through Trujillo and Cajamarca, Peru.
Some paramo may occur in the eastern cordillera south of
this line as well ... (Salomon 1986:24).
Frost conditions highland agriculture in the puna Andes, and
Salomon documents a subsistence focus on tuber cultivation
16
that contrasts with the maize-based agriculture
characteristic of the paramo Andes.
Although the puna Andes
environment appears to correspond to the Central Andean
region, Burger (1984a) emphasizes the lack of a close fit
between cultural affiliation and environment in northern
Peru.
As Salomon observed, the paramo extends well into the
Central Andes.
East of the Central Andean continental divide, highland
waters feed tributaries of the Amazon River.
The greatest
ecological variation is encountered where the largest Amazon
tributaries such as the Marafion, Mantaro, Apurimac and
Urubamba Rivers cut deep canyons into the cordillera.
Areas
of dry, xerophytic vegetation in valley bottoms below
approximately 2,300 mare called temple (Rosas and Shady
1974) or yunga in native terminology (Pulgar 1987).
Relatively moist intermontane valleys of the quechua between
2,300 and 3,200 m support cultivation of many modern
highland staple crops, but the greatest variety of
prehistorically important tubers is grown above 3,200 min
the jalca or suni.
The higher and wider Andean cordillera
farther south (below 10° south latitude) includes wide
expanses of puna environments (above 4,000 m according to
Pulgar 1987) where cultivation was undertaken at substantial
risk and camelid pastoralism was a principal pre-Hispanic
economic activity.
The southern Central highlands
surrounding Lake Titicaca constitute a broad plateau with
17
precipitous western and eastern flanks.
These environments
supported cultural developments of only secondary relevance
for this study.
The Central Andean Tropical
Montane Forests
The tropical montane forests are woodland environments
found at a maximum altitude of approximately 3,700 m around
the peripheries of the Central Andean region.
Based upon
"topography and predominant climatic and biogeographical
influences"
(Young and Leon 1993:237}, the greater tropical
montane forest has been divided into highland, northern,
western-slope and eastern-slope forests
(Fig. 2}.
In Peru,
the implementation of the Holdridge System to classify
ecological zones (Tosi 1960; ONERN 1976} has heightened
awareness that tropical montane forests were once more
extensive than they appear today (Guillet 1985}.
At
present, the montane forest is most clearly seen as two
parallel, altitudinally-restricted belts extending north and
south along the western and eastern flanks of the Andean
cordillera.
On the western side the forest is most clearly
defined from the 7th parallel south latitude (near Peru's
Chicarna River valley} to northern-most Colombia, while the
eastern forest stretches from northernmost Argentina to
Venezuela.
The western montane "dry" forest has been
divided into macrotherrnal (lower: below 1000 m}, mesotherrnal
(middle: 1200-2000 m) and oligotherrnal (upper: 2500-3200 m}
18
forests (Valencia 1992).
Mesothermal forests like the upper
Zafia montane forest described by Dillehay and Netherly
(1983) do not occur below the 7th parallel, while
oligotherrnal cloud forest fragments extend farther south.
The eastern montane forest shades into the lowland tropical
forests of greater Amazonia below 1,500 m.
Fragmented
northern montane forest joins the eastern and western strips
at the low saddle separating the Northern and Central Andean
massifs.
Of the four divisions, the highland forest has suffered
the greatest impact from human depredation.
Its extent
prior to heavy anthropogenic deforestation in the distant
past has not been adequately appreciated in the
archaeological literature, mostly due to lack of
paleoenvironmental field studies.
Guillet (1985) describes
several processes that contributed to their destruction
during both historic and prehistoric eras.
Young and Leon
(1993:239) estimate that only five to ten percent of the
original surface cover remains of Peru's highland Andean
montane forests.
Continuous expanses of northern and western-slope
montane forest become increasingly fragmented south of the
Ecuadorian border reflecting both human impact and natural
processes (Ibid. :237; Valencia 1992).
M. Shimada (1982,
1985) reports ethnohistoric and archaeofaunal evidence for
more extensive woodlands in the Cajarnarca Basin during
19
prehistory.
She emphasizes the importance of forest-
browsing white-tailed deer in the intermontane valley's
early subsistence economy.
Perhaps little more than 25
percent now remains of the northern montane forest (Young
1992a:58).
The western slopes of the northern Peruvian Andes show
remnants of previously more extensive montane forest, and
the best known forest relic lies in the upper Zafia River
valley (Valencia 1992; Alva 1988a).
Dillehay et al.
(1979)
report that the upper Zafia montane forests once supported
endemic fauna that included parrots, monkeys, jaguars and
boa constrictors.
Valencia (1992: Fig. 2) notes that modern
western-slope human populations cluster most densely in the
oligotherrnal forests around 3200 m.
The eastern-slope
montane forest is of primary interest for this study.
It
differs both structurally and physiographically from the
western "dry" forest, and it is separated from the western
and northern montane forests by the Marafion River Canyon
(Young 1992a) .
The Eastern Montane and
Premontane Forests
While scholars in the past maintained simplistic views
of the eastern montane forest as an environmental and
cultural transition separating the highland punas from the
lowland tropical forests, geographers and biologists have
recently recognized it as an exceptional and singular
20
environment (Gentry 1992; Leon et al. 1992; Young 1992b,
1993; Young and Leon 1993; Young and Valencia 1992), worthy
of study in its own right.
Young (1992b:121) characterizes
the Central Andean eastern-slope montane forest as "a
complex, but fairly natural geographical and biological
unit."
Geographers and botanists view this Andean forest as
a disappearing repository of unparalleled species diversity
and local endemism (Gentry 1992:11).
That so many floral
and faunal species are endemic and restricted to the montane
forest underscores its distinctive character and begs the
question: what are the implications of this new awareness
for interpretations of prehistoric human utilization and
settlement?
The highest elevations of the forested eastern slopes
are often referred to as "cloud forests" because of the
dense mists which shroud them almost daily.
Moist air
moving east to west over the Amazon lowlands rises and cools
as it strikes the cordillera, dropping much of its moisture
on the eastern flanks,
rain forests.
thus watering one of earth's wettest
As previously noted, the eastern-slope
montane forest lies between 1,500 and 3,500 min altitude
(Young 1990, 1992b, Young and Leon 1993).
Within Peruvian
territory, the eastern montane forest stretches 1,500 km
north to south, and its width varies from 50 to 250 km
(Young 1992b).
Young estimates a typical gradient as 3,000
to 4,000 vertical meters within a distance of less than 100
21
km.
Only sixty to seventy percent remains of eastern-slope
montane forest cover as a result of deforestation (Young and
Leon 1993:239).
Peru's eastern montane forest can be divided into upper
(2,500 to 3,500 m) and lower (1,500 to 2,500 m) tiers (Young
1992b) .
Together they are bounded by subalpine grasslands
above 3,500 m, and premontane and subtropical forests below
1,500 m.
These subalpine grasslands constitute the southern
extension of Troll's paramo Andes that stretches down the
eastern face of the Central Andean massif.
Above this
paramo lie barren and rocky alpine life zones, while below
it, a belt of fragmented forest (often found between 3,400
and 3,700 m) provides the transition into continuous montane
forest.
The distinctions which researchers draw between the
montane and premontane forests, like other life zones, are
based on bioclimatic characteristics such as mean annual
biotemperature and precipitation and how these vary
seasonally (Holdridge 1967; Hartshorn 1983).
While details
of biogeographic taxonomies are of marginal importance to
this study, Holdridge life zones do often correspond to
environments preferred for certain economic activities.
For
example, the premontane forest between 1,500 and 500 m is
the preferred environment for coca cultivation (Plowman
1984).
The spatial distributions of eastern slope life zones
are complex and locally variable.
Many tributaries of the
22
Amazon draining the Peruvian Andes flow south to north (e.g.
the Urubamba, Apurimac, Huallaga and Marafion Rivers)
complicating what might otherwise be an orderly clinal west
to east transition from highlands to lowland tropical
forest.
Traveling down-slope one expects mean temperatures
to rise and the forest canopy to occur at progressively
greater heights.
However, dendritic river systems create
extensive rain shadows (e.g.
the upper Tarma and upper
Huallaga River valleys) and localized, open highland-like
environments.
Peculiar juxtapositions of montane forest and
dry forest {xerophytic) vegetation are not uncommon.
Any
study of eastern slope resource utilization or settlement
must take many environmental variables into account and
focus on local conditions particular to the area in
question.
Furthermore, present site environments in
proximity to the ecotones may have differed substantially
during periods of occupation because of ancient or modern
deforestation practices and/or paleoclimatic change.
For descriptive purposes, Young {1992b: Fig. 1) has
divided the eastern slope montane forest of Peru into six
"physiographic provinces or subregions" characterized by
differing geologies and life zone distributions.
From south
to north, these are: Madre de Dios, Urubamba, Tambo, upper
Pachitea, western Huallaga and Chachapoyas.
One problem
with adopting Young's subregions for heuristic purposes is
the tendency of montane forest settlements to straddle river
23
divides, thereby straddling subregion boundaries.
For this
thesis, broader distinctions drawn between southern, central
and northern subregions are sufficient.
These not only
correspond to subregions utilized by Leon et al.
(1992: Fig.
1} to analyze the eastern montane forest's floristic
composition, but they match the subregions originally
recognized in the first synthetic descriptions of montane
forest archaeology (Bonavia and Ravines 1967}.
The southern
subregion corresponds to the former departamentos (now
politically realigned in regiones} of Puno, Madre de Dios
and Cuzco; the central to Ayacucho, Junin, Pasco and
Huanuco; and the northern to San Martin and Amazonas.
To
avoid awkward terminology in the remainder of this thesis, I
propose to refer to the southern, central and northern
subdivisions of the eastern montane forest simply as the
southeastern, central and northeastern montane forests
respectively (Fig. 2}.
Along the length of the Central Andes, montane forest
timberline has responded to global climatic events at the
end of the Pleistocene epoch, and to local, periodic
phenomena such as droughts (Rodbell and Hansen n.d.}.
Within the forest,
the upper slopes experience periodic
landslides leaving scars that are rapidly colonized by
secondary vegetation, while down-cutting by turbulent waters
in the canyon bottoms frequently destroys the narrow river
terraces during rainy-season floods (especially between
24
October and May).
By the early Holocene, human agents had
begun to contribute to the ecological dynamics of the
eastern slopes.
Although topography, elevation, and especially frost
are commonly considered the principal environmental limiting
factors determining the elevation and configuration of
timberline, modern-day highland villagers living adjacent to
the eastern montane forest periodically burn back the forest
vegetation for a variety of reasons (Young 1990, 1993; Leo
1992).
In Young's opinion (personal communication), few if
any regions remain in which forest timberline can be
observed in equilibrium and unaffected by burning and other
human activities.
Vegetation at the forest's edge is burned
every few years in order to eliminate woody species and
improve pasturage for cattle.
He notes "that evidence of
burning, while ubiquitously present in the form of scorched
bases of tussock grasses and dead woody plants, is not
necessarily obvious to a short-term visitor to the area"
(1990:28).
Young also believes that the first indigenous
peoples to inhabit these grasslands burned them
intentionally simply because clearing the scrubby vegetation
aids foot transit so markedly.
According to ONERN (1976:128), life zones of the upper
montane rain forest are devoid of permanent human habitation
because of insurmountable topographic and climatic obstacles
to traditional economic activities.
However, reconnaissance
25
by explorers and archaeologists has revealed evidence that
prehistoric human activities dramatically altered the
montane forest landscape.
Abandoned agricultural terraces,
ruined settlements and paved roads within the upper forest
attest to the presence of substantial pre-Hispanic
populations that cleared large expanses of forest for
intensive agriculture.
In much of the montane forest,
native ecosystems have returned to some semblance of
equilibrium following Spanish conquest and abandonment by
most permanent forest dwellers.
From the early Spanish
occupation to the present day, however, highland villagers
of the eastern cordillera have maintained access to forest
resources (Brush 1977; Camino 1977; Gade 1972; Murra 1972;
Thompson 1980).
During the mid-twentieth century there has been a
renewed effort to colonize the eastern slopes by landless
Peruvians.
Meanwhile, pioneering agricultural and lumbering
activities moving up the western tributaries of the major
lowland rivers have been altering forest composition where
not eliminating the forest altogether.
Young observes that
the Tambo Province shows the most substantial impact from
highland settlement and use.
The Urubamba Province likewise
shows heavy human impact.
Interpretive Frameworks for Andean Migrations
This section will provide a brief description of the
popular frameworks that have shaped scholarly perception of
26
the montane forest.
The "jungle barrier" vision of the
montane forest that predominated until the mid-twentieth
century was outlined in the previous section.
Here I will
briefly address the theoretical frameworks that underlie
population movement and colonization hypotheses described
individually in Chapter 2.
These can be characterized as
"ruling frameworks," analogous to Rouse's {1986:3)
theories" or Snow's {1995:59)
"ruling
"controlling models," except
that they are fundamental sets of assumptions at higher
levels of abstraction from which ruling theories or
controlling models are derived.
The first framework is the
unitary origins paradigm from which the postulated AndeanAmazonian population movements sprang.
The second is the
popular model utilized to describe Andean cultural ecology
and political economy termed "verticality" or
"complementarity."
Unitary Origins and Population Movements
This dissertation focuses on the span of centuries
between 2000 B.C. and A.D. 700.
During this time, Central
Andean cultures experienced a sporadic series of
transformations leading to increasingly complex social
organization and the eventual emergence of politicallycentralized states.
Specific cultural transformations that
have captured the interest of scholars interested in the
origins of South American civilizations include the adoption
of agriculture, the appearance of sedentary villages, the
27
introduction of ceramic technology, the rise of ceremonial
centers and cities, and the emergence of socio-political
complexity and its associated art and ideologies.
Eastern Asia (e.g. Meggers et al. 1965), Mesoamerica
(Spinden 1917; Uhle 1922b; Rivet 1968 [1924]; Meggers and
Evans 1963), the Peruvian coast (Larco 1941; Lanning 1967),
the Central Andean highlands (Tello 1929, 1943, 1960) and
the Amazon Basin (Tello 1942; Lathrap 1970, 1977) have
alternately been touted as nuclear areas or "hearths" during
the decades prior to 1970 when evolutionary theories
invoking migration and diffusion were common.
More
recently, Lathrap (1977) has resurrected Spinden's (1917)
hypothesis to advocate a "unitary model" for agriculture's
invention and subsequent spread from Amazonia.
Agriculture
was invariably believed to be the foundation of complex
cultural development, and population movements, along with
diffusion, were considered principal modes of spatial
transmission for such innovations.
A basic but erroneous
assumption underlying most migration theories was the belief
that race,
language and culture are not independently
variable, but move across the landscape in fixed
association.
Hence, Uhle, Rivet, Meggers, Lathrap and
others have proposed series or "waves" of long-distance
population movement as pivotal events in archaeological
sequences, and especially in the "founding" of
civilizations.
28
Archaeological data gathered over recent years have led
to the realization that Central Andean civilization was not
a package of culture "traits" appearing and diffusing in
synchrony (Lumbreras 1981; Moseley 1983).
Evolutionary
milestones such as the adoption of agriculture, pottery and
sedentary lifeways often have independent histories, and
cultural development proceeded in piecemeal rather than
revolutionary fashion.
That the accretion of the elements
which together constitute "civilization" was a slow process
in the Andes is illustrated by the span of nearly three
thousand years separating the first construction of public
monuments and the emergence of the first Andean regional
states.
Nowadays, few scholars believe that a single hearth for
Andean civilization can be indicated.
For example, many of
the earliest pottery technologies found on the South
American continent apparently developed independently
(Raymond et al. 1994; Hoopes 1994).
The earliest Andean
socio-economic hierarchies probably arose at several
localities participating in long-distance interaction
sparked by the Chavin cult's spread from the north-central
highlands (Burger 1992, 1993).
The first expansionist
states arose independently on the coast (Fogel 1993) and in
the highlands (Isbell and Schreiber 1978).
Migration-origin
theories have thus been discredited first on empirical
grounds, and thereafter for paradigmatic reasons.
Their
29
legacy however, has been the identification of suggestive
material patterning that still warrants archaeological
analysis and interpretation.
Andean Political Economy, Verticality
and Colonization
Since the mid-1960s, Murra's (1964, 1967, 1975)
ethnohistorically-derived model of highland Andean political
economy has dominated ethnographic and archaeological
studies of Central Andean interaction and exchange.
Murra
observes that Andean societies developed singular techniques
of procuring products from a variety of distinct,
vertically-stacked ecozones in the highlands through direct
colonization, rather than relying upon exchange systems that
potentially compromised a community's economic selfsufficiency.
The Chupachu and Lupaca evidence for
colonization extracted from sixteenth century census
documents provides Murra with two examples of Andean "ethnic
groups" with differing degrees of socio-political complexity
sharing this same economic strategy.
The former occupied
the high elevations above the upper Huallaga River valley,
while the latter inhabited the shores of Lake Titicaca on
the southern Peruvian altiplano.
Both groups reportedly
maintained distant colonies in warm lowland ecological
zones.
Squarely within the substantivist tradition of economic
theory, Murra's model of "vertical control" or "zonal
30
complementarity" emphasizes non-market systems based upon
traditional modes of reciprocal exchange at the community or
"folk" level, and redistribution at complex politicaladministrative or "state" levels.
More specifically, Murra
hypothesizes that the desire of highland polities for access
to maize, coca and other goods available only in the
lowlands spurred the "conquest and colonization" of
appropriate ecological zones (Murra 1975:59).
To Murra, the
imperative of establishing vertical control through
colonization was a "pan-Andean" characteristic (Murra 1967),
and that the well-known Inca practice of relocating
populations as mitimae colonies represented "a late and
altered manifestation of an ancient Andean pattern"
(1975:60).
Maintenence of community self-sufficiency
through intra-ethnic exchange is considered a uniquely
Andean alternative to the interregional, inter-community
trade economies that characterize other prehistoric world
areas like Mesoamerica and the Near East (Morris 1978:318;
Murra 1981:52; Van Buren 1996:341).
The degree to which
occupational specialization and mercantile activities took
place in the Central Andes (Rostworowski 1977) remains
unresolved, and has received relatively little attention
from archaeologists.
Murra's conclusions that verticality was an ancient
pattern common to Andean societies at all levels of sociopolitical organization from Colombia to Argentina has had
31
wide-reaching implications for studies of Andean cultural
development and exchange.
Archaeologists have applied
Murra's model to interpretations of prehistoric Andean
cultural evolution (e.g. Moseley 1983, 1992), especially in
the south-Central Andes (e.g. Mujica 1985; Rice et al. 1989;
Kolata 1983, 1991; Aldenderfer and Stanish 1993; Goldstein
1993).
Moseley (1992:43) views verticality as an adaptive
strategy supplying highlanders throughout the Central Andes
with needed "marine salt, seaweed, fish,
fruit, beans,
maize, coca and cotton" from the lowlands.
There is a
consensus (e.g. Moseley 1983, 1992; Burger 1985a; Stanish
1992) that access to produce from lower elevations was and
still is a biological imperative for Andean populations on
the puna.
Moseley (1992:46) hypothesizes that verticality
strategies emerged during Preceramic times at the household
and community level as a consequence of highland population
growth coupled with an expanding diversity of domesticated
lowland crops.
Moseley theorizes that the domestication of the llama
and alpaca as cargo beasts, and the application of
irrigation technology within the intermontane basins
signaled the appearance of "true verticality adaptations" by
the beginning of the Initial Period (Ibid. :100, 142).
However, Miller and Burger (1995) present evidence for the
relatively late utilization of the llama for cargo
transport, during the Early Horizon in most areas.
In
32
Moseley's view (1983:194-196, 1992:46), the highlanders'
continual drive to directly control lowland agricultural
lands generated "a marked downward thrust" manifest in
repeated prehistoric colonization and periodic military
incursions into the lowlands.
Thus, highland occupations
frequently extended down to 1000 m on the Andean slopes
(1992:45).
He observes (Ibid.)
that the verticality
strategy is "least pronounced" in the north, and more
intense and elaborate in the south-Central Andes.
Whereas Moseley views verticality/complementarity as an
adaptive strategy, Stanish (1992) regards it as a politicaleconomic strategy.
In a recent study of south-Central
Andean cultural development, Stanish (Ibid.:SO) argues that,
"the most successful criteria for addressing the prehistory
of the region are political and economic ones as understood
within the general framework of zonal complementarity."
Stanish is among many scholars (see Masuda et al. 1985;
Lynch 1981) that have implicitly or explicitly advocated
widening the concept of complementarity beyond "directcontrol" resource procurement through colonization, to
include "indirect control" through trade between autonomous
polities.
In effect, all investigations of Central Andean
exchange have become macro-scale analyses of highland
political-economic evolution, emphasizing core-periphery
relationships.
Archaeologists have already registered complaints that
33
the concepts of verticality/complementarity had been
stretched to include virtually all Andean interactions (e.g.
J. Topic and T. Topic 1985:56; Patterson 1988:217).
Burger
(1985a:275-277) has contested Moseley's position and
indicated the need for an "interregional model" for
exchange.
The Topics (J. Topic and T. Topic 1985:58)
rejected the verticality model as a conceptual aid to
interpret long-distance interaction between coastal and
highland societies in their north-Central Andean study area.
They observe that implementation of the verticality model
obscures "the range of goods and services considered
necessary for self-sufficiency, and the means used to obtain
them ... at different times and places within the Andean
area."
Stanish acknowledges points of controversy
surrounding the complementarity model (1992:6), but
maintains that gathering "virtually all" Andean exchange
under the umbrella of complementarity provides an internally
consistent framework for explaining Andean cultural
evolution, at least in the south-Central Andes.
In the most recent theoretical examination of
verticality, Van Buren (1996) questions some of the most
basic assumptions underlying arguments by Murra, Moseley and
many other Andeanists.
Van Buren examines the south-Central
Andean case of the Lupaca kingdom showcased by Murra as an
example of verticality colonization, and observes that
intensive archaeological investigations on the western
34
slopes have failed to produce solid evidence for a preHispanic colony-type occupation.
Drawing on post-Conquest
historical documents, she is able to show that the primary
economic activities at the Colonial Period (and possibly
Inca) colony of Terata Alta most clearly served the needs of
a small group of highland political elite rather than the
greater community.
Thus Van Buren demonstrates that
verticality colonization cannot be interpreted as the
archetypal Andean ecological adaptation a priori.
Further,
she has sown seeds of doubt that verticality colonization
even existed as a pre-Hispanic institution, political or
otherwise.
Archaeologists examining interaction in the northCentral Andean highlands have made limited use of
verticality and complementarity conceptual frameworks,
invoking these only for local-level exchange (e.g. Shady and
Rosas 1979; J. Topic and T. Topic 1985).
Network analysis
(J. Topic and T. Topic 1983), "peer polity interaction"
(Burger 1984a, 1993), and envoy or "expedition" exchange
(Lumbreras 1993:364) are among alternative methods and
models yet to be adequately explored for interpretation of
regional and interregional levels of interaction.
The Problem of Tropical Montane Forest
Prehistory: Migration or Interaction?
The research problem confronted in this thesis entails
determining whether archaeological material patterning in
35
the eastern montane forest constitutes evidence of
prehistoric migrations, or indicates local cultural
development and interaction.
According to Rouse, who has
detailed the most extensive methodology for "inferring"
migrations from archaeological data (Rouse 1958, 1986},
these are the two basic alternatives which must be weighed
one against the other if a prehistoric migration involving
population replacement is suspected.
The challenge is to
distinguish the movement of culture from the movement of
people.
Both kinds of movement were constant features of
human prehistory world-wide,
yet only the movement of
culture through "diffusion" and "trade" has generated a
large body of theoretical literature promoting diverse
frameworks to guide interpretation (Adams et al. 1978;
Schortman and Urban 1987}.
Adams et al.
(1978} have traced the intellectual
history of migration explanations in anthropology, observing
that "migration theory" is neither a paradigm, nor a formal
interpretive framework.
As they wrote,
"migrationism" was
said to be in full retreat as a viable explanation for
change in the sub-fields of Archaeology and Physical
Anthropology, while less so in Linguistic Anthropology.
During the mid-twentieth century, many migration theories
have failed to withstand critiques backed by enlarged data
bases and improved analytical techniques.
Other migration
theories have been "obviated" by American Archaeology's
36
shifts toward nee-evolutionist and nee-Marxist agendas
intent on toppling culture-historic interpretations deemed
unscientific (Ibid.:1978:504; Trigger 1989:421).
al.
Adams et
(1978:487) observe that most migration theories had
never been much more than "ad hoc and somewhat mechanical
explanations" for prehistoric change, especially for
developmental discontinuities and anomalous trait and site
distributions.
On the other hand, they argue that the
theoretical bias against "migrationism" is deleterious to
the discipline.
Two compilations of culture-historical studies bearing
directly on the problem of discriminating prehistoric
migration and diffusion were presented in the 1950s by
Thompson (ed. 1958) and Lathrap (ed. 1956).
Lathrap and his
colleagues coined the term "site-unit intrusion"
(usually
evidence of migration) to contrast with "trait-unit
intrusion"
(usually evidence of diffusion) .
However, Rouse
(1958 [in Thompson ed.]) presented the first explicit set of
guidelines for "inferring" prehistoric migrations.
The
fifth and final of his five analytical steps contended that
"it is incumbent upon the person who wishes to demonstrate
migration to consider and eliminate the possibility that
some other hypothesis may better fit the facts at his
disposal"
(Rouse 1958:66).
This statement called for
rigorous scrutiny of the evidence and clearly placed the
burden of proof upon advocates of migration explanations.
37
Almost 30 years later Rouse himself assumed "the
burden" with the publication of Migrations in Prehistory, an
explicit and detailed methodology for inferring prehistoric
migrations with four case studies (Rouse 1986).
In addition
to reiterating and developing points argued in 1958, Rouse
emphasizes differentiation of kinds of migrations, kinds of
cultural development and kinds of interaction.
He maintains
that much of the disillusionment with migration explanations
has stemmed from the failure to discriminate between kinds
of migrations, and failure to systematically test hypotheses
(1986:17-18).
Thus Rouse delineates specific systematic
techniques for controlling chronological and spatial aspects
of the problem in order to establish temporal priority at
hypothesized sources, and to identify stages and routes of
movement.
Also, issues of mobility including modes of
transportation, environmental transitions and potential
barriers must be considered.
He maintains that evidence
from historical linguistics and physical anthropology that
archaeologists typically muster to corroborate their
hypotheses must be utilized only as independent checks.
Rouse's work sparked a flurry of migration literature
authored by archaeologists representing American processual
and European historical research traditions (as these are
defined by Binford and Sabloff 1982).
Anthony's (1990)
processual approach borrowed concepts from contemporary
disciplines of Geography and Demography to demonstrate how a
38
nomothetic, behavioral approach supports a postulated
westward migration of Yamna Kergan pastoralists from the
Eurasian steppes during the fourth millennium B.C.
European
archaeologists Otte and Keeley (1990) simultaneously called
for the revival of migration and diffusion concepts to
interpret upper Paleolithic Period culture change in the Old
World, arguing that both are valid, historically documented
processes.
Ironically, Anthony's study was denounced as
ahistorical by European scholars (Chapman and Dolukhanov
1992), while Otte and Keeley's position was criticized by
processualists (Clark and Lindly 1992) who categorically
reject analogies between historical processes and
Paleolithic site formation processes.
Meanwhile, Marxist
critics (e.g. Patterson 1991; Sued-Badillo 1992) have
condemned Rouse's methodology, at least as it has been
employed in the Caribbean, as the reductionist
trivialization of local historical processes typical of a
"hegemonic framework"
(Patterson 1991:4).
Clearly,
"migrationism" has not fit comfortably within any single
contemporary archaeological paradigm.
Rouse (1986:158) observes that the relative success
ascribed to his four case-studies was owed in part to their
prehistoric recency and the feasibility of tracing material
patterning back through time and space using the directhistorical approach.
Thus, Snow (1995) employs Rouse's
39
formula to reassert previously discredited theories of
Iroquoian migration.
He denounces the anti-migration bias
of the "controlling in situ model," and contends that major
anomalies in reconstructions of prehistoric Iroquoian
language, kinship, ceramic manufacture and site distribution
support a migration explanation.
The oft-cited Four
Corners-Rio Grande pueblo migration is currently being reexamined utilizing insights gained from the debates prompted
by Rouse's work (Cameron 1995).
Those concepts deemed
useful for this particular study will be described in the
following section.
Concepts and Strategies for Evaluating
Migration Hypotheses
Selected methodological points and procedures developed
by Rouse and others provide a theoretical basis for
evaluating postulated montane forest population movements
and colonization.
Unfortunately, partisan scholarship has
produced parallel sets of operative concepts and
definitions.
Consequently, the choice of concepts and terms
has tended to reflect each scholar's inclination toward
either culture-history or processual paradigms.
For
example, Snow (1995) relies on Rouse's methodology, while
Cameron (1995) chooses to modify Anthony's criteria.
The
following theoretical discussion will rely principally on
Rouse's original formulations, but will also draw from the
most useful of Anthony's and Cameron's additions and
40
modifications.
Operational definitions of basic concepts such as
"population movement" and "colonization" are essential
because, from an archaeological perspective, they entail
investigation of distinct sets of material correlates.
Rouse's term "population movement" is analogous to Anthony's
(1990:902)
"long-distance migrations" that
characteristically cross ecological or cultural boundaries,
and Cameron's (1995:113)
"long-range, short-term movements
of large segments of a population."
Rouse (1986) observes
that such population movements may transpire as the
"peopling" of unoccupied areas, or as the replacement of
extant populations by rapid processes of expulsion or
absorption.
Population movements resulting in replacement
must be distinguished from imperial conquests with agendas
of territorial expansion.
Archaeologists ostensibly have the methodological tools
to detect short-distance and long-distance migrations at the
levels both of communities and societies (Adams et al.
1978:489).
Rouse and proponents of Andean-Amazonian
population movements are clearly concerned with the "society
level."
Anthony (Ibid.)
remarks that "long-distance
migration should result in changes that would have distinct
effects on the archaeological record," although Adams et al.
point out that these population movements are the least
common "in the course of recorded history."
Well-known
41
archaeological examples of postulated society-level
migrations resulting in replacement include the "wave-ofadvance" spread of neolithic farmers across Europe (Ammerman
and Cavalli-Sforza 1973) and the westward Indo-European
migration of horse-mounted societies from eastern Europe
(Anthony 1986, Renfrew 1987).
Rouse notes that population movements should not be
confused with "immigration" which is localized migration of
individuals or social groups that results in assimilation
rather than population replacement.
Snow (1995:74-75)
postulates a series of Iroquois community-level migrations
that ultimately amount to a society-level population
movement, while Cordell (1984, 1995) postulates series of
family and community moves that likewise constitute a
society-level population movement.
Neither of these cases
provide good analogies for this particular investigation as
they both might be considered intra-regional from an
ecological point of view.
The Four Corners-Rio Grande
migration as described by Cameron and Cordell appears to be
a case of massive immigration and assimilation within a
broad pre-existing interaction sphere.
Because the term "colonization" is utilized in Andean
studies to refer to at least three separate processes, a
working definition for this study must be carefully
specified.
Andean colonies in Murra's verticality model
appear on the landscape as "archipelagos" or enclaves
42
implanted far from the colonizing community's homeland.
The
colony is an intrusive settlement, perhaps accompanied by
similar colonies from other homelands, surrounded by an
indigenous population at its new locality.
Confusion stems
from additional academic and colloquial usages of
"colonization" to refer to either: 1) the peopling of an
unoccupied frontier, or 2) the frontier installation of
administrative armature by a conquest state in order to
exploit local natural and human resources.
In this study,
the terms "colony" and "colonization" refer strictly to
Murra's archipelago variety in which colonists expend their
own labor to exploit local resources.
The other two so-
called colonizing processes are referred to in this study
simply as "peopling" and "conquest."
Comparison of material patterning from suspected
homelands, along routes of movement and at final
destinations,
is a procedure fundamental to the evaluation
of population movement and colonization hypotheses.
Some
general guidelines serve equally for the archaeological
identification of both processes.
For example,
there is a
consensus that a culture-historical discontinuity or anomaly
in both time and space is common to all hypothetical
migrations.
To evaluate similarities across temporal and
spatial dimensions, Rouse (1986:160) and Stanish (1992)
emphasize the importance of tracing contexts rather than
isolated attributes or "traits."
They agree that sites and
43
site "structures" with behavioral correlates must be the
primary units of analysis.
Such comparative analyses of
site structures focus on activities, activity areas and
specific behaviors associated with artifacts.
Despite general resemblances in overall strategies,
specific approaches for recognizing population movements and
colonization must differ in details.
Basically, this is
because of the differing intentions of migrants in each
process.
The protagonists in most hypothesized prehistoric
population movements are societies with organizational
and/or technological advantages that encourage encroachment
and predation (cf. Sahlins 1961; Davis 1973; Rouse 1986;
Snow 1995).
Archaeologists working within cultural-
ecological frameworks often contend that migrants are
"pushed" out of their core areas by population pressure.
Cameron notes that southwestern archaeologists are
increasingly considering "pull" factors such as economic and
ideological incentives.
One important variable that Cameron formally introduces
into the archaeological migration literature is the "unit of
migration."
She (1995:112-113) argues that, because
decisions to migrate were probably made by kin groups or
social units larger than individuals or nuclear families,
Anthony's application of modern demographic concepts may be
inappropriate.
In the pre-Conquest New World, extended kin
groups served as the basic organizational units for most
44
social, economic, political, and religious institutions.
The implications of such large units of migration are that:
1) population movements were relatively rare occurrences; 2)
only under certain conditions of stress ("push") and social
disintegration did individuals and nuclear families migrate
and 3) under benign or favorable ("pull") conditions, large
units of migration may be expected to have left evidence
more closely resembling a site-unit intrusion.
As Cameron
suggests, it behooves the analyst to investigate specific
social, political and economic circumstances promoting or
preventing migration.
Most postulated population movements cross relatively
homogeneous landscapes with few natural barriers (Adams et
al. 1978:501).
Rouse argues that archaeological complexes
cannot be simplistically followed across the landscape
because migrants carry only a subset which alters
concomitantly with migrants'
comparing site assemblages,
"adaptive radiation."
In
the analyst examines internal
contextual relationships by noting the co-occurrence of
artifacts and features.
Most important is identification of
the behavioral correlates of sites, complexes and
assemblages relating to environmental habits, preferred site
locations, subsistence strategies, settlement layouts,
household organization and food selection (Rouse 1986:166174) .
For Rouse, the selection of appropriate units for
45
analysis and classification is crucial to the success or
failure of correct inference.
Combining Willey's (1945)
concepts of archaeological "horizons" and "traditions,"
Rouse and Cruxent (1963:23) developed the concept of the
"series" to describe development of a style or complex
across space as well as through time.
The series concept
has clear advantages for inferring population movements as a
means of diachronically interpreting spatial dimensions of
stylistic developments.
Series and subseries labels carry
the suffixes "-oid" and "-an", and Rouse (1987:14) likens
these to language families and subfamilies.
Because series
and subseries under conditions of migration represent
population segments carrying only a portion of their
culture, he (1987:10) compares the establishment of a new
line of development to the "founder's principle" concept of
population biology.
Rouse contends that archaeological
classifications should be modeled after the linguistic
classification schemata which have most successfully
revealed historical relationships.
Unlike migrating populations, Andean verticality
colonies are deliberately implanted into ecological zones
differing from the homeland.
In theory, such colonies
should be clearly distinguishable as site-unit intrusions.
Morris (1978) was among the first scholars to consider the
kinds of archaeological patterning expected under conditions
of verticality colonization.
He (1978:318) theorizes that,
46
"if [Murra's] model is correct, we should have a situation
whereby interregional exchange of subsistence goods is
handled by intra-community mechanisms."
Morris suggests
that such intra-community exchange should produce tightly
circumscribed spatial distributions of pottery style
attributes.
Pottery styles in colonized regions should be
interdigitated with others like "the islands of an
archipelago"
(Ibid.).
In theory, each style represents one
node within a single intra-community exchange system
spanning multiple ecological zones.
Despite the popularity of Murra's model for theorizing
and interpreting archaeological data, little field research
has been directed specifically towards the identification of
Andean verticality colonies until recently.
Archaeological
work on the western slopes of the south-Central Andes by
Aldenderfer and Stanish (1993; Stanish 1992) has attempted
to identify such colonies through synchronic evaluation of
similarity between contexts found in differing ecological
zones.
Domestic contexts, and especially households, they
argue, are most appropriate units of analysis for
identifying the special intra-"ethnic" relationship between
colony and homeland.
Kitchen middens, non-elite residences
and storage structures are among domestic contexts said to
directly reflect a resident population's ethnic
affiliations.
Non-domestic contexts such as tombs, elite
residences and corporate architecture are considered more
47
likely to contain exotic elements reflecting external
relations such as exchange and alliance systems (Stanish
1992:9).
Stanish lays out a specific set of archaeological
correlates for identifying Andean colonies.
First, a colony
may show non-local artifacts in both domestic and nondomestic contexts.
A lack of any exotics, however, might
suggest no complementary relationship.
The domestic
component of a colony should be identical to that in the
homeland, showing only "ecological impact."
Actually the
theoretical literature includes little discussion of the
potential effects of a colony's new environmental and
cultural surroundings on households and domestic economies.
According to Murra and Stanish, colonies are mono- or multiethnic enclaves that may be distributed among other mono- or
multi-ethnic colonies in a particular valley.
Thus,
verticality colonization introduces heterogeneity at the
level of sites, localities and regions (Ibid.:44-45; Morris
1978:318).
Some of Stanish's criteria consider colonization in
differing regional socio-political contexts.
He notes that
the degree to which social organization within colonies
mirrors that of the homeland remains unknown.
However, he
hypothesizes that colonies established by centralized Andean
states would house low-status individuals in a servile
capacity.
According to Bonavia (1978:400-401), colonies
48
established under the auspices of Inca or Huari imperial
interests should show little or no evidence for occupational
differentiation, social structure or political structure.
"Absolute" intra-site functional homogeneity, he argues,
characterizes settlements that he observed in the central
and northeastern montane forests.
Bonavia theorizes that
these colonies were occupied by communities of full-time
agriculturalists fully dependent upon their parent highland
polities (Ibid.).
Utilizing Stanish's and Bonavia's assumptions, we can
hypothesize that montane forest colonies established under
conditions of imperial hegemony would lack architectural
features reflecting social heterogeneity and political
autonomy.
We might even expect a dearth of architectural
evidence for the integration of family and other social
units, and for socio-political hierarchies.
Potential
analogies might include the Inca settlements Callachaca A
(Niles 1987: Fig. 2.1, 2.3) and Raqay-Raqayniyoq (Ibid.:
Fig. 2.8) that apparently supplied housing for small family
units (perhaps rendering their share of m'ita labor tribute
to the Inca state).
The lack of common courtyards, or even
facing doorways, suggests that intra-community social
interaction that potentially interfered with work or
fostered organized resistance was discouraged.
The Inca cases just described provide exceptionally
clear examples of hypothetical colonies where colonists
49
divested of their ethnic identities (at least temporarily)
provided labor tribute to the Inca state.
As previously
mentioned, Murra views these as late, altered variants of
Andean verticality colonies.
Ostensibly they conform to the
verticality model because non-local populations were
installed for production directly benefitting the homeland
(in this case, Cuzco).
Implicit in Murra's model is the
assumption that, regardless of the colonizing polity's
degree of socio-political integration, the implanted
population is directly engaged in production rather than in
the administration of local populations.
Evidence that
local rather than colonizing groups were the producers
suggests that analytical frameworks of conquest and imperial
administration are more appropriate than verticality
colonization models.
Stanish hypothesizes that colonies established by
Andean polities at lower levels of social integration would
house groups of status equal to their homeland counterparts
(Ibid.).
Stanish interprets the post-Tiwanaku highland
colonization of the western-slope Otora Valley as multiethnic in composition.
According to his reconstruction,
initial colonization was followed by the emergence of a
local elite and subsequent independent cultural development.
In this case, colonization has, in effect, become a
population movement.
Stanish's data has potential for
revealing much about general processes of migration and
50
consequential environmental impacts and social changes, and
their archaeological interpretation.
In cases of both hypothetical population movements and
colonization, it is usually the degree of perceived
similarity between assemblages at sources and destinations
that supports or contradicts the migration argument.
Of
course the lack of any similarities render questions of
migration moot.
The evaluation of postulated population
movements requires the diachronic analysis of change, while
the evaluation of postulated colonization involves detailed
synchronic comparisons.
Because interpretations of
population movements must take temporal as well as spatial
variables into account,
they ultimately require a greater
number of theoretical assumptions.
Hence, population
movements, as they are presently understood by
archaeologists, should be far more difficult to demonstrate
than Andean colonization.
Perusing the theoretical literature devoted to
identification of prehistoric migration, readers are led to
believe that, provided adequate knowledge of spatial and
temporal variables, as well as a variety of behavioral
contexts, archaeological evidence of population movements
and colonization should be easy to interpret.
Anthony
(1990:897) insinuates that his methodology, unlike Rouse's,
yields "conclusive" results.
However, it would seem that
only certain kinds of migrations can be inferred with any
51
confidence.
Although Stanish has published only selected
sets of data, it would appear that his implementation of
context-specific strategies for comparative analysis has
succeeded in identifying outlying western-slope settlements
composed of highland social groups.
His methodology should
also be useful in identifying the early stages of population
movements, especially given their purported propensity for
bypassing nearby territories in favor of more distant areas,
or "leap-frogging"
(Anthony 1990:902-903).
Contexts for Inquiry: Manachaqui Cave and
the Greater Eastern Montane Forest
In this thesis I attempt to evaluate the evidence for
population movements and colonization from increasingly
broad archaeological contexts beginning with Manachaqui Cave
and extending to the greater eastern montane forest.
Different levels of inquiry require different strategies
appropriate to the kinds of data available.
Furthermore,
the utility of certain kinds of archaeological data may be
regarded as questionable considering the emphasis on
contextual approaches advocated in the literature.
Immediately suspect is my primary source of data, Manachaqui
Cave.
Manachaqui Cave is, of course, a rockshelter and not a
homeland or hypothetical "colony."
However,
the rockshelter
lies within Chachapoyas, an archaeological culture area
known from documentary and archaeological evidence to have
52
occupied montane forest environments.
Manachaqui Cave is
also strategically placed between two postulated migratory
"highways," the Marafion and Huallaga River valleys (see
Chapter 2); and beside an east-west pre-Hispanic road
connecting highland and montane forest destinations (see
Chapter 3).
The archaeological remains at Manachaqui
include deposits from time periods corresponding to the
population movements and colonization postulated by
archaeologists investigating montane forest prehistory.
Regardless of the aforementioned features favoring the
utility of Manachaqui Cave, issues surrounding the site's
functions and assemblage comparability must be addressed.
Stanish (1992) dev0tes extensive discussion to the potential
for misinterpretation inherent in "artifact-based
approaches" where contexts are not controlled.
Archaeological assemblages recovered in the highland,
montane forest and lowland regions surrounding Manachaqui
Cave represent a variety of contexts, usually cemeteries,
caves and rockshelters, early ceremonial structures and late
prehistoric administrative complexes.
Unfortunately,
relationships between site function and assemblage
composition are seldom addressed in useful detail.
Nonetheless, potential sampling biases should be recognized
to the degree possible and taken into account.
In the Andes and elsewhere, caves and rockshelters are
settings for limited sets of human activities.
As Straus
53
(1990:279) points out,
"the natural features in caves and
rockshelters structure activities that were conducted at
these types of places."
Manachaqui Cave is a small
an~
rockshelter that became smaller through time as natural
cultural deposits accumulated under foot.
The range of
activities suited to such a space was always limited.
From a broader perspective, Manachaqui Cave represents
one "place" in a larger system of places, each corresponding
to particular sets of activities or behaviors (Ibid.;
Binford 1982).
Likely functions suggested by Manachaqui
Cave's small size, and by analogies found in the Andean
ethnographic, ethnohistoric and archaeological literature,
include: 1) a permanent habitation for a small family,
2) a
temporary hunting camp, 3) a herders' camp, 4) a huaca or
sacred site, 5) a wayside station, and combinations of any
of these.
A "domestic" assemblage representing the entire suite
of household activities in a permanent cave habitation might
be the most desirable of data sets for interregional
comparisons.
"base camps"
However, only hypothesized Preceramic Period
(e.g. Rick 1988) and one Initial Period herding
camp {Lavallee et al. 1985) have been interpreted as
permanent or semi-permanent habitations.
Unfortunately,
ceramic components from hunting and herding camps have been
reported only superficially, but these assemblages might be
characterized as specialized subsets of a larger domestic
54
repertoire.
As repositories of offerings valued as exotic,
sacred cave sites would presumably present the greatest
danger for misleading interpretations.
If Manachaqui Cave frequently served as a wayside
station as its roadside location suggests, then an important
question is whether it served migrating populations or
groups engaged in interregional exchange.
In either case, a
wayside station would contain assemblages of a domestic, or
"secular" utilitarian nature, although concerns with
transportation and portability should impose constraints on
the structure of pottery and other artifact assemblages.
DeBoer {n.d.) reports that vessels utilized by modern
Central Ucayali Shipibo travelers are simply smaller
versions of the large vessels utilized within domiciles.
Under conditions of population movement and/or colonization,
there is no reason why archaeological assemblages from a
wayside station should not reflect the stylistic norms of
the migrant or colonizing populations.
At Manachaqui Cave, a prehistoric population intrusion
of any kind should leave a disconformity in the cultural
sequence representing the sudden and complete imposition of
an intrusive assemblage.
Under conditions of population
movement and replacement, Manachaqui Cave would be
positioned to service the migratory counterstream {Anthony
1990), and the post-migration cultural sequence should
reflect only the development of the new population.
Ceramic
55
assemblages, while they present some interpretive problems
(cf. Adams 1979), do yield useful information for evaluating
migration hypotheses.
Snow {1995) points to anomalous
shifts in ceramic technology in support of a migration
explanation, while Raymond et al.
{1994) and Hoopes {1994)
argue that salient differences in earliest New World pottery
assemblages discredit migration theories.
might
d~scern
Simultaneously we
evidence for shifts in diet and/or subsistence
technologies.
Under conditions of colonization as Murra describes it,
we might find a temporary co-occurrence of local and
intrusive artifacts and artifact styles within single phase
assemblages.
Likewise, evidence for subsistence could
appear as the co-occurrence of local and intrusive sets of
food remains in single faunal and ethnobotanical phase
assemblages.
The interpretive challenge would be to
distinguish such co-occurrences from assemblages of food
remains left by single populations with generalized
subsistence patterns.
In order to approach such
interpretive problems within a single archaeological site,
high levels of stratigraphic and temporal resolution are
crucial.
The interpretation of co-occupation from food
remains would rely on the identification of a temporary
shift from a relatively focused subsistence pattern to an
ostensibly generalized pattern and back again.
Considering
the harsh realities of site formation and assemblage
56
preservation in most sites, these interpretive avenues
present considerable difficulties.
In broader montane forest valley or multi-valley
contexts, predictive statements by Morris, Stanish and
Raymond concerning style distributions and settlement
patterns are potentially useful for detecting prehistoric
population intrusions.
The reader will recall from previous
discussion of criteria that verticality colonies in regional
contexts should be identifiable by circumscribed,
archipelago-like spatial distributions of distinctive
domestic assemblages, artifact styles and style attributes
(Morris 1978; Stanish 1992:43).
Conversely, regional
homogeneity should indicate local autonomy, and "a discrete
'culture area' definable by a relatively consistent set of
archaeological indicators" (Stanish 1992:45).
Substantial
stylistic diversity exhibited within site ceramic
assemblages (i.e. abundant "foreign" pottery) might
constitute additional evidence for local autonomy and
interregional exchange rather than verticality colonization.
Influenced by Brush's (1974:280) use of Murra's (1972)
framework, Raymond (1988:290) stresses the importance of
local geography in determining other characteristics of
montane forest colonization: "the permanency and size of the
settlements in the montana are a function of the distance,
relative economic importance of the resources, and the
extent to which it is necessary to protect the lands from
57
neighboring ethnic groups - of both highland and lowland
origin."
He offers the following additional predictions
based primarily on the distance separating highland and
"montaiia"
(montane or premontane forest) resource zones:
If the resource zones are near one another, i.e., within one
or two day's walk, then all the resources may be exploited
from a single highland village, and settlements in the
montana may be either non-existent or small and scattered.
Resources which are widely spaced and require several days'
walk lead to a pattern of settlement which Murra (1972) has
called the Archipelago type. The principal villages are in
the highlands but satellite communities are established in
the montafia ... If the resource zones are large and continuous
but strung out over a long valley system, then an extended,
multi-ethnic settlement pattern may develop with products
moved via a trade and market system ... Under such
circumstances permanent settlements of highlanders would be
established in the montana (Ibid. :290-291).
Raymond's formula is based upon Murra's ethnohistoric
descriptions of montane forest colonization by the Chupachu
of Huanuco, and the Lupaca of Puno.
Murra (1975:70-71)
reports that Chupachu colonies consisted of three or four
homes and were located within four day's walk from the
highland nucleus.
The highland Lupaca "kingdom" reportedly
maintained colonies with over a hundred homes, ten or more
days' walk distant (Ibid.:79-80).
Raymond's third
prediction for long, ecologically diverse valleys is based
upon Brush's and Gade's
(1975) interpretations of land-use
in the Vilcanota (or Urubamba) Valley of eastern Cuzco
Department.
Such geographical and ecological formula, backed by
ethnohistorical evidence serve Raymond to interpret the
archaeology of the montane forest.
Mayer (1985:66) observes
that settlement locations are also partly determined by the
58
need to be close to the production zone requiring "the most
intense care."
In addition to interpretive utility,
Raymond's predictive model provides a baseline from which to
measure deviation from expected patterning both in the
northern subregion, and in other eastern montane forest
subregions.
The recognition of such deviations should
prompt a search for historical and other non-ecological
variables.
The Alternative Hypothesis: Local
Development and Interaction
In the remainder of this thesis, I hope to demonstrate
that the archaeological data from both Manachaqui Cave and
the Central Andean montane forests support interpretations
of autochthonous cultural development by populations
habitually engaged in long-distance exchange between
neighboring Andean and Amazonian regions.
In order to
proceed with this study, some factors expected to complicate
evaluation of the hypothesized migrations should be
considered.
The following discussion is most pertinent to
the postulated Andean-Amazonian population movements, but
implications for the evaluation of colonization hypotheses
are likewise evident.
Clearly, population movements with the highest
probabilities of archaeological detection are those which
"people" sparsely inhabited or uninhabited regions like
Polynesia and east-Arctic North America (Rouse 1986).
As
59
bounded gene pools with tightly delimited spheres of
interaction, island settings seem ideal for tracing
population movements, especially where migrants carry a
distinctive and complex technology as Rouse's Caribbean
migrants did.
However, changes in artifact attributes and
behavioral contexts that occur during slowly unfolding
population movements (e.g. multi-directional population
expansions,
"waves-of-advance" or prolonged immigration)
into varied social and physical environments may leave
material patterning sufficiently complex to confound
archaeological identification of the process.
Of course the
inability to document migration archaeologically does not
mean that migration and population replacement have not
occurred.
Linguistic and physical anthropological evidence
might still provide evidence for population replacement.
Physical anthropological evidence for population replacement
is frequently unavailable.
Portentous for this study, is the fact that even the
Caribbean Island evidence that Rouse showcases may leave
ample room for contention (e.g. Drewett 1993:456).
One even
begins to suspect that prehistoric society-level population
intrusions into occupied social and physical environments,
at least as archaeologists have envisioned them, are ideal
theoretical constructs with few counterparts in
archaeological reality.
Unlike the postulated migrations to
be evaluated in this thesis, Rouse's four "successful" cases
60
had historical components and were therefore amenable to
application of the direct-historic approach.
Rouse warns
that interpretation becomes increasingly tenuous moving
backward through time, and Snow (1995:60} flatly states that
"attempts to correlate archaeological and linguistic lines
of evidence for periods prior to recent millennia are
specious."
It is not surprising that Anthony's and Otte and
Keeley's postulated Copper Age and Paleolithic Period
migrations received harsh criticisms.
The postulated Andean-Amazonian population movements
purportedly surmounted an imposing natural barrier.
sheer magnitude,
For
the environmental contrast between the
Andean and Amazonian regions has few if any peers.
Perhaps
unrivaled is the quantity and compaction of ecological
regimes entirely replaced across short horizontal distances.
Investigations of cultural geography by Steward (1948;
Steward and Faron 1959} and other anthropologists assure us
that the Central Andean-Amazonian environmental transition
has coincided, and continues to coincide, with a major
cultural boundary of long standing.
The degree to which
Central Andean and Amazonian populations were culturally,
linguistically and biologically discrete at different
moments in antiquity remains to be adequately addressed with
archaeological evidence.
However,
this shifting cultural
interface (or series of interfaces} has probably ranged
within the approximate geographic limits of the montane and
61
premontane forests for at least two and a half millennia
(since the Chavin horizon according to Lathrap 1971:97).
The antiquity and persistence of Andean-Amazonian
boundaries undoubtedly structured and conditioned every
class of border interaction.
Steward envisioned "migratory
waves" breaking against the base of the Central Andes,
creating what Lathrap calls "linguistic balkanization" at
the western edge of Amazonia.
A complex history of language
divergences and convergences in this cul-de-sac is likely
matched by complex archaeological and biological
distributions.
Cultural interaction has been sustained
along the Andean-Amazonian interface for millennia as the
archaeological data unequivocally demonstrates (Lathrap
1973a; Raymond 1988).
This continuous interaction
inevitably included processes of immigration, acculturation
and replacement.
The well-founded assumption that
opportunities for the exchange of material goods,
information and genes were constant, and always changing,
strongly implies that postulated migrations would have
occurred against a backdrop of intense and incessant social
interaction and trade.
Preliminary archaeological research (e.g. Allen 1968;
Lathrap 1970; Raymond et al. 1975; Hastings 1985)
indicates
that the Andean-Amazon interface is characterized by complex
distributions of styles and stylistic attributes.
al.
Adams et
(1978:503), observe that migration theories are most
62
controversial "in areas where there is a diversity and
complexity of cultural styles."
Apparently, style is a
frequently cited, but seldom convincing tool for migration
advocates.
This is because stylistic resemblances also
serve archaeologists to identify networks or spheres of
interacting societies.
The frequent failure of migration
theories to withstand close scrutiny strongly suggests that
stylistic evidence for intense cultural interaction and
exchange is prone to frequent misinterpretation as evidence
of society-level population movements.
Such interaction
characteristically creates "noise" in what Rouse and Anthony
imply should be a relatively straightforward procedure.
Archaeologists have observed that this noise or "ambiguity"
is typical of social boundaries where styles become fluid,
often merging to render "hybrids," as they are manipulated
to serve different ends (Ericson and Meighan 1984; Lightfoot
and Martinez 1995:482).
Lightfoot and Martinez advocate
analyses of material culture, stylistic behaviors and
information exchange at boundaries to elucidate cultural
processes of potential significance both locally and in the
adjoining regions.
It would appear that there is seldom sufficient
evidence to convincingly infer or entirely discount
hypothetical prehistoric migrations in contexts of intense
and sustained interaction.
Indeed, exchange of goods and
information along the Central Andean, Northern Andean and
63
Amazonian boundaries must have been accompanied by some
population interbreeding and genetic exchange.
Under these
circumstances it seems most appropriate to adopt an
inclusive definition of "boundary interaction" that
encompasses the exchange of material goods, information,
linguistic elements and genetic material.
From this
perspective, Manachaqui Cave's probable function as a
wayside station offers a unique opportunity to take the
pulse of interregional interaction at a major prehistoric
crossroads.
While migration explanations constitute
convenient and occasionally elegant solutions to temporal
and spatial problems, a framework emphasizing "boundary
interaction" provides promising new avenues for future
investigations.
CHAPTER 2
CIVILIZING MIGRATIONS, VERTICALITY COLONIZATION
AND THE EASTERN MONTANE FOREST
"Los guerreros emplumados, los c6ndores y las cabezas clavas
que ... en el Pajaten aparecen en forma primaria, en Chavin se
replican y alcanzan su perfecci6n. cEntonces? Las
investigaciones deben encaminarse en este sentido. En el
Gran Pajaten debemos empezar a buscar las raices de Chavin,
o sea las huellas del hombre cavernario que se transfor.ma en
hombre-pensante, racional, humano ... ;Si hubo un Eden
americano, lo hallaremos en el Gran Pajaten o cerca de el!
Tal vez eso quiso decirnos el sabio medico llamado Julio C.
Tello (Micerino the wise professor in Teofilo Maguina
Cueva's Chavin: La Epopeya Jamas Contada 1990).
This chapter constitutes a survey of the major theories
that bear either indirectly or directly on tropical montane
forest prehistory.
In Chapter 1, I described the earliest
notions which implicitly or explicitly characterized the
montane forest as an empty frontier or barrier separating
pre-Hispanic Andean and Amazonian worlds.
I also described
the "unitary origins" and "verticality" frameworks
responsible for hypotheses invoking: 1) society-level
population movements and 2) colonization.
For descriptive
purposes, these two migration scenarios can be labelled
"civilizing migrations" and "verticality colonization,"
titles that reflect their theoretical underpinnings.
In the
first two sections of this chapter, I will describe
prominent civilizing migration theories followed by the
64
65
major verticality colonization hypotheses.
The third
section will note alternative viewpoints expressed by a
small group of "dissenting" archaeologists.
A map is
provided (Fig. 3) to locate the sites mentioned in the
following text.
Civilizing Migrations and the
Eastern Montane Forest
As contemporary scholars who frequently referred to one
another's work, Meggers (1971) and Lathrap (1970) have been
central figures formulating "civilizing migrations" that
reputedly introduced "advanced" cultural innovations to
their Andean or Amazonian destinations.
However, the stage
was set for the development of such frameworks by the early
twentieth century theorizing of Spinden (1917), Uhle
(1922b), Rivet (1968 [1924]), Steward (1949) and Tello (1967
[1921]).
These and other scholars have worked under the
assumption that a New World neolithic or "formative" stage
complex of pottery, sedentism and agriculture had a single
origin or hearth, and subsequent spread.
Meggers and
Lathrap are scholars who have most clearly labored under the
influence of unilineal evolutionary "ruling frameworks."
Meggers' framework is derived from her mentor Leslie White,
the "intellectual heir" to L. H. Morgan (Willey and Sabloff
1974:154; Trigger 1989:290).
Lathrap's is an eclectic blend
of culture history and evolutionary theory (Oliver 1992)
influenced by a diverse range of scholars including Charles
66
Darwin, Edward Sapir, Alfred Kroeber, Carl Sauer and Julio
C. Tello.
While working within the popular evolutionary
paradigm of his era, Tello was inspired by nationalist and
regionalist fervor to substantiate autochthonous origins for
highland Peruvian civilization particularly to counter
Uhle's (1922b} theory of "foreign" Mesoamerican origins
(Lumbreras 1974:8).
Although these population movement theories seem
obsolete in the face of enlarged data bases and newer
paradigms, Tello's inductively reasoned model reminds us
that they persist as explanations for anomalous material
patterning and hypothesized "site unit intrusions" in the
eastern Central Andes and western Amazonia.
Tello's theory
persists in Peruvian popular culture as Maguifia Cueva's
quotation at the beginning of this chapter attests.
Rivera's (1975, 1991, Rivera and Rothhammer 1986} relatively
recent presentation of evidence for a Preceramic Period
"trans-altiplano" population movement from lowland Amazonia
to the Chilean coast demonstrates that prehistoric
population movements up and down the Andean slopes remain as
viable hypotheses, and should not be rejected out-of-hand.
Discussed below are major "civilizing migration"
hypotheses that can be separated into two groups.
Representing the first set are European and Peruvian
scholars who argued for long-distance, society-level
population movements.
During the early twentieth century,
67
Tello and Rivet promoted theories envisioning population
intrusions from the lowlands north and east of the Central
Andes.
Tello (1967 [1921]; 1942) postulated a precocious
role for migrant Amazonian populations as pioneers of
agriculture and donors of essential components of high
culture to the Central Andes.
Rivet (1968 [1924]) indicated
the "oriente" along with more distant and diverse sources
including Mexico, Central America and Polynesia for the
succession of ancient Peruvian "civilizations."
Both
scholars relied on perceived distributions of material
culture and native languages for supporting evidence.
As Tello influenced generations of Peruvian
archaeologists and inspired the more recent proposal of
Amazonian origins by Lathrap, his hypothesis and its
implications for montane forest prehistory will be outlined
in some detail.
Rivet's impact on montane forest prehistory
has been less direct, and is apparent in subsequent
theorizing by French scholars like Langlois (1939, 1940a,
1940b) and the Reichlens (1950), who first attempted to
characterize ancient settlement in the northeastern montane
forests.
The French group represents a long tradition of
European scholarship rooted in evolutionary assumptions
(Binford and Sabloff 1982; Clark 1994).
A more recent
evaluation of northeastern montane forest prehistory by
Peruvian archaeologist Kauffmann (Kauffmann et al. 1989) is
explicitly rooted in L. H. Morgan's unilinear evolutionary
68
framework and its basic assumption of unitary origins
(Kauffmann 1983:80-81).
The more contemporary set of "civilizing migration"
theories has been framed by the U.S. scholars Meggers,
Lathrap and Lathrap's student Isbell (1974).
Although
Lathrap is known for his hypothesis of unitary Amazonian
origins for the New World Formative Stage, he also proposed
an additional, independent population movement for the
peopling of the eastern slopes (1970:171-179) that was
further elaborated by Isbell (1974).
Isbell linked
Lathrap's hypothetical movement of Quechua-speakers to the
Central Andean developmental sequence and, coincidently, to
the Middle Horizon expansion of the highland Huari &mpire.
Although the intellectual pedigrees are clearly distinct,
Lathrap's and Isbell's population movements, like Tello's,
Rivet's and Meggers' all assume the synchronous movement of
race,
language and culture.
Primordial Migrations and the
Eastern Montane Forest
Much of Julio C. Tello's writing may be regarded as a
counterpoint to Uhle's thesis that Peruvian civilization
resulted from migrations of Mayoid populations spreading
southward from Central America (Uhle 1922a, 1922b) .
Tello
argued that the earliest civilizations originated within the
modern-day borders of Peru, and that the mother civilization
which gave birth to Mochica, Nazca and other later
69
"civilizations" developed in situ at the highland site of
Chavin de Huantar (Tello 1942, 1943, 1960).
However,
Chavin's population and many of its cultural traditions
derived directly from the tropical forest.
During what Tello first termed an "Era Primordial,"
migrants of "low culture" (Tello 1967 [1921] :199) from the
tropical forests arrived in the highlands, first entering
the humid northeastern Andes and subsequently adapting to
the dryer central Peruvian Andes as they moved west and
south.
These early forest migrants subsisting on
agriculture, hunting and fishing, and bearing cult beliefs
and myths featuring the jaguar-deity, formed a basal
cultural stratum in the northeastern Andes (Ibid.:209,
1960).
1929;
They also brought pottery and cultigens such as
cotton, coca, maize, chili peppers, gourds, squash, beans,
peanuts, sweet potatoes and manioc (1929:21, 1960:12).
However, agricultural technologies and animal domestication
first developed in the "healthy" highland environment.
The
fully-developed subsistence complex then descended to the
arid and insalubrious western yunga and coast (1929:21, 166;
1960) .
Tello presented ecological, archaeological, linguistic
and iconographic arguments, coupled with observations based
on eastern tropical forest ethnology and myth, to support
his view that Chavin civilization drew constituent elements
from the tropical forest.
Like other scholars of his time,
70
Tello regarded the edge of the tropical forest as the
natural limit of Andean civilization (1942:621).
However,
he was aware of the ecological complexity of the forest's
edge (Ibid.:633), and believed that much of the cultural
interchange between tropical forest and highland cultures
took place in regions of ecological interdigitation,
especially where warm tropical east-west, or "cis-Andean,"
valleys deeply penetrated the highlands.
Tello's ecological
argument for the transfer of cultigens from lowland to
highland ecozones (1942, 1960) actually anticipated some of
Sauer's ideas (1952:50) by a decade.
In Tello's model,
human colonization and the adaptation of lowland cultigens
to colder environments occurred at moist intermediate
elevations between the eastern lowlands and highlands where
irrigation was unnecessary.
The Marafion river valley in particular attracted
Tello's attention as a strategic area for colonization by
tropical forest populations exploiting both forest and cisAndean or yunga resources, and where lowland cultigens
became adapted to colder and more varied environments
{1960:18, 23).
On both sides of the Marafion River, Tello
initially identified settlements of a culture currently
known as Recuay, which he alternately termed Recuay, RecuayPasto, Huaylas, Marafion and Archaic North Andean.
This
culture exhibited distinctive monumental stone architecture
and ceramics, an elaborate lithic sculptural tradition and
71
shaft tombs.
In his early work, Tello (1960) considered
these as representative of the first Andean civilization.
However, he eventually became aware of the stratigraphic
superposition of Recuay over Chavin remains at Chavin de
Huantar (Tello 1943:155), and his later writings emphasized
the primacy of Chavin civilization (1942, 1943).
To argue for the temporal priority of Chavin, Tello
relied on stratigraphic and ceramic evidence.
The fine,
dark monochrome incised and polished pottery which he
attributed to the Chavin style lay beneath temple
constructions at Chavin de Huantar, and also underlay the
archaeological remains of other cultures throughout both
highland and coastal regions (1943:158-159).
Tello also
invoked ceramic evidence to advance his hypothesis of
Chavin's tropical forest affiliations.
He believed that
sherds from the deepest strata at Chavin de Huantar
resembled "the incised and carved types of the Amazon"
(1943:152).
The importance that he attached to the major
Amazonian river valleys as hypothetical migration routes
prompted Tello to visit the headwaters of the Marafion and
Huallaga Rivers.
At an elevation of approximately 2,000 m
within the upper Huallaga River valley surrounding the town
of Huanuco, Tello found evidence to support his tropical
forest origins thesis.
Artificial mounds at Kotosh and
Shillacoto contained what he interpreted as "incised and
carved" Amazonian pottery in strata below those containing
72
Chavin and Marafion types (Tello 1942:710).
Tello was familiar with three pottery "types" found in
"Anti Suyo," for which he provides cursory descriptions: 1)
the early dark monochrome excavated at Kotosh, and reported
from the localities of Matibamba (20 km downriver from
Chavin de Huantar) and Ong6n (near present-day Tayabamba in
southeastern La Libertad Department), Monzon (northeastern
Huanuco Department) and Satipo and San Luis de Shuaro (Junin
Department), 2) bichrome and polychrome pottery with
geometric and cursive decoration from the upper Utcubamba
and lower Marafion river valleys and 3) glazed, polychrome
pottery like that produced by the Conibo and other Panoanspeaking groups in the Ucayali River valley (1942:635-636).
A close look at Tello's schematic view of Andean cultural
development (1942:715) reveals that he envisioned these
three types as sequentially representative of the first,
second and third ages or epochs.
First and second epoch
pottery styles cluster in and around the eastern montane
forests.
Other lines of evidence utilized by Tello to argue for
tropical forest origins are of varied nature.
He perceived
historical relationships between linguistic groups stretched
along the same latitudinal parallel like the coastal Muchik
and the Amazonian Amuesha, Chol6n and Hivito languages as
evidence for the westward expansion of Arawakan speakers
{Tello 1942:629).
Also indicative of the northeastern
73
Andean cultural substratum formed by Tello's early forest
migrants are ethnographically recorded myths featuring a
dragon, the moon and sun and their twin offspring
(Ibid.:629), and the jaguar cult (1929:168).
The tropical
forest jaguar, Tello argued, is the principle deity depicted
in Chavin sculpture (1960:160), and appears as the god of
rain in the highland Wira-kocha myth (1929:168).
Secondary
deities in Chavin art include condors, bats, serpents, fish
and other fauna common to Andean habitats (Ibid.:168).
Tello also observed a number of Andean cultural traditions
which he believed originated in the eastern forests.
Among
these were the taking of trophy heads and the attached
belief systems, cranial deformation, urn burial and other
funerary practices, the manipulation of feathers for
sumptuary display and the use of stone axe heads.
In summary, Tello argued that Andean civilization was
conceived in the highlands.
However, its basic constituents
such as populacions, cultigens and technologies invariably
originated farther east in Amazonia.
A review of Tello's
maps and diagrams (Tello 1943: Lam. I, p.645; Lam. VII,
p.715), and the archaeological sites specified in his
discussions, indicates that the area which he alternately
referred to as the "floresta," "Andes Orientales," and
"Antisuyo" lies in the sub-tropical forest at the eastern
base of the Andes, or Lathrap's upper montana (1970).
As
Bonavia (1981) points out, Tello never actually worked in
74
the eastern montane forest proper.
The middle elevations of
the Marafion and Huallaga river valleys served as loci for
the "colonization" or upslope expansion of lowland
populations and cultigens in the process of adapting to new
environments.
The headwaters of the Marafion and Huallaga
rivers ultimately yielded pottery and other kinds of
evidence which Tello utilized to construct his argument.
Criticisms of Tello's theoretical positions aimed at
his interpretation of Chavin as the first pan-Andean
civilization (e.g. Larco 1941; Kroeber 1944), rather than
his population movement hypothesis.
In fact, he makes
reference to Rivet's support for his ideas regarding
tropical forest origins (1942:633).
Uhle, Jacinto Jij6n y
Caarnafio (1927), Rivet and Tello all agreed on migrations
from a northerly direction.
Tello parted company with this
group by proposing exclusively Amazonian, as opposed to
Central American origins.
Civilizing Chachapoyas
The upper elevations of the northeastern montane forest
between the Marafion and Huallaga Rivers correspond to the
Inca Province of Chachapoyas.
The name Chachapoyas has been
retained by historians, anthropologists and archaeologists
referring to the ethnographic and archaeological culture
area surrounding the Utcubarnba River Valley.
Unlike other
"civilizing migrations" postulated to account for
prehistoric events in the heart of the Central Andes, these
75
hypotheses reflect an early scholarly fascination with
Chachapoyas.
As we will see, hypothesized migrations
converge on Chachapoyas from all directions.
Major
"civilizing migrations" impacting Chachapoyas from the north
were postulated during the early twentieth century.
An
early consensus for northern and/or eastern origins, based
largely on intuition rather than rigorous analyses, lasted
until the mid-1960s when population movements from the west
were proposed.
These theories are described below.
Population Movements and
"Cuelap Civilization"
Adolf Bandelier (1909) wrote the first monograph on
archaeology and ethnography of the northeastern montane
forest Utcubamba Valley.
Like scholars arriving before and
after him, Bandelier devoted considerable attention to the
spectacular Chachapoyas site of Cuelap (also Kuelap) .
Setting the stage for subsequent investigations in the Inca
Province of Chachapoyas, Bandelier noted similarities
between the Chachapoyan Quechua dialect spoken and the
southern Andean Aymara language, as well as an abundance of
native names of places and individuals which could not be
related to Quechua.
He concluded,
"the country of
Chachapoyas was once inhabited by a tribe or tribes, that
belonged to a stock different from the Peruvian
mountaineers, their western neighbors"
(1909:10).
After his
survey of Chachapoyas archaeology, Langlois (1939, 1940a,
76
1940b) hypothesized population movements
"from the north or
the northeast" (1940b:226; this and subsequent translations
from Spanish are mine)
that recall Tello's claim for the
importance of the northeastern Peruvian Andes as the
entryway for the tropical forest populations ultimately
responsible for Chavin civilization.
However, Langlois'
interpretations were most strongly influenced by Uhle's
( 1922b) ·and Jij 6n y Caamafio' s
( 1927) arguments for long-
distance migrations of "archaic cultures" from Central
America.
Langlois utilized evidence (mostly toponymic) of
intrusive linguistic elements from Ecuador and Mexico to
bolster his suggestion.
With the encouragement of Rivet in the 1940s, Henry and
Paule Reichlen reported evidence for three successive
"civilizations" spanning five centuries of continuous preHispanic occupation based on reconnaissance and excavations
in and around Cuelap.
They were most impressed by
resemblances between pottery of the "Cuelap Civilization"
and styles of the Amazon, Orinoco, Ecuador and Colombia
(Reichlen and Reichlen 1950:240).
Neither the Reichlens nor
Langlois cite Rivet's theory of multiple migrations into the
Central Andes from the north and east (Rivet 1968 [1924]),
yet Rivet's vision likely influenced their interpretations.
Nine years after the Reichlens' publication, Horkheimer's
synthetic review of Utcubamba valley prehistory expressed
support for Langlois' original thesis of "primitive
77
immigrations from the north or northeast"
(Horkheimer
1959:89), although he declined to cite more specific points
of origin.
Regarding the "immigration" process, Horkheimer
theorized:
The topography shows us that the entrance to the Utcubamba
valley is relatively easy from the north and that, in
contrast, natural communication with the south and west is
relatively difficult ... The immigrants stayed within the
blind alley of the Utcubamba where the arrivals from lowland
and tropical regions found a series of ecological elements
to which they were accustomed and that other regions lacked.
The difficulties of communication with the south and west
caused, for centuries, an isolation during which the
established immigrants developed their own cultural
characteristics without much outside influence. Later, with
increasing population and changing needs and aspirations,
contact between the inhabitants of the Utcubamba and their
neighbors, especially in the region of Cajarnarca, was
accentuated (Ibid.:89-90).
Apparently, Horkheimer's "immigrants" intruded upon local
"civilizaciones desconocidas"
(unknown civilizations) that
he places at the bottom of his "scheme of cultural
succession in the Utcubamba region"
(Ibid. :Fig. 1).
He
postulates that "Cuelap Civilization," characterized by
influences from nearby highland Cajamarca, was in place by
A.D. 1000.
Most recently, Lerche (1995} has restated the same
hypothesis for which he finds supporting evidence in the
province's restrictive geography and its ceramic styles,
iconography and historical linguistics.
Lerche believes
that the "Chachapoyas cultural complex" intruded from the
north or east around A.D. 900, following a 400 year period
during which the province was only lightly inhabited
(Ibid. : 3 7} .
The undated Huayabamba ceramic complex
78
excavated in the central Huallaga Valley lowlands by Ravines
(1978) purportedly represents the eastern limits of
Chachapoyas (Ibid.:21).
Like Ravines (1978:531; 1981a:178),
Lerche believes that the lowland Huayabamba complex is
ancestral to Cuelap styles (Ravines declined to specify
migration or any other mode of style transmission) .
A
carved wood statue reputedly from a Chachapoyas tomb located
east of Leimebamba bears the representation of a cayman,
thereby providing Lerche with additional evidence for
lowland origins (Lerche 1995:75).
By "north and east,"
Lerche clearly refers to Amazonia.
"Highlandizing" the Montane Forests
Since 1980 Kauffmann has investigated Cuelap, Gran
Pajaten (also called Abiseo) and other archaeological sites
of the northern and northeastern montane forests between the
Ecuadorian border and Huanuco Department.
However, he
refers to the greater northern and eastern montane forest
from Ecuador to Bolivia comprehensively as the Andes
Amaz6nicos or "Amazonian Andes"
(Kauffmann 1986, 1987,
1992b; Kauffmann and Ligabue 1990; Kauffmann et al. 1989).
During the past two decades Kauffmann has reported on field
investigations funded by Italy's Centro Studi Ricerche
Ligabue, and presented a "working hypothesis" that he terms
"serranizaci6n de la selva," or the "highlandizing of the
jungle, " in reference to a proposed prehistoric eastward
expansion of both highland populations and anthropogenic
79
paramo environments produced by deforestation (1986, 1987,
Kauffmann and Ligabue 1990; Kauffmann et al. 1989).
The highlandizing process as envisioned by Kauffmann
was preceded by a long period of prehistoric population
growth and agricultural intensification in the Andean coast
and highlands.
Kauffmann subscribes to a hypothesis of
unitary origins for New World "High Culture" in coastal
Ecuador, and spread to the Central Andes (1992a:96-102).
In
numerous publications, Kauffmann has expressed his view that
virtually every aspect of Central Andean cultural
development was caused and conditioned by the constant state
of crisis rendered by overpopulation, insufficient
agricultural land and the shortage of food (Kauffmann 1976,
1983:81, 1991, 1992b; Kauffmann et al. 1989:8).
The
eastward population movement began when carrying capacity
was reached in the Central Andes.
Kauffmann observes that
virtually nothing is known of indigenous montane forest
inhabitants who were either absorbed, displaced or
exterminated during the eastward population movement
(1986:6; 1987:6).
Contact with these populations,
adjustments to the environment and relative isolation imbued
the resulting cultures of the Amazonian Andes with Amazonian
features and "singular formal aspects" although they
remained fundamentally Andean in character (1987:5,
1992b:108).
Kauffmann's eastward population movements reportedly
80
began by A.D. 1 (1992b:108; Kauffmann and Ligabue 1990:336;
Kauffmann et al. 1989), and took place in "systematic form
and administered by coercive powers"
that extensive terracing systems,
(1987:6).
He theorizes
"cultist-administrative
centers" like Gran Pajaten, and elite cemeteries provide
evidence for centralized political control of the
highlandizing process (1987:6; Kauffmann et al. 1989:7-8).
Rock art and lithic sculpture documented in the northern
montane forests near Jaen indicate "pre-agricultural" and
"formative" occupation respectively (1991:47).
However, it
remains unclear whether Kauffmann regards these earliest
finds as indigenous or intrusive.
He attributes the
sculptural style of lithic carvings from the Tabaconas River
valley (also near Jaen) to influence from "formative"
highland centers like Kuntur Wasi (Ibid.).
Kauffmann's hypothesis is a broad-brushed and panoramic
scenario fashioned to a large degree for consumption by a
lay audience, especially Peruvians and Italians.
In
support, he typically provides a general discussion of
archaeological features encountered in the northern and
northeastern Amazonian Andes.
He does not differentiate the
indigenous from the highlandizing intrusive populations
utilizing archaeological evidence.
largely hypothetical.
These remain,
therefore,
It is the amalgamation of the two
hypothetical populations, or its material remains,
Kauffmann customarily describes.
that
His hypothesis dimly
81
echoes Murra's verticality-type colonization coordinated by
"coercive" authorities, yet his unitary origins framework
sets Kauffmann's theory squarely within the tradition of
"civilizing migrations."
Neolithic Migrations and the
Eastern Montane Forest
Most influential to the work of Lathrap were Tello's
notions that lowland to highland migration characterized
early relations between the Amazon and Andean areas, and
Sauer's view that the botanically-rich lowland tropical
forests were the likely hearth where agricultural systems
first developed.
Domesticated cultigens, sedentary
agricultural economies, ceramic technology, a complex
cosmology and related art, Lathrap argues, are among the
tropical forest's contributions to Andean civilization.
He
assumed Tello's position that the first pan-Andean
civilization sprang from the center at Chavin de Huantar,
and that Chavin's residents were agriculturalist migrants
from the eastern tropical forest, perhaps arriving via the
Marafion valley (Lathrap 1973b:92-93).
Lathrap also followed
Tello in the belief that the migrants brought with them a
complex mythology based on tropical forest cosmology.
Yet
while Tello argued for a highland crystallization of Chavin
civilization, Lathrap (1970, 1971, 1973b, 1974, 1977)
contended that the entire Chavin complex arrived with
intrusive Arawak-speaking populations bearing a fully
82
neolithic economy and complex social organization developed
many centuries prior on the Amazon River floodplains.
Much of Lathrap's work aimed at discrediting the
proposition advanced by Steward, but elaborated by Meggers
and Clifford Evans, that prehistoric cultural evolution in
Amazonia was retarded by the region's uniformly low
agricultural potential (Meggers 1954).
Meggers and Evans
(1983:330) have consistently argued that "the lowlands of
South America were the recipients rather than the
originators of major technological innovations, especially
agriculture and pottery making."
From their perspective,
Amazonian societies contributed little, if anything, to the
development of Central Andean civilization.
During his
fieldwork on the Central Ucayali, however, Lathrap observed
cultural and ecological processes on the river floodplains
and proposed mechanisms for Amazonian population growth
based on tropical forest subsistence and demographic
patterns.
A "phenomenally efficient system of cultivation
of bitter manioc" supplied the motor, churning out
populations in excess of local floodplain carrying
capacities and driving them upstream into every available
floodplain niche and up through the eastern montane forests
(Lathrap 1970, 1977:744).
Chavin's immigrant populations
expanded from their Amazonian homeland southward "along the
eastern face of the Central Andes" where they penetrated the
north and south-trending Andean valleys of the Marafion and
83
Huallaga rivers (Lathrap 1982:312).
The interpretation of Chavin iconography is central to
Lathrap's argument for the "Chavin pattern's" tropical
forest origins (1971, 1973b, 1974, 1982).
The images
evoking lowland Amazonian cosmology and staple foods
depicted on the carved stone Tello Obelisk (1973b, 1982)
represent "a celebration of the prior [tropical forest]
agricultural system which allowed the Chavin elite to
achieve its current dominance"
(Lathrap 1973b:103).
Thus,
the depicted mythological event and inventory of lowland
cultigens constitute a material record of the subsistence
economy that fueled the up-stream expansion of floodplain
populations, the intrusion into the eastern highlands and
settlement at Chavin de Huantar.
At the time of Lathrap's
investigations (prior to Burger's reinterpretation of the
Chavin horizon), scholars believed that Chavin's sphere of
influence had disseminated as far as the south coast Ica
valley by 1000 B.C.
(Patterson 1971).
To debates regarding Andean civilization's origins,
Lathrap brought new ceramic evidence derived from the
earliest and most complete stratigraphic sequence from the
upper Amazon Basin (Lathrap 1958, 1962).
His Early and Late
Tutishcainyo Phases from the Central Ucayali drainage
roughly correspond to the Central Andean Initial Period,
while the Shakimu Phase (later sub-divided) was thought
coeval with the spread of Chavin.
Like Tello, Lathrap
84
sought to buttress his population movement hypothesis by
searching beyond Chavin to find archaeological localities
that served as migratory links.
the Japanese
He took special interest in
excavations at highland Kotosh in the upper
Huallaga Valley where the earliest ceramics comprised a
large sample of the same early pottery first identified by
Tello as "Amazonian"
(Izumi and Sono 1963).
Lathrap
utilized the new evidence to restate Tello's argument for an
early intrusive Amazonian population movement into the upper
Huallaga River drainage.
This contention hinged on a series
of stylistic similarities shared between the Initial Period
Wairajirca Phase pottery, Tutishcainyo Phase ceramics, and
the geographically and stylistically intermediate Cave of
the Owls fine wares from Tinge Maria (Lathrap and Rays
1963).
The Wairajirca material directly overlay preceramic
temples of the Kotosh Mito Phase.
Meggers and Evans were likewise interested in the upper
Huallaga and Ucayali sequences.
They proposed that
Tutishcainyo and Wairajirca were both manifestations of a
cultural sub-stratum characterized by pottery featuring
zoned-hachure decoration and flanged rims found beneath
Amazon Basin sequences ranging from the Rio Napa of Ecuador
to Maraj6 Islru1d of eastern Brazil (Meggers and Evans 1961,
1983; Meggers et al. 1965; Evans and Meggers 1968).
The
"Zoned-hachure Horizon/Tradition" was said to have
originated during the third millennium B.C. Tesca Phase on
85
the Caribbean coast of Colombia.
Meggers and Evans
hypothesized that the tradition was carried from the Andes
into Amazonia during an episode of increased aridity between
2,000 and A.D. 1 that facilitated population movement by
opening forest-free corridors (Meggers 1975, 1977, 1979).
Accordingly they viewed lowland Tutishcainyo as a derivation
of the highland Wairajirca style (Evans and Meggers
1968:92).
In response to Meggers and Evans' skepticism regarding
his claims for Tutishcainyo•s great antiquity and temporal
priority, Lathrap (1971) constructed a complex argument
utilizing cross-dates in lieu of lacking radiocarbon
evidence in order to demonstrate the temporal priority of
the Tutishcainyo tradition.
To demonstrate how the pottery
styles reflected lowland to highland population movement
rather than the reverse, he listed a series of Early and
Late Tutishcainyo stylistic attributes that appear to be
intrusive within the Wairajirca assemblage.
Lathrap (1965,
1970, 1971) argued that, at the same time, Central Andean
ceramic traditions had exerted no visible influence on the
lowland Tutishcainyo pottery.
Lathrap described the Wairajirca Phase assemblage as "a
blending of two distinct cultural traditions, one of which
was very like Early Tutishcainyo, and relates to an upward
expansion of tropical forest peoples to the eastern slopes
of the Andes; the other of these traditions appears to be
86
indigenous to the Central Andes or Peruvian coast"
1971:93).
(Lathrap
While the Wairajirca neckless ollas typify
earliest Central Andean traditions, a hypothetical
"collateral relative" of the Amazonian Tutishcainyo style
donated carinated vessel shapes, and zoned-incised and postfire painted decorations deemed intrusive at Kotosh.
Lathrap emphasized that all Wairajirca vessel shapes and
decorative modes, except for the neckless olla, have Early
Tutishcainyo "homologies."
Similarly shaped double-spout-
and-bridge bottles appear in both assemblages.
This
evidence, combined with the lack of neckless ollas and other
signs of highland-derived influence in the Tutishcainyo
assemblage, provided Lathrap with a powerful case for
upslope rather than downslope intrusion (Lathrap 1970:106107, 1971:94).
Lathrap aligned the following Kotosh Kotosh Phase with
the final years of the Ucayali Late Tutishcainyo and
subsequent Early Shakimu phases.
Kotosh Kotosh radiocarbon
dates are 890 ± 170 B.C., 920 ± 230 B.C. and 1120 ± 110 B.C.
(Izumi 1971), while his Early Shakimu Phase yielded the date
of 650 ± 100 B.C.
(Lathrap 1971).
The Early Shakimu Phase
appearance of slipped and burnished bowls with excised
decoration similar to those of the Kotosh Kotosh Phase
suggested a historical relationship to the spreading Chavin
art style (Ibid. :92) and a shift to downslope diffusion
(Ibid.:88).
Stylistic similarities between vessels of the
87
Kotosh Sajara-patac Period and Late Shakimu Phase ostensibly
demonstrated continuing communication between the highlands
and lowlands that ended abruptly around 200 B.C. with the
Central Ucayali appearance of the Barrancoid pottery
tradition (Lathrap 1970:117).
The coeval Higueras Phase at
Kotosh represents the Andean Early Intermediate Period.
Lathrap theorized that the early Kotosh sequence
documents a "site-unit intrusion"
(1971:94) of Amazonian
populations that dominated the indigenous Andean populations
of the upper Huallaga River valley.
for the Tello-Sauer hypothesis
Hence, he found support
(Lathrap 1970:107),
explicitly rejected Meggers and Evans' downslope hypothesis,
and concluded that,
"The earliest sedentary communities with
elaborate masonry structures so far discovered in the
Central Andean highlands appear influenced by, and indeed
partially derived from,
tropical forest peoples"
(1971:94).
The fundamental assertions coursing through all of Lathrap's
publications are: 1) neolithic economies and stratified
societies had developed fully by 3,000 B.C. in the tropical
lowlands and 2) that the fully-developed complex penetrated
the Central Andes intact via population movements from
Amazonia.
Some of his bold and provocative assertions have
been sustained, although far less dogmatically, by the small
group of scholars that he trained.
Support for Lathrap's interpretation of developmental
events at Kotosh appeared with Kane's conclusions derived
88
from study of material culture excavated at Kotosh and
nearby Shillacoto (Kano 1972, 1979).
At Shillacoto,
archaeologists uncovered a temple similar to the preceramic
Kotosh Mito Phase temples, yet utilized during the Initial
Period Wairajirca Phase (Burger and Salazar-Burger 1980:30).
Kano utilized stylistic and iconographic evidence to
demonstrate that the Chavin "feline cult" had a history of
pre-Chavin stylistic development in the Huanuco basin, and
still earlier roots in as yet undiscovered Amazonian styles.
He admitted, however, a lack of supporting evidence for
concurrent shifts in construction styles, subsistence
economy and social organization expected to accompany such a
population replacement during the Mito/Wairajirca transition
(Ibid.).
Most criticisms of Lathrap's position came from
scholars arguing for Andean civilization's temporal priority
on the Central Coast.
Lanning dismissed the possibility of
any historical relationships between Kotosh and Tutishcainyo
pottery, suggesting instead "some interchange of stylistic
ideas"
(1967:87-88).
Fung (1982 [1971] :457-458) remarked
that evidence for socio-political complexity is lacking at
Kotosh.
Lathrap's lowland immigration hypothesis received
close scrutiny from proponents of Moseley's theory of Andean
civilization's maritime origins.
Moseley questioned the
validity of Lathrap's analytical procedure by dismissing
Initial Period pottery styles as useful analytical units:
89
In ceramic assemblages from the early part of the Initial
Period, ornamentation is generally based upon a limited
repertoire of incised geometric and curvilinear motifs, and
the vessel-shape categories are not numerous. Styles are
therefore relatively unelaborated, and often share
superficial resemblances because of their simplicity
(Moseley 1983:209).
Furthermore, Moseley contested Lathrap's conceptual linkage
of pottery style, tropical forest agriculture and complex
social organization, thereby rejecting the validity of
utilization of pottery styles as indices reflecting "the
invention and diffusion of the economic foundations of New
World civilization" (Ibid.).
Moseley pointed out that
corporate architecture (which he regards diagnostic of
complex social organization) pre-dates pottery in the Andes.
Finally, he echoed Kane's observation that Kotosh produced
no evidence that changes in subsistence practices and
architectural styles accompanied the appearance of
Wairajirca pottery during the Mito/Wairajirca transition
(1983:209, 210).
More recently, Moseley (1992) has acknowledged the
temporal priority of the Amazonian Tutishcainyo tradition
around 2000 B.C., but he provides an alternative to
Lathrap's population movement hypothesis.
To account for
the stylistic resemblances, Moseley considers tropical
forest influences on Wairajirca the result of verticality's
"downward thrust"
(Ibid.:46) and highlander exposure to
lowland ceramic technology during regular forays into the
lowlands (Ibid.:lOO, 144).
Moseley's idea suggests highland
"procurement" rather than lowland "bestowal," and highland
90
developmental superiority is unstated, yet strongly implied.
Moseley's critical remarks recounted above reflect the
relatively recent consensus that agriculture, sedentism and
monumental construction each have their own developmental
histories in the Central Andes dating from the Preceramic
Period.
Chavin civilization is presently characterized as
an amalgamation of cultural traditions drawn from several
regions,· the tropical forest included (Burger 1992;
Lumbreras 1993).
The current perception of Chavin as nexus
has rendered moot any consideration of Amazonian site unit
intrusions.
Lathrap's interpretation of the Initial Period
U-shaped coastal monuments as representations of the
cayman's jaws (Lathrap 1985) constitutes a parting shot at
Lanning's and Moseley's maritime origins theory.
Many scholars acknowledge tropical forest sources for
the imagery carved in Chavin stone sculpture, yet its
significance has rarely been interpreted as more than
evidence of Chavin's amply sized interaction sphere (e.g.
Lumbreras 1981; Raymond 1988).
Nevertheless, Raymond still
regards Lathrap's postulated migration as a valid hypothesis
(1988:293).
Raymond emphasizes the temporal priority of
tropical forest agriculture and pottery technology, and
reiterates Tello's and Lathrap's views regarding the
importance of the Marafion River valley as a likely route
through which early lowland cultigens penetrated the
highlands.
Transhurnance or "limited migration" are cited as
91
possible modes of agriculture's early diffusion from the
north highlands to the Peruvian coast (1988:294).
Lumbreras
(1993:354-355) also continues to emphasize Chavin de
Huantar's auspicious location at the headwaters of the
Amazon River's Marafion and Huallaga affluents.
In a recent consideration of Chavin sculpture's
tropical lowland imagery and Chavin de Huantar's origins,
Burger (1992:154-155) rejects Lathrap's population movement
hypothesis for two reasons.
First, he questions Lathrap's
use of sculptural iconography and its mythic representations
to reconstruct historical sequences and population
movements.
Second, Burger submits that if Lathrap's
migration scenario were accurate, then the earliest pottery
styles from Chavin de Huantar's cultural sequence should
reflect tropical forest ancestry.
Instead, Burger
interprets the earliest Chavin pottery, assigned to the
Urabarriu Phase (Burger 1984b), as an amalgamation of
diverse Central Andean styles that reflects Chavin's
economic and social interaction with distant coastal and
highland regions (1992:155).
Burger also pointed out that
subsistence data from Chavin de Huantar show a mixed
highland agricultural economy,
the product of a long history
of local Andean development (Ibid., 1995; Burger and van der
Merwe 1990; Miller and Burger 1995).
To Burger, the
tropical forest imagery that Lathrap utilizes to bolster his
hypothesis most likely represents the importation of foreign
92
esoteric knowledge from the lowlands by leaders at Chavin de
Huantar's Old Temple (1992:157-158).
Such sacred knowledge
could have been obtained during the long journeys undertaken
by shamans, and its lowland source would have facilitated
acceptance and respect as a source of supernatural power.
Quechuas, Maize and the Eastern
Montane Forest
Like the Langlais-Lerche hypothesis of northeastern
montane forest peopling, Isbell's (1974) elaboration of
Lathrap's hypothesis postulates population movements
originating in distant premontane or subtropical forest
settings to the north.
Isbell's hypothesis combines
interpretations of archaeological, ecological and linguistic
data to generate a theoretical model for the peopling of the
eastern montane forest by Quechua-speaking populations.
Isbell took Lathrap's model one step further however,
describing a population intrusion with widespread impact
throughout the Central Andes.
His argument is complex and
requires acceptance of a series of assumptions beginning
with several proposed initially by Lathrap.
At the onset of
the Higueras Phase in the upper Huallaga valley near
Huanuco, Lathrap noted a sudden shift of site locations from
valley bottoms to defensible promontories between 3,500 and
4,000 m, accompanied by a shift to slope agriculture on
artificial terraces.
The simultaneous adoption of coarse,
brown ware ceramics featuring large jars with flaring rims
93
and red slip, crude modeling and applique decoration
purportedly indicates "a total replacement of population"
sometime between A.D. 100 and 600 (Lathrap 1970:173).
This
intrusion of Quechua-speakers from the Bolivian montane
forests far to the south ostensibly marked the beginning of
the Central Andean Early Intermediate Period (Ibid. 174176).
Examples of coarse wares purportedly associated with
eastern-slope Quechua speakers include pre-Tiahuanuco Chulla
Pampa ceramics from Bolivian Cochabamba, Late Intermediate
Period Killke from highland Cuzco, Chacra de Giacomotti in
lowland Junin, and pottery of the historic Chupachu of
Huanuco.
In effect, Isbell borrows Lathrap's theoretical
approach linking Amazonian ecological zones, language groups
and pottery styles, and applies it to the highland Central
Andes.
Of central importance to Isbell's model of Quechua
expansion is Murra's postulated association between
terracing systems and maize cultivation (Isbell 1974:142;
Murra 1956 [dissertation cited by Isbell], 1980:7, 12).
Maize served as the staple for an expanding Quechua
population which developed terrace agriculture as one of
several stages accommodating its agricultural technology to
the Andean slopes.
Second, Isbell reasons that,
"the origin
of the adaptation to the valley slopes should be found in an
area where dry cultivation of maize was possible without the
necessity of terraces"
(Isbell 1974:150).
Discounting the
94
dryer and colder regions south and west, he proposes that
the evolution of maize cultivation systems should have taken
place in the warm, moist valleys between Macas, Ecuador and
Chachapoyas, Peru (Ibid.).
Here dry maize cultivation could
be adapted to gentle slopes with a simple system of banking
earth, and to steeper slopes by employing masonry walls to
retain the thin soils typical of the eastern flanks.
The
development of terracing techniques permitted initial
Quechua settlement on the moist eastern slopes.
The
subsequent addition of irrigation to terrace systems allowed
the system's transfer into dry Andean environments further
south which, in turn, facilitated continued Quechua
expansion.
The principal types of evidence employed by Isbell are
ceramic attributes, radiocarbon dates and historical
linguistics.
In order to conceptually link the coarse brown
wares to Quechua population movements he implements Rouse's
classificatory terminology to refer to a coarse brown or "CB
series."
In addition to jars with flaring rims, he judges
strap handles and hemispherical bowls to be diagnostic
attributes of the CB series.
These features appear most
notably in pottery from Lambayeque, Huamachuco (terminal
Early Intermediate Period), Gran Pajaten, Huanuco (Kotosh
Higueras Phase and historic Chupachu), Huancayo (mixed with
Early Horizon San Blas), Middle Horizon Ayacucho, Cuzco
(Killke and Inca), the Ica valley (Ocucaje Phases 8 and 9)
95
and Eastern Bolivia.
Isbell describes radiocarbon evidence
for the appearance of the CB series at eastern Andean sites
noting dates of A.D. 70+200 for the Higueras Phase and an
additional date "in the first century A.D." associated with
the Bolivian Chullpa Pampa style.
The presence of the CB
series in the south coast Ica valley at the end of the Early
Horizon (Ocucaje Phase 9) suggests to Isbell that the CB
series "could have been introduced in the Central Andes
towards the second half of the first millennium B.C."
(1974:146).
Isbell regards suggestions of a Central Andean protoQuechua hearth by linguists Parker (1963) and Torero (1964)
as subject to alternative interpretation.
Most important to
the region and time period pertinent to Isbell's study is
the present-day distribution of Quechua A speakers in the
former Departments of Cajarnarca, Lambayeque, Amazonas,
Ecuador, San Martin and Loreto lying north and east, and
circumscribing Quechua B dialects in the central Peruvian
Andes.
According to Isbell, early eastern Andean
occurrences of the proposed CB series such as the Higueras
style relate to upslope population movement of proto-Quechua
A/B speakers (1974:140).
He proposes a subsequent
divergence of the two Quechua sub-groups within the Central
Andean highlands immediately prior to the Middle Horizon
(Ibid. :148).
Isbell's interpretation of the distributions of coarse
96
wares and Quechua languages was published four years after
the hypothesis of his mentor.
When later confronted with
alternative models for the prehistoric spread of Quechua
from the central coast (e.g. Torero 1974; Bird et al. 198384), Isbell restated his position (Isbell 1983-84:253) and
pointed to corroborating linguistic evidence for eastern
Ecuadorian Quechua origins presented by Louisa Stark (fully
published in 1985).
In a lengthy and complex discussion of the origins of
Amazonian Shipibo art, Lathrap unreservedly endorsed
Isbell's model which should therefore be viewed as
superseding his own south to north hypothesis (Lathrap et
al. 1985).
For the first time, Lathrap explicitly outlines
the assumptions central to virtually all of his work on
Amazonian culture history in a series of nomothetic
statements:
" ... on a pre-state level of organization ...
language does not spread geographically unless there is a
concomitant movement of appreciable groups of people"
(Lathrap et al. 1985:43).
Furthermore, he states that
populations (and their languages) do not move unless forced
to migrate due to population growth or environmental
degradation, and population movements are enabled by
technological innovations.
Finally,
" ... people cannot
preempt the land of their neighbors unless there is some
kind of competitive superiority either in sheer numbers,
more integrated organization, or in other factors giving a
97
military edge" (Ibid.).
These same assumptions provide the
foundations for Isbell's hypothesis.
Like Isbell, Lathrap et al.
(Ibid.:76,
83) places the
hearth of proto-Quechua in southeastern Ecuador.
Arguing
for Quechua influence in Curnancaya assemblages of the upper
Ucayali, Lathrap et al. utilize evidence from modern
Shipibo-Conibo narratives referred to as the "Inca cycle."
The Curnancaya style appearing on the Central Ucayali by A.D.
800 purportedly corresponds to the arrival of Quechuaspeaking intruders that subsequently dominated Panoanspeaking lowlanders (Lathrap et al. 1985:74).
The authors
regard Curnancaya as derived from the red-banded-incised
pottery recovered near Macas in Ecuadorian Amazonia.
Ostensibly providing further evidence of Quechua
affiliation, the applique faces on the upper portion of
Sivia jars purportedly link the Curnancaya tradition to
"plain brownware styles of the northeastern Peruvian Andes,"
and especially "the style known as Cuelap"
1985:68).
The authors conclude that,
(Lathrap et al.
"the hearth of lowland
Quechua expansion must be placed in the deep archaeological
sites of the Macas region of Ecuador, and the initial
expansion toward Peru must have been down the Upano-Zarnora
trough" (Ibid.: 83).
The composite Isbell/Lathrap model postulates a hearth
of proto-Quechua speakers on the major tributaries of the
Santiago River south of Macas, and southward population
98
movements along the base of the Central Andes.
Migration
occurred either as a wide swath or as a multi-pronged
advance which by A.D. 1 incorporated the eastern Peruvian
highlands (e.g. Kotosh Higueras and Chulla Pampa styles).
By A.D. 800 the Quechua-speakers reached the lowland Ucayali
and Apurimac drainages (e.g. Cumancaya and Sivia complexes).
The CB series constitutes the salient material evidence for
the Quechua spread into the "previously unoccupied niche on
the eastern slope of the Andes"
(Ibid.:76), while pottery
derived from the Ecuadorian red-banded-incised style
accompanies the intrusion of a Quechua-speaking population
deep into the eastern lowlands.
Criticisms bearing on southward Quechua population
movements target the conceptual linkage between language,
pottery and mythology assumed by Lathrap and his colleagues
(DeBoer and Raymond 1987:120-121).
DeBoer and Raymond note
that Stark's (1985) postulation of an eastern Ecuadorian
hearth for proto-Quechua provides a point of support for the
northern origins of the language, but they protest Lathrap's
liberal application of untested presumptions.
Lathrap's
(Lathrap et al. 1987:237-238) rejoinder argues tersely that
the proposed linkage of language and pottery styles is not
only justified, but that every archaeologist should be
manipulating linguistic data in conjunction with
archaeological data analysis.
An additional criticism of
Lathrap's original 1970 argument by Hastings (1985:555)
99
finds the suggestion of north and south population movement
"difficult to accept" and "counterintuitive" given the
imposing natural barriers presented by eastern slope
topography.
The larger scholarly community seems to have
shared little interest in these hypotheses and subsequent
debate.
Verticality, Colonization and the
Eastern Montane Forest
The previous section described how the eastern montane
forest has been envisioned as a threshold traversed by waves
of migrating populations moving upslope or downslope.
Yet
even prior to 1911 and the "discovery" of Machu Picchu it
was clear that pre-Hispanic populations once inhabited this
seemingly inhospitable region.
The immense fortified
settlement of Cuelap had already been visited during the
nineteenth century by Bandelier (1906), Raimondi (1942),
Middendorf (1895), Wiener (1884) and others.
Concerted
efforts to investigate archaeological occupations in and
around Peru's eastern montane forest began only three
decades ago.
The pioneering explorations of Bonavia (1964,
1968b, 1968-69, 1972b), Thompson (1973), Lathrap (1970) and
Lathrap's students Allen (1968), Isbell (1968) and Raymond
(1972) marked an era of growing interest in the montane
forests which had become known within the archaeological
community by a number of terms, especially the Peruvian
vernacular ceja de selva (Bonavia 1964) and ceja de montana
100
(Lathrap 1970).
Bonavia and Ravines (1968:153) first
conceived of the ceja de selva as synonymous with Tosi's
(1960)
"bosque muy humedo montana" or "very humid montane
forest" Holdridge life zone (3,800 - 2,500 m).
However,
Hastings describes a ceja de montana between 1,500 and 4,300
m, while Lathrap employs the term "ceja" even more liberally
to refer to habitats down to 300 m (1970:95).
Regardless of
the terms utilized, research in recent decades has
demonstrated that prehistoric montane forest settlement
typically straddles multiple ecological zones.
Unlike the somewhat anachronistic "civilizing
migration" hypotheses, colonization theories based upon
Murra's verticality model have been on the cutting edge of
Andean studies for the past two decades.
The precedent for
these studies was set by Bonavia's (1964, 1967-68, 1968b,
1972b) early investigations in the central and northeastern
montane forests.
His fieldwork provided the basis for his
synthetic overview of eastern montane forest prehistory, and
the formulation of a seminal working hypothesis, co-authored
with Ravines, that attributed montane forest settlement
complexes like Caballoyuq and Gran Pajaten to "tropas de
colonizaci6n agricola"
(Bonavia
ar.~
(agricultural colonization troops)
Ravines 1967:62; Bonavia 1968b:74).
The
theoretical statements expressed by Bonavia and Ravines
incorporated core elements of Murra's verticality thesis,
especially the "ideal economic pattern in which each group
101
had access to natural resources from other ecological
levels"
(Bonavia and Ravines 1967:62).
The authors cited
the Visita Hecha g la Provincia de Chucuito por Garci Diez
de San Miguel en el Afio 1567 (Diez de San Miguel 1964
[1568]) published by Murra, along with Nufiez del Prado's
(1958) study of the Q'ero.
Bonavia later directly cited
Murra's model of vertical settlement distributions to
support his interpretations of field observations in eastern
Ayacucho (Bonavia 1972b:29).
Estimations of dates, site functions and cultural
affiliations, coupled with a series of ecological
assumptions, constitute the foundations of Bonavia and
Ravines' theory.
The authors (1967:67, 1968:155-156, 157)
hypothesized that, prior to Inca imperial expansion, the
ceja de selva was increasingly utilized on a temporary basis
by highland communities expanding agricultural production
into eastern slope ecozones.
They characterized initial
colonization as "gee-cultural" and "a spontaneous step," and
categorically discounted demographic pressure as the
principal cause (1967:67).
Ceja de selva colonization by
these same populations became permanent under the direction
of the Inca state at the end of the fifteenth century
(Ibid.:67).
Bonavia and Ravines cite the Inca's interest in
controlling access to coca as the most probable motivation
(Ibid. :62) behind the sixteenth century expansion of the
agricultural frontier to its maximum eastern limit
102
(Ibid.:68).
The collapse of supporting infrastructure with
the Spanish conquest of the Inca empire purportedly
precipitated the abandonment of ceja de selva colonies which
were, by definition, dependent.
Analyzing broad patterns of prehistoric settlement
along the eastern frontier of the Andes, Bonavia and Ravines
consider a list of "ceja de selva" sites within varied
eastern slope environments.
They group these sites
geographically into southern, central and northern
categories that exhibit differing characteristics (Bonavia
and Ravines 1967:61).
Their southern group includes sites
surveyed by Isbell (1968) in the upper Inambari River valley
near the Bolivian border, and sites east of Cuzco examined
by Tello (1942:633), Fejes (1944) and Bingham (1930).
Sites
in eastern Ayacucho Department surveyed by Bonavia (1964,
1967-68, 1972b) constitute the central group, and their
northern group consists of Gran Pajaten (Bonavia 1968a),
Bambamarca, Tantarnayo and Chachapoyas sites.
Most of these
ceja de selva sites are not confined to the humid montane
forest, but occupy ecological zones on both sides of montane
forest timberline (Church 1994:282-283).
To Bonavia and
Ravines, the term ceja de selva refers to cultural as well
as environmental distributions.
In repeated statements, Bonavia and Ravines concur with
Tosi's negative assessment concerning the montane forest's
soils, their stability and their potential to sustain
103
permanent agriculture (e.g. Bonavia and Ravines 1967,
1968:153-154; Bonavia 1972a:92).
They specify terracing as
the critical soil conservation technique that allowed
agricultural production (Bonavia and Ravines 1967:67).
To
account for the perceived abandonment of settlements along
the eastern frontier, they cite the diminished capability of
maintaining the artificially created agricultural microenvironments after the fall of Cuzco:
"Following the
collapse, control over the peripheral centers naturally
disappeared and meant the abandonment of these places in
which, without certain conditions imposed by a strong
organization, it is practically impossible to subsist
(Bonavia 1968b:75).
Bonavia and Ravines do not directly address the
possibility of indigenous forest-dwelling ceja de selva
inhabitants prior to the proposed colonizations episodes.
They ambiguously mention "traditional groups settled on the
eastern slopes"
(1968:157), yet it remains unclear if these
populations represent their postulated first wave of
highland colonists maintaining temporary residences or to
some antecedent indigenous population.
Bonavia offers his
first explicit statements on the nature and antiquity of
pre-Inca settlement of the ceja de selva in later
publications (e.g. Bonavia 1978:398-399) in which he
endorses Lathrap's Quechua expansion hypothesis.
The descriptions presented below demonstrate that
104
virtually all attempts to model settlement of the eastern
montane forest have featured late prehistoric (post-Chavin
horizon) downslope migrations and colonization from adjacent
highland Andean regions.
Scholarly views of montane forest
prehistory contrast sharply with developmental scenarios
proposed for the Andean coast and highlands where Preceramic
Period populations began clustering in and around production
zones of·increasing importance to their developing
agricultural economies.
Preceramic Period occupations
within the continuous montane forests have not been
documented.
Rather,
incursions from adjacent areas are
invariably viewed as signaling the beginning of montane
forest prehistory.
Before describing individual hypotheses regarding
eastern montane forest colonization, some common problems
affecting the investigations and the data base should be
noted.
Because of the logistical difficulties of working in
unpopulated areas, very little survey, systematic or
otherwise, has been done.
The primary sources for
archaeological data include sketchy reports of brief
exploratory forays into the montane forest.
Rarely has
sustained research focused on cultural remains within the
continuous montane forest, and broad generalizations often
stem from surface observations at sites situated at the
subalpine-montane forest ecotone.
The archaeological evidence most frequently described
105
within and adjacent to the montane forest is architectural.
Ruined settlement complexes often consist of stone buildings
set on prominent ridge tops and surrounded by artificially
terraced slopes.
Surface scatters of artifacts are scarce
where sites are not heavily disturbed, and the harsh climate
rapidly erodes the surfaces and decorations of exposed
potsherds.
If nothing else, the following discussions
should demonstrate that, as a consequence of the
aforementioned problems, theoretical discussion of eastern
slope prehistory has far out-distanced the size and quality
of the data base.
The Southeastern Montane Forest
According to Bonavia and Ravines, the southeastern
montane forest (their southern ceja de selva division)
includes the upper Inambari Valley where Isbell noted two
architectural styles that he attributed to Inca and Late
Intermediate Period occupations (Isbell 1968:110).
dating of the Inambari sites
~s
His
largely conjecture, as
pottery and other temporally diagnostic features were rare
or absent.
While settlements were distributed between 4,000
and 1,800 m, agricultural terracing continued down to 800 m.
Isbell's views regarding the origin and antiquity of eastern
montane forest settlement (described earlier in this
chapter) were published several years after this fieldwork.
In the Urubamba River drainage east of Cuzco, Bonavia
and Ravines observe that imperial Inca architectural styles
106
attest to the cultural affiliations and dates of occupation
of sites listed for the southeastern montane forest.
Most
of the sites lack published descriptions, nor have they been
probed for earlier subsurface components.
The authors note
substantial functional variability between sites, and
suggest that defensive considerations account for their
strategic locations.
The southern pattern differs from that
of the central and northern groups which show more equivocal
evidence for Inca involvement in the hypothetical
colonization.
The Central Montane Forest
Bonavia's evidence for colonization in the central
"ceja de selva" was gathered during three brief field
excursions during 1963, 1964 and 1970 in eastern Ayacucho
(Bonavia 1964, 1967-68, 1972b).
Most of the data stems from
surface evidence gathered at settlements and dispersed
habitations featuring circular dwellings, vast terrace
complexes and occasional funerary chambers called chulpas.
Bonavia encountered low quantities of eroded sherds at these
sites, and even test pits yielded disappointing quantities
of diagnostic material.
Nevertheless, his three reports
document extensive and densely occupied settlements.
For
example, at Caballoyuq (3,570 m) Bonavia counted 289
circular structures in the portion of the site not concealed
by forest, and estimated that the settlement complex
extended five kilometers down the ridge (1964:13).
These
107
structures ranged between 3.90 and 7.70 min diameter.
The
overall distribution of settlements like Caballoyuq extends
from 4,000 m above the eastern banks of the Mantaro River
down through humid subalpine paramo and montane forest
eastern slope environments to perhaps 1,600 m (Bonavia
1972b:23, 29).
Evidence of prehistoric intensive
agriculture continues down to approximately 900 m where some
highlanders presently cultivate coca destined for transport
and eventual sale in Huanta (Ibid.).
Based upon his architectural observations and his study
of approximately one
~housand
sherds and a few whole and
reconstructed vessels collected during the three field
trips, Bonavia concluded that all of the sites reconnoitered
pertained to the same people ("gente")
68:256).
(Bonavia 1967-
An apparent lack of architectural evidence for
intra-settlement functional variation led him to suggest
that these were specialized agricultural villages
("villas").
Bonavia did not locate any unequivocal evidence
for chronology on any of his explorations. After his third
excursion to eastern Ayacucho however, Bonavia concluded
that " ... the villages and other agricultural works
discovered in the area are nothing other than products of
the same phenomenon: the Inca colonization of an
economically important area ... " (1972b:29).
The data, he
argued, conformed to Murra's model of vertical control
(Ibid.).
Bonavia recognizes the site of Taipi in the
108
subalpine paramo, which alone displays rectilinear
architecture, as the lone exception to his pan-regional
generalizations.
A Huari-style ceramic figurine from a
grave context (Bonavia 1967-68:Lamina 10), coupled with
Isbell's finds at the Huari site of Jargampata (Isbell 1971)
ostensibly supported Bonavia's suggestion that Taipi dates
to the Middle Horizon.
Consequently, the archaeology of the
region was thought to reflect the Inca emulation of
practices begun during the prior Huari imperial expansion
(Bonavia 1972b:26).
Colonization as Huari
Political Economy
Not far south from where Bonavia had described eastern
slope settlement between the middle and lower reaches of the
Mantaro River, two of Lathrap's students had begun survey
and excavations in the Pampas and Apurimac River valleys by
the late 1960s.
In the Pampas River valley, Isbell (1971,
1977) excavated at Jargampata, a site that he describes as
strategically situated at 2,500 m elevation between puna and
eastern lowland production zones, and interprets as an
administrative complex serving the Huari empire.
Raymond
(1972, 1976, 1985, 1992) meanwhile worked in the premontane
forests (500 to 1500 m) of the Apurimac River canyon below.
In his study of the Apurimac premontane and subtropical
forests, Raymond utilized ethnohistoric references and
ethnographically documented settlement distributions as well
109
as archaeological evidence to argue for "repeated,
systematic colonization of the lowland area" by highland
groups during the Middle Horizon (Raymond 1976:206, 1992).
Raymond begins by citing Murra's model and its premise
that "control of the eastern slopes of the Andes was a
critical component of the economic systems"
1992).
(1976:205,
He identified three "late prehistoric cultural
complexes" named Sivia, Quimpiri and Simariba that he
characterizes in terms of their ceramic assemblages,
settlement distributions and inferred economic orientation
(Raymond 1985).
Ceramic assemblages were found together and
in isolation, and correlations between the three assemblages
and site locations suggest Amazonian riverine, premontane
forest interfluve and highland-derived occupations
respectively.
Modern-day Panoans like the floodplain-
dwelling Conibo of the Ucayali provide an analogy for Sivia
complex peoples.
Shallow deposits on ridge and bluff tops
set back from the river contain Quimpiri complex deposits
believed to represent an orientation to interfluvial
resources similar to the present-day Campa's.
Quimpiri
sites line both sides of the river below 600 m but extend
farther upstream only on the eastern side.
On the western
side adjacent to the highlands, the Simariba complex is
found at sites often lying distant from the river between
600 and 800 m on the same valley slopes extensively
colonized by highlanders in historic times.
Several lines
110
of evidence suggest a history of hostile relations between
peoples of the Amazonian Quimpiri and Andean Simariba
complexes in competition for interfluvial agricultural land.
The interspersed Quimpiri and Simariba site distributions
are interpreted as evidence of fluctuating territories and
buffer zones.
Among the sites containing Simariba pottery, Raymond
located two especially large ones (over 30 hectares} with
architectural evidence for highlander occupation (Raymond
1985}.
At Palestina and Vista Alegre, stone alignments may
have served as footings for adobe houses and/or field
partitions while other constructions are clearly rooms with
walls up to two meters high.
Raymond cross-dates these
sites "on the basis of the few diagnostic ceramics,
settlement plan and architectural features" to the Middle
Horizon and hypothesizes that they served as "outposts of
the Huari empire"
(1985:42}.
He suspects that the Simariba
complex also includes Early Intermediate and Late
Intermediate Period pottery (Ibid., 1976:209}.
The absence
of Late Horizon material he attributes to the lack of
chronological control and research.
The Simariba pottery illustrated is a small sample from
the Granja de Sivia excavations (Raymond 1972: Fig. 53, 54a;
Raymond et al. 1975: 64a-d}.
Both "formal characteristics"
and "stylistic affinities" are said to ally this complex to
Huari ceramics in the neighboring highland Andes (1985:42,
111
1976:208, 1992:28).
Raymond (1985) notes the correspondence
between Simariba site locations and the ideal coca growing
elevations between 600 and 900 m.
Palestina and Vista
Alegre he interprets as "frontier settlements built to
define a territorial boundary, administrate coca production
and trade with tropical forest peoples (1988:298, 1992:29).
He reasoned that these sites probably linked the highland
Huari empire to Panoan-speaking lowland floodplain-dwellers
20 km away at the site of Granja de Sivia, which in effect
afforded Huari access to commodities circulating within
lowland riverine trade networks.
Raymond offers two summary positions regarding the
origin and antiquity of ceja occupation: one specific to the
Apurimac valley, and a second broadly generalizing.
Based
upon his fieldwork, Raymond believes that there is little
evidence that highland utilization of the forested slopes of
eastern Ayacucho pre-dates the Middle Horizon (1988:298).
However, he points out that the Ayacucho site complexes
examined by Bonavia remain undated (1976:210-211).
The
distribution of some poorly preserved settlements with
"circular, oval, D-shaped and rectangular structures" of
unknown cultural and temporal affiliations reportedly
resembles the Simariba distribution (1992:27), but these
were only cursorily examined.
Raymond's fieldwork included
a "survey transect" up the western side of the Apurimac
River valley in which he found "no signs of prehistoric
112
settlement" between 1,500 and 3,000 m (1985:40}.
Although
he does not specify its exact location, his transect was
probably not far south of Bonavia's (1972b} route from the
middle to the lower Mantaro River courses.
The lack of
evidence for "settlement" contrasts with the abundant
evidence described by Bonavia, and the evidence recovered by
Hastings (1985} farther north in the Tarrna River valley.
Raymond's broad characterizations of tropical forest
settlement systems outline features of three generalized
types (Raymond 1988}, apparently based upon his
interpretation of the Sivia, Quimpiri and Simariba cultural
complexes.
In addition to Amazonian floodplain and
interfluvial settlement systems, Raymond notes a third
occupying the ceja that,
"is basically an extension of the
highland agrarian system" and,
"is not self-sufficient but
is tied closely to social, political and economic networks
in the highlands"
(Ibid.:286).
He concludes that such
"colonization of the steep terrain of the upper montana
apparently began during the Early Intermediate Period"
(1988:296), and cites Lathrap and Isbell's assessment of
eastern slope prehistoric time depth.
Raymond theorizes
that "the expansion and contraction of highland colonies in
the Montana corresponds to the rise and fall of imperial
states in the highlands" {1976:211, 1992).
Yet even during
the post-Huari period of political fragmentation and
territorial contraction, highland polities maintained a
113
colonial presence in the forests because of the continued
importance of coca and lowland produce (1988:299).
His
discussions of eastern slope human occupations include no
specific references to indigenous ceja populations.
To summarize Raymond's view, the intensity of eastern
slope occupation, as well as Andean-Amazonian interaction
corresponds closely to highland political developments.
Lathrap assumed Raymond's position by considering the
fleeting emergence of social complexity interpreted at
lowland Cumancaya sites as a secondary development
stimulated by interaction with the Huari empire (Lathrap et
al. 1985:89).
Bonavia (1991:380) also subscribed
unreservedly to Raymond's interpretations, and especially
his dating of systematic colonization of the eastern slopes
to the Middle Horizon.
Finally, Raymond views the Inca
presence in the ceja as imperial conquest and co-option of
pre-established highland-based political and economic
systems (LeMoine and Raymond 1988:125-126).
Verticality Colonies
in the Tar.ma Canyon
Sites of the lower Tarma River valley farther north
also lie within Bonavia and Ravines' central ceja de selva
group, yet they are physically separated from Bonavia's and
Raymond's Apurimac valley sites by the substantial mountain
range dividing the Apurimac/Ene and the
Tarma/Chanchamayo/Perene river watersheds.
This same range
114
combines with local topographic features to create a "weak
rain shadow"
(Hastings 1987:147) and considerable ecological
complexity in the Tarrna River valley.
Hastings' work
focuses on the distribution of "ceja de montana"
( 1, 500-
4,300 m) settlements and their ecological and temporal
contexts in order to reconstruct economic and political
systems across eastern slope "vertical" gradients.
A wide
transect joining adjacent portions of previously surveyed
highland puna and lowland tropical forest constitutes
Hastings' study area (1985:270, Map 6-1).
Thus he is
presented with the opportunity to place his investigations
within a broad regional context, and to examine evidence for
prehistoric ethnic and culture area boundaries.
Hastings is
clearly influenced by Murra's work as the explicit
incorporation of "verticality" as a conceptual framework for
his study attests (1985:4).
He also attempts to interpret
evidence for economic articulation between the Andean
"vertical" system and neighboring Amazonian economies.
Based on interpretations of survey and excavation data,
Hastings reconstructs two settlement patterns corresponding
to two phases of occupation in the lower Tarrna valley
montane forest termed the Tranca and Paraupunta Phases.
Distinctive architectural styles and statistically-derived
ceramic complexes characterize each of the two
patterns/phases, which fall relatively late in prehistory.
His lower Tarrna valley work failed to turn up unequivocal
115
evidence of human occupation prior to the late Middle
Horizon.
Therefore, Hastings surmises that the eastern-most
distribution of the Early Intermediate Period Huacrapuquio
ceramic complex in the upper Tarma valley (down to
approximately 2,700 m, yet absent below present montane
forest treeline) comprised Andean civilization's eastern
frontier prior to the first occupation of the lower Tarma
valley (1985:553).
The Tranca Phase, beginning between A.D.
800 and 900, is said to represent "the first substantial and
widespread colonization of the Lower Tarma ceja de montana"
or montane forest (Ibid. :693).
The Tarma Phase occupation features ridgetop
settlements with square and rectangular masonry buildings,
many with internal dividing walls, and pottery assigned to a
complex termed Malambo.
The Tranca settlement pattern
(defined by 33 sites) emphasizes the subalpine zone near
treeline, the ridge-tops between 2000 and 2300 m protruding
into the forest, and to a lesser extent the slopes below.
The inordinately large site of Tranca in the subalpine zone
may have served as an administrative hub overseeing economic
activities in the region.
Hastings suggests that the paramo
emphasis may indicate the dietary importance of potatoes,
while relatively small herds could be tended nearby.
Lower
ridgetop settlements provided access both to lowland produce
and to exchange with Amazonian groups.
Hastings' survey
revealed an eastward "nearly continuous site distribution
116
from the culturally Andean Tranca pattern into the more
Amazonian-affiliated Camonal [ceramic complex] zone of
occupation" (Ibid.: 697).
Hastings did not recover stratigraphic evidence to
bolster his argument for the Tranca Phase's temporal
priority within his two-phase sequence.
A few corrected
radiocarbon measurements suggest dates of A.D. 855 to 1235
for the Tranca Phase (Ibid.:682).
However, Hastings admits
that efforts at both relative and absolute dating of the
Tranca Phase encounter substantial problems that hinder
precise temporal placement.
To establish a basis for the
relatively early placement of Tranca, he cross-dates the
associated Malambo pottery complex using comparisons with
Early Intermediate Period and Middle Horizon pottery from
the upper Mantaro valley.
He also notes the presence of a
Huari style cup or kero recovered from one site tentatively
assigned to the Tranca Phase (Ibid.:792).
Hastings suggests that the Camonal Complex was
Malambo's contemporaneous Amazonian counterpart during the
Tranca Phase.
The Malambo and Camonal Complexes overlap
stylistically sharing such attributes as specific rim forms
and strap handles (1985:520, 589, 674), but comparisons
between the two are also hampered by problems of inadequate
chronological control.
Hastings characterizes Camonal as an
"Amazonian-Andean hybrid"
(1985:589), yet he is aware that
it bears no explicit resemblance to any one Amazonian style.
117
He also grants the possibility of greater antiquity for the
Camonal Complex (Ibid.:678).
Unfortunately, ceramic
sequences in both the neighboring highlands and lowlands
also suffer inadequacies in chronological precision.
Hastings declines to promote a single hypothesis
regarding the origins of Tranca Phase "colonists"
(1985:721).
He observes that the predominant pottery,
Malambo Luster Ware, appears to have been locally produced.
Lack of clear antecedents in the surrounding regions leaves
Malambo pottery "an anomaly"
(Ibid.:553).
Hastings notes
that apparent similarities to pottery assemblages from the
Chupachu territory of the Pachitea Andina "raise intriguing
questions about possibly widespread colonization north and
south along the eastern Andean margins"
view,
(Ibid.:721).
In his
the possibility of these two complexes representing
"variants of a single general tradition prevalent for a time
on the eastern flanks of the Central Peruvian Andes"
warrants investigation (Ibid.:554).
Hastings in effect
departs from his verticality framework to entertain the
possibility of a northern progenitor for the Malambo and
Pachitea Andina complexes.
He observes that similarities
between the two are to some degree paralleled by
architectural similarities (Ibid.:652).
In addition to the Tranca architectural and settlement
patterns, Hastings identified another "architectural
pattern" labelled Peruruhuay.
It is characterized by small,
118
semi-square, semi-subterranean structures.
Although the
sites remain undated, ceramics are said to be "weakly
diagnostic of the Malambo Complex"
(Ibid. :635).
The
settlement distribution of the relatively small site sample
approximately matches the Tranca pattern (Ibid.: Table 115).
Neither the Peruruhuay nor the Tranca architectural
patterns have similar highland counterparts (Ibid.:642).
The San Blas ceramic complex and an architectural
tradition featuring circular masonry buildings characterizes
"a second period of colonization" of the lower Tarma River
valley during the Late Intermediate Period and Late Horizon
termed the Paraupunta Phase.
Because San Blas Red-on-Buff
"seems to develop earlier on the puna than in the valleys"
(1985:724), and because of similar architecture on the Junin
plateau, Hastings interprets the Paraupunta Phase as the
result of colonization from the adjacent puna (Ibid. :727).
Various lines of evidence lead Hastings to identify
Paraupunta settlement with the historically documented
Chinchacocha, best known as puna herders.
The Paraupunta
settlement pattern is similar to the Tranca pattern with one
clearly dominant site (the site of Paraupunta).
Less
ambiguous spatial and temporal resolution facilitates
Hastings' interpretation of the Paraupunta economic system
in terms of Murra's framework (Ibid.:731-732; Parsons and
Hastings 1988:218).
More masonry structures per hectare,
rather than a greater number of sites assigned to the phase,
119
ostensibly indicates population growth (Ibid.:701).
Fortification and defensible locations become a greater
concern.
The mixing of Tranca and Paraupunta assemblages at
some sites suggests to Hastings that the Paraupunta
settlement and subsistence system's replacement of the
Tranca system was "a gradual process"
(Ibid. :727).
Hastings describes Paraupunta Phase Andean and coeval
Amazonian settlements as separated by as much as 1000
vertical meters (Ibid.:702),
and Paraupunta sites along the
Andean-Amazonian cultural interface as heavily fortified
(Ibid. :730).
However, San Blas sherds in Amazonian site
contexts dominated by the coeval Amazonian Chanchamayo
Complex, and Chanchamayo sherds in San Blas contexts
demonstrate exchange relations.
Close resemblances between
the Chanchamayo Complex and Raymond's Quimpiri complex
prompt Hastings to propose a strong association between
these two styles and the historic Campa or Amuesha.
The
overlapping distribution of the Camonal and Chanchamayo
complexes precludes interpretations assigning either to
floodplain or interfluvial ecological niches.
Hastings also
reports that sites with remains of these ceramic complexes
invariably lack stone architecture.
Hastings concludes that the Tarma ceja de montana
occupations, especially during the Paraupunta Phase, show a
continuous distribution of occupation down the eastern
slopes spanning 3,000 vertical meters.
His interpretations
120
recognize a substantial degree of political and economic
autonomy obtained by the lower Tarma valley subgroup, and
increasing economic centralization overall, yet he considers
the possibility that the Tranca and Paraupunta systems
developed locally within the ceja de montana unlikely:
It might, however, be misleading to perceive structural
changes at this subgroup level as exemplary of cultural
evolution in the ceja de montana. The emergence of such
subgroup centralization in the Andean margins could be as
much an externally introduced change as a locally evolved
process. The lower Tarma ceja de montana is now and always
has been a remote, sparsely populated area with a foul
climate for permanent occupation. It should not be
surprising if local communities were more often recipients-or direct outgrowths--of regionally evolving change rather
than instigators of such change. Cultural developments in
the Lower Tarma can only be seen as part of a regional
system, and for most of prehistory these developments have
probably been frontier manifestations of that system
(1985:730-731).
Arguing against the possibility of local evolutionary
processes, Hastings asserts that the "drastic shift in
settlement locations" at the beginning of the Paraupunta
Phase reflects "economic changes not previously underway
during the Tranca Phase"
(Ibid. :731).
He believes that a
perceived lack of developmental continuities from one phase
to the next illustrates an absence of internally generated
evolution.
The Northeastern Montane Forest
Bonavia's brief field investigation at Gran Pajaten in
1966 provided much of the evidence for tropas de
colonizaci6n agricola that figured prominently in the
theoretical statements co-authored with Ravines (Bonavia and
Ravines 1967; Bonavia 1968a).
However,
the first scientific
121
reports describing Gran Pajaten's archaeology were published
by Pimentel (1967) and Rojas (1966, 1967) who directed
various aspects of the government-sponsored 1965 and 1966
expeditions to the site.
At 2,850 m, Gran Pajaten is
situated on a ridge top above the Montecristo River, a
tributary of the Abiseo River which in turn empties into the
Huayabamba (also Huallabamba) and Huallaga rivers.
The site
is relatively small, yet it has not been completely mapped
nor the total number of its constructions tallied.
While
only eighteen buildings appear on published maps (Pimentel
1967; Bonavia 1968a), some of the circular stone structures
are elaborately embellished with cornices, stone mosaic
friezes,
tenoned heads, and large staircases serving
elevated entryways.
Based on information gathered during both expeditions,
Pimentel and Rojas (a student of Tello) concluded that Gran
Pajaten belongs to the Huaylas culture that, according to
Tello (1942:712-713), surrounded the Marafion River basin
during the third epoch dated A.D. 800 - 1321.
They both
recognize an Inca (Late Horizon) component at the site
identified by a copper knife and a fragment of Inca-style
pottery.
Rojas (1967:17) evaluates their significance:
These two objects are strong evidence for contact with the
Inca empire.
Presumably this contact was of a commercial
nature, since the architectural elements lack any trace of
Inca influence ... The influence of this ancient community
must have been great, perhaps dominating a society which
occupied much of the upper forest land.
Thus Rojas envisions Gran Pajaten as a local and independent
122
development that presumably began with the arrival of
Tello's "primordial" Amazonian migrants.
Ultimately, it
participated in Late Horizon Inca-orchestrated interaction.
Although Bonavia participated in the same governmentsponsored investigation in 1966, he did not address Rojas'
hypothesis.
Architectural observations, surface collections
and small-scale excavations in and around two of the site's
most elaborate structures (Building Nos. 1 and 2} provide
the basis for Bonavia's independent conclusion that Gran
Pajaten pertains to the Late Horizon.
Bonavia postulates
that, like the eastern Ayacucho sites, Gran Pajaten
represents a process of eastward expansion directed by the
Inca.
His theoretical statements emphasize evidence gleaned
from a sample of 1,645 sherds, 400 of which derived from a
controlled 2 x 2 m excavation within Building No. 1.
These
sherds were classified into three types labelled A, B and C.
Type A predominates and consists of a single locallyproduced coarse ware featuring two distinct decorative
modes.
To the local decorative style based on applique
Bonavia gave the name "Abiseo," while he referred to a
polychrome-painted decoration as "Inca-derived."
Necked-
jars with flaring rims and shallow bowls dominate the
inventory of Type A vessel shapes which Bonavia classified
according to categories established for Inca pottery of
Cuzco by Pardo (1957}.
A coarse ware distinguished from
Type A chiefly by paste characteristics, Type B contains 14
123
sherds.
Three sherds of fine kaolin ware pottery constitute
Type C.
Neither Type B nor C figures in Bonavia's
interpretations.
Based upon his ceramic analysis, Bonavia concluded that
the two Type A stylistic modes represent the co-existence of
two cultures at Gran Pajaten {Bonavia 1968a:70, 157-158).
The Abiseo style "corresponds to a traditional style, linked
with the highlands and that represents a local colonizing
culture"
(Ibid. :74).
The Inca-derived style "would be the
style of prestige that would indicate the Incaic imposition"
(Ibid.).
Bonavia regarded the Late Horizon date for the
settlement as clearly demonstrated by the Cuzco-related
vessel shapes and by the "Inca" polychrome decoration on
sherds (Ibid.:107, Lamina 12).
Bonavia recognized that evidence for various phases of
occupation might be revealed by future investigations at
Gran Pajaten (Ibid.:66).
Building No.2 in particular was
believed to show evidence of refurbishment by the haphazard
use of carved, decorated sandstone slabs that presumably
carne from older, dismantled structures.
As a "working
hypothesis", Bonavia suggested that the sandstone slabs
which served as media for mostly curvilinear motifs, like
the Abiseo style pottery decoration, represented a prior
site occupation by people of a local highland cultural
tradition (Ibid.:73-74).
The rectilinear slate mosaic
decoration may have represented the imposition of the "new
124
norm," manifest also in the presence of the Inca-derived
pottery style.
In effect, Bonavia was suggesting the
possibility of pre-Inca construction at the site by a local
"colonizing" highland culture.
In a recent publication,
Bonavia (1991:527) dates colonization of the northeastern
montane forest to the beginning of the Late Intermediate
Period.
Bonavia's monograph on Gran Pajaten's archaeology
generated critiques by two reviewers (Ravines 1967-68:335336; Isbell 1970:237) who questioned his methodology and his
conclusions.
Ravines noted that both the settlement layout
and the ceramic sample showed potential for the
differentiation of more than one temporal unit.
However,
Bonavia's analytical emphasis on technological aspects of
the pottery had rendered his study insensitive to the
detection of chronologically significant attributes.
Ravines' criticism was echoed by Isbell who also complained
that Bonavia's use of Pardo's vessel shape classification
masked insights on ethnic groupings as well as temporal
developments intimated by changing vessel shape preferences
within Bonavia's stratigraphy.
Isbell (1974:139-140) his
aforementioned Quechua expansion hypothesis as an
alternative to Bonavia and Ravines'
(1967) postulated Late
Horizon west to east expansion.
Bonavia's interpretations of Gran Pajaten satisfied
other scholars.
An excerpt from Moseley's (1992:190) recent
125
synthesis of Peruvian prehistory shows support for the
colonization interpretation:
Refinement of high montane adaptations facilitated a
northward thrust into the Ecuadorian sphere of tropicalinfluenced sierra populations, known as Kuelap. These
colonized mouncain settings were in a sense more tropical
than Andean.
Remarks to a journalist by Meggers implicitly question the
need for further research:
A full-scale investigation will be interesting, but of
predominantly local significance ... There is no indication
that the culture at Gran Pajaten spread to other areas. The
inhabitants of the city were probably recipients of cultural
innovations from Inca centers in southern Peru, not
originators (Science News 1985: 117).
The views of Moseley and Meggers clearly reflect their
respective "ruling frameworks" that dictate responses
characterizing montane forest sites like Cuelap and Gran
Pajaten as highland-derived implants.
Alternative Viewpoints
The interpretations of some archaeologists who have
worked in the northeastern montane forest Utcubarnba Valley
over recent years coincide with Langlois' suggestion of far
northern linkages, although they argue unequivocally for
autochthonous origins and in situ development (Ruiz 1972;
Shady 1987a, 1987b) .
Based upon analysis of pottery remains
excavated from Cuelap, Ruiz emphasizes similarities between
Cuelap's earliest styles and El Salado Phase styles from
lowland Bagua, Garbanzal pottery from far north coastal Peru
and Guangala pottery from southwestern Ecuador (1972:181).
Ruiz and Shady work under the assumption that settlements
126
between the Huallaga and Marafion Rivers like Gran Pajaten
and Cuelap "give evidence for singular and distinctive
Central Andean cultural expressions in their urban and
architectural design as much as in their diverse known
manifestations; in no way can they be explained as derived
from colonizing advances from the nearby highlands" (Shady
1987a:86).
Also in reference to the northeastern montane forest,
Lumbreras observes that: "this zone as a whole exhibits
closer relationships with the little-known northern Andean
area than with the Central Andes ... " (1974:149).
He
characterizes this area as the "focus of high-level social
development ... where peoples flourished with a high level of
development that now we measure partially with architectural
accomplishments as important as those of so-called Pajaten
in the northeast of PerU with its neighbors and relatives
Cuelap and others in Chachapoyas ... " (Lumbreras 1981:31).
Although Lumbreras does not directly address the issue of
human occupation's time depth in the montane forest, he
clearly envisions in situ development.
In a more recent publication, Morales (1993:646-653)
assumes an even more extreme posture favoring autochthonous
cultural development within the greater northern and eastern
montane forests.
By labeling this region Amazonia Andina,
or "Andean Amazonia," Morales implicitly contradicts
Kauffmann's highland-centric interpretations of the Andes
127
Amaz6nicos.
In Morales' view, ancient montane forest
settlement cannot be comprehended utilizing Andean or
Amazonian perspectives.
He states that not only do
scenarios eliciting conquest and/or colonization impose
inappropriate analytical frameworks, but that applications
of Andean periodization schemes presupposing Huari and Inca
hegemony are spurious.
Rather, Amazonia Andina societies
engaged in interregional interaction independently and
irrespective of political developments in the neighboring
highlands since the late Preceramic Period.
Cuelap, he
regards as the fortified center of an upper Utcubamba Valley
state-level polity (Ibid.:652).
During later prehistory,
encroaching Andean empires were forced to adapt to local
montane forest social, economic and political conditions.
Alternative views of local development are typically
expressed in unambiguous terms, yet provocative assertions
are seldom buttressed by more than perfunctory references to
monumental architecture at Gran Pajaten or Cuelap.
Not only
is the montane forest data base rarely addressed in
systematic fashion, but archaeologists rarely address each
other's hypotheses.
Langlois' hypothesis of northern
origins has been ignored by all but the European scholars.
Bonavia (1978, 1991) and Hastings (1985) refer to Lathrap's
Quechua-expansion hypothesis despite its replacement by
Isbell's, and Tello's postulated migrations from the
northeast is seldom cited by anyone concerned with montane
128
forest settlement.
As a result, potentially viable yet
conflicting hypotheses for montane forest settlement have
tended to accumulate, rather than to supersede one another.
This chapter has summarized numerous distinct
hypotheses that conceive of the prehistoric montane forest
as an empty migratory corridor, a migratory cul-de-sac, a
sparsely populated buffer zone, a remote ecological zone
servicing highland polities and an archaeological culture
area.
Isbell's, Moseley's and Morales' interpretations
could scarcely be more at odds with one another.
The
following chapters will present documentary and
archaeological data confirming the presence of autochthonous
and autonomous polities in the northeastern montane forest,
and evaluate each of the aforementioned hypotheses in light
of the new evidence.
CHAPTER 3
THE PATAZ-ABISEO STUDY AREA:
PRESENT AND PAST
The Chachapoyans, whom I thought to be the builders of
the [Gran Pajaten] ruins, promised to be an exciting new
culture of Amazonia. Could they have been indigenous or had
they come from the eastern lower jungles? Did they come
from the northern Andes to the west? Only further
exploration and study could answer these questions. The
site appeared to be a ceremonial center or important temple
building of a civilization far more advanced than that of
the builders of Vileabamba the Old. The circular buildings,
more difficult to construct than rectangular structures,
were vastly superior to anything I had seen in the montanas
of southern Peru.
If the ruins did belong to the
Chachapoyas civilization, then they certainly had a culture
far higher than that of the Incas during the same period.
(Gene Savoy 1970:143)
The previous two chapters outlined the hypothetical
prehistoric migrations most frequently cited as salient
developmental events in eastern montane forest prehistory.
This chapter introduces the study area and provides
background information on present and past cultural
geography and economic activities.
The study area straddles
the high cordillera separating the upper Marafion and central
Huallaga river watersheds (Fig. 4), and includes
northeastern montane forest life zones.
Both rivers
originate in the snow-capped peaks of Huanuco and Pasco
departments and flow north-northeast roughly parallel to one
another until the Huallaga swings east and exits the Andean
129
130
foothills near 7o south latitude.
eastward farther north at
so
The Marafion veers
south latitude.
More precisely, this study will focus upon the narrow
strip of intervening cordillera between 7c and go 30' south
latitude,
just south of where the two rivers diverge.
At
these latitudes, the forested foothills of the Andes flank
the narrow central Huallaga floodplain, but the entrenched
Marafion does not emerge from its rocky inter-Andean canyon
until it approaches its confluence with the Utcubamba River
much farther north.
Scholars usually conceive of societies
above the eastern banks of the Marafion, and on the crest of
the Marafion-Huallaga divide, as "Andean"
(e.g. Rowe 1946:Map
3), while those along the west banks of the Huallaga are
regarded as "Amazonian"
(e.g. Steward 1948:Map 5).
However,
little is truly known of ancient population distributions in
the area, especially within the unexplored montane forests
of the Huallaga valley.
The immensity of the Marafion Canyon is truly aweinspiring.
It is over 3,000 m deep from the summits of the
Marafion-Huallaga divide (4,500 m) to the valley bottom
(1,200 m), and approximately 40 km wide.
Less than 20 km
separate the course of the Marafion and the top of the
divide, rendering an average gradient between 15 and 20
percent for the Marafion valley slopes.
Along this stretch
the Marafion is not a reliable means of transportation as it
is navigable by canoe or balsa raft for short distances only
131
during the brief dry season.
Its gradient within the lower
canyon is three percent (Pefiaherrera del Aguila 1986:131).
Below the mouth of the Utcubamba, the Marafion enters a 150
km stretch of cataracts (or pongos), punctuated at beginning
and end by the great pongos of Rentema and Manseriche.
None
of the Marafion's east-bank tributaries south of the
Utcubamba are navigable.
At the same latitudes, the Huallaga River lies much
lower at approximately 350 m.
Its banks are 70 to 80 km
distant from the Marafion-Huallaga divide.
While the
gradient of the Huallaga valley's slopes averages around
five percent, most of the vertical drop occurs abruptly in
the precipitous upper montane forests.
The river descends a
relatively mild gradient of approximately one percent
(Ibid. :143).
Historically, canoes and balsas plied the
central Huallaga below Tingo Maria (elevation 650 rn)
Casas 1930, 1934; Pefiaherrera del Aguila 1987).
{de las
From Tingo
Maria down to the mouth of the western tributary Huallabamba
River (also Huayabamba), riverine traffic contends with
numerous rapids including the pongos of Cayumba, Sabaloyaco
and Cachihuafiusca (Raimondi 1876:423).
Below the
Huallabamba, the widening river has accumulated a floodplain
up to 5 km wide {DeBoer 1984), and traffic flows relatively
unimpeded until reaching the next series of whitewater
obstacles below Shapajo {near Tarapoto) .
Below Chazuta, the
river passes into the lowlands via the great pongo of
132
Aguirre.
Some small pongos along the central Huallaga have
been "pacified" with dynamite in historic times (de las
Casas 1930:182-183).
Many of the west-bank tributaries of the Huallaga south
of 7c latitude can be navigated for short distances on their
lower reaches.
However, Weberbauer (1945:98) characterizes
the largest of these, the Huallabarnba, as "the axis of a
great fluvial system."
The Huallabarnba River may be
ascended in canoe at least 60 km to the abandoned mission at
Jesus de Pajaten (Savoy 1970), and perhaps up to 100 kms
(Torres Calderon 1903:296).
include, from north to south,
The Huallabamba's tributaries
the Huarnbo, Jelache,
Huabayacu, Bombonaje, Jepelache, Pajaten and Abiseo (or
Apisoncho) rivers.
Since the beginning of this century,
geographers and engineers have recognized the Huallabamba as
an extraordinary westward intrusion of lowland waterway
that, if reached by road from the Andes, would most
efficiently establish a commercial artery connecting the
economies of Andean northwestern Peru and the Amazon River
port city of Iquitos (Torres Calderon 1903; Weberbauer 1920;
de las Casas 1934).
Despite such early interest, the
Huallabamba drainage remains lightly occupied and virtually
unexplored by outsiders.
Politically, the study area corresponds to the
districts of Pataz (La Libertad Department) on the west side
of the divide, and to Huicungo District (San Martin
133
Department) on the east side.
Manachaqui Cave lies at Pataz
District's northern border with Condormarca district.
Since
1983, the Abiseo river drainage in Huicungo District has
been administered as the Rio Abiseo National Park (RANP) .
Together the Pataz and Huicungo districts (including the
RANP) encompass a west to east transect beginning at the
bottom of the Marafion canyon, spanning the cordillera and
ending in the Huallabamba valley lowlands.
This wide
transect will be referred to as the "Pataz-Abiseo area"
(Fig. 5).
The eastern addition of Pachiza District on the
banks of the Huallaga completes this transect.
The borders
of these modern political-geographic units are not expected
to coincide with past cultural or ethnic distributions.
Studies of modern, historic and prehistoric highland
Andean economies rely on classification schemes to
differentiate ecological zones that served to provide
desired resources.
I begin this chapter by detailing a
scheme of biogeographic zonation based on botanical
inventories compiled in the Pataz-Abiseo area, and I
describe modern land-use within each zone.
Analogies
derived from modern ethnographic sources can be
indispensable for archaeological interpretation as long as
their utility is evaluated rather than assumed.
Ethnohistoric evidence presented next portrays rapidly
changing demographic and interregional economic conditions
during the Colonial Period.
A third section marshals
134
archaeological evidence for pre-Hispanic settlement
distributions and boundaries in the Pataz-Abiseo area, and
relates them to their biogeographic (ecological) contexts.
While this chapter provides basic background
information regarding the study area environment and socalled Chachapoyas culture, it also emphasizes change and
the need for an informed historical perspective in order to
approach problems of prehistoric economic activities in the
area.
As the first section on modern land-use demonstrates,
the study area is relatively isolated from the mainstream
Peruvian national economy.
Readers informed about Peru's
modern political and economic geography may regard the
selection of this area for studies of prehistoric highlandlowland relations as misguided.
In recent years, opponents
of llregionalizaci6nll and the proposed administrative union
of
II
Andean
II
La Libertad and
II
Amazonian
II
San Martin
departments point to the complete lack of economic
articulation across the natural barriers separating two
worlds with disparate natural, social, economic and
political histories.
At first glance, there is no reason to
suspect that conditions have ever been otherwise.
However, a review of early Colonial Period documents
demonstrates that the study area's present economic
conditions are not indicative of those of its past.
fact,
In
the Pataz-Abiseo area was a locus of particularly
intense highland-lowland exchange, and several historical
135
processes and events that contributed to its present
"marginal" economic status can be identified.
By
contrasting present and Colonial Period information on
cultural geography with what is known archaeologically of
prehistoric settlement distribution, we may reconstruct
major economic and demographic transformations.
The early
Colonial Period stands out as a time of rapid, even chaotic,
change although thin strands of economic ties survived only
to dissolve slowly but completely by the early twentieth
century.
The following narrative will demonstrate how modern
ethnographic observations may fail to illuminate prehistoric
conditions, and thereby illustrate the weakness of
archaeological studies that rely uncritically upon
ethnographic analogies.
Present economic articulations {or
lack thereof) between Andean highland and Amazonian lowland
regions along the length of the eastern slopes cannot be
mechanically projected into the prehistoric past.
The
widespread idea among archaeologists of a highland-lowland
dichotomy reflects the current national social, political
and economic reality of Peru.
A brief summary will
highlight points interpreted as crucial for understanding
Pataz-Abiseo history and prehistory, and the dynamics of
long-term regional economic transformation.
136
Life Zones and Land-Use in the
Pataz-Abiseo Area
Natural and anthropogenic fragmentation of the montane
forest renders the Marafion-Huallaga divide an ecologically
complex orographic region.
From south to north summit
elevations diminish, and fragmented montane forest covers
increasing percentages of the valley slopes.
The east-west
distribution of ecological zones found between the
headwaters of the Rio Utcubamba and the modern town of
Huancaspata (including the study area) conforms to Brush's
"compressed" type of zonation (Brush 1977) .
The repetitive
sequence of dry Marafion-side to moist Huallaga-side life
zones is most notably broken where the alpine grasslands
broaden north of Bolivar, and south where the intermontane
valley created by the Cajas River partitions the cordillera
near Tayabamba.
Unfortunately, our understanding of eastern
slope environments is hindered by the lack of field studies
and meteorological stations which would normally provide
useful information for evaluating the agricultural potential
of given life zones.
It is my intention in this study to clearly distinguish
between kinds of site environments, yet recognize that the
ecotone separating the highland grasslands from the montane
forests is wide and complex, and that it shifted over time.
Andeanists studying historic and prehistoric regional
economies on the eastern slopes frequently apply taxonomies
derived from folk tradition (e.g.
Pulgar Vidal 1987),
137
western biological sciences (e.g. Holdridge 1967) or hybrid
combinations of these to differentiate study area ecological
zones (Mayer 1985:49-50; Church 1994:282).
In the Pataz-
Abiseo area, information regarding pre-Hispanic production
zones cannot be obtained through modern land-use analogies
because all but the western-most portion of the area is
virtually uninhabited today.
Furthermore, an
ethnographically-derived zonation scheme would be of
questionable validity for dealing with past situations in
which unknown socioeconomic systems organized production of
unknown crops.
Because it does not impose biases on
interpretation of pre-Hispanic land-use patterns, Young's
(1992b, 1993) scheme developed for the Pataz-Abiseo area
utilizing the Holdridge system provides the foundation for
archaeological investigations (Church 1994).
Young's (1990) classification of ecological zones
encountered in the west to east transect between the Marafion
and Huallaga Rivers was developed in part to examine issues
of conservation.
His categories have heuristic value for
biological inventories and biogeographic taxonomy, but they
do not necessarily coincide with pre-Hispanic eastern slope
production zones utilized for potatoes, maize and coca.
Exactly how the extreme environmental conditions within the
continuous montane forest of the central Huallaga watershed
affected the distribution of pre-Hispanic production zones
requires further investigation.
Young observed,
inventoried
138
and collected specimens of plant species encountered within
Pataz District life zones, and from the upper reaches of the
Rio Abiseo National Park from 4,200 m down to approximately
2,300 m (Leon et al. 1992; Young 1990, 1992b, 1993; Young
and Leon 1988, 1990).
Therefore, his descriptions cover the
ecological zones from which the archaeological data
pertinent to this study derive.
Within the Pataz-Abiseo area, Young classifies
environments into four "macro-ecological zones"
6)
(Figs. 5 and
loosely corresponding to the folk taxa used locally:
temple, Quechua, jalca and puerta del monte (Young 1993).
From west to east these are the Dry Forest, Moist Montane,
Tropical Alpine and Montane Rain Forest zones (Ibid. :271).
For each of the four macro-ecological zones, Young provides
inventories of native plant species and common crops, as
well as a brief summary of land-use practices and the
agricultural calendar for Pataz.
Published descriptions are
not yet available for each of the montane forest life zones
that Young has identified (Young, personal communication).
Also, the lower montane forest and premontane forests of the
Abiseo drainage remain mostly unexplored by western
scientists.
Virtually all of the following biogeographic
information specific to the Pataz-Abiseo area has been
distilled from Young's doctoral thesis and recent
publication of Pataz's ecological zonation (Young 1990,
1993).
139
The Dry Forest Zone
At the western extreme of the Pataz-Abiseo area, the
Marafion valley rain shadow produces the Dry Forest macroecological zone (1200 - 2300 m) .
The Dry Forest zone
corresponds to Holdridge Tropical Thorn Steppe and Tropical
Dry Forest life zones.
year.
Rainfall averages only 500 mm per
According to Young, human exploitation has eliminated
the original character of the forest,
leaving cactus and
scrub vegetation to further proliferate on the steep slopes
of the Marafion canyon.
Land-use patterns today probably
bear little resemblance to those in prehistory because of
the introduction of sheep and goat grazing, and industrial
gold and silver mining.
Ecological degradation has
eliminated wild populations of large mammals.
Only five
percent of the land area is suitable for cultivation, and
irrigation is necessary to grow maize, avocados, onions,
tomatoes, limes, pineapples, bananas, passion fruit, guavas,
cherries, mangos, oranges, papayas, sweet potatoes, manioc,
Trujillo coca, sugar cane as well as assorted crops of minor
economic importance (Young 1993:271, Table 2).
Informal
placer mining by native villagers on the shores of the
Maranon yields small amounts of gold.
Residents of the Dry
Forest Zone sell products such as fruit to passing truck
drivers, and coca to highland villagers.
The Moist Montane Zone
Travelling eastward and upslope one encounters the bulk
140
of the human population between 2,300 and 3,400 meters in
the Moist Montane macro-ecological zone.
Young notes that
all of the staple crops except for coca are cultivated
between 2,400 and 3,400 meters (Young 1990).
The town of
Pataz (incorporated in 1770) is situated on the southern
slopes of the Frances River valley for easy access to
mineral resources at 2,700 m.
Village residents still
utilize shafts abandoned by industrial mining interests in
the 1940s.
Pataz interacts commercially with Trujillo where
gold is exchanged for cash.
Agricultural produce from the
highlands surrounding Huamachuco makes up the deficit of
foodstuffs not obtainable from the cultivated slopes above
and below Pataz.
Pataz and adjacent districts have long been recognized
as among Peru's richest in mineral resources (Tarnawiecki
1927).
Above town, miners have followed veins bearing gold
and silver sulfides within a quartz and pyrite matrix deep
into the mountainside.
Modern mining activities have
traditionally centered on the mountainsides from Zarumilla
south to Buldibuyo.
Mines located between 3,500 and 4,000 m
above Parcoy and Llacuabamba have produced extremely high
yields of over 100 grams of gold and silver per ton of
processed mineral (Ibid. :182).
According to Tarnawiecki
(1927:178) prehistoric and early Colonial Period mining
activities took advantage of gold and silver oxides (placer
deposits) occurring 40 to 60 meters below primary exposures
141
on the lower valley slopes of the Moist Montane and Dry
Forest zones ( "sitios de climas templados") .
Los Alisos and other agricultural hamlets above Pataz,
like the self-sufficient towns of Condorrnarca and Uchucmarca
farther north, occupy locations at approximately 3,000 m for
efficient access to principal crop production zones (Brush
1977:82-83).
The Moist Montane Zone receives 750 - 1000 mm
of precipitation annually, most between November and May.
Native vegetation has been completely altered by human
exploitation, and now consists of shrubs and small trees,
and forest remnants in gullies, ravines and along hedgerows.
Only approximately 10 percent of the steep terrain can be
cultivated.
Common crops include maize, beans (P.
vulgaris), broad beans (V. faba),
squash (Cucurbita maxima),
peppers (Capsicum annum), wheat, alfalfa, barley, tarwi,
peas, potatoes, olluco, oca, agave, chirimoya, prickly pear,
figs, apples, peaches, elderberries, cabbage, carrots,
cayenne and oregano.
From Pataz, the Marafion can be reached
after a four to five-hour walk west, and the edge of the
continuous eastern montane forest in the Abiseo River
watershed lies a long day's walk over the cordillera.
Above Pataz between 3,300 and 3,600 m lies a
discontinuous belt of dense closed forest which is usually
fog-bound during the rainy season (Young 1993:272. Table 4).
It is a habitat for deer and provides a source of firewood.
142
The Tropical Alpine Zone
Young describes the Tropical Alpine macro-ecological
zone, occurring above 3,600 m on the western side of the
Marafion-Huallaga divide, as divided equally between rocky
peaks and slopes, and wide U-shaped valleys.
Tussock
grasslands characterize the valley floors with sedges and
shrubs thickly covering areas of poor drainage.
The western
side of the cordillera (Tropical Subalpine Wet Paramo life
zone) receives 1,500 mm of precipitation, but is relatively
dry between May and September.
The eastern side (Tropical
Subalpine Rain Paramo life zone) receives over 2,000 mm of
rain and experiences fog almost daily.
Frost, sleet and
snowstorms are not uncommon at the peak of the dry season.
Highland villagers utilize the Tropical Alpine zone for
cattle grazing today.
Cattle owners typically maintain
rustic huts, tend a potato patch and hunt waterfowl, deer
and predators such as Andean foxes, spectacled bears and
mountain lions during periodic visits to the area.
Within the past 25 years, economic activities in the
Tropical Alpine zone have been altered by the reduction of
cattle-herding resulting from 1970s agrarian land reform,
and by the introduction of trout which are caught with a
baited hook or by hand.
The voracious trout eliminated much
native aquatic life prior to identification by specialists
(Leo 1992), although there were probably few native
vertebrate species.
Economic activities on the eastern side
143
of the divide were again transformed in 1983 by the creation
of the Rio Abiseo National Park and new prohibitions of
traditional activities such as cattle grazing and hunting.
A state-sponsored agricultural program recently introduced a
herd of alpacas tended by the Comunidad Campesina of Los
Alisos.
Long range plans include harvesting and exporting
the alpaca wool as a cash-generating industry.
The Tropical Montane Rain
Forest Zone
Scientific studies in the forested Abiseo River
drainage have long been hindered by the absence of pertinent
and reliable cartographic information.
Modern map coverage
by the Institute Geografico Nacional (IGN) cartographic
(1:100,000 series) does not include the eastern montane
forests of San Martin Department.
Comparison of recent
LANDSAT photos to the 1:500,000 departmental map (1985)
reveals that the latter errs in its location of several
rivers.
Aerial photographs have proved useful for ground
surveys, but coverage by the AF-60-17 series (1962-63)
currently available from the IGN does not extend much below
the upper montane forest.
These mapping problems await
rectification by the Peruvian government.
Although technically belonging to the Tropical
Subalpine Rain Paramo life zone, Young and Leon's (1988)
"fragmented" montane forest constitutes the ecological
transition between 3,700 and 3,400 minto "continuous"
144
montane forest.
Young utilizes the term "timberline forest"
to indicate both the fragmented forest and the uppermost
strip of continuous forest.
Forest patches occur on the
slopes of the U-shaped valleys, especially under cliffs,
around boulders, within gullies and surrounding standing
water.
In order to catalogue vegetation, Young and Leon
(1990) divide the continuous montane forest into
approximations of the lower (1,500 - 2,500 m) and upper
montane forest
(2,500 - 3,500 m) zones described earlier in
this chapter.
The lower montane forest is less important
for the purposes of this study because it remains
archaeologically unknown.
Young identifies two principal Holdridge life zones
within the upper montane forest.
These are the Tropical
Montane Wet Forest life zone (2,500 - 2,900 m) and the
Tropical Montane Rain Forest life zone (2,900 - 3,400 m).
He believes that generally low temperatures and frost,
low
light levels, acidic soils with high aluminum levels, and
the very high humidity are factors which individually or in
combination discourage crop cultivation in the Tropical
Montane Rain Forest life zone (Young, personal communication
1992).
Hastings (1985:70, 73-74) likewise reports that the
"cloud forest sub-zone" at elevations between 2,500 and
3,300 m in the Tarma River canyon has only minimal economic
importance.
Rather than indicating that agriculture never
extended into this zone, these observations suggest that
145
agricultural production focused on more favorable higher and
lower elevations.
The boundaries, and even the presence and
absence of the Tropical Montane Rain Forest and other life
zones should be expected to vary regionally and temporally.
A higher forest canopy allows easier travel within the
Tropical Montane Wet Forest life zone.
drainage,
Within the Abiseo
this zone contains abundant evidence for dense
prehistoric settlement and economic utilization that will be
outlined later in this chapter.
Bonavia noted maize
thriving during his visit to Gran Pajaten, and in 1986
University of Colorado archaeologists found potatoes that
locals had planted in their 1985 excavation backfill.
While the lower montane forest remains virtually
unexplored, it has attracted small groups of highland
agriculturalists that have recently colonized Jucusbamba
(2,000 m) on the Abiseo (UC and FPCN 1991:34; Young et al.
1994).
Agricultural systems of the lower montane forest
remain undescribed.
Equally little is known of the
premontane forest life zones between 500 and 1,500 m, except
that similar environments north and south of the RANP
currently produce much of the world's coca crop.
Many non-agricultural resources may lie concealed
beneath the montane forest vegetation.
During his attempt
to reach the Huallaga from Pataz in 1919, Weberbauer
(1920:8) complained that some residents of Pias jealously
withheld information regarding the upper forest location of
146
a source of salt.
According to Tarnawiecki (1927:177}, the
gold and silver-bearing deposits mined above the Marafion
continue into the "montaiia," yet their extent can only be
crudely estimated.
De las Casas claims that mineral
deposits at the abandoned settlement of Achiras (1,000 m} in
the Abiseo valley premontane forest yielded high gold
content (1935b:342}.
Some western affluents of the Huallaga
also contain placer deposits (de las Casas 1930:173}.
The Huallabamba and Hua1laga Valley
Premontane Forests
De las Casas documented the predominant agricultural
activities in the premontane and subtropical forests of the
central Huallaga valley lowlands in a series of monographs
promoting the region's economic potential (de las Casas
1930, 1932, 1933, 1935a, 1935b}.
He details a wide range of
cultivated produce and plant resources including beans,
potatoes, maize, manioc, assorted fruits, cotton, coca,
vegetable dyes and medicinal plants (de las Casas 1932}.
The town of Pachiza was founded above the banks of the
Huallabamba by Franciscan Padre Sobreviela in 1790 as a
place for canoe travelers to rest and re-supply during their
descent to the Huallaga from the now-abandoned mission of
Jesus de Pajaten (Unanue and Sobreviela 1963 [1791] :154}.
How far canoes may navigate up the Abiseo river system
within the study area proper remains unknown.
At the time
of de las Casas' report (1935b:346}, Pachiza villagers
147
exported agricultural staples including livestock via the
Huallaga to Iquitos in canoes and balsas.
Numerous sources of salt are available along the course
of the central Huallaga.
These include several near
Tocache, and a source named Cachihuafiunca near El Valle (de
las Casas 1930:179).
A source producing salt that
circulated widely in pre-Hispanic and Colonial Period
Amazonian trade lies farther down the Huallaga at Pilluana
and Chazuta (DeBoer 1984; Reeve 1994).
Rubber was of
substantial economic importance at the time of de las Casas'
study, but DeBoer noted that coca has now become the area's
principal cash crop (DeBoer 1984).
The social and political
problems that have accompanied narcotics trafficking
presently render this region virtually inaccessible for
scientific study by outsiders.
Future investigations of east slope zonation and human
land-use within the study area might include matching life
zones and crop requirements, thereby suggesting a range of
hypothetical prehistoric crop production zones optimal for
potato, maize, manioc and coca under present climatic
conditions.
Such an effort would be severely hindered by
the lack of information pertinent to soil conditions as well
as prehistoric crop varieties and agricultural techniques.
Of primary interest is the relationship between the vertical
distribution of potential production zones on the eastern
slopes and the cultural boundary separating the
148
monument-building, nucleated montane forest groups known to
have once occupied the study area (Lennon et al. 1989) from
the upper Amazonian groups dwelling in the Huallaga valley
bottom.
Ethnohistory of the Pataz-Abiseo Area
The following discussions utilize the available
documentary information to explore how late prehistoric and
Colonial Period populations were distributed over this
landscape.
According to Espinoza's (1967) ethnohistoric
analysis of Chachapoyas, prehistoric populations on the
Marafion-Huallaga divide were organized in numerous
autonomous political units.
Utilizing Colonial Period
administrative and ecclesiastical records, it can be
demonstrated that some of these polities interacted
habitually with neighboring Amazonian groups of the central
Huallaga River valley. The latter may be thought of as
border intermediaries (Reeve 1994) connecting Amazonian and
Andean interaction spheres.
The documentary evidence also reveals that the early
Colonial Period was a time of demographic chaos in both
highlands and lowlands.
Sudden depopulation and demographic
disintegration contributed to the breakdown of traditional
communication linkages, and ultimately led to the complete
disarticulation of the two areas by the mid-twentieth
century.
This decline was merely one aspect of the pan-
regional economic collapse that accompanied the Spanish
149
imposition of new political and economic orders.
The
descriptions and interpretations to follow first focus on
the highland populations occupying the crest and upper
slopes of the Marafion-Huallaga divide, and then on tropical
forest groups occupying the lower slopes of the Huallaga
valley.
Southern Chachapoyas
The Pataz-Abiseo area lies within the southern half of
the Inca province of Chachapoyas (Cieza de Leon 1976 [1553];
Garcilaso de la Vega 1966 [1609]; Rowe 1946; Schjellerup
1984).
According to Espinoza's ethnohistoric analyses of
several surviving documents (1967:333), so-called Chacha or
Chachapoyas ethnic groups were distributed north to south
between Levanto near the modern city of Chachapoyas and
Huancaspata at the modern departmental border of La Libertad
with Huanuco.
The Marafion canyon separated Chachapoyas from
groups subsumed within the Inca provinces of Cajamarca,
Huamachuco and Conchucos to the west.
Espinoza draws an
eastern boundary incorporating the upper montane forest.
Inca provincial boundaries have traditionally provided the
principal basis for determining the spatial extent of
Chachapoyas ethnic groups.
While accepting Espinoza's
frontiers for the purposes of discussion, it should be
remembered that the Inca provinces were political constructs
imposed upon population sets for the purpose of efficient
administrative control.
Local ethnic boundaries were
150
disregarded where they were considered irrelevant to
imperial interests (Zevallos 1987).
The term "southern Chachapoyas" will be used to refer
to the Spanish-administered province (or corregimiento) of
Cajamarquilla.
The Spanish sub-divided Cajamarquilla into
five repartimientos (or encomiendas) listed as Leimebamba y
Cochabamba, Chilchos y Laya, Cajamarquilla y Condormarca,
Sucos y Puymal and Collay from north to south.
The three
southernmost encomiendas are of primary interest to this
study.
Ethnohistoric research in southern Chachapoyas
suffers from the failure to locate the pertinent Toledo era
Visita compiled between 1572 and 1574 by Diego Alvarez
(Espinoza 1967:225).
Thus we lack the kind of early
detailed census information utilized to examine demographic
and organizational problems in other regions like Huanuco.
Also, the synchronic ethnographic profile offered below
should be regarded as specific only to Chachapoyas on the
eve of Inca conquest.
Chachapoyas Culture
Documentary evidence confirms archaeological reports
regarding Chachapoyas settlement patterns.
Chachapoyas
natives utilized all four of Young's macro-ecological zones
and the bulk of the population centered on the upper slopes
and high promontories near the interface of the Moist
Montane and Tropical Alpine zones around 3,500 m (Brush
1977:82-83).
In this respect, Chachapoyas settlement
151
patterns were little different from coeval patterns in other
Central Andean regions where defense and efficient access to
both tuber and pastoral production zones were primary
considerations (Ibid.:83).
Chiefly interested in maximizing
control over indigenous labor resources, both the Incas and
Spaniards centered their administrative activities at
highland Chachapoyas settlements.
Inca-installed mitimaes
from the coast inhabited the Dry Forest Zone near Balsas
(Espinoza 1967:230), and both Thompson (1976:99) and Curtin
(1951:60) report archaeological evidence of occupation
(albeit undated) at Pusac and Matibamba respectively.
Chachapoyas occupation of the Tropical Montane Rain Forest
Zone will be addressed later within this chapter.
Current conceptions of Andean commodity movement across
a "vertical" landscape lead us to expect a linear west-east
distribution of intra-ethnic exchange partners in the PatazAbiseo area.
Even today, Bolivar (Cajamarquilla) interacts
commercially with Cajamarca, just as Pataz communicates
almost exclusively with Huamachuco and Trujillo.
Yet
considering perceived similarities between archaeological
assemblages distributed north-south along the eastern slopes
of the Central Andes (e.g. Lathrap 1970; Isbell 1974;
Hastings 1985:554) the longitudinal orientation of shared
Chachapoyas cultural characteristics is significant.
While
topography may dictate the elongated shape of Chachapoyas to
some degree (see Rowe 1946:185, Map 3), continuous pre-Inca
152
interaction along a north-south axis is also a likely
factor.
Zavallos (1987) observes that "there should have
been powerful reasons for the Lima government to consider
the zone [Pataz and Cajamarquilla] as part of the polar
jurisdiction of the city of Chachapoyas since the 16th
century."
The fact that Chachapoyas' continued to exercise
jurisdiction over Pataz until well into the 19th century is
probably an artifact of ancient north-south interrelations.
Although some have hypothesized state-level political
integration for pre-Inca Chachapoyas (Ravines 1972:218;
Brush 1977:43-44; Morales 1993:652), most analyses of
ethnohistorical and archaeological evidence have concluded
that Chachapoyas populations had never been politically
unified prior to Inca conquest (Langlois 1939a:229; Espinoza
1967:233-234; Zevallos 1987; Schjellerup 1990:237; Lerche
1995).
Regarding socio-political organization, Espinoza
(1967:233-234) argues that the primary unit was the ayllu
led by a chief (or curaca) and a council of elders.
Although each ayllu was independent, hostile threats
prompted the creation of temporary alliances.
Each ayllu
maintained a principle village with secondary hamlets
comprising a small-scale village hierarchy which applied the
name of the ayllu to the entire group.
Espinoza reports
that ayllu leadership was traditionally inherited, and that
ethnic unity throughout Chachapoyas was expressed in a
common principal deity named Curichaculla, as well as a
153
common language, dance and music.
Chachapoyas settlements typically consist of clusters
of circular stone buildings between six and 15 m in
diameter, arrayed on hill or ridge tops in a seemingly
random fashion.
family.
Each building likely housed a nuclear
Some ruined settlements such as Cuelap are famed
for their elaborate systems of walled fortification.
A few
structures bear stonework mosaic friezes similar to, yet
less elaborate than, those reported at Gran Pajaten.
Espinoza remarks that ample land for cultivation engendered
pacific relations between Chachapoyas ayllus.
He views the
fortifications as providing defense against the abundant
jaguars, pumas and bears (Ibid.:234), and against lowland
marauders from the north and east (Ibid.:235-236).
However,
Bandelier noted oral traditions describing hostilities
between settlements such as Cuelap and Levanto (1907:19).
An oral history of hostilities between the Chachapoyas
settlements of Cunturrnarca and Coben survives today in
Condorrnarca (personal observation, 1986).
Espinoza concurs with Langlois'
(1939a:233) conclusion
that, because there was no political unity, there were no
"organized temples" nor attendant religious specialists, and
that coordination of religious activities was charged to
shamans (Espinoza 1967:235).
Jesuit priest Jose de Acosta
(1958 [1577] :236) commented on the activities of powerful
Chachapoya shamans: "grandes hechizeros y herbolarios," some
154
of whom were coaxed into giving up their practice of
poisoning their enemies.
Bandelier observed shamanic curing
with cornmeal, coca and an herb called shayr at the end of
the nineteenth century, and remarked that "these medicinemen
are for good as well as for evil"
(Bandelier 1907:13-15).
Fifty years later, the Reichlens (1950:221-222) found little
remaining of native curing practices and other indigenous
traditions.
Chachapoyas Under the Inca
Following the Inca conquest led by Tupac Inca Yupanqui,
the Chachapoyas Province was divided into hunos or decimal
units of 10,000 tributaries for the purpose of expedient
administration.
A provincial Inca administrative center was
built at the site of Cochabarnba and a curaca installed to
govern the huno of Cochabarnba y Leimebarnba (Schjellerup
1979-80:302).
Tupac Inca Yupanqui charged the curaca
Chuquipundio of Cunturmarca with the governance of the
southern Chachapoyas huno of Cunturmarca y Collai (Espinoza
1967:240).
The seat of administration may have been located
at approximately 3,500 m several kilometers upslope from the
modern village of Condormarca where local residents refer to
a ruined settlement of circular and Inca-style stone
buildings as "Cunturmarca" or simply "Pueblo Viejo"
(personal observation, 1986).
At Cunturrnarca, a geometric
frieze on the facade of one rectangular structure
illustrated by Izaguirre (1923:194) was still partially
155
preserved in 1986.
Schjellerup (1984) describes Inca dominance of
Chachapoyas as always tenuous.
Conquest was achieved only
after a series of fierce battles at several Chachapoyas
settlements like Cunturrnarca and Cajamarquilla, and heavy
losses suffered by Inca militia (Garcilaso de la Vega
1966:479).
Huayna Capac quelled one rebellion (Ibid.:554)
and natives were punished by the relocation of some
populations, several of which were sent to Cuzco.
The Inca
simultaneously implanted foreign mitimaes composed of
Huanca, Cajamarca and Chimu-Lambayeque groups from the south
and west.
In 1532, the unforgiving Chachapoyas offered
their military services to Pizarro's troops in order to rid
themselves of Inca domination.
The Spaniards rewarded them
with independence until Alonzo de Alvarado and his troops
marched into Chachapoyas in 1537 to claim it for the Crown.
Spanish Conquest and Administrative
Units in Southern Chachapoyas
The complexities involved in utilizing colonial
administrative units such as encomiendas or repartimientos
to locate places and interpret ethnic and/or political
boundaries in such a poorly understood region are
substantial.
Nevertheless, the utility of this analytical
approach has been demonstrated by Julien (1985) and will be
pursued in the following discussion.
The earliest well-
known published maps that provide detail on Chachapoyas and
156
the Pataz-Abiseo area were drawn by Franciscan missionaries
in the late 18th century (e.g. Sobreviela's and Amich's maps
in Izaguirre 1923:105, 237).
forest place names.
These provide few montane
Sorely lacking are detailed maps and
other published documentation pertinent to Jesuit activity
in the 16th and early 17th century.
Unfortunately,
Espinoza's map of ethnic groups and places on the MaraftonHuallaga divide (1967:333) lacks sufficient supporting
bibliographic information with which to evaluate its
accuracy.
Maps currently printed by the Institute
Geografico Nacional (IGN)
furnish few clues to locate the
pertinent places.
According to pre-Toledo era census information compiled
in 1538, 1548 and 1561 (Rivera Serna 1956-57; Loredo 1958;
Hampe 1978), two southern Chachapoyas encomiendas of primary
importance to this study can be isolated, although their
boundaries are difficult to establish.
The encomienda of
Sucos y Puymal charged to encomendero Honorato Juan Bautista
Esteban includes places and/or ayllus named Sucos, Puymal,
Piax and Baldeboyo.
The modern highland towns of Pias and
Buldibuyo appear on modern maps on the Marafton side of the
cordillera bordering the Rio Abiseo National Park.
The
encomienda of Cajamarquilla y Condormarca held by Ines Nieto
and Juan Garcia Sanmames abuts Sucos y Puymal to the north,
and Collay held by Juan Montenegro lies to the south.
That settlements at Sucos and Puymal were located deep
157
within the Tropical Montane Rain Forest will be demonstrated
below.
During the Viceroy Toledo's population relocation
program in the 1570s,
at highland "Vchupiax"
"Andaraca"
the forest inhabitants were resettled
(probably modern highland Pias) and
(location unknown) where presumably they were
easily manipulated (Miranda 1906) .
Spanish administrators
torched the abandoned villages so that resettled families
would not return (Espinoza 1967:237).
Nevertheless, some
families did surreptitiously return to the montane forest by
1593 as Mogrovejo's visita implies (Ibid.).
Demographic Collapse
Garcilaso de la Vega (1966:476) reports an estimate of
more than 40,000 inhabitants in the Chachapoyas Province
prior to Inca conquest, a sum which he most likely took from
Chachapoyas-native Blas Valera's lost manuscript.
This
figure seems far too conservative given Julien's calculation
of over 43,000 individuals for the Colonial Period
corregimientos of Cajarnarquilla, Pacllas and Luya y Chillao
in the 1570s (Julien 1985:25-26, Table III).
The basis for
Garcilaso's estimate remains unknown, but the sum of 40,000
probably refers only to tax-payers
(tributaries).
Lerche's
(1995:36) estimate of 300,000 to 500,000 individuals in preInca Chachapoyas seems more reasonable.
Cook (1981:110)
confirms that "major epidemics swept Peru in the 1520s,
1530-2, 1546 and 1558-60," and estimates a total population
decline in the Andes between 1520 and 1620 as "approximately
158
93 percent"
(Ibid.: 114) .
According to Cook (1981:195), the northeastern Peruvian
Andes experienced a demographic rate of decline "generally
more than double the north highland average, " and
"paralleled and in some cases exceeded the rates of the
coast."
Regarding the population decline in the southern
Chachapoyas corregimiento of Cajarnarquilla, Cook (1981:183)
observes that,
"the populations of the corregimiento were
relatively large, but the tributary population was declining
at a rate much higher than that of the other 'highland'
repartimientos of the north ... "
During the 27 years between
1575 and 1602, the number of individuals paying tribute in
the repartimientos of Cajarnarquilla and Buldibuyo
{corresponding to the encomiendas of Cajarnarquilla y
Condorrnarca and adjacent Sucos y Puyrnal) fell 45 percent and
56 percent respectively {calculated from Cook 1981:185,
Table 40).
In 1576, Jesuit priest Cristobal Sanchez died of
fevers after administering last rites to victims of a
smallpox epidemic in nearby Leimebarnba {de Acosta 1958:238).
"Few Indians" were left in Moyobarnba that year because of
"high mortality"
(Ibid. :234).
To summarize the fate of southern Chachapoyas
populations, Cook cites Chronicler Vasquez de Espinoza's
observation:
Caxamarquilla y Collay ... has at present few Indians, because
many have died, and others have retreated to the land of war
that they call los Aucaes; all of the province is heavily
forested and very warm ... (Vazquez de Espinosa 1969
159
[1626] :281).
The term Auca derives from the Quechua "Auka" meaning
"enemy" or "rebel"
(Cooper 1946:690).
In Ecuador it was
reportedly used by the Quijos to refer to "the pagan tribes
of the lower forest regions"
(Steward 1948:653).
Espinoza's
report (1967:237) of Toledo-era refugees in the montane
forest lends support to the idea of an eastward flight from
the area of Spanish control.
Espinoza (1967:230) and Zevallos (1987) both emphasize
the population decline caused by the abuses of early Spanish
landlords responsible for organizing native laborers.
Most
of the Spaniards controlling Chachapoyas in 1548 held mines
(Loredo 1958), but a scarcity of labor ultimately left many
of these idle (Vasquez de Espinoza 1969:281).
(1921 [1593] :70) counted eight "negros"
Mogrovejo
(probably African
slaves) and some "yanaconas" working the mines of Bartoleme
Gutierrez near Condormarca.
Slave labor at the mines was
not unusual in Colonial Peru, but here the importation of
valued laborers from elsewhere may reflect local scarcity.
Cook reports that demographic collapse precipitated the
decline in gold production in the nearby region of Jaen
(Cook 1981:191).
Apparently, such logistical difficulties
overshadowed and ultimately extinguished Spanish interest in
Chachapoyas' mineral wealth as well.
The Central Huallaga Valley
The prehistory and early colonial history of the
160
central Huallaga River valley is of secondary importance to
this study only because Manachaqui Cave lies squarely within
southern Chachapoyas.
Unfortunately, most historical
descriptions portray Amazonian societies already altered by
decades of disruptive European influence.
Epidemic
diseases, and Spanish and Portuguese slaving reportedly
prompted demographic instability and intensified hostil1ties
throughout the region (Reeve 1994:109-111).
Differing
colonial administrative policies exercised by the Spaniards
mean that much information pertinent to the years preceding
Spanish conquest commonly provided to civil census-takers in
the highland Andes will never be forthcoming for most of
upper Amazonia.
Much of our knowledge of 16th to 18th
century central Huallaga populations in and adjoining the
study area is gained from written reports by missionaries
and soldiers, but the period prior to Franciscan missionary
interest beginning in 1670 remains obscure.
The following
account provides only a cursory glimpse of pre-contact
central Huallaga cultures abutting Southern Chachapoyas.
Early Missionary Contacts
According to the earliest documentary reports
(Mogrovejo 1921 [1593]), the Cholon and their northern
neighbors the Hivito occupied the central Huallaga lowlands
within and adjacent to the study area.
Seventeenth century
missionaries describe their homelands as behind the
corregimiento of Cajamarquilla (ARSI 1636-37:fos.116v, 121:
161
transcription by Mary-Elizabeth Reeve).
By 1600, itinerant
priests, apparently working out of highland Condorrnarca, had
"reduced" substantial numbers of Cholones and Hivitos, and
established churches at five of nine listed settlements
(Mogrovejo 1921:68-69; de la Riva Herrera 1907 [1655] :290).
According to seventeenth century Jesuit documents, the
Cholones and Hivitos were conquered by the Incas (ARSI 163637:fos.116,121).
Subsequent to the Spanish conquest and
administrative partition of Peru, the two groups were
included within the Chachapoyas encomienda and doctrina of
Cajamarquilla y Condorrnarca (Ibid.).
pastoral visita in 1593 (1921)
Archbishop Mogrovejo's
lists the number of
tributarios at the nine Chol6n and Hivito settlements.
These lowlanders were among the first upper Amazonian groups
to be "missionized" by virtue of their incorporation within
the southern Chachapoyas encomienda/doctrina.
Information regarding Chol6n and Hivito residence
patterns is also provided by Mogrovejo•s pastoral census.
At the Chol6n "pueblo" of Quisupay (where there was a
church), the archbishop encountered at least 120 natives
living in three houses ("Casas apartadas").
At Chamal
(where there was no church) , two houses sheltered at least
88 persons.
same.
Hivitos residence patterns were apparently the
Historically such residence patterns are common to
tropical forest societies of the western Amazon basin in
which extended families reside communally in malocas, large
162
houses constructed of perishable materials (Steward 1948;
Meggers 1971).
Mogrovejo may have been describing typical
mission dwellings in which long houses were divided into
single family apartments (Metraux 1963:649), but apparently
the Chol6n and Hivito villages were not proper mission
settlements (reducciones).
All of the settlements except
for San Joan de Ulat retained their solely indigenous names.
Mogrovejo's description of an individual named Don Joan
Momuman as headman of the "montafia" provinces ("cacique de
toda la dicha montafia") indicates the presence of some
overarching native authority.
According to Reeve's analysis of documentary evidence
(1994:112-113), the Cholones and Hivitos acted as "border
intermediaries" mediating exchange which linked lowland
floodplain-dwellers such as the Tupian-speaking Cocama to
highland Chachapoyas populations.
prior to Spanish contact,
Reeve observes that,
"the Chol6n and Hivito appear to
have represented the uppermost extent of tropical forest
peoples into the Andean region along the Upper Huallaga
River"
(1994:112-113).
The desire to control this well-
established commercial conduit for Amazonian produce likely
constitutes the rationale behind Inca conquest of this
forested region.
The first century of Spanish and Portuguese contact
brought an undetermined number of epidemic diseases from
both the highlands and lowlands (Cook 1981, 1992; Myers
163
1988; Denevan 1992; Reeve 1994).
Studies by Myers (1988:65-
66) conclude that the Cocama on the lower Ucayali and
Huallaga rivers suffered an epidemic between 1558 and 1560.
Measles and small-pox spread along the course of the
Huallaga between 1638 and 1645.
Contact period trading
activities evident all along the eastern slopes undoubtedly
accentuated the transmission of contagious diseases between
the Andes and Amazonia.
Intensive interaction at particular
locations would have created an effective nexus of disease
vectors with dire local consequences (Dobyns 1992).
Denevan
(1992:220-221) regards both the population densities and the
impact of disease as relatively light in the Amazonian
"forest uplands"
(excluding the ceja de montafia).
Considering the uplands as a whole his low counts may
accurately reflect demographic change, but his level of
generalization does not take into account local
environmental, social and historical factors that rendered
some populations more vulnerable to contagion than others.
Forest upland groups linked into disease vectors centered in
the Andes and upper Amazon River floodplains by habitual
interaction probably suffered disproportionately.
The
sudden demographic disintegration at Bagazan on the route
connecting San Juan de la Frontera (Chachapoyas) and
Moyobamba documented by Cook (1981:195-196) may illustrate
the heavy price paid for intermediating highland-lowland
exchange.
Population clustering around the churches
164
established early in Cholon and Hivitos country must have
provoked similar consequences.
Early seventeenth century competition for converts
between Jesuit missionaries from Borja near the bend of the
Marafion, and Franciscan missionaries from Huanuco widely
disseminated coveted trade goods which acted to amplify and
distort patterns of trade and warfare in the intervening
area (DeBoer 1981; Reeve 1994).
Increased raiding to obtain
slaves to exchange with the Spanish for iron tools came to
characterize contact between the central Huallaga, the lower
Ucayali and the upper Amazon (Reeve 1994:112).
Already in
1593, missionary observers noted that travel between
settlements was dangerous, and that the Cholones and Hivitos
defended themselves from attacks by the Motilones (or Lamas)
in a "fortress"
(Mogrovejo 1921 [1593] :69).
Missionary
records state that the Cholones and Hivitos suffered attacks
by the Chocoltos (the Choltos listed by Steward 1948:600)
from the north, and also that the two groups warred with one
another (ARSI 1636-37:fol.116v, 121).
Between 1630 and 1636, Jesuit missionaries from
Trujillo reportedly catechized upper Amazonian natives east
of Moyobamba and Cajamarquilla (Vargas Ugarte 1941:29-30).
Missionaries devoted to newly proselytizing the Cholones and
Hivitos were deemed particularly successful.
In three years
the Jesuits documented native languages and established two
mission settlements: La Concepcion with 709 Cholones and San
165
Javier with 624 Hivitos.
By 1640, the Hivitos had endured
hostilities perpetrated by the Cocama and Aguano of the
lowland river floodplains to the northeast (Reeve 1994:111;
Lopez de Alvarado 1907 [1656]:362).
At some point, the
lowland chain of violence spilled into southern Chachapoyas
where raids from the central Huallaga "destroyed" the
Chachapoyas settlements of Condormarca and Collay (Amich
1854 [1768] :75; Steward 1948:601).
At about this same time,
the Trujillo Jesuits abandoned the central Huallaga doctrina
settlements (de la Riva Herrera 1907 [1655]:290), perhaps
because of violence, epidemics and/or neglect by central
offices.
Some Cholones and Hivitos later testified that the
priests were driven out by "some old caciques" who
threatened to kill them (Ibid.).
The names of the
settlements listed by Vargas Ugarte and Mogrovejo do not
appear in subsequent documents.
In August of 1654, General Don Martin de la Riva
Herrera began a campaign to "pacify" the lowland populations
along the central and lower Huallaga river course.
De la
Riva Herrera planned to enter the northeastern montane
forest via Condormarca and Capillania with over two hundred
soldiers.
According to de la Riva Herrera's (1907) reports,
a party of sixty lowlanders anticipated his arrival and
climbed to welcome the soldiers at the highland "puesto de
Cillangat
foods
(?)"
[sic] with gifts of cassava and other lowland
(de la Riva Herrera 1907:286-287).
Five days' march
166
north from Capillania,
this "puesto" is clearly the Tarnbo de
Callangate (3,500 m) in the Tropical Alpine Zone between
Bambamarca and Cajamarquilla (Raimondi 1900a:36).
With his lowland guides, De la Riva Herrera likely
descended a route later identified by Raimondi (1900a:39)
that parallels the Tubaybal and upper Tepna rivers,
affluents of the Jepelache River.
A five day journey
brought the company to a settlement of the Porontos.
Steward (1948) does not mention the Porontos who may have
occupied territory either north or west (upslope) of the
Hivitos.
De la Riva Herrera claimed the Porontos province
for the king, ordered the building of a church and renamed
the settlement San Antonio de Porontos.
After installing a
government and a resident priest, he distributed axes,
machetes and knives (de la Riva Herrera 1907:288-289).
Then
he marched two days to Hivitos country.
The Hivitos also peacefully received de la Riva
Herrera, who renamed their village Limpia Concepcion de
Jivitos and sent for Chol6n caciques to participate in the
proceedings.
After receiving assurance from the gathered
natives that the trouble-making "old cacique" had since been
eaten by a "tigre," de la Riva Herrera distributed more
gifts of tools and left Spanish-sanctioned authority,
including a priest, before departing.
The soldiers built
rafts in Hivitos country and, despite difficult passages,
they descended the "Xivitos River"
(the Jepelache and/or
167
Huallabamba?) 12 leagues (his estimate) to the Huallaga in
approximately three days (de la Riva Herrera 1907:293).
At this point, De la Riva Herrera marched on to lead a
military fiasco in Jivaro country, where he eventually
distinguished himself by his abusive tactics (Jouanen
1941:417-426; Reeve 1994:133).
Ultimately these later
Cholon and Hivito settlements also failed for unknown
reasons.·
To Jesuit authorities in Quito, the doctrina of
Cundurmarca represented their most distant southern outpost,
one which perhaps they no longer cared to support.
Priests
were obligated by law to accompany conquistadors regardless
of an order's interest in supporting the endeavor (Reeve
1994:120).
Any initial enthusiasm demonstrated by the newly
installed missionaries must have been quickly dampened by
epidemics reported in the nearby upper Huallaga region in
1662 and 1670 (Steward 1948:597), and by the widespread
"reign of terror" incited by rebellious Cocamas who attacked
mission settlements downriver between 1663 and 1666 (Reeve
1994:122).
In 1670, lowlanders once again emerged from the forest
to request missionaries and resumption of peaceful trade
with the highlands.
This time the Franciscan order based in
Huanuco filled the void left by the Jesuits and established
their own permanent missions among the Cholones and Hivitos
in 1676 (Amich 1854:76; Steward 1948:600-601).
The Hivitos
were settled at Jesus de Pajaten and Jesus de Monte-Sion,
168
and the Cholones further south at San Buenaventura del Valle
and Pampa Hermosa (Amich 1854:78).
Amich provides sketchy
descriptions of the Cholones and the Hivitos who he found
scattered throughout the forest by small-pox epidemics and
no longer answering to any single political authority
(Ibid.:19).
He reports that fishing and the cultivation of
typical lowland crops provided central Huallaga subsistence,
while cotton and coca were also grown.
Little documentary
information exists regarding customs and ritual activities.
Steward (1948:605) states that,
"the Cholones were reputed
to be powerful doctors in 1830," but he provides no
indication of his source for this information.
Ethnohistoric Evidence for HighlandLowland Boundaries
An important question formulated during Julien's
research (1985) on Chachapoyas ethnohistory concerns the
explicit justification of "highland" boundaries in the
eastern tropical forests.
Documentary evidence may be
utilized not only to justify Espinoza's placement of
Chachapoyas' eastern boundary, but also to show that the
upper montane forest served as the locus of social and
political activities in the Pataz-Abiseo area.
Two
historical references identify Sucos and Puyrnal as forest
settlements on the Huallaga side of the cordillera.
To
conduct his pastoral visita of 1593, Archbishop Mogrovejo
entered the montane forest from "Cundumarca"
(presumably
169
Condormarca) and traveled to,
"some villages inside the
montana, that they call los montes de Puymal"
1921:68).
(Mogrovejo
Mogrovejo's report claims that "all of the
Indians of the montana" gathered at the "pueblo" of "Yare de
Puymal" or "Yare que llaman Puymal," to be counted.
Espinoza (1967:237) speculates that Yare is the original
name of the site dubbed Gran Pajaten by Savoy (1970), and
Abiseo by Bonavia (1968b).
After descending from Yare de
Puymal an arduous 16 leagues ( "se va con ha.rto trabajo"),
Mogrovejo's entourage reached the first of several Hivito
settlements where the inventory of residents and converts
continued.
The brief report provides no information
regarding settlement in the intervening country.
The second geographic reference to Sucos and Puyrnal is
contained in writs dated 1653 and addressed to the king
touting Condormarca as the most appropriate point of entry
for General de la Riva Herrera's lowland pacification
campaign (Bautista de Escobar 1899 [1653] :81-83).
Entering
"tierra de montafia" from Condormarca, Bautista de Escobar
notes, one encounters the abandoned "pueblos" of "Xucos,
Puymal, Cholones, Xivitos, San Geronimo de Chinch6n" and
others.
The details provided in Mogrovejo's and Bautista de
Escobar's testimonies confirm that Sucos and Puymal were
montane forest settlements situated above territory occupied
by Cholones and Hivitos, and that early southern Chachapoyas
encomiendas included both highland and upper montane forest
170
territories and/or populations.
If Espinoza is correct in assuming that the ayllu
constitutes the principal Chachapoyas socio-political unit
from which the principal settlement takes its name,
then it
follows that Sucos and Puymal were forest-centered ayllus
and not colonies dependent upon larger, more powerful
highland polities.
Pre-Toledo era lists of corregimientos
and encomiendas within and beyond the province of
Chachapoyas are likely to contain the names of other eastern
montane forest settlements and/or ayllus.
Pinpointing their
locations with the available documentary information
constitutes a challenge for future ethnohistoric
investigation.
The native populations settled in and around the PatazAbiseo area probably belonged to the encomienda of Sucos y
Puymal rather than to Cajamarquilla y Condormarca.
This
conclusion is based simply upon the distribution of known
locations of place and/or ayllu names grouped within each.
That Mogrovejo and Bautista de Escobar speak of gaining
entrance to the forested sites of Sucos and Puymal via
highland Condormarca seems to contradict this inference.
However, the Condormarca entry to the forest likely provided
the direct access to the settlements at Sucos and Puymal
from the north (and Cajamarca) prior to Spanish
colonization.
Once within the montane forest, Mogrovejo's
entourage may have journeyed southeast to encounter Yaro de
171
Puymal in the upper Montecristo River where ruined
settlements including Gran Pajaten are found.
Alternatively, Mogrovejo may have found Yaro de Puymal where
modern Condormarca residents claim that there are ruins, due
east in the upper Pajaten River valley.
Unambiguous evidence that the lower montane forest was
occupied during the late 18th century derives from
Franciscan documents.
In 1791, Padre Sobreviela listed 28
"Indios y Neofitos" at Jucusbamba in the Abiseo River valley
(Unanue and Sobreviela 1963:157).
The information at hand
does not clarify whether the missionaries established the
settlement with a local population or brought in natives
from above and/or below.
The ruined foundations of a church
found in association with Inca structures near La Morada
(2,100 m) in the Bombonaje River valley by Schjellerup
(1992:359) may represent an earlier (16th or 17th century)
reducci6n.
Pre-Hispanic settlement on the eastern slopes of the
Marafion-Huallaga divide may have been continuous, or perhaps
broken by lightly occupied territory utilized for coca
production in a manner like Raymond (1985) and Bonavia
(1972b) describe for the premontane forests of the Apurimac
and lower Mantaro valleys.
The evidence just recounted
places the Porontos, Hivitos and Cholones within the
premontane forests where the latter two groups reportedly
cultivated coca during the seventeenth century (Amich
172
1854:78-79; Steward 1948:601).
Central Huallaga groups may
have moved upslope to occupy desirable land vacated by
beleaguered highlanders.
At the same time they likely
absorbed many highland refugees.
Also, increased raiding by
the Aguanos and Cocama along the course of the Huallaga
early in the seventeenth century probably rendered river
terrace locations undesirable for settlement.
In sum,
paltry documentary information distorted by high disease
mortality, population dislocations and forced relocations
does not offer an adequate basis for definitive
interpretations of highland-lowland frontiers in the study
area.
Ethnohistoric Evidence for HighlandLowland Interaction
Much documentary evidence for pre-Hispanic commercial
relations between southern Chachapoyas and the central
Huallaga groups is indirect.
First, it seems that the
incorporation of the montane forest settlements and lowland
"provincias" of the Cholones and Hivitos into the highland-
based Jesuit "doctrina de Cundumarca" prior to 1593
(Mogrovejo 1921:67-70) indicates early, perhaps preHispanic, linkage between these Andean and Amazonian
populations.
Priest Rodrigo Alonso's doctrina stretched
from Calemar and Bijos (now Vijos) on the banks of the
Marafion into the central Huallaga valley lowlands, and
included Condormarca and Puymal.
The Condormarca doctrina
173
is apparently unique as lowland forest communities do not
figure into any other highland doctrina census counts south
of Moyobamba.
Other clues suggesting linkages between southern
Chachapoyas and central Huallaga populations have been
offered.by Julien (1985).
Based on her interpretation of
Loredo's pre-Toledo era (1548) list of tributaries, Julien
believes that the terms Imigas, Ancimgas and Animigas
(Loredo 1958:259-260) refer to lowlanders.
She points out
that these terms are contrasted in the lists with "Serrano"
(highlander).
Juan Lopez Montenegro (or Montero) held "unos
Imigas" in his highland encomienda of Collay (Ibid. :260) .
One of the two caciques subject to the encomienda of
Chilchos y Laya on the upper Huallabamba is referred to as
an Ancimga (Ibid.:259).
Encomendero Juan de Rojas was
charged with "otros tantos Imigas" in the lowland Bagua area
(Ibid. :264).
Highland Collay and nearby Tayabamba sit atop
a route descending the Mishollo and Tocache Valleys to the
Franciscan mission of Pampa Hermosa.
An approximation of
the route between Tayabamba and lowland Tocache was utilized
by Franciscan missionaries
(Raimondi 1876:428; Unanue and
Sobreviela 1963:154), and is still travelled today.
After 1670, the Franciscans maintained access to Cholon
and Hivito settlements by two additional trails.
One
connected Capillania above Condormarca and the mission of
Jesus de Pajaten (approximately 400 m) where travelers
174
embarked down the Pajaten River in canoes.
It probably fell
into disuse when the mission burned in 1801.
Accounts of
lowland traders appearing in Condormarca persist in oral
narratives (personal observation, 1986) .
The other route
descended the Abiseo River valley from the Franciscan
convent at Huaylillas, passed by the installment at
Jucusbamba, and terminated at the mission settlements of El
Valle and Si6n near the banks of the Huallaga.
Only its
upper portion is traveled today by ten highland families
that have recently resettled Jucusbamba (Young et al. 1994).
Along these routes, the Cholones and Hivitos from the
Franciscan missions habitually traded lowland produce,
especially coca, to highlanders for tools and clothing
(Amich 1854:78-79).
Nineteenth century commerce brought coca and tobacco to
the highlands (De las Casas 1935a:345).
However, Raimondi
(1900a:40) noted that traditional trade between southern
Chachapoyas and the lowlands had ceased despite the
potential economic advantages of obtaining cheap salt from
central Huallaga sources.
According to Rios et al.
(1982:88), the Abiseo River route through Jucusbamba and
Achiras was abandoned with the opening of the central
highway connecting Huanuco and Tinge Maria during the mid20th century.
This leaves only the southernmost Tayabamba-
Tocache route still in sporadic service to foot traffic.
175
Archaeology of Southern Chachapoyas and the
Central Huallaga Valley
This section presents archaeological evidence for
ancient settlement on the Marafion-Huallaga divide.
Discussion focuses on regional chronologies, subsistence
economies and exchange.
Clear distinctions between
archaeological assemblages from southern Chachapoyas and the
central Huallaga valley correspond to expected differences
between Andean and Amazonian cultural traditions.
A
narrowed focus on the Pataz-Abiseo area highlights data
pertinent to site distribution by ecological zone, and
evidence for chronology and economic activities.
Most
important to note is the evidence for greater population
concentration on the eastern slopes of the divide, a pattern
that contrasts sharply with that observed today.
Southern Chachapoyas has attracted few systematic
archaeological studies to date.
Because of material
similarities shared within greater Chachapoyas, there is a
temptation to assume that southern Chachapoyas'
developmental trajectory differed little from that of the
better-known northern region.
It should be useful therefore
to consider briefly the results of investigations throughout
Chachapoyas prior to focusing on the Pataz-Abiseo area.
The
central Huallaga has received only three archaeological
studies to date, but only one of these bears directly on
prehistoric developments within the Pataz-Abiseo area.
A
summary overview of archaeological work on both sides of the
176
divide will offer background information that should
ultimately aid evaluation of data from the study area.
The history of investigations into the archaeology of
Chachapoyas province reflects archaeologists' unwavering
fascination with Chachapoyas architecture, especially the
spectacular "fortresses" of the upper Utcubamba.
However,
the bias that this emphasis on architectural contexts has
imposed upon reconstructions of the region's prehistory
remains largely under-appreciated.
Some investigations have
been confined to cleaning and recording architectural
details, while others describe excavated contexts within and
around buildings.
The cultural sequences for which there is
stratigraphic evidence derive from investigations in and
around Cuelap (Reichlen and Reichlen 1950; Horkheimer 1959;
Ruiz 1972).
Other projects center attention on late
prehistoric settlements around Uchucmarca (Thompson 1973,
1976) and Chuquibamba (Schjellerup 1990, 1992).
These
projects have rendered some appreciation of architectural
variability and settlement patterns, and an approximate age
for Chachapoyas architectural traditions.
To date there has
been no concerted attempt to investigate the antiquity of
human occupation in Chachapoyas territory.
With the exception of the recent report of
archaeological investigations in and around the Rio Abiseo
National Park (Lennon et al. 1989), the archaeological
literature from Chachapoyas lacks references to possible
177
Preceramic Period occupations.
The cultural sequence
outlined for the upper Utcubamba valley begins with Ruiz's
Early Intermediate Period Cancharin Phase (Ruiz 1972).
Archaeological remains dating to the Initial Period and
Early Horizon have not been reported on the Marafion-Huallaga
divide between Bagua and Tantamayo in Huanuco Department
(Bonnier 1983).
The Reichlens (1950:300) noted that,
although Tello's 1937 explorations on the divide stopped
just south of the Utcubamba drainage, he placed Chachapoyas
within the "area of distribution" indicated for his ChavinKotosh Civilization (Tello 1942:711, Lam. III).
Ruiz
(1972:181-183) perceived a close historical relationship
between Cuelap's Cancharin Phase styles and Early Horizon
styles from lowland Bagua, but did not attempt to evaluate
Tello's migration hypothesis.
Ruiz's interpretation of the excavated Cuelap sequence
emphasizes local cultural development and stylistic
continuity in pottery design attributes from the Cancharin
Phase through the Late Horizon Cuelap-Inca Phase.
A coarse
brown culinary ware dominated by simple jar and bowl shapes
often exhibits applique decoration and incised or notched
rims.
Cancharin Phase bowls with negative resist and white-
on-red geometric painting may indicate connections farther
north.
Schjellerup has identified coeval "Cajamarca I"
ceramics between Cuelap and Pataz in Chuquibamba (1992:362).
The architectural tradition at Cuelap featuring
178
circular stone dwellings appeared by the end of the Middle
Horizon (A.D. 800 according to Narvaez 1988:139).
At this
time Cajamarca cursive style painted bowls were locally
imitated (Ruiz 1972:188), and widely imported across the
Marafion between Cuelap and Cajamarquilla (now Bolivar)
farther south (Ibid.:187; Vega 1978:19).
Lerche's (1995:24-
25) suggestion of a Middle Horizon hiatus in the Cuelap
sequence is not supported by Ruiz's data.
During the 1967
clearing activities at Cuelap, Ruiz (1969) recovered sherds
that he considered Middle Horizon 2B imports from Ayacucho
or Ica.
Shady (1987a:86) assigns these same sherds to Nazca
Phase 9.
Surface architecture at the most elaborate Chachapoyas
hilltop settlements seems to date to the Late Intermediate
Period and Late Horizon (Narvaez 1988; Kauffmann 1991;
Schjellerup 1992; Lerche 1995), as do chamber tombs set into
cliff faces.
Shady (1976:585, 1987a:86) reports that
pottery of the corresponding Cuelap phase has been found in
Cajamarca and Bagua, indicating long-distance exchange with
the highlands to the west and lowlands to the north.
Schjellerup's (1985) studies of agricultural earthworks such
as slope terraces between 3,200 and 3,800 m postulate
extensive cultivation of potatoes and other high altitude
tubers in the Chuquibamba area.
These activities, in
combination with camelid herding, constitute a subsistence
orientation much like that of other highland Central Andean
179
regions south and west.
Its prehistoric development in
Chachapoyas remains unclear.
Excluding the study area, virtually all of the
radiocarbon dates from Chachapoyas have been reported by
Schjellerup (1992), and none pre-date the Late Intermediate
Period.
These and other ancient settlements span the upper
elevations of Young's Moist Montane Zone and the lower
elevations of his Tropical Alpine Zone.
On a visit to the
village of La Morada near 2,000 m within the lower montane
forest, Schjellerup located a small complex with classic
Inca architectural features and a bath that she named
Pukarurni.
Lerche (1995) provides a sketchy description of
additional Inca remains at similar elevations within the
Huallaga watershed.
Most of the central Huallaga river valley remains
archaeological terra incognita.
DeBoer (1984) surveyed
selected areas of the floodplains and dry tropical forest
along the valley bottom north of the study area between
Bella Vista and the Pongo of Aguirre.
Significant for this
study is his interpretation of several sherd scatters
located not far from the Huallaga-Huallabarnba confluence
(HUA-4, HUA-5) as remains left by a multi-family house, or
maloca.
Late ceramics with crushed-sherd temper are much
like the material that central Huallaga potters produce
today.
An older style from Chazuta (Ibid.:114-115) and
Tarapoto (Myers 1981b) shows some design similarities to the
180
Curnancaya ceramics described further south (Raymond et al.
1975).
DeBoer concludes that central Huallaga styles are
typical of an ancient and wide-spread upper Amazonian
potting tradition.
featuring bowls,
A shared tri-partite shape inventory
jars and ollas probably has "great time
depth," and may be "indigenous to the montaiia"
1984:114).
(DeBoer
Excavated collections from the study area
described by Ravines (1978, 1981b) will be addressed in the
following section.
Archaeology of the Pataz-Abiseo Area
Before attempting to derive some understanding of
population distributions and regional economic relations
from archaeological data so far recovered in the PatazAbiseo area, the data's limitations should be understood.
The study area has not been extensively surveyed on either
side of the divide, although some site localities have been
recorded prior to Rio Abiseo National Park Research Project
(RANPRP) studies in 1985.
The sites that we visited in
Pataz and in the upper montane forest are typically those
with standing stone architecture well-known to the local
villagers.
The villagers are intimately familiar with sites
on the Marafion side of the divide, but, collectively at
least, they know little about sites within the Huallaga-side
montane forest.
The chief biases distorting perceptions of
the local prehistoric panorama are those posed by certain
construction materials (perishable vs. stone), and forest
181
cover.
The following paragraphs represent the current
understanding of selected features of Pataz-Abiseo
archaeology.
Pataz District
Several archaeologists passing through Pataz (usually
en route to Gran Pajaten) have observed some archaeological
features such as pictographs, chamber tombs and terraces.
During investigations in areas higher up and farther east
between 1985 and 1990, RANPRP archaeologists described and
mapped some of these features, and photographed the
collections of local farmers.
The following observations on
the archaeology of Pataz District stem primarily from these
observations.
In the Dry Forest Zone, only one site discovered
fortuitously in 1990 has been documented between the Marafion
and the town of Pataz.
A series of small terraces leading
up to a single ruined stone enclosure lie at approximately
1,400 m, 200m above the Marafion.
A few thin sherds from
fine painted bowls may indicate an Early Intermediate Period
occupation.
Clearly the Dry Forest Zone was utilized, but
it still awaits systematic research.
The lower portion of the Moist Montane Zone below the
town of Pataz has not yet received attention from
archaeologists.
In the surrounding heights, however,
villagers are familiar with two small clusters of ancient
stone structures crowning promontories that extend westward
182
into the Marafion canyon on either side of the Frances River
valley.
Rojas'
A visit by RANPRP archaeologists in 1990 confirmed
(1967:9) report of ruins above Zarumilla on the crest
of Cerro Cuyrnuy (3,380 m).
This site, called El Ushnu by
Zarumilla residents, consists of four oblong ascending
terraces reinforced by low field-stone walls that delimit a
total area of approximately 1.2 hectares.
Remains of at
least seven semi-circular stone walls can be discerned
through the vegetation, but recent fence construction seems
to have borrowed much of the ancient construction material.
All but one of the 24 sherds collected from the surface
pertain to large jars with long flaring rims.
The paste is
sand-tempered, and rim morphology is identical to that of
Late Horizon pottery illustrated by Bonavia (1968b) at Gran
Pajaten.
The odd rim sherd belongs to a neckless olla.
Zarumilla residents also report sherd scatters farther up
the ridge at Pampa Verde near 4,000 m.
El Ushnu's extraordinary view up and down the length of
the Marafion canyon suggests that it was one of several sites
established at strategic locations for geo-political
control.
From El Ushnu,
the stepped profile of Cerro Alto
Las Pircas (4,100 m) across the Frances valley (behind
Pataz) can be seen.
Pataz residents report that a small
cluster of stone walls is found there, but archaeologists
have yet to examine the site.
The Institute Nacional de
Cultura's inventory of archaeological monuments (INC
183
1983:Hoja 16-g) specifies late dates of occupation for a
ruined settlement on Cerro Tinajera (3,102 m) above the
south side of the Lavasen River.
RANPRP archaeologists did
not visit the site, nor could additional published
information on the settlement be located.
Of the known sites within the study area, only the site
on Cerro Tinajera at the northern extreme seems to be a
settlement, and therefore potentially indicates something of
pre-Hispanic Pataz population distributions.
Pottery that
has surfaced in fields surrounding Los Alisos (3,100 m)
appears to be late prehistoric in age.
Nearby funerary
chambers and pictographs previously described by Perez
(1969) and visited in 1990 on the slopes of Cerro Colpar may
also date to the Late Intermediate Period or Late Horizon.
The preliminary assessment of prehistoric settlement in the
Moist Montane Zone of Pataz District is that it may have
been lighter than expected.
However,
the bias favoring
reports of intact stone constructions may encourage specious
impressions.
Pataz settlement also appears sparse compared with
information from adjacent southern districts gathered during
surface surveys by Curtin (1951).
Pre-Hispanic settlements
documented in Parcoy, Buldibuyo and Chillia further south
typically feature circular stone constructions crowning
hilltops and ridgetops between 3,200 m and 3,800 m.
Among
several classes of observed architectural features, Curtin
184
observed several small terrace systems.
Building counts are
not provided, but some of these sites covered an area of
several hectares.
Stylistic evidence from fine painted
kaolin bowls and relief sculpture on stone slabs at
Nufiamarca suggest communication linkages south and west with
the Callej6n de Conchucos, the upper Santa Valley and the
Santiago de Chuco region during the Early Intermediate
Period (Campana 1988).
Aside from the Pataz-Abiseo area,
the archaeology between Parcoy and Uchucmarca remains
unstudied.
Some late sites in Cajamarquilla/Bolivar
Province have been briefly described (Vega 1979; Lynch
1992).
Virtually nothing is known of the archaeology
surrounding Bambamarca and Condormarca.
For the Tropical Alpine Zone at the crest of the
Marafion-Huallaga divide, Lennon et al.
(1989) describe four
classes of sites, none of which represent proper
settlements.
The first class consists of administrative
structures and bridge remains associated with the paved
"Inca" road purportedly linking Huanuco and Chachapoyas (Von
Hagen 1955), and the road spurs striking eastward toward the
montane forest.
These structures date to the Late Horizon,
but the road system is probably much older.
Rock shelters
such as Chirimachay (Cueva Negra) and Manachaqui Cave
comprise the second class.
Travellers and herders used
these for protection into the pre-Inca past.
Highly
dispersed and isolated house footings and windbreaks,
185
usually appearing as low stone circles or semi-circles, dot
the valley slopes and remain undated, as do the low (less
than 50 em) stone-faced platforms constructed near water
sources that constitute the fourth class of archaeological
remains.
The Abiseo Drainage and Central
Huallaga Lowlands
East of the Marafion-Huallaga divide only some upper
portions of the Abiseo's tributary Montecristo river valley
have been explored, but these upper montane forests
apparently harbored the densest pre-Hispanic settlement in
the Pataz-Abiseo area.
Sites such as Gran Pajaten, Cerro
Central, Las Papayas, La Playa and El Encanto (Pimentel
1967; Rojas 1967; Bonavia 1968b; Savoy 1970; Deza 1975-76;
Kauffmann 1983; Lennon et al. 1989) exhibit architectural
and ceramic styles similar to those reported elsewhere in
Chachapoyas.
While La Playa dates only to the Late Horizon
(Deza 1975-76; Cedr6n 1989), recent excavations by RANPRP
archaeologists at Las Papayas have produced radiocarbon
evidence for a Late Intermediate Period occupation (Church
1994:293).
Simultaneous investigations at Gran Pajaten yielded
pottery and radiocarbon dates confirming occupation during
the Early Intermediate Period, and perhaps the Early Horizon
(Ibid.: Table 1).
Maize was consumed, and perhaps
cultivated in the site vicinity, as evidenced by charred
186
maize kernels recovered from a charcoal lens dated A.D. 40 ±
60 (uncorrected) within Building No. 1's construction fill
(Ibid.).
Fine kaolin ware bowls imported from adjoining
highland regions indicate early long-distance interaction.
Culinary pottery from both Early Intermediate Period and
Late Horizon contexts features the necked jars with flaring
rims and semi-hemispherical bowls.
It has been suggested that similar ruined settlements
and terraces should be found on all of the hilltops
surrounding Gran Pajaten (Savoy 1965:4; Rojas 1967:11)
RANPRP archaeologists were able to confirm the presence of
an extensive settlement with an estimated 150 to 200
circular constructions across the valley at Cerro Central
(Lennon et al. 1989).
Between La Playa and Gran Pajaten
lies Las Papayas with an estimated 100 buildings resting on
a series of large terraces.
The system of terraces below
Gran Pajaten covers over 50 hectares, but the total extent
of terrace and settlement systems in the valley cannot be
estimated without more systematic survey coverage.
Also,
the Montecristo valley below 2,600 m remains unexplored by
archaeologists.
Ravines'
(1978) survey and excavations in the lower
Abiseo, Huallabamba and Pachicilla valleys constitute the
only archaeological investigations in the central Huallaga
valley premontane forest.
The report provides little
information regarding site environments and how the survey
187
and excavations were conducted.
and 440 m.
Most sites lie between 390
Ravines excavated three ceramic "complexes" and
stratigraphic evidence for a relative chronology at the site
of Santa Rosa (S-Huay-5).
Sand-tempered pottery with
excised decoration lay beneath sherds from vessels with
applique decoration assigned to the Huayabarnba Complex.
The
Huayabarnba Complex is divided stylistically and temporally
into Huayabarnba 1 and the more recent Huayabarnba 2.
Ravines
regards Huayabarnba 1 pottery ancestral to both the overlying
16th century Jerusalen Complex, and Andean styles of
Chachapoyas (1978:531).
Crushed-sherd temper characterizes the later complexes,
while potters utilized fine sand temper to fabricate the
earliest complex that Ravines and DeBoer (1984:117) agree is
related to the Central Ucayali Shakimu styles (circa 650
B.C.).
Limited information on a small sample of vessel
shapes impedes judgment regarding whether or not the undated
Huayabarnba 1 and 2 complexes participated in the upper
Amazonian tradition described by DeBoer.
The Huayabamba
shape inventory appears to exhibit variability exceeding the
upper Amazon tradition's.
Three of the four most clearly
identified vessel shapes belonging to the 16th century
Jerusalen Complex (Ravines 1978:Laminas l:a, 2:1 and 2:2)
more strongly suggest inclusion within the upper Amazonian
tradition, but the small sample size still hinders
evaluation.
188
Considering the Pataz-Abiseo area as a whole,
it seems
clear that settlement density by the time of Spanish
conquest was weighted to the Huallaga watershed Tropical
Montane Rain Forest Zone.
The extensive and sophisticated
terrace and building constructions at forest sites like Gran
Pajaten represent investments in time, labor and materials
that were not expended in potentially productive areas of
Pataz.
Nor does the topography of the Montecristo valley's
upper montane forest favor settlement, at least for
agricultural purposes
(Church 1994).
Sampling problems
restrain attempts to place the Pataz-Abiseo settlement
distributions into a regional context.
Most important is
the question of whether or not Montecristo valley settlement
density is representative of densities in neighboring
valleys north and south.
Summary
This chapter has considered events during particular
time periods in the Pataz-Abiseo area in an attempt to
identify economic and demographic transformations that have
obscured perception of this area as an active node in preHispanic Andean-Amazonian commerce.
Most remarkable is the
demographic shift of populations from the Tropical Montane
Rain Forest Zone of the Abiseo drainage to the Moist Montane
Zone of Pataz that accompanied the European incursions.
Achieved against a backdrop of high mortality and eastward
flight, Toledo's forced relocation of these populations into
189
nucleated highland settlements probably constitutes the
single most important demographic event.
It also seems to
represent the moment at which previous highland-lowland
linkages were rendered structurally untenable.
The following paragraphs will attempt to abstract other
salient points from the ethnographic, ethnohistoric and
prehistoric data recounted within this chapter.
First,
these data confirm that a highland-lowland boundary did
separate cultures with markedly disparate developmental
features on the eve of Spanish conquest.
The review of
ethnohistoric and archaeological evidence provides a view of
loosely organized highland societies with small-scale
settlement hierarchies, residence patterns emphasizing
nuclear family households, a subsistence pattern typical of
the Central Andes, and a pottery style commonly associated
with highland Quechua-speakers.
While the socio-political
organization of central Huallaga groups like the Hivitos
remains virtually unknown, both ethnohistoric and
archaeological evidence demonstrate that residence patterns
emphasized extended and/or multi-family households and that
subsistence activities focused on manioc and other lowland
produce.
Ravines' Jerusalen Complex may represent the
associated lowland Amazonian pottery.
Early church interest in extending the highland
doctrina of Cundurmarca into the eastern lowlands likely
indicates a desire to tap into the pre-Hispanic tradition of
190
interaction between southern Chachapoyas and central
Huallaga societies as the Inca had previously.
Reports that
both the Inca and Spanish officials similarly incorporated
the Amazonian Cholones and Hivitos within the administered
Andean province of Chachapoyas support the notion of an
ancient and durable exchange linkage between highlands and
lowlands in the northeastern montane forest.
Scholars
suggesting that the Inca failed to conquer the Amazonian
forests overlook the fact that the Inca province of
Moyobamba (Rowe 1946:187) is covered by tropical montane and
premontane forests, and Moyobamba lowlanders probably had
similar ancient traditions of trading (DeBoer 1984:15) and
mediating trade (Reeve 1994).
Montane forest settlements such as Sucos, Yaro de
Puymal and Gran Pajaten seem to have been politically
independent nodes in a pre-Hispanic communication network
that linked Central Andean populations to the expansive
lowland interaction spheres described by Lathrap, Reeve,
DeBoer (n.d.) and others.
Savoy's account of finding
circular stone constructions at the abandoned Franciscan
mission of Jesus de Pajaten, and a paved road near the
junction of the Jepelache and Pajaten Rivers at
approximately 400 m (Savoy 1970:148) provides provocative
but sketchy evidence suggesting "highland" settlement at a
strategic point of river navigation situated on the
Huallabamba River waterway.
The Huallabamba, easily reached
191
by Huallaga River trade expeditions originating downstream,
probably served as an important destination offering access
to a large number of potential highland "consumers."
For
Central Andean societies west and south of the study area,
the Huallabamba could have provided the closest and most
direct access to lowland Amazonian riverine interaction
networks.
Thus the Hivitos and southern Chachapoyas natives
assumed ostensibly lucrative intermediary positions
channelling goods to these Central Andean societies.
Demographic collapse in the study area undoubtedly
sparked breakdowns in local political economic structures
and kinship-based exchange obligations, contributing to a
rapid disintegration of long-distance exchange networks.
The prospect of administering a sparse and isolated native
population diminished Spanish economic interest in the area,
despite its purported mineral wealth.
Nor did the Jesuits
see fit to continue support for the distant doctrina of
Cundurmarca.
Ultimately, the sudden population decline
exacerbated the seemingly complete acculturation of
remaining southern Chachapoyas populations into Peruvian
mestizo culture, and the virtual disappearance of native
languages.
While Colonial Period southern Chachapoyas populations
attempted to cope with sudden soaring mortality,
the
political authority within Cholon and Hivito territory noted
by Mogrovejo in 1593 was likewise undermined by contagious
192
diseases, intensified warfare and slave raiding.
The first
Franciscan missionaries in 1676 found the Cholones and
Hivitos "scattered throughout the forest neither recognizing
cacique nor superior other than their elders to whom they
confer special respect"
(Amich 1854:76).
Hopes that highland-lowland commerce could be restored
dimmed slowly during the Colonial Period.
Bautista de
Escobar's 1653 attempt to direct De la Riva Herrera's route
of conquest through Condormarca, Sucos and Puymal represents
one such effort.
The importance that the central Huallaga
lowlanders attached to regaining access to highland trade is
demonstrated by their emergence twice in the highlands (in
1654 and 1670) attempting to secure a missionary presence in
their homeland.
By then however, the Hivitos' highland
trade partners had either died, fled or been forcibly
removed from the upper montane forests.
After the burning
of the mission at Jesus de Pajaten in 1801, both traditional
and newly forged communication routes bypassed Condorrnarca,
Sucos and Puymal.
Highland-lowland interaction consequently
diminished to a pale remnant of its former character by the
beginning of the 19th century.
Epidemics that continued to
strike highland communities like Huaylillas and
Cajamarquilla during the mid-nineteenth century (Raimondi
1900a:37, 1900b:l39) denied highland populations the
opportunity to recover.
According to Mariategui, the
marginalization of Peruvian Amazonia was exacerbated by the
193
end of the rubber boom after World War I
(1971:162, Note 4).
Mid-twentieth century construction of roads descending the
eastern Andean slopes through Moyobamba to the north, and
Huanuco to the south, finally assured the consignment of
this region to an economic oblivion mitigated only
periodically by mining booms.
Information gleaned from the small quantity of
available Spanish documentation does not clearly elucidate
the nature of pre-Hispanic nor early Colonial Period
economic activities within and adjacent to the study area.
However, interaction at the time of Spanish conquest
probably occurred on several scales.
Local and regional
exchange, which are usually imbedded in kinship networks in
the Central Andes (Morris 1978), emphasized the east-west
flow of goods.
In the Pataz-Abiseo area, the geographic
distribution of life zones which, like the cordillera, are
aligned north to south, promotes movement east and west as
the most efficient means of accessing the widest variety of
ecozones apt to serve as desirable production zones for
certain crops or other resources.
The east-west economic
interaction recorded by Brush (1977) at Uchucrnarca may be
considered a modified version of prehistoric local or
regional exchange that once characterized the MarafionHuallaga divide.
North-south interaction for which there is
indirect evidence probably involved the exchange of
information and goods unrelated to subsistence.
Within the
194
study area, montane forest populations at Yaro de Puymal and
Gran Pajaten may have maintained production zones both deep
in the forested lowlands and in drier Marafion-side
ecological zones, although confirmation requires further
research.
Opportunities for controlling interregional exchange
along an important traditional trade route probably
represented a powerful inducement to settle the montane
forests east of Condormarca, although exploitation of
montane forest resources was also a likely consideration.
Interregional interaction extended beyond kinship bonds and
connected centers of political power across boundaries
separating coastal, highland and tropical forest cultural
traditions.
With Spanish conquest, the collapse of Andean
political and economic structures led to the virtual
elimination of interregional interaction.
Networks across
the Andes suddenly contracted or disappeared leaving
atrophied fragments of local and regional exchange that
mostly served domestic economies.
Ancient settlements like
Sucos, Puymal and Gran Pajaten that figured importantly in
interregional interaction rapidly became functionally
irrelevant in the scheme of emergent Spanish hegemony.
CHAPTER 4
MANACHAQUI CAVE AND ITS ENVIRONMENTAL CONTEXT
There is said to be a superb view of the Montana from
the summit, but the clouds {almost within reach of the hand)
boiling up from the great deep below, effectually cut it off
and we could see nothing. When we had got some distance
down and obtained a view through an opening in the thick
growth of the mountainside, we looked down upon the most
rugged country I have ever seen. There seemed to be no
order or regularity in the hills which were thickly covered
with forest; but the whole had the appearance of the surface
of a vast boiling caldron suddenly stricken motionless.
{William L. Herndon 1952 [1854] :66)
Archaeological investigations at Manachaqui Cave
represent the most recent stage of continued research in the
vicinity of the Rio Abiseo National Park begun in 1985 by
archaeologists from the University of Colorado and the
National University of Trujillo, Peru under the direction of
Thomas Lennon, Miguel Cornejo Garcia and Segundo Vasquez
Sanchez.
During 1985 and 1986, survey, mapping and
excavations focused on Montecristo valley sites situated
between 2,600 and 2,900 m within the Tropical Montane Rain
Forest Zone.
During these campaigns, Manachaqui Cave served
as a convenient place for archaeologists, field crews, muledrivers and cargo animals to spend the night near the crest
of the Marafion-Huallaga divide during the two-day trip
between Pataz and La Playa staging areas.
Most of this
route follows pre-Hispanic paved roads that pass close to
195
196
rockshelters such as Chirimachay (Cueva Negra) and
Manachaqui Cave.
Raimondi (1900b:127) obliquely refers to
Manachaqui Cave as a local landmark within his description
of travel from Condormarca to Pataz in 1860, and the Gran
Pajaten expedition of 1965 sponsored by the Peruvian
government was the first of many to camp at Manachaqui
(Rojas 1967:9; Deza 1975-76:45; Kauffmann 1980:27).
One of the research priorities of the Rio Abiseo
National Park Research Project was to unearth stratified
cultural sequences in order to date the antiquity of human
presence in the montane forest.
While the archaeological
potential of the many rockshelters in the Tropical Alpine
Zone was always suspected, Manachaqui Cave's promise became
particularly obvious in 1986.
Between the 1985 and 1986
University of Colorado-University of Trujillo field seasons,
villagers from Pataz cut sod blocks from the ground surface
in front of the cave in order to plug openings at the cave
rear,
thereby creating a dryer, more comfortable interior
space to sit, cook and sleep.
The sod removal exposed
sherds and lithic remains that the archaeologists observed
and collected at the end of the 1986 season.
Cursory examination of the artifacts confirmed
Manachaqui Cave's potential to yield a sequence of human
occupation.
Although the project would have preferred
recovery of an archaeological sequence from within the
continuous montane forest, budget constraints prompted
197
Lennon to target Manachaqui Cave and other rockshelters for
test excavations in 1988.
Despite the success of these
excavations in recovering a surprisingly early sequence
dating from the early Preceramic Period, the project was
indefinitely suspended because of Lennon's competing
commitment to archaeological research in the U.S.A.
I had
participated in the Rio Abiseo National Park Project from
its inception as a graduate student at the University of
Colorado-Boulder, and in 1989 Lennon offered me the
opportunity to take over the 1990 Manachaqui Cave
investigations for my dissertation research.
This chapter focuses on Manachaqui Cave's immediate
natural surroundings, and its present and past usage as a
shelter.
The first of two sections situates Manachaqui Cave
within the Pataz-Abiseo study area, and characterizes its
changing natural surroundings, especially the Holocene
geology, climate, flora and fauna.
Modern floral and faunal
data serve as important baseline information with which to
identify ecological zones and habitats from which the
archaeological flora and fauna of Manachaqui Cave were
procured.
The second section includes a description of the
Manachaqui Cave site complex and the network of pre-Hispanic
paved roads with which it appears to be
associated.
f~~ctionally
The problem of Manachaqui's prehistoric
function or functions is more precisely defined, and a
working hypothesis presented.
198
The Changing Manachaqui Valley Environment
Archaeological investigations at Manachaqui Cave
benefit from several years of multi-disciplinary fieldwork
in and around the Rio Abiseo National Park.
Young's
contribution to study area zonation has already been
described.
Between 1985 and 1990, the Rio Abiseo National
Park Research Project also sponsored geological,
palynological and zoological field studies.
These provide a
rich set of data on Manachaqui Cave's surrounding
environment lacking from most archaeological localities.
Most importantly, they provide baseline information with
which to evaluate evidence for past environmental
alterations.
The Manachaqui Cave site complex lies in the Tropical
Alpine Zone at 3,625 m, 7= 42' south latitude and 77c 30'
west longitude, above the south bank of the Manachaqui River
(Fig. 7).
The 4,300 m peaks of Cerro Suitacocha and Cueva
Negra flanking the Valley to the north and south are typical
of the jagged Marafion-Huallaga cordillera which Weberbauer
(1945:98) describes as weathered into fantastic shapes of
"towers, horns and needles."
Pleistocene epoch glaciation
has carved Manachaqui into a "text-book" example of a
shaped valley.
u-
The Manachaqui River descends gradually for
13 kms from Laguna Brava at the Marafion-Huallaga divide and
continues westward through small lakes and marshes to the
terminal moraine at Laguna Baja.
Below Laguna Baja,
the
199
river enters the steep, V-shaped valleys below and combines
with other streams tumbling out of the Suitacocha and
Chirimachay valleys to form the Lavasen River tributary of
the Marafion.
Geology and Pleistocene Glaciation
Despite some efforts by Wilson and Reyes (1964) to map
the geologic strata in Pataz quadrangle during the 1960s,
Birkeland et al.
(1989) regard the bedrock of the Manachaqui
valley as poorly known.
Apparently Wilson and Reyes devoted
minimal attention to the less accessible Tropical Alpine
Zone east of the Marafion, and none at all to the Tropical
Montane Rain Forest Zone east of the Marafion-Huallaga
divide.
Wilson and Reyes' stratigraphic profileD includes
part of the Tropical Alpine Zone in the western portion of
the Pataz-Abiseo study area (1964:Fig. 5).
According to
these studies, Manachaqui valley bedrock belongs to the
Lavasen Formation and consists of Tertiary age volcanic
materials.
The Lavasen Formation caps both the pre-Cambrian
age Marafion Complex, which consists of mica, schist, metaandesite and shale, and upper Cretaceous granodiorite
intrusives that can be seen in outcrops in the Moist Montane
Zone below.
The Geological Map of Peru (Institute de
Geologia y Mineria 1975) shows the Lavasen volcanics and the
mostly metamorphic Maranon Complex as predominant bedrock
strata beneath the montane forests of the Abiseo River
drainage.
However, the observation of sedimentary
200
formations containing beds of limestone, sandstone and slate
surrounding Gran Pajaten at 2850 m attest that Huallaga
basin geology remains for all intents and purposes unknown.
The Lavasen Formation upon which Manachaqui Cave sits
is composed of pyroclastic material, especially dacitic and
rhyolitic tuffs with andesites and volcanic breccias (Wilson
and Reyes 1964:48).
However, Birkeland et al.
(1989:56)
observed mostly granites and quartzites during their work in
the glacial valleys of the Tropical Alpine Zone.
A pebble
count from Manachaqui valley alluvium by Rodbell (1991: Fig.
2.1) showed a predominance of quartz accompanied by
granodiorite and volcanic rocks.
These disparate
observations await reconciliation and exploration by
systematic geological survey.
Landforms of the Tropical Alpine Zone are largely the
product of Pleistocene epoch glacial scouring and Holocene
epoch glacial deposition and erosion.
Terminal Pleistocene
glaciation covered Manachaqui Cave's current location and
extended down to elevations of 3,600 in western-facing
valleys, but it was more substantial on the eastern side of
the divide, reaching down to 3,200 m.
Radiocarbon dating of
organic material in Manachaqui Valley glacial deposits by
Birkeland et al.
(1989a, 1989b) documents the first retreat
of glacial ice by 12,100 ± 190 B.P.
Based upon additional
indices, Rodbell (1991:16) estimates that deglaciation began
by 13,500 B.P., and that all cirques (and, hence, all
201
valleys) were ice-free by approximately 10,000 B.P.
Deposition of loess and the formation of soils followed
rapidly after deglaciation, and landforms have remained
relatively stable to the present (Birkeland et al. 1989:63).
Holocene Environmental Change
According to ONERN (1976),
the uppermost slopes west of
the Marafion-Huallaga divide are classed within the Tropical
Subalpine Wet Paramo life zone.
However, fragments of
Abiseo watershed forest spill westward over the divide into
the upper Manachaqui valley and surround the Laguna Brava
cirque.
Young's studies of forest relicts thriving in
protected Tropical Alpine Zone micro-environments indicate
that "the spatial distribution of these timberline forests
is essentially produced and maintained by the grassland
fires, which can kill unprotected trees and shrubs ... "
(Young 1991:28).
He estimates that, because of past and
present burning practices,
"the elevational limit of closed
forest on most valley bottoms has probably been lowered by
about 500 m"
(Young 1993:277).
If the hypothetical limit of
closed forests reached 3,800 or 3,900 m, then Manachaqui
Cave lay within tropical montane forest prior to
anthropogenic deforestation.
The question then becomes,
When exactly was Manachaqui Valley deforested?
To reconstruct the local paleoenvironment, sediment
cores were extracted from Laguna Baja at Manachaqui Valley's
terminal moraine by Rodbell and Hansen (1990; personal
202
communications 1993-94}.
Hansen's analysis of pollen
samples from the cores provides new information on upward
and downward shifts in ecological zonation during the late
Pleistocene and early Holocene epochs.
Preliminary results
from these unpublished studies indicate that Montane Wet and
Montane Moist Forest life zone species advanced into the
valley immediately following the glacial retreat and
climatic amelioration around 12,000 B.P.
A reduction in
forest species accompanied by an increase in paramo species
between 12,000 and 10,000 B.P. may signal a period of drier,
cooler conditions.
During the early Holocene (after 10,000 B.P.} Montane
Wet Forest species again predominated in pollen counts until
approximately 6,000 B.P.
A subsequent decrease in forest
species coincides with a period of greater regional aridity
reported from the eastern lowlands of Ecuador (Bush and
Colinvaux 1988}.
Evidence from ice cores recently extracted
from Mt. Huascaran in the north-Central Andes near Huaraz
document similar periods of early Holocene Amazon forest
expansion followed by a cooling trend and lowering
timberline after 5,000 B.P.
(Thompson et al. 1995}.
After
4,000 B.P,. the Manachaqui core shows increased quantities
of charcoal co-occurring with pollen indicators of
agricultural activities that provide evidence of burning and
extensive anthropogenic environmental alteration.
Pollen
evidence for subsequent environmental changes is not easily
203
interpreted because human activities appear to have
increasingly influenced the composition of local vegetation.
While timberline forest may have periodically expanded and
contracted, it is reasonable to conclude that Manachaqui
Cave has stood in relatively open subalpine paramo since
4,000 B.P./2,000 B.C.
A record of later Holocene environmental change is
lacking for the Pataz-Abiseo area.
Cardich (1985} considers
the effects of Eddy's (1977} postulated global climatic
oscillations on the upper limits of cultivation and cultural
development in the Central Andes.
In Cajamarca, frost
limits the present elevational limit to 3,700 m although
Cardich (1985:305} noted evidence of past cultivation up to
3,900 m.
In Cajamarca and the Pataz-Abiseo area, warmer
temperatures may have facilitated cultivation up to 3,900 m
during the Late Preceramic, Initial, Early Intermediate and
Late Intermediate Periods (Ibid.: Fig. 6.14}.
How severely
cultivation limits would have been depressed during
relatively cool intervals is more difficult to determine.
Cardich's studies concur with investigations farther south
in Junin Department that suggest periods of cooling during
the Middle Horizon and terminal Late Intermediate Period
(Wright et al. 1989; Seltzer and Hastorf 1990}.
Flora and Fauna
The recent floral and faunal inventories in the Rio
Abiseo National Park were the most systematic and thorough
204
yet attempted in Peru's montane forests.
General
characteristics of Young's Tropical Alpine macro-ecological
zone were described in the previous chapter, and its floral
composition is detailed elsewhere (Young 1993:Table 5).
Graminoids and shrubs typical of Andean paramo environments
are the most conspicuous elements of the zone's flora.
Trees found in protected micro-environments are Gynoxys spp.
and Weinmannia microphylla.
These species presently occur
in closed montane forests on both sides of the MarafionHuallaga divide.
Grasslands,
forest remnants and other
habitats recently sampled by biologists harbor a diverse
fauna, much of it endemic to the eastern slopes, as well as
numerous species new to science.
Fauna of the Tropical Alpine and Tropical Montane Rain
Forest zones in and around the Rio Abiseo National Park was
inventoried by biologists in 1987, 1988 and 1989 (APECO
1988a, 1988b, 1991).
Collecting by various methods took
place at broadly defined locations such as lakes and river
valleys along a transect beginning on the western side of
the divide and terminating near 2,000 min the lower montane
forest.
Specialists worked independently each season to
sample birds, mammals, amphibians and reptiles in a variety
of habitats such as dry grasslands, floodable grasslands,
rocky areas, isolated forest patches and continuous forest.
Fish were sampled along the same transect.
Overall
methodologies have always been geared to logistical
205
feasibility in this remote region.
The authors emphasize that a number of constraints
precluded the collection of truly representative samples
from each locality.
The Manachaqui Valley was not
extensively sampled although collections at nearby
Chirimachay and La Plap valleys augment the total sample
from the Marafion-side Tropical Alpine Zone.
During his 1919
expedition to Pataz, Weberbauer (1920:6) noted the abundance
of bear and deer for hunting in the "puna."
Among the most
conspicuous open grassland fauna today are deer, fox,
armadillos and a wide variety of rodents and herpitofauna.
Biologists did not tabulate species according to ecological
zone although species distributions were occasionally
estimated.
They conclude in their final report that the
large data base obtained during the three field seasons
remains insufficient to adequately characterize the region's
fauna (APECO 1991:52).
The Manachaqui Valley and Manachaqui Cave
Between 1986 and 1990, RANPRP archaeologists surveyed
the Tropical Alpine Zone valleys of Chirimachay, Suitacocha,
Manachaqui, La Plap and Chochos at the northwest corner of
the park.
The complex of archaeological features at
Manachaqui Cave has been referred to in reports as M-1, the
first of a series of sites recorded during systematic
archaeological reconnaissance of the Manachaqui Valley's
bottom and accessible lower slopes (Lennon et al. 1989).
206
The most frequent site type occurring in all but La Plap and
Chochos valleys are rockshelters.
In Manachaqui Valley,
survey during the 1988 and 1990 seasons documented 12 sites.
Six are rockshelters.
Two of these (M-7 and M-8) were
sampled by test excavation, but their cultural remains have
yet to be analyzed.
The six additional sites in Manachaqui Valley include
the remains of a shrine or "huanca"
(M-5); a small,
naturally occurring niche between stones on the valley slope
that contained a nearly complete Abiseo Style vessel (M-3);
the remains of a bridge foundation (M-4); and a pair of low
platforms on either side of the Manachaqui River where it
breaks through the moraine that dams Laguna Baja (M-9).
In
1986, Birkeland reported a band of gray and green chert
crossing the pre-Hispanic road approximately 2 kms east of
Manachaqui Cave.
M-11 in 1990.
We recorded this suspected quarry site as
Also in 1990 we designated M-12 the stone
masonry foundations of a bridge crossing the Manachaqui
River's deep ravine below Laguna Baja.
crossing this bridge (1900b:127).
Raimondi describes
Data from these sites
will be detailed only when deemed pertinent to
interpretations of Manachaqui Cave's archaeology.
Manachaqui Cave
The rockshelter traditionally referred to by local
villagers as Manachaqui Cave is formed by an accumulation of
large granite boulders that apparently fell from the heights
207
of the valley's southern slopes during late Pleistocene
glacial ablation (Plate I) .
These boulders sit on a mound
of gravely alluvium beneath a network of gullies draining
the south slope.
The principal gully runs with water year-
round.
Because Manachaqui Cave consists of fortuitously
clustered boulders, it differs from well-known highland
Andean caves like Guitarrero Cave (Lynch 1980), Pacharnachay
(Rick 1980) and Pikimachay (MacNeish et al. 1981) that were
shaped by mechanical or chemical weathering of exposed
bedrock.
The boulders are clustered in such a way that
large interstices between and beneath them form three small
"shelters" labeled M-1A, M-1B and M-1C facing north, west
and east respectively (Fig. 8).
M-1B and M-1C both yielded
small surface collections of sherds, but M-lA constitutes
the most spacious and intensively utilized of the three
shelters.
It therefore received virtually all of our
attention.
The mouth of M-1A opens to the northeast over an ample
berm (Plate II).
M-1A's interior is approximately four
meters wide and three meters deep with a surface area of
nearly 13
m~.
The shelter's ceiling is presently little
more than one meter from the ground at its highest point.
Spaces between the boulders leave openings at the rear of MlA and on the eastern side.
Deeper within the cluster of
boulders is a large open space partially exposed to rain and
208
dripping water.
A 1 x 1 m excavation unit revealed shallow
deposits and few cultural remains, indicating infrequent
use.
During prehistory, M-1A's inhabitants periodically
sealed the principal or anterior shelter from wind, rain and
draining water with stones and sod.
The shelter's berm (or
talus slope) approaches two meters in height and spreads
over an estimated area of 130 m2 •
It contains trash midden,
stone rubble and the scant remains of eroded living
surfaces.
In total, M-1A's cultural deposits cover a
horizontal area of nearly 145 m2 •
The pre-Hispanic road and a rockshelter designated Site
M-1D (formerly M-2) lie approximately 40 yards due north of
M-1A.
A great 9 x 7 m boulder nearly 10 m high sheltered
brief occupations sampled with a 1 x 1 m excavation unit in
1990.
Between M-1D and the base of M-1A's berm, an area
eight meters in diameter was leveled, and two concentric
stone circles or semi-circles were laid.
referred to as Site M-5,
Previously
this feature was re-labelled M-1E
in 1990 because of its presumed functional association with
Site M-1.
Its construction apparently occurred late in M-
1A's history of occupation.
The builders cut into the base
of the berm to lay the outer stone ring.
Excavations
uncovered a long, prismatic stone lying inside the inner
circle that once stood upright in the center of the stone
arrangement as the previously mentioned shrine or huanca.
Other nearby and somewhat enigmatic archaeological
209
features may have been functionally associated with Site M1.
On the slope due west of M-1A, a small, but prominently
situated shelter commands clear views up and down the
valley.
Several stone alignments approximately 40 m north
and west of M-1A are difficult to delineate and interpret,
but suggest that portions of the slope below the M-1A were
modified for unknown reasons.
Across the pre-Hispanic road
and situated on the valley bottom nearly 100 m northeast of
M-1A is a low L-shaped wall.
The long portion running east
and west continues approximately 15 m, and from its western
end a much shorter wall extends northward.
been partially dismantled.
These walls have
About 100 meters west-northwest
of the L-shaped wall is an isolated 2 x 1 m rectangular
stone platform less than half a meter in height.
Without
excavation the functional assessment of these features is
highly speculative.
Some of them may not be prehistoric.
The Pre-Hispanic Road System
For decades Andean scholars have been aware that a preHispanic paved road (commonly referred to as the Inca Road
or Highway) runs along the crest of the Marafion-Huallaga
divide (Raimondi 1900a:36-40, 1900b:123-130; Weberbauer
1920:7; Von Hagen 1955; Vega 1979:25).
This north-south
artery may have once connected centers of Inca
administration at Huanuco Pampa and Levanto near the modern
city of Chachapoyas (Von Hagen 1955).
For modern villagers
the pre-Hispanic road remains the principal route of travel
210
between towns strung along the divide such as Pias, Pataz,
Condormarca, Bambamarca and Bolivar.
Residents of
Condormarca utilize the road to run herds of cattle in the
Tropical Alpine Zone, and they continue to maintain it
through periodic communal labor.
RANPRP scientists and
personnel have habitually utilized the stretch above Pataz
to travel north as far as Chirimachay.
Hyslop (1984:233) observes that "stone paving is
primarily a water-control devise on Inka roads."
Accordingly, roads crossing the study area are more often
paved than not.
The main pre-Hispanic road and associated
spurs now show varied states of preservation depending to a
large degree upon whether or not they have been refurbished,
regularly used and maintained following the Spanish
conquest.
The principal north-south artery has received the
most attention from modern villagers, but bridges have not
been repaired.
The degree to which the road's pre-Hispanic
width has been maintained is not clear, but it generally
does not exceed three meters.
Narrow drainage channels of
original Inca design cut the road into segments where
drainage is needed.
Artificial causeways have not been
observed but the road often takes advantage of glacial
moraines to cross swampy terrain.
The ancillary roads
branching eastward toward the forest are more difficult to
follow as they are typically covered by vegetation and often
by soil.
211
At least two ancillary roads or road systems departed
eastward from the Huanuco-Levanto artery in the Pataz-Abiseo
area.
Weberbauer (1920:8, map on p.13) mentions a road east
of Pias that descends the valley of Pampa Hermosa to the
upper Tumac River valley.
While this road has not yet been
examined by archaeologists, RANPRP scientists have travelled
a second road system composed of two east-trending spurs
that join at Laguna Empedrada before continuing down the
upper Montecristo valley, ultimately to disappear under
dense vegetation.
These two spurs provide access to the
Montecristo valley from the north and south.
From the north, the spur henceforth referred to as
Tragaplata-Empedrada departs from the Huanuco-Chachapoyas
artery at Laguna Tragaplata and descends southeastward into
the Pefia Blanca Valley.
Entering the Chochos Valley, it
curves due south and gradually climbs to the junction at
Laguna Empedrada.
From the south, the Chirimachay-Empedrada
spur departs from the artery northeastward at Chirimachay
[Cave].
Then it crosses a high ridge and descends into
Manachaqui Valley, passes Manachaqui Cave and climbs over
the Marafion-Huallaga divide to Laguna Empedrada.
Remains of
rectangular stone buildings are situated along the road at
points of high visibility.
Site C-8, Los Paredones and
Mirador (Lennon et al. 1989) are "control sites" as defined
by Hyslop (1984:315-316).
Like the north coast sites that
Hyslop views as "the best evidence for a series of control
212
points" (Ibid.) these stations are found at the entrances of
valleys.
They are invariably positioned to view movements
along the roads from optimal vantage points.
Test excavations at C-8, Los Paredones and a larger
complex of rectangular structures, C-1 near Chirimachay,
produced small numbers sherds exhibiting late prehistoric
local and Inca styles (Lennon et al. 1989:47) suggesting
that the entire complex of control installations was built
under Inca direction.
Indirect evidence hints at greater
antiquity for the paved road.
The confirmation of camelid
remains in Manachaqui Cave's Early Intermediate Period
strata prompts the observation that loaded caravans could
not have repeatedly crossed the swampy alpine grasslands and
entered the muddy montane forest without the benefit of
paved surfaces like those that characterize the extant
"Inca" road.
Hypothetical Functions of Manachaqui Cave
One of Manachaqui Cave's most striking characteristics
is the integrity of its cultural deposits.
Until 1985-86
they remained essentially undisturbed since the shelter was
last utilized during the Late Horizon.
In contrast,
rockshelters along pre-Hispanic roads examined in the Pefia
Blanca valley and east of Bolivar have been shoveled out and
"renovated"
(personal observation, 1989).
The most altered
rockshelters are those in Tropical Alpine Zone valleys where
particular families have long held grazing rights.
At M-1A,
213
the only artifacts recovered from the upper layers and
clearly post-dating the Late Horizon were left by recent
scientific and tourist expeditions camping en route to Gran
Pajaten.
Lack of Colonial Period artifacts suggest that the
Chirimachay-Empedrada spur was not habitually utilized by
missionaries or traders between the 16th and 20th centuries.
In fact,
the Montecristo valley road may descend the eastern
slopes no further than the upper montane forest, although it
conceivably connects to other road networks reported deep
within the forest (Savoy 1970; Muscutt et al. 1994:6;
Muscutt, personal communication 1994) .
Modern Uses of Highland
Rockshelters
In recent decades, villagers from Pataz have sheltered
at Manachaqui Cave while tending livestock, hunting, fishing
and guiding outsiders on the three-day journey to Gran
Pajaten.
Between the modern towns of the Marafion-Huallaga
divide, a number of well-known rockshelters serve as
convenient refuges for travelers.
The names of some appear
in the accounts of Raimondi and Weberbauer, although none
have been investigated archaeologically.
In north highland
Peru only Guitarrero Cave in the Callej6n de Huaylas (Lynch
1980) has been intensively studied.
In the puna Andes
farther south, caves and shelters above 4,000 rn are still
sporadically, seasonally and even permanently utilized by
herders, and occasionally by their animals (Ravines 1971:18;
214
Cardich 1973:16; Lavallee 1977:69).
Flores-Ochoa (1979:104)
notes that traveling herders leading alpacas in the southern
highlands of Puno Department habitually camp in caves during
trading excursions.
During fieldwork in the Junin puna,
Rick investigated a small shelter known to modern herders as
Hotelmachay de Pampacancha (literally "Hotel Cave of
Pampacancha") in reference to its hotel room-like appearance
(Rick 1980:240) .
While the name "Hotelmachay" implies overnight visits
by journeying highlanders, the cave at Pampacancha is also
known to herders as Pintadomachay because of pictographs
within.
The painting and/or pecking of rock surfaces to
shape images with sacred meaning is a particular form of
ritual activity performed in and around highland Andean
caves during prehistory.
Rick (1980:56) reports that
natives traditionally left ritual "offerings of coca leaves
and liquor to the huamanis (cave spirits)" at the
rockshelter Pachamachay.
Caves figuring in Central Andean
mythology are pacarinas, considered portals to Ucu Pacha,
the underworld from which humans, plants and animals once
emerged to populate the surface world or Cay Pacha (Arriaga
1968:24; Valcarcel 1964:137, 144).
In Pataz, residents
refer to a cave in nearby Suitacocha valley as the "cueva
del bruj o" (the sorcerer' s cave) .
Oral traditions recall
its secret use by local shamans for rituals, and local
villagers still will not willingly enter.
215
Ancient Oses of Highland
Rockshelters
Highland Andean rockshelters may have served
prehistoric functions similar to those of today.
However,
during the intervening centuries, Andean societies have
undergone such massive structural transformations that
substantial qualitative and quantitative changes in material
patterning associated with functions like herding and
hunting should be expected.
The previous chapter documented
sudden and radical 16th century changes in southern
Chachapoyas local and regional economies.
During
prehistory, differences in material culture and animal
species associated with hunting and herding activities
should correspond to periodic modifications of subsistence
priorities.
Changing socio-political contexts presuppose
differing behavioral correlates for hunting during the
Preceramic Period, Late Horizon and late 20th century, and
we may expect concomitant alterations in Manachaqui Cave's
material patterning.
Most useful interpretive accounts of hunting and
herding activities during the Initial Period and subsequent
periods derive from excavations of rockshelters in the Junin
puna (e.g. Lavallee 1977, Lavallee et al. 1982; Matos and
Rick 1981).
Lavallee's studies incorporate analyses of the
entire range of material and associated activity areas to
generate conclusions, and therefore contribute useful
216
comparative information for this study.
At a glance the
relatively small faunal collections, and paltry lithic
assemblages from Manachaqui Cave's earliest ceramic-bearing
levels suggest activities other than those commonly
performed during comparable time periods at Telarmachay.
Salient differences between collections of organic remains
and artifact assemblages from the Junin caves and Manachaqui
Cave will be detailed in Chapter 10.
Unfortunately there
are no detailed studies of Andean caves interpreted as
shrines and wayside stations to provide comparisons for this
analysis.
Manachaqui Cave as a Wayside Station
Manachaqui Cave's roadside location and its modern
function as a shelter for villagers traveling to and from
the upper montane forest present an ethnographic analogy
that serves as a working hypothesis for this study.
Evaluation of this hypothesis offers a special challenge
because Andean archaeologists have yet to investigate
rockshelters with a comparable primary function.
In fact,
my search of the world-wide archaeological literature failed
to locate any references to caves or rockshelters utilized
specifically as wayside stations in similar contexts of
regional or interregional transit.
Hence, adopting the
wayside station working hypothesis necessitates comparisons
of Manachaqui Cave's archaeological remains to those from
highland cave sites interpreted as hunting and herding camps
217
in search of analogous artifact assemblages and activity
areas.
Reports of excavations at Chobshi Cave, Guitarrero
Cave, Telarrnachay, and Pachamachay offer serviceable data
for such comparisons.
The latter two especially may render
useful functional analogues because they are interpreted as
camps for hunting and herding, and also because they span
the Preceramic Period-Initial Period transition when the
occupants adopted the use of pottery.
However, the
comparisons drawn may provide negative correlations at best.
They can not directly confirm the working hypothesis.
Whether or not Manachaqui Cave was regarded as a
pacarina may never be determined, but the associated huanca
(M-1E) clearly indicates past ritual activities at the site
during late prehistoric times.
Excavations of M-1E revealed
shallow deposits and a small sample of late ceramics
indicating that the huanca was probably erected during the
Late Horizon.
That the name Manachaqui (or mama-chaki)
might be translated literally in Quechua (using the popular
dictionary compiled by Lira 1979) as "mother-foot" or
"mother-dry"
(dependent upon whether a glottal stop
distinguishes ch'aki) also hints at the attachment of
religious meaning to the rockshelter.
The name Manachaqui
may thus have signified "mother of the traveler."
In the Andean literature, Inca way-stations, or tambos,
represent the closest functional analogues to Manachaqui
Cave.
However, their free-standing masonry construction and
218
the internal diversity consonant with their complex
administrative functions (Hyslop 1984:279-280) hinder
comparisons to Manachaqui Cave.
The Tambo de Tunscucancha
in Huanuco, the most intensively studied to date, consists
of 10 or 11 buildings with artifact distributions suggesting
domestic and ritual functions, as well as status differences
(Morris 1966).
Less elaborate complexes like Callangate and
Frailetambo have yet to be thoroughly investigated.
Rowe
(1946:270-271) reports that the Inca tightly regulated
travel and trade.
Pre-Inca interaction in southern
Chachapoyas was undoubtedly governed by other sets of
constraints and opportunities left for archaeologists to
discover.
The basic technologies involved in trade and travel
were probably quite simple.
Personal belongings and goods
for trade may have been wrapped in a mantle and slung onto
the traveler's back as Rowe (1946:237) describes.
Llamas
carried up to 100 lbs. in cloth saddle bags (Ibid.:239) or
perhaps net bags.
Such technologies are probably as old as
textile technology and camelid domestication.
The former
was in place by the Late Preceramic Period (Bird 1963;
Adovasio and Maslowski 1980; Grieder et al. 1988), but
intensive utilization of domesticated camelids in north
highland Peru may only date to the Chavin horizon (Miller
and Burger 1995).
Material remains of transport
technologies may be difficult to identify.
Textile remains
219
in particular do not normally preserve in humid tropical
archaeological contexts.
Manachaqui's Ethnographic
Analogy
Consideration of how Manachaqui Cave is utilized today
aids the generation of a working hypotheses to model
Manachaqui's ancient functions.
literature search,
Again, a world-wide
this time for ethnographic descriptions
of modern cave and rockshelter utilization, failed to locate
useful analogies for Manachaqui Cave.
Shanidar Cave in
northern Iraq is a well-known example of a modern cave
habitation, but it is not functionally comparable to
Manachaqui as its enormous interior seasonally shelters a
community of approximately 45 Kurds, their huts and even
their livestock (Solecki 1979).
Comparisons of Manachaqui
Cave's present surface activity areas to those within deeper
strata should help support or refute a wayside station
working hypothesis.
Today up to eight individuals sleep
huddled together for warmth on a bed of ichu straw against
the eastern interior wall of the shelter where they find
shelter from damp easterly winds (Fig. 9).
Occupants
maintain a hearth for cooking and/or warmth against the
western wall and close to the cave mouth where the same
winds carry smoke out into the open.
Young spent long periods of time traveling with Pataz
villagers through the Tropical Alpine Zone at the western
220
border of the Rio Abiseo National Park, and reports most of
the following observations (personal communication, 1994).
Nowadays, a trip to the "jalca"
(Tropical Alpine Zone)
social occasion greatly enjoyed by the men.
is a
Young noted
that meals are invariably prepared prior to departure,
thus
obviating the need for cooking fires, collecting or
transporting fuel and the great patience required for
boiling food at high altitudes.
On the first day, several
varieties of small previously-boiled potatoes are eaten.
Sufficient portions of cooked guinea pigs (or cuy) are
brought so that each traveler has his own portion, and chili
peppers are added to most foods.
Young's description agrees with ethnographic
observations from the northeastern Andes reported by Morales
(1995).
According to Morales (1995:131), pre-prepared
guinea pig serves as the usual travel "fiambre" or "cooked
food provision."
After consuming the guinea pig on the
trail, its skull may be carried horne to assure successful
propagation of the household livestock, or it may be offered
with coca to the mountain spirits for protection from storms
and other travel hazards.
Morales also observes that
travelers may keep the right foot to carry as an amulet for
good luck and to give strength for walking.
Parched maize (or cancha) and/or nuna beans accompany
the main fare.
Gade (1975:126) reports that cancha is
habitually carried as "trail food" in Cuzco.
Cobo's
(1890-
221
93, I:343; Coe 1994:222) chronicles of traditional native
customs note that portions of cancha, or cancha flour mixed
with cold water, were carried as travel meals.
Nowadays,
food is typically carried in a pair of metal or plastic
bowls lashed one over the other and transported in a small
woven saddle bag on mule-back, or over the traveler's
shoulder.
Travelers also catch trout and hunt white-tailed
deer (outside of the park boundaries) if time permits.
we
observed that they leave nothing when they abandon
Manachaqui Cave except matted straw and ashes from a small
fire maintained for warmth.
Andean foxes would be expected
to scavenge any discarded bones or other food remains.
Working Hypothesis
In order to evaluate the working hypothesis that
Manachaqui Cave functioned as a wayside station, negative
evidence for hunting and herding functions may be combined
with ethnographically and theoretically-derived sets of
expectations for archaeological material patterning.
First
we would expect that most, if not all, artifacts would have
been produced elsewhere.
"imports."
In a sense, they would be
At a wayside station frequented by itinerant
individuals and groups, the range of functional types might
constitute only a fraction of the total range of functional
categories produced or used at home.
In view of the
ethnographic analogy, we might hypothesize that large
cooking vessels would be rare or absent from a wayside
222
station assemblage.
When ceramics occur, small sizes should
be characteristic of culinary pottery.
Portability would be
a primary concern, and small vessels also suffice to service
fewer individuals.
Vessel types associated with special
ritual activities should be rare or absent.
Such types
include decorated bottles presumably utilized for ritualized
consumption of corn beer (chicha)
(Burger 1992:155, 166).
Identification and quantitative analyses of Manachaqui
Cave's organic remains can provide insights into the
shelter's ancient functions. Prehistoric "trail food" might
consist of grains, legumes or tubers either pre-cooked or
easily prepared at the site.
While travelers presumably
hunted and trapped game for consumption en route, a wayside
station's deposits would not include the great quantities of
bone typically unearthed at hunting and herding camps (cf.
Miller 1979).
Hunting and herding refuse at Chobshi Cave,
Guitarrero Cave, Telarmachay and Pachamachay included many
kilograms of bone material, mostly from large game animals
like deer and camelids.
At a wayside station we might
instead expect to see the remains of small animals or
portions of larger animals that could be conveniently
packed.
Large game animals like deer may have been
preferred targets, but most of the meat would likely have
been prepared for transport and carried away, perhaps to
feed families at home.
Consequently, Manachaqui Cave's
refuse should reflect disposal of small portions of large
223
animals consumed at the shelter.
In her analysis of Manachaqui Cave's botanical remains,
Pearsall (Appendix F) lists a set of assumptions governing
expected patterns of food charring and deposition within a
rockshelter setting.
Based on these assumptions, and the
known regional distributions of ecological zones and plant
habitats, she suggests some indicators for distinguishing
the site·' s function.
The first concerns the abundance of
charred food remains relative to the abundance of wood
charcoal, and aims to determine the degree to which cooking
of plant foods was done at Manachaqui.
Food remains should
be more plentiful relative to charcoal at habitations where
food is inadvertently charred during repeated cooking
activities.
Disproportionately large quantities of charcoal
would result from fires built for warming rather than
cooking.
The utility of this indicator rests upon the
assumption that travelers typically carried pre-prepared
food.
Changes in charred food:wood ratios may signal shifts
in Manachaqui's function.
Of course warming fires could
fortuitously char small amounts of food.
Additional indicators suggested by Pearsall concern the
quantity of food items gathered or grown locally relative to
the quantity of "exotic" food items brought from distant
ecological zones.
Changes in local:exotic food ratios
through time may indicate shifts in Manachaqui's function as
a habitation (relatively high values for local foods) or
224
wayside station (relatively high values for exotic foods).
Caveats to be considered include the fact that hunting and
plant gathering in the rockshelter's vicinity may have
provided fresh food when distances to be traveled were
great.
Also, the tending of some high altitude crops could
have routinely occupied Manachaqui Cave's habitual users.
A third indicator proposed by Pearsall relates to the
abundance of corn kernels relative to cobs.
probably "exotic" to Manachaqui Valley.
Corn is
A higher proportion
of kernels relative to cobs may indicate that corn was
shelled and parched at the site.
Pearsall hypothesizes that
shelling and parching may more likely occur in habitations.
The higher percentage of cobs relative to kernels may
indicate a tendency of travelers to consume corn directly
from the cob.
However, the inadvertent carbonization of
cancha brought to Manachaqui Cave by travelers carrying preprepared food could result in misleading ratios and
erroneous interpretations.
The results of this study
suggest that the kernel:cob ratios are not effective
indicators of Manachaqui Cave's function, but they are
included so that readers may judge for themselves.
All of
the ratios are susceptible to alteration by a variety of
outside factors as well as site formation processes.
Regional subsistence shifts with the introduction and
abandonment of certain plant foods more or less likely to be
preserved during Manachaqui's occupational sequence
225
potentially alter ratio values, as would changes in food
preparation and consumption practices.
Therefore, the
changes in this and other ratios must always be evaluated in
conjunction with other lines of evidence.
Furthermore,
Manachaqui Cave may have served multiple functions at any
given time.
The various kinds of material patterning
hypothesized above presuppose ideal situations that may
seldom conform closely to prehistoric reality.
Summary
The two most important problems addressed within this
chapter involve the identification of changing
paleoenvironmental contexts during periods of Manachaqui
Cave's intensive prehistoric use, and the determination of
what those uses were at given prehistoric moments.
Interpretations of function are to some degree dependent
upon knowledge of corresponding paleoenvironmental contexts
if subsistence-related functions such as herding and hunting
are to be considered.
Timberline forest environments may be
appropriate to sustain game such as white-tailed deer, paca
and armadillos, but herds of llamas and alpacas are best
adapted to open puna grasslands.
Paleoenvironmental reconstruction suggests that open
grasslands predominated in the Manachaqui Valley during the
time period under study, although fragmented and continuous
upper montane forest habitats always remained within reach
by short (no more than a half-day) excursions from
226
Manachaqui Cave.
Timberline forest probably encroached on
the M-lA locality periodically under favorable climatic
conditions, and in the absence of sustained burning.
Manachaqui Cave's ecotone location offered favorable
circumstances for
hum~~
utilization of a great diversity of
resources, yet poor conditions for the sustained
exploitation of single species.
The working hypothesis of Manachaqui Cave as a wayside
station was sketched in the preceding paragraphs.
M-lA's
modern usage provides one basis for this hypothesis, and the
presence of the pre-Hispanic road provides additional
plausibility.
Also, the distribution of southern
Chachapoyas settlements in the highlands and montane forests
described within the preceding chapter invites the
assumption that people frequently traveled between the two
ecological zones.
As a node in east-west transit,
Manachaqui Cave conceivably sheltered migrants, pilgrims,
trading expeditions, war parties, hunters and herders during
the periods under study.
Because this study of function has
few if any theoretical predecessors, the data must be
interpreted through series of expectations or multiple
working hypotheses, each taking environmental and historical
contexts into consideration.
However, precise
identifications of Manachaqui Cave's prehistoric functions
are not necessarily pivotal to evaluations of other
hypotheses regarding Andean-Amazonian interactions addressed
227
within this thesis.
The dearth of ancient settlements in
these Tropical Alpine Zone valleys permits the conclusion
that many individuals sheltering in Manachaqui Cave arrived
from one side of the Marafion-Huallaga divide for the express
purpose of crossing to the other.
CHAPTER 5
EXCAVATIONS AT MANACHAQUI CAVE
As noted in previous chapters, excavations at
Manachaqui Cave were undertaken in 1988 and 1990 in order to
remedy the dearth of detailed cultural sequences in the
northeastern Peruvian highlands.
As repositories of
sequentially stacked primary contexts, caves and
rockshelters have served as "the essential data source for
the construction of cultural stratigraphies and histories in
many regions of the world"
(Straus 1990:265).
Like sites
with architecture, cave sites with occupation floors may
provide diachronic evidence of economic or other activities.
Unlike sites with architecture, it is less probable that
cave floors were altered subsequent to deposition by renewed
construction and the addition of potentially confusing mixed
fill brought from elsewhere.
Under ideal conditions, the
stratigraphic order encountered reflects consecutive stages
of human occupation.
But in reality, caves and rockshelters
present problems such as "imperceptible depositional
hiatuses, stratigraphic disturbances" and "high rates of
palimpsest deposition"
(Ibid. :256).
Fortunately, resulting
interpretive ambiguities can be minimized by implementing
certain methodological adjustments which will be outlined
228
229
within this chapter.
Analysis of Manachaqui Cave's stratified ceramic
remains provides the basis for the chronological sequence
that serves as the core of this thesis.
Andean archaeology
has a long history of reliance upon pottery styles as
indicators of temporal and spacial relationships (Adams et
al. 1978:492).
At Manachaqui Cave, ceramics constitute by
far the most abundant and potentially informative cultural
remains.
The primary depositional contexts of Manachaqui's
strata offer opportunities to define archaeological phases
with a high degree of temporal precision.
The regional data
base compiled by other investigations affords possibilities
for cross-dating and the identification of interaction
networks.
Hence, this thesis emphasizes Manachaqui
ceramics, although it also integrates information rendered
by analyses of lithic and other cultural remains.
Cave and rockshelter sites frequently yield wellpreserved organic remains that are rapidly weathered or
destroyed by exposure at open sites.
The dry soils within
Guitarrero Cave (Lynch 1980) and the Ayacucho caves
investigated by MacNeish et al.
(1981) contained a diversity
of botanical and faunal remains, and even relatively intact
assemblages of Preceramic Period basketry and textiles.
Excavations at Pachamachay (Cave) in Junin (Rick 1980)
encountered damp soils in which carbonized macro-botanical
remains were useful for identification and sourcing by
230
ecological zone and habitat (Pearsall 1980).
Like
Pachamachay's, Manachaqui Cave's soils are humid.
Nevertheless, bone and carbonized botanical remains are
preserved throughout the site, although the interior
deposits yielded larger, less weathered samples.
The study
of Manachaqui Cave's stratigraphic record of organic remains
presents an unprecedented opportunity for the reconstruction
of subsistence activities and changing patterns of cultural
interaction in the humid northeastern Peruvian Andes.
This chapter begins by outlining the techniques
employed for 1988 and 1990 excavations at Manachaqui Cave.
A second section provides descriptions of the stratigraphy,
occupation floors, and associated hearths and architectural
elements.
The radiocarbon evidence from these contexts
permits the bracketing of archaeological phases with dates.
Interpretations of site formation are offered where
pertinent to the understanding of Manachaqui's cultural
sequence.
A third section details changes in hearth
morphology, and the final section briefly describes
analytical methodologies utilized for the study of
Manachaqui's artifacts and organic remains.
The 1988 Excavations
The 1988 test excavations at Manachaqui Cave codirected by Lennon and Segundo Vasquez Sanchez aimed to
determine the depth of the shelter's deposits, evaluate
material preservation and assess the site chronology.
Work
231
took place during 21 days between February 20th and March
12th.
This is the rainy season in the eastern Andes, and
working conditions were therefore less than ideal.
On many
days it was necessary to work under tarpaulins, although
these were removed in order to examine and draw features and
stratigraphy.
In order to expose a cross-section of the site's
strata, as well as to sample both the interior and exterior
deposits, a trench comprised of six 1 x 1 m2 units (Units 16) was situated to extend outward from the rear of the
shelter along the eastern interior wall (Fig. 10).
Two
additional units aligned with the trench were excavated near
the base of the berm (Units 7 and 8), and a single unit
(Unit C-1) was placed downslope beyond the berm to examine
culturally sterile soil deposits from the surrounding
slopes.
Digging by trowel then proceeded utilizing 5 ern
arbitrary levels measured from a datum assigned the
arbitrary value of 100.0 m.
Levels for all units except 7
and 8 maintained a strictly horizontal orientation
irrespective of surface slope.
In Units 7 and 8, each
arbitrary level did maintain the slope and undulation of the
berm surface in hopes of approximating tilting sub-surface
cultural strata.
Unlike the level numbering system employed
for Units 1 through 6 which reflects absolute elevation with
reference to a fixed and independent datum, the system
utilized for Units 7 and 8 differs by simply assigning level
232
numbers in relation to the surface beginning with level
number one.
Soil samples were systematically collected from
each level of selected units prior to passing all of the
excavated dirt through 1/4" or 1/16" screens.
The excavation of Units 1 through 8 provided
information with which to formulate plans for future work.
First, Manachaqui Cave was found to contain cultural
deposits to a depth of one meter within the shelter
interior, and two meters in the deepest part of the berm.
The berm deposit appeared equally divided between strata
with ceramic period remains, and layers with Preceramic
Period remains beneath (Plate IV).
Second, there were few
visible strata extending throughout the site with which to
differentiate occupations and/or phases.
For the site as a
whole, visible stratigraphic distinctions consisted of
differing shades of black and dark brown soil, the presence
or absence of color mottling, and varying sediment size.
The more visibly complex interior stratigraphy showed
horizontally distributed bands and lenses of charcoal, ash
and a yellowish brown soil as ephemeral indications of
layered floors, hearths and activity areas.
The berm
excavations yielded fewer visible strata, but changing
patterns in artifact content became obvious as the
excavations proceeded.
During 1988, preliminary analysis of the excavated
pottery confirmed that, despite the paucity of visible
233
stratigraphy at Manachaqui Cave, a temporal sequence of
pottery styles does remain intact.
The principal
methodological concern resides in how to identify and
isolate the temporal components.
The method employed here
to distinguish the cultural sequence resembles that utilized
by Lathrap for ceramic complexes of the Central Ucayali site
UCA-6 that he ultimately assigned to the Early and Late
Tutishcainyo Phases (Lathrap 1962).
During excavation, we
observed that particular pottery attributes were
characteristic of some levels and entirely absent in others.
The quantity of lithic artifacts per level also varied in a
suggestive manner.
Consequently, the frequency of those
pottery attributes were charted by excavation unit and level
in order to illuminate potentially discrete vertical
distributions.
Lathrap (1962:46-47) refers to this
utilization of frequency distribution as a
"macrochronology."
Further refinement of each isolated
cultural component produces what he terms a
"microchronology."
Pottery attributes chosen for a macrochronological
study of Manachaqui Cave's cultural sequence were sherds
with 1) basal, medial and shoulder carinations and/or
applique ribs, 2) fine-line incised decorations and 3) fine
kaolin pastes.
Observations from ceramic sequences in
neighboring northern Andean regions suggest that these
isolated components most likely pertain to the Initial
234
Period, Early Horizon and Early Intermediate Period
respectively.
The resulting macrochronological
distributions appear in Appendix D.
Units 4, 5 and 6 show
the clearest separation of temporal components, while Units
1, 2 and 3 demonstrate far less.
This can be explained in
part by the tendency of the 1988 horizontal, arbitrary
levels to approximate the orientation of the horizontal
stratigraphy within Units 4, 5 and 6, and their tendency to
cross-cut the sloping stratigraphy within Units 1, 2 and 3.
A fourth temporal component within the uppermost layers at
Manachaqui lacks any of the three ceramic attributes, and
probably corresponds to the Late Intermediate Period and/or
Late Horizon.
A primary objective of the 1990 excavations
at Manachaqui Cave was to confirm, or to modify and complete
this tentative and sketchy macrochronological sequence, and
ultimately to produce a more refined chronological sequence.
The 1990 Excavations
The 1990 investigations at Manachaqui Cave (directed by
the author) took place during a period of 38 days between
September 25th and November 4th.
The project's start was
delayed almost two months by Peru's major economic crisis
and the reduced purchasing power of the dollar.
Consequently, we were forced again to work during the rainy
season.
Again we were obliged to use tarpaulins for
protection.
Troweling through black soils under tarpaulins
and leaden-gray skies challenged everyone's eyesight, and
235
the illumination provided by a Petromax gas lantern greatly
facilitated the identification of soil details in the
shelter interior.
The preliminary analyses of Manachaqui Cave's 1988
pottery sample suggested that the east-side interior
deposits were highly disturbed.
A concavity in the east-
side ground surface matched by undulations in the underlying
strata indicates that soil may have been repeatedly scooped
out or moved around, probably to prepare sleeping surfaces.
Because the west profile of the test trench showed evidence
of more intact occupation floors,
it was decided to devote
special attention to careful excavation and sampling of the
interior's west side.
The 1988 excavations had produced
small quantities of bone and charred botanical remains that
promised useful economic information if larger samples from
less disturbed contexts could be obtained.
Utilizing the same 1988 grid system, additional 1 x 1 m
units were selected for excavation.
With the goal of
understanding deposition and site formation, we proceeded to
expose stratigraphic profiles by excavating trench-like sets
of 1 x 1 m squares oriented east to west.
A strategy of
maximizing the number of exposed profiles has been
recommended by Straus
(1979, 1990).
This technique
facilitates correlations and comparisons of stratigraphy
across the site, although admittedly we were simply
searching for more useful stratigraphic evidence than that
236
which the 1988 excavations had exposed.
The trench-like
sets consisted of Units 12-13, 18-23 and 29-32 (Plate III).
Units 35, 38 and 39 connected 1988 Units 1-6 and 7-8.
The
subsequent opening of additional squares completed the
sampling and exposed architectural features.
In order to discuss procedures and results, it becomes
expedient to differentiate between the shelter interior and
the exterior berm deposits with the labels Sector A and
Sector B respectively (Fig. 8).
In Sector A, the exposure
of relatively intact strata visible in Unit 12's and 13's
north profiles (Plate V) led to the decision that Units 14,
15, 16 and 17 should be excavated simultaneously as a 2 x 2
m block with the hope of uncovering successive occupation
surfaces.
It was suspected that hearths and evidence of
economic activities might be found on the west side of the
interior just as they occur on the west side of the site
surface today.
Thus, the 2 x 2 m block was excavated by
stratigraphic levels or occupation surfaces where possible,
and by 1, 2 or 5 em arbitrary levels where neither strata
nor surfaces could be detected.
Stratigraphic excavation of
the block relied on tactile as well as visible observation
because occupation surfaces were invariably more compact
than surrounding layers.
Additional information on Sector A
excavations and the cultural stratigraphy will be offered
later within this chapter.
The decision to excavate unit levels that maintained
237
the slope and undulations of the ground surface constitutes
an important change in 1990 excavation procedures for both
sectors.
In 1988, only Units 7 and 8 were excavated in this
fashion.
This technique was intended to diminish the
likelihood that sloping sub-surface strata would be crosscut and "mixed."
Where it was discovered that the slope of
sub-surface strata did not match the surface slope, we
attempted to adjust excavation levels accordingly.
Arbitrary 5 em levels were most frequently employed in
Sector B, although stratigraphic levels were preferred when
they could be discerned during excavations.
Readers should
be advised when consulting the appendices that, because
levels within some units were adjusted to accommodate
natural strata, floors and localized features,
they do not
always correspond to similarly numbered levels in other
units.
Two methods were utilized to sample organic remains
from both sectors.
First, unscreened soil samples were
systematically collected from selected units.
A second
method first implemented in 1988 involved passing soil
samples through 1/16" screens with water at the site.
The
stream located a short distance west of the rockshelter
supplied running water for this procedure which was applied
in 1990 only to samples from Sector A's occupation surfaces.
After surfaces were exposed, mapped, photographed and bulk
sampled, they were "peeled off" by troweling and then passed
238
through the 16" screen at the stream to remove fine
sediments.
The remaining screen contents (both the heavy
fraction and charred botanical material) were then bagged.
In Trujillo, selected screened and unscreened soil samples
were processed by water and chemical flotation techniques
after removing charcoal for potential use in radiocarbon
dating.
The heavy fractions typically contained stones,
lithic detritus and bone fragments that were sorted and
bagged separately.
The contents of hearths encountered during excavation
were bagged in their entirety.
In Trujillo, these samples
underwent water and chemical flotation after removing some
wood charcoal for dating.
Heavy fractions were then sorted
and bagged by class of material.
A final category of soil
sample was collected from each stratum for laboratory
evaluation of pH value and sediment characteristics.
Those
results will be presented below.
Site Formation and Stratigraphy
This section will provide descriptions of Manachaqui
Cave's stratigraphy and depositional sequence.
As
previously noted, it is helpful to conceptualize the shelter
interior and the exterior berm separately as Sectors A and
B.
Sector A contains Units 1-3, 9-17 and 25.
The division
separating Sectors A and B approximates the shelter's drip
line.
The stratigraphy and the ceramic period cultural
sequence generally repeat themselves across the site.
239
Excavations in both Sectors A and B encountered occupation
surfaces, hearths and architectural features, yet the two
sectors represent differing behavioral contexts.
Sector A
served as Manachaqui Cave's principal focus of activity,
while Sector B in effect constitutes Sector A's trash
midden.
Much of its broken pottery and lithic detritus
originated in the shelter interior where relatively clean
living surfaces were preferred and maintained.
The organic remains and activity areas preserved in
Sector A offer the most useful information with which to
reconstruct economic activities at Manachaqui Cave.
Sector
B's deposits have suffered weathering by exposure to rain
and moving groundwater, but they remain less disturbed by
trampling, hearth construction and other post-depositional
cultural activities.
cultural sequence.
Sector B contains the most intact
Together, sectors A and B offer
complimentary information with which to approach problems of
Manachaqui Cave's occupational sequence and changing
functions.
Site Formation
The following constitutes a general description and
interpretation of Manachaqui Cave's deposits in terms of
their constituents, and their culturally and naturally
induced transformations.
The term "transformation," coined
by Schiffer (1983), refers to the cultural and natural
processes which affect archaeological remains and their
240
distributions following initial deposition.
At Manachaqui,
both sectors were subjected to characteristic sets of
transformations, although some particular transformations
were common to the archaeological deposits throughout the
site.
These are discussed following a brief description of
Manachaqui's depositional processes.
The sediments comprising Manachaqui Cave's deposits may
be conceived of as both "endogenous" and "exogenous"
(Farrand 1985:21).
Endogenous sediments originate in the
shelter's ceiling and walls as the result of frost action
and collapse.
They frequently occur in Sector A's deposits
as stones and gravel.
Exogenous sediments include those
introduced by wind, water and animals (including humans).
The natural sediment matrix in Manachaqui Cave, and the
principle constituent of post-glacial soils in Manachaqui
Valley, is loess in the form of dark-colored silt and clay
sediments.
On Manachaqui Valley's Pleistocene-age moraines,
Rodbell (1991:93-95) reports caps of loess 20 to 40 em thick
overlaying glacial till.
Our excavation of Unit C-1 a few
meters below the rockshelter revealed 15 em of dark, finegrained sediments between a five em thick humic layer above,
and bedrock below.
Small amounts of inorganic sediment must
have been inadvertently introduced to Sector A by human
carriers.
Also a light brown silty soil layered among the
floor deposits appears to have been intentionally laid down.
It may have been brought from the stream bed west of the
241
site.
Sector A, the shelter interior, contains a complex
stratigraphy comprised of thin and tightly stacked layers
corresponding to prepared and unprepared occupation surfaces
or "floors."
These are occasionally interspersed with
layers of disturbed and undifferentiated floor remains.
The
first series of floors identified at Manachaqui Cave was
found in Units 12 and 13, and designated Floors 1 through 4.
Floor remains also appear in Unit 25.
A careful excavation
of the adjacent 2 x 2 m block {Units 14-17) revealed the
complex series of occupation floors identified by the
letters A through Z, and AA through GG from top to bottom.
Sector A's archaeological deposits range in thickness from
0.80 min the rear to 1.1 m near the mouth.
That the
shelter's interior and exterior stratigraphic and cultural
sequences maintain good vertical correspondences throughout
the ceramic period suggests that Sector A has never been
subjected to massive soil disturbances resulting from
attempts to clean or renovate the interior.
Sector B's deposits range from 1.1 m deep at the mouth
of the shelter to nearly 2.0 mat the crest of the berm, and
0.90 m deep in excavations near the berm's foot.
In the
deepest portions of Sector B, the ceramic and pre-ceramic
deposits share a nearly equal thickness of approximately one
meter.
Sector B's deposits contain less organic material
than Sector A's {1.54% of total bulk sediment for the
242
average level in Unit 36 compared to 4.13% in Unit 15
levels), but a much greater volume in artifacts.
The five
hundred cubic centimeters of sediment within each of Sector
B's arbitrary levels frequently yielded over 250 sherds, or
at least one sherd for every two cubic centimeters.
Thus,
the thickness of Manachaqui's ceramic period strata, and the
height of the berm in general, are primarily functions of
artifact, rather than soil deposition.
Straus (1990) emphasizes the importance of
understanding the rates of sedimentation at caves and
rockshelters for stratigraphic interpretation.
Utilizing
data from Unit C-1 and Rodbell's (1991) observations
regarding local rates of loess deposition, a clearer
understanding of Manachaqui Cave's site formation processes
may be approached.
A radiocarbon date of 9700 ± 270 B.P.
(Birkeland, Rodbell and Short 1989:58) from organic
materials lying atop "glacial or fluvial sediments" located
less than half a kilometer up-valley from Manachaqui Cave
provides a minimum estimate for glacial retreat above the
shelter.
If we assume that the 20 em of loess and humus
have accumulated on top of Unit C-l's gravel deposits over
the following span of ten thousand years,
then we may derive
an average deposition rate of 2 em for each thousand years,
or 2 mm for each one hundred years.
Certainly the simplicity of this formulaic description
of Manachaqui Cave's natural sedimentation masks the
243
dynamics of a complex depositional process.
For example,
early Holocene flooding would have governed both erosion and
deposition of the lowest strata.
deforestation
~n
Also, burning and
the Manachaqui valley probably accelerated
loess deposition after 4,000 b.c.
Regardless, it is safe to
assume that the tempo of cultural deposition at the shelter
outpaced that of natural deposition during most of
Manachaqui's stratigraphic history.
As Straus predicts
(1990:269), this results in the close juxtaposition and
complex stacking of activity residues characteristic of
Sector A.
Where occupation floors can be isolated, they are
difficult if not impossible to interpret in terms of
longevity and intensity of use.
Clearly both factors
operated to determine the volume and density of Manachaqui
Cave's deposits, but their relative importance remains
obscure, at least at present.
Furthermore, with such a slow
natural sedimentation rate it should not be surprising if we
cannot stratigraphically identify occupational hiatuses
lasting several hundred years.
may seem directly juxtaposed.
Pre- and post-hiatus strata
At Manachaqui Cave the
processing of series of radiocarbon dates as recommended by
Straus facilitates interpretation of the changing dynamics
of site formation.
Water was an important agent of sediment transport at
Manachaqui Cave.
Having seeped into the shelter interior
through the numerous interstices between boulders, it has
244
acted as agent of both sediment deposition and postdepositional transformations.
Prior to human efforts to
seal the openings in Sector A's rear and east walls, runoff
generated by periodic heavy rains would have periodically
washed, and perhaps scoured, the shelter interior.
The
earliest evidence for attempts to seal the interior consists
of piled field stones.
Modern practices suggest that sod
brought in from outside was also utilized to plug the
openings.
This procedure would have transferred exogenous
soils (and perhaps cultural materials) into the shelter
interior.
Subsequent erosion of the sod plugs would have
introduced secondary deposits into Sector A's otherwise
primary depositional sequence.
This kind of process may be
responsible for much of the depositional mixing evident in
Sector A, and especially in Unit 11 which abuts a gap
between two boulders comprising the interior east wall.
While no evidence has appeared to suggest massive
efforts to clean or renovate the shelter interior by the
removal of soils, there are indications that it was kept
relatively free of heavy trash accumulations.
Sherd counts
for units within Sector A are considerably lower than those
for Sector B.
Laboratory retro-fitting of Sector A and B
sherds suggests that trash such as broken
routinely pitched out onto the berm.
potte~J
was
Other culturally
induced transformations at Manachaqui Cave include
sequential episodes of activity area preparation and
245
trampling.
The preparation of hearths and sleeping areas
certainly caused the translocation of sediments and
associated cultural remains,
from earlier strata.
typically pulling deposits up
Trampling is the most ready
explanation for the small size of potsherds at Manachaqui
Cave, particularly in Sector A.
Other disturbances include a large basin-shaped pit
that was·dug into Unit 16's upper strata, and a tubular
vertical hole, perhaps from a post, found in the northern
half of Unit 13.
Leo {personal communication 1992) reports
that rodent species frequenting Manachaqui Valley do not
burrow, although they do take advantage of pre-existing
burrow-like spaces.
Several burrowing centipedes were
encountered during excavations at Manachaqui Cave.
Most
were in Sector B, but the extent of disturbance that they
may have caused is difficult to evaluate.
Stratigraphy
During the excavation of Manachaqui Cave's
archaeological deposits, several factors hindered the
precise identification of individual strata.
All of the
strata vary subtly in chromas and hues from dark brown and
gray to black, and difficulties distinguishing these were
aggravated by the aforementioned lighting problems.
Color
variation within strata further complicated
characterization.
In many cases sediment grain size and the
presence or absence of color mottling served as defining
246
characteristics.
The high number of large stones impeded
stratigraphic differentiation in the lowest levels,
especially in Sector B.
Soil sediment analyses performed at
the National University of Trujillo Department of Chemical
Engineering document stratigraphic change by measuring
percentages of clay, silt, sand and gravel
(Tables 1 and 2).
Samples analyzed from Units 15 and 36 show rising quantities
of gravel and diminishing amounts of silt and clay with
increasing depth.
Soil pH was also measured for each
stratum.
Manachaqui Cave's deposits are divisible into four
principal stratigraphic units.
through 4 (Figs. 11-21).
These are labeled Strata 1
Stratum 4 is bedrock.
In certain
locations Strata 1, 2 and 3 are sub-divisible by visual
criteria.
Stratum 3 corresponds to Manachaqui's Precerarnic
Period occupation and its gravely sediments throughout the
site confirm that water was an important depositional and
transformational agent.
Stratum 3 and its contents will not
be described in detail, except as necessary to illuminate
the sequence of site formation.
The upper, ceramic-bearing
Stratum 2 and Stratum 1 provide the data for this study.
The following paragraphs constitute descriptions and
interpretations of Manachaqui Cave's stratigraphic units.
Strata and sub-strata present throughout the site are
detailed, but floors, hearths and architectural features are
described by sector.
Figures 22 and 23 illustrate the
247
relationship between the stratigraphy and arbitrary levels
in typical Sector B unit profiles.
The reader may use these
in conjunction with the pertinent macrochronological
distributions presented in Appendix D to locate temporal
components.
Radiocarbon evidence is also introduced.
Stratum 4
Bedrock at Manachaqui Cave is a yellow-brown color and
consists of large rocks embedded in a compact, gravely soil
matrix.
Many loose rocks lie directly on the bedrock soil,
especially near the foot of the berm.
Most of these rocks
may have been removed from the shelter interior and tossed
downslope during the earliest occupations.
Stratum 3
Stratum 3, the gravely sediments overlaying bedrock and
containing Preceramic Period remains, is distributed
throughout both Sectors A and B.
However, the densest
cultural deposits left by Preceramic Period occupations at
Manachaqui (mostly lithic refuse) lie within the thicker
Stratum 3 layers of gravely soil at the bottom of Sector B,
the berm.
Although Stratum 3 within the shelter contains
some lithic material, there are virtually no diagnostic
pieces to compare with those from the artifact-rich
Preceramic levels in Sector B.
Radiocarbon evidence
presented below suggests that only the final centuries of
the Preceramic Period are represented in the shelter
248
interior while the thicker berm deposits yield much earlier
dates.
The differences between these two depositional
contexts can be explained as the result of flooding and
precipitation runoff that repeatedly washed Sector A's
earliest deposits into Sector B until the shelter's users
began habitually to divert water away from the interior.
Therefore, Sector B's Stratum 3 deposits, especially the
lowest and most gravelly portions, should be regarded as
mixed with secondary deposits from Sector A.
Of course we
might also expect Manachaqui's users to perform wasteproducing activities such as stone tool working outside,
rather than within the confined sleeping and cooking areas.
Toward the rear of Sector A, Stratum 3 can be subdivided by a subtle difference in color into Stratum 3A and
Stratum 3B.
While Stratum 3A is black (lOYR 2/1 black),
Stratum 3B has a brownish chroma (7.5YR 3/2 dark brown),
probably from mixing with the bedrock soil.
The distinction
between the two strata blurs and becomes imperceptible
toward the shelter mouth, but again becomes evident outside
of the shelter.
A report of the cultural content of Stratum
3 falls beyond the purview of this thesis.
Stratum 2
As previously noted, Stratum 2 contains the organic and
inorganic remains of first importance for this study.
Its
deposits pertain to cultural occupations from the terminal
Preceramic Period through the Late Horizon.
Stratum 2
249
features a higher silt to gravel ratio than Stratum 3.
However, like Stratum 3, it can be subdivided on the west
side of the shelter interior into a nearly black Stratum 2A
above a dark brown Stratum 2B.
It is also sub-divisible
north of Unit 6, beyond the crest of the berm.
Although Strata 2A and 2B cannot be distinguished from
one another at the deepest part of the berm, a layer of
angular stones ranging from five to ten centimeters in
length or diameter lies just beneath the projected 2A/2B
interface and probably represents an eroded occupation floor
or surface.
This layer is most clearly visible in the east
and west profiles of Units 4, 5 and 6 where lies between 70
and 75 ern below the ground surface.
An associated hearth
(Feature R-4) will be described below.
Why the stones
should appear so concentrated on this surface remains
unclear.
Stratum 2A varies from 35 to 45 ern in thickness and has
the highest percentages of silt and clay of the Manachaqui
strata.
Where it can be isolated, it contains Early
Intermediate Period materials.
Within Stratum 2B the
percentage of gravel steadily increases with depth.
Its
thickness ranges from 45 to 55 ern, and its contents include
the late Precerarnic, Initial Period and Early Horizon
materials.
Stratum 1
Stratum 1 is the uppermost of the Manachaqui Cave
250
strata, consisting of grasses, rootlets and humus.
Associated cultural remains pertain to the Late Intermediate
Period and/or the Late Horizon.
Stratum 1 is only
discernible around the perimeter of the berm because a thick
layer of sod from the crest was removed during the
villagers' 1986 efforts to "clean" the site.
Some of this
sod was cut into bricks and crammed into the open
interstices at the rear of the shelter.
Soil was also
tossed over the edge of the berm to form the layer
designated Stratum 1A.
Stratum 1B constitutes the original
pre-1986 superficial layer.
find Stratum 1 intact.
Only in Units 18 and 19 did we
Stratum 1 can be distinguished from
Stratum 2 because of its organic content rather than because
of a markedly different color.
In stratigraphic profile it
has been observed that this layer dries more rapidly than
Stratum 2.
Archaeological Features, Floors and Hearths
Within this and the following section reporting Sector
B's cultural deposits, several classes of non-portable
artifacts will be described.
walls.
Most are floors, hearths and
The floors within Sector A are labelled as such.
Labelled as "features" are several prehispanic walls,
isolated fragments of partially preserved floors and a
single hearth in Sector B, and a drainage duct in Sector A.
To help the reader locate floors and features in threedimensional space, pertinent unit and level numbers are
251
offered, as well as the stratum number and depth below
ground surface (bgs).
For each floor described, the depth
below the ground surface is reported at the vertical
division between Units 1 and 2 seen in the west profile of
Units 1-3 (Fig. 11).
This is the common point shared by
Units 1, 2, 13 and 15.
Figure 19 indicates the location of
some floors in relation to Unit 15's stratigraphy.
Sector A Features and Floors
The floors excavated within the 2 x 2 rn square formed
by Units 14, 15, 16 and 17 were difficult to identify and to
excavate as discrete entities.
Many appeared as compact
lamina 0.5 to 2.0 ern thick that we attempted to isolate and
peel off in succession.
Straus summarizes the interpretive
problem presented by tightly stacked occupation surfaces,
what he refers to as "complex palimpsest cave deposits"
(Straus 1990:264):
The stacking effect, often with few or no discernible
sterile stratigraphic zones, means that archaeologically
even the finest levels are palimpsests ... It is unreasonable
and unnecessary to try to isolate individual occupation
residue surfaces (Ibid.:265 with original italics).
While the isolation of "moments in time" may not be
theoretically possible or even necessary, Sector A's
"palimpsest" strata did serve our purposes as stratigraphic
alternatives to arbitrary excavation levels.
Given the soil
conditions however, this excavation method still did not
enable us to discern whether or not some tightly stacked
Sector A floors had truncated earlier floors, or if minor
252
cleaning episodes had eradicated portions of one or more
floors.
In excavating the so-called floors of Manachaqui Cave's
Sector A, the primary goals were to evaluate: 1) the
chronology of site occupation through stratigraphic sampling
and 2) the site's functions through diachronic analysis of
the behavioral contexts that the floors and other
archaeological features represent.
Straus promotes the
continued use of cave sites to gather information on
regional cultural chronology as long as careful
stratigraphic excavation methods acknowledge the palimpsest
problem.
In order to understand the shelter's functions, he
councils that caves and rockshelters be understood as
"places" within larger systemic contexts where limited sets
of activities repeatedly took place: "even if our thinnest
archaeological levels and lenses in caves and rockshelters
are palimpsests, it is this redundant, structurally
conditioned use of space that allows us to obtain meaningful
behavioral information from these kinds of sites"
1990:279).
(Straus
Behaviorally meaningful levels, Straus argues
(Ibid. :268), are best isolated by "anchoring" them with
features such as hearths and other activity areas.
To interpret every floor within tightly stacked series
as a separate occupational event would imply that each can
be isolated, that its rate of cultural deposition can be
accurately accessed and, therefore, that each individual
253
"floor" has knowable temporal and behavioral meaning.
Since
this is clearly not the case, the only Stratum A floors
reported individually within the following paragraphs are
those which appeared in stratigraphic isolation from other
floors.
These are the lowest floors in Stratum 2B.
The
series of floors in Stratum 2A occur as stacked and
virtually indistinguishable palimpsests, and many were only
discerned in the southern half of the 2 x 2 m block (Units
14 and 15).
Beyond the shelter's drip line, which crosses
Units 16 and 17, erosion has eliminated most traces of
ancient surfaces.
Other interpretive problems related to
floor excavation will be discussed during description of the
floors and features to follow.
Only Feature R-5, a small drainage canal, and the
remains of one floor are associated with Stratum 3 in Sector
A.
The canal constitutes the first evidence of human
modification of the shelter interior.
The preservation of
the first identifiable floors may be attributed to its
efficacy in diverting water.
Feature R-5 (Stratum 3 [70 em bgs];
Level 16 in Unit 12)
Feature R-5 is a small drainage channel shaped with
stones and clay resting directly on Sector A's bedrock (Fig.
24).
This drain was found in Unit 12 and designated "Rasgo
5" or R-5.
It was built to divert runoff entering the rear
of the shelter away from activity areas on the east side of
254
the shelter.
R-5 begins at the rear southwest corner under
a ceiling protuberance that we observed collecting and
dripping water into the interior during each storm.
A semi-
circular ridge of clay apparently forms part of a catchment
basin.
The basin drained through a narrow channel comprised
of parallel ridges of stones embedded in clay.
The water
ran through a gap between two large boulders protruding from
Sector A's bedrock that functioned as a natural conduit
directing the water out of the western side of the shelter's
mouth.
Floor GG (Stratum 2/3 [99 em bgs];
Level 68 in Units 15 - 17)
One layer designated "possible" Floor GG was identified
as a compact surface in three of the four units comprising
the 2 x 2 m block (excavation of Unit 14 had previously
exposed bedrock).
Like the floors above, GG was probably
compressed by prolonged human activity.
It lies at the
interface of Stratum 2 with Stratum 3 and contained abundant
charred organic material, but lacked artifacts except for
micro-debitage recovered in the 1/16 inch screens.
No
radiocarbon dates were run because of lacking associated
diagnostic material.
Floor FF (Stratum 2B [93 em bgs];
Level 66 in Units 15-17)
This compact surface was uncovered within Stratum 2B in
Units 15, 16 and 17.
An
irregularly shaped rock-filled
255
hearth straddles Units 15 and 17 (Fig. 25a).
Its length is
50 em, its width is 34 em and it varies from 4 to 7 em deep.
Charcoal from the hearth rendered the date 3520 ± 100 B.P.:
1570 b.c.
(all dates reported in the text are uncalibrated).
Several angular stones lie embedded within its fill which
was bagged for transport and processing in Trujillo.
One
chipped-stone scraper and one intrusive sherd from
overlaying strata lay on FF's surface.
Floor EE {Stratum 2B [88 em bgs];
Level 63 in Units 15 - 17)
Floor EE includes a large, at least partially
disturbed, rock-filled hearth in Units 16 and 17 {Fig. 25b).
It appears to have been roughly circular with a diameter of
approximately 50 em, and 10 to 13 em deep.
The western edge
was probably blurred by exposure to rain and roofline
dripping.
Hearth charcoal produced the date 3830 ± 100
B.P.: 1880 b.c.
Floor DD {Stratum 2B [85 em bgs];
Level 60 in Units 14 - 17)
At the eastern edge of Unit 15 lies a circular hearth
cut during excavation of Unit 3 in 1988.
Its diameter
measures 26 em and it is only three em deep.
Like Floor
EE's hearth, Floor DO's also contains numerous angular
stones.
256
Floor CC (Stratum 2B [81 em bgs];
Level 57 in Units 14 - 17)
Floor CC yielded a large rock-filled hearth 50 em in
diameter and five em deep in Unit 17 (Fig. 26a; Plate VI).
Floor BB (Stratum 2B [77 em bgs];
Level 54 in Units 14 - 17)
Floor BB appears to be the latest of the Preceramic
Period floors preserved at Manachaqui Cave.
An oblong
hearth 38 em long by 24 em wide in Unit 15 is only two em
deep (Fig. 26b).
Several stones found resting above the
eastern and western rims of Floor BB's hearth may have been
pushed aside during hearth preparation, or extracted from
the hearth prior to re-use.
The hearth's fill yielded the
date 3670 + 100 B.P.: 1720 b.c.
A few sherds collected at
this depth (only in Unit 15) are most likely the product of
stratigraphic mixing toward the eastern side of Sector A.
Feature R-7
(Units 9, 10, 11, 25)
Feature R-7 is a stone wall uncovered in Units 9, 10
and 25 that appears to be continuous (Figs. 27, 28).
The
excavation of Unit 11 revealed the remains of a similar and
probably contemporaneous wall.
The term "wall" is applied
here to connote architectural function.
In reality, Feature
R-7 is a pile of deliberately placed fieldstones.
The base
of R-7 rests directly upon the aceramic/preceramic deposits
at the bottom of Stratum 2B, yet the presence of potsherds
embedded within the stone masonry indicates Initial Period
257
construction.
As mentioned previously, this wall impeded
the flow of water into the shelter interior from the slopes
behind.
It may have been refurbished repeatedly for
centuries, but we found it completely covered by soil.
Floor AA (Stratum 2B [58 em bgs];
Level 43 in Units 14 - 17)
Floor AA is the earliest Initial Period floor, although
sherds begin to appear in substantial quantities in all four
units of the 2 x 2 block above level 50.
A circular hearth
in Unit 17 measures 26 em in diameter and two em deep (Fig.
29a).
A date from the hearth charcoal is 2850 ± 90 B.P.:
900 b.c.
Floor Z (Stratum 2B [56 em bgs];
Level 41 in Units 14 - 17)
Floor Z contains two hearths labeled Hearth 1 and
Hearth 2 in Units 16 and 17 respectively (Fig. 29b).
1 is the largest encountered at Manachaqui Cave.
Hearth
In plan
view it is slightly quadrangular, and it is 48 em long (east
to west) and an estimated 45 em wide.
The northern extreme
was removed during the excavation of Units 20 and 21 leaving
this and Hearth 2 both visible in profile (Fig. 20).
Hearth
1 measures 11 em deep and, unlike the other Sector A
hearths, it contained many sherds.
Hearth 1's charcoal
rendered the date 2800 ± 90 B.P.: 850 b.c.
Hearth 2 was
also cut by prior excavations, but it has a circular shape,
measures approximately 30 em in diameter and has a depth of
258
nearly five em.
The 1988 profile (Fig. 11) revealed a stone
in the center of Hearth 2's charcoal fill.
Floor Y (Stratum 2B [50 em bgs];
Level 38 in Units 14 - 17)
Floor Y was identified only by its compaction and the
presence of horizontally dispersed sherds.
There is no
hearth, but charcoal gathered from the floor produced the
date 2450 ± 90 B.P.: 500 b.c.
Floor X (Stratum 2B [46 em bgs];
Level 35 in Units 14 and 15)
Floor X was only identified in the southern half of the
2 x 2 m block.
Floor charcoal yielded the date 2560 ± 100
B.P.: 610 b.c.
(this charcoal sample was small and required
additional counting time during processing).
Sector A's
palimpsest strata between Floor X and Floor H (Level 11) in
Units 16 and 17 were not sufficiently preserved to permit
stratigraphic excavation.
The corresponding levels in these
two units were excavated arbitrarily in five centimeter
increments.
Floor X shows increasing color mottling,
perhaps from the infiltration of light brown soils found in
more substantial quantities within overlying strata.
Floor W (Stratum 2B [38 em];
Level 30 in Units 14 and 15)
Floor W appears as a compact layer best preserved in
Unit 15.
It corresponds to the lowest of the four floors
first identified in Units 12 and 13 (Floor No. 4).
A
259
radiocarbon date from Floor W charcoal is 2110 ± 80 B.P.:
160 b.c.
Floor V (Stratum 2B [35 em bgs];
Level 28 in Units 14 and 15)
Floor V was also identified by its compaction, and
otherwise lacks distinguishing cultural features.
Because
of its proximity, it might be grouped with the series of
Floors U through H just above.
Within Floor V and the
overlaying floors, a narrow fissure in the soil trending
northwest to southeast becomes visible in Unit 15.
The
crack closely traces the outline of the large rock
protruding from bedrock below.
It does not correspond to
any stratigraphic divisions that we could detect visually,
but is clearly a post-depositional soil feature.
The
mechanics of this "micro-faulting" remain unclear but may
have something to do with the differential rates of soil
drying and matrix contraction and/or compression originating
above.
This fissure is a likely perpetrator of
stratigraphic mixing.
Floors U through H (Stratum 2A [17-34 em bgs];
Levels 26 - 11 in Units 14 and 15)
This series of 16 "palimpsest" floor layers seems to
represent centuries of intensive, repeated use of the
shelter.
During these centuries, Manachaqui Cave's users
periodically laid down the coarse yellowish-brown fill
described earlier.
Subsequent disturbances have left brown
260
mottles and large brown patches in otherwise black,
charcoal-laden silty soil.
We excavated hearths in Floors
T, P, L, J and H in Units 14 and 15.
The shape of Floor P's
hearth may indicate repeated uses (Fig. 30).
Excavation of
Unit 16 by arbitrary level uncovered two hearths.
One is
most likely associated with FloorS (Fig. 31a), while the
other is less securely matched to Floor M (Fig. 31b).
In
Units 12 and 13, floors identified as 3, 2 and 1 roughly
correspond to Floors T, Nand I in Units 14 and 15.
these floors are associated with Stratum 2A.
All of
Hearth
dimensions and pertinent radiocarbon dates from Sector A are
listed in Table 3.
Floors G- A (Stratum 2A [2-16 em bgs];
Levels 10 - 2 in Units 14 - 17)
Floors G through A constitute a continuation of the
same series of palimpsest occupations as Floors U through H.
They are distinguished only because of change in the
attributes of its associated pottery.
contain six hearths.
Floors G through D
Five of these are irregularly shaped
and no more than one ern deep,
indicating haphazard
preparation and perhaps only a single usage.
The sixth
hearth, in Floor D (Unit 16), is 21 ern in diameter, six ern
deep and contains an irregular stone embedded in the
underlying soil.
Floors C through A have three irregular,
shallow hearths.
Floors G through A yielded the largest
amounts of ceramic and lithic remains of all of Sector A's
261
floors, and they most likely pertain to the Late
Intermediate Period and/or the Late Horizon.
Sector B Features and Floors
As previously mentioned, Sector B constitutes Sector
A's trash midden.
Although potsherds in Sector B are most
often found vertically or diagonally arrayed, isolated
"pavements" of horizontally-oriented sherds representing
fragments of eroded occupation surfaces indicate that refuse
was also generated by activities on the berm surface.
Where
these surface fragments were preserved, we assigned them a
feature number, exposed them as feasibility permitted, and
described their salient characteristics.
Possible
correlations between Sector B's and Sector A's floor
fragments and architectural elements are discussed later in
this chapter.
Features R-6 (Level 16 [47 em bgs] in
Units 23 and 28) and R-8 (Unit 18)
Features R-6 and R-8 are crude stone walls that extend
from the mouth of Manachaqui Cave onto the berm, presumably
for additional protection from wind and rain (Fig. 32a, b,
28).
Directly abutting the shelter exterior, Feature R-6
projects 1.2 m northward from the east side of the shelter's
mouth.
It is 1.1 m high and 0.80 m wide.
The wall's size
and form were obtained by piling large and irregularly
shaped stones three courses high and two courses wide.
base of the wall corresponds to Unit 22's Level 16 (the
The
262
lower portion of ceramic-bearing Stratum 2), and therefore
indicates Initial Period construction.
In Unit 18, Feature R-8 seems to be a similar yet
poorly preserved construction projecting from the western
side of the cave mouth.
The wall was formed by placing
several large stones over a boulder that protrudes from
bedrock.
The top of Feature R-8 is 30-40 em higher than the
top of R-6.
Because only a small area of the stone feature
was exposed, its dimensions could not be assessed.
It also
remains unclear whether or not the wall abuts the shelter
exterior, although it likely does.
Feature R-8's
stratigraphic associations are likewise difficult to
establish.
Its construction dates to the late Preceramic or
the Initial Period.
Feature R-4 (Stratum 2 [69 em bgs];
Level 16 in Unit 31)
Feature R-4 represents the only post-Preceramic Period
hearth found in Sector B (Fig. 33a, b).
It is associated
with the layer of stones just beneath the projected 2A/2B
interface previously described.
A circle of stones with a
diameter ranging from 30 to 35 em delimits the hearth's
circumference.
Poor preservation and stones within the
surrounding soil matrix hindered precise definition of the
base and walls of the basin-shaped hearth, but its depth is
estimated at three to four ern.
Charcoal from the hearth
contents produced the radiocarbon date 2810 ± 100 B.P.: 860
263
b.c.
At the bottom of the hearth lay a horizontal cluster
of sherds from the same carinated vessel.
A thin layer of
charred organic remains lies in association with the hearth
surface.
Feature R-3 (Stratum 2 [41 em bgs];
Level 8 in Unit 21)
Feature R-3 is a small horizontal distribution of
sherds that represents the remains of an occupation surface
within the lower half of Stratum 2.
Many of the fragments
belong to two vessels with incised-line decoration and
assigned to the Early Horizon.
Feature R-2 (Stratum 2 [35 em bgs];
Level 6 in Units 26, 29, 30)
Feature R-2 is a scatter of sherds and one ground slate
point distributed horizontally in Stratum 2.
concentration lay in Unit 30.
The principal
The association of kaolin
wares suggests and Early Intermediate Period date.
Feature R-1 (Stratum 2 [15 em bgs];
Level 4 in Unit 19)
Feature R-1 is another horizontal distribution of
sherds within Stratum 2.
Remnants of this floor may have
been removed in all but Unit 19 during the 1986 cleaning
episode.
Material remains of this occupation were probably
incorporated within sod bricks utilized to seal the shelter
interior, or were tossed off of the berm's edge with loose
soil.
Because of ancient surface undulations, R-1 may
264
correspond to the uppermost floors A or B in Sector A, and
perhaps to Feature R-9, an identical distribution of
horizontal sherds uncovered within the first excavation
level in Unit 26.
further detail.
Feature R-9 need not be described in
Both Features R-1 and R-9 date to the Late
Intermediate Period or Late Horizon.
Changing Hearth Morphology and Function
The 24 hearths excavated within the 2 x 2 block may be
classed into three groups based upon salient attributes.
It
is likely that each class served a different function.
These classes are: 1) rock-filled hearths, 2) hearths with a
single stone embedded in the bottom, and 3) simple basinshaped hearths with no stones.
The five rock-filled hearths uncovered in Floors FF,
EE, DD, CC and Z are generally largest with a mean diameter
of nearly 44.8 em and mean depth of 7.8 em.
Archaeologists
have interpreted hearths like these as roasting pits for
meat based upon ethnographic analogies (e.g. Lavallee
1977:87-88).
The traditional native practice of roasting
meat in a pit filled with hot rocks is called a pachamanca.
The six "single-stone" hearths in Floors D, H, M, P, S
and T average 30.2 em in diameter and 6.4 em in depth.
The
embedded stones are irregularly shaped and vary from 10 to
15 em in length or diameter.
Floor Z's Hearth 2 may be the
earliest of this variety, but its accompanying stone
appeared to be floating in the hearth fill rather than
265
embedded within the soil beneath.
In the absence of
ethnographic analogies, it may be difficult to establish
whether or not the single-stone hearths were used for
cooking.
The stone may have absorbed heat, acting to
radiate warmth for the shelter's inhabitants after the
fire's coals had cooled.
The 13 simple basin-shaped hearths found in Floor BB
and floors above are highly variable in shape and surface
extent.
Their depth averages only 1.9 em.
Perhaps these
served merely to warm chilled wayfarers for short periods.
A functional analysis of Manachaqui Cave's hearths
should provide clues regarding the shelter's changing
function, and changes in the site function may reflect
transformations in regional economic systems.
In viewing
the relative stratigraphic positions of each hearth class,
the most striking change is the virtual disappearance of
rock-filled hearths above Floor CC.
Floor Z's Hearth 1 may
constitute a late exception, although it contains few rocks
in comparison to earlier examples.
Associated radiocarbon
dates suggest that the shift in hearth preference occurs
during the Preceramic Period/Initial Period transition and
coincides with the introduction of pottery technology in the
Peruvian north highlands.
The study of organic remains
associated with Manachaqui Cave's three hearth types and
their corresponding floors potentially provides information
regarding relationships between subsistence and technology.
266
These data are addressed in later chapters.
Radiocarbon Evidence for Chronology
In order to date the occupational sequence at
Manachaqui Cave, a strategy of processing radiocarbon dates
from stratigraphic columns in each sector was implemented.
The series of radiocarbon dates from the 2 x 2 m block
(Units 14-17) in Sector A provides a preliminary
chronological assessment of the shelter's history of use and
disuse.
Radiocarbon dates from Sector B aid stratigraphic
correlation of the two sectors, and simultaneously serve to
date the macrochronological components within their most
intact stratigraphic contexts.
The radiocarbon dates from
each sector are discussed below, and all of the dates are
listed in Table 4.
Sector A Radiocarbon Evidence
Radiocarbon dates from Sector A show consistent overall
trends.
Some ordinal inconsistencies may be discounted by
considering the standard deviations of the corrected dates
(Pearson and Stuiver 1993; Stuiver and Pearson 1993).
For
example, the corrected sigmas of the seemingly inverted
dates from Floors H and L (A.D. 544 - 660 and A.D. 646 - 776
respectively) overlap.
The same is true of the seemingly
inverted dates from Floors X and Y (A.D. 801 - 413 B.C. and
761 - 391 respectively) .
Recall that Floor Y supplied a
small sample of charcoal requiring special treatment.
267
More problematic are inconsistencies between dates
rendered by Floor P (110 b.c.) and Floor T (a.d. 490).
The
corrected sigma from Unit 14 Floor P's earlier date (A.D.
129 - 341) does not overlap with that from Unit 15 Floor T
(A.D. 595- 672), even when taking two sigma for both dates
into account.
This problem may stem from post-depositional
transformations such as the partial removal of layers while
preparing or "cleaning" activity areas.
The Floor P date
was extracted from a hearth, and the Floor T date was
processed from charcoal embedded in the floor.
An effort at
refurbishment that involved the removal of soil from the
center of the shelter's interior could result in such a
stratigraphic disconformity.
It might be hypothesized that
the "cleaning" occurred sometime between circa a.d. 110 and
circa a.d. 450, and involved the destruction of Floor P
remains that would have been contained within Unit 15.
On
the other hand, the early Floor P hearth date was
conceivably produced by the burning of comparatively old
wood.
In general, we found disturbances extremely difficult
to detect and delimit given Sector A's stratigraphic
conditions.
Dating inconsistencies could also have resulted
from a tendency of our stratigraphic excavation to
inadvertently cross-cut tilting and undulating prehistoric
palimpsest surfaces where they are not well-preserved.
is improbable that alternative techniques utilizing
It
268
arbitrary levels would have alleviated such mixing problems.
A single radiocarbon date that requires explanation
because it is later than dates from the two floors above
belongs to Floor FF.
The corrected sigma of FF's date (1913
- 1677 B.C.) overlaps with BB's (2135- 1789 B.C.), but not
withEE's (2397- 2038 B.C.).
The stratigraphy at this
depth shows every visible indication of being intact, and it
seems most logical to disregard Floor FF's date as aberrant.
Manachaqui Cave's Sector A radiocarbon dates may be
grouped into five sets corresponding to the Andean periods
and horizons defined in Rowe and Menzel's sequence (1967).
The sets of dates also corroborate the periodization
proposed for the 1988 ceramic macrochronology.
Accordingly,
phase names taken from local toponyms have been assigned to
the terminal Preceramic Period component as well as the four
ceramic period macrochronological components.
Radiocarbon
dates from Floors BB through FF (1880 to 1520 b.c.)
represent the Lavasen Phase, the final centuries of the Late
Preceramic Period.
Dates from Floors AA and Z (900 and 850
b.c.) pertain to the Initial Period, and the associated
ceramic complex featuring vessels with carinated and semicarinated wall profiles is assigned to the Manachaqui Phase.
Radiocarbon dates from Floors Y and X (610 and 500
b.c.) fit within the Early Horizon, and the corresponding
ceramic complex with its fine-line incised decoration
pertains to the Suitacocha Phase.
Dates from Floors W
269
through H (160 b.c. to a.d. 570) coincide with the Early
Intermediate Period when the use of kaolin ware pottery
became widespread throughout the north-central Andes.
These
dates can be separated into an earlier component represented
by dates from Floors W and P (160 b.c. and a.d. 110) and a
later component represented by dates from Floors H, L and T
(a.d. 450 to 570).
These temporal components are termed the
Colpar and Empedrada Phases respectively.
Manachaqui Cave's
radiocarbon dates and their relative stratigraphic positions
also suggest extended periods of the shelter's disuse, but
the problem of sequential hiatuses will be considered in the
following section.
Sector B Radiocarbon Evidence and Stratigraphic
Correlations with Sector A
Three Preceramic Period dates were processed from
Sector B's berm deposits.
An AMS date derived from a small
charcoal sample from Stratum 3C in Unit 31 Level 35 is
10,270±60 B.P.: 8,320 b.c.
However, the problem of early
Preceramic Period occupations at Manachaqui Cave awaits
future study.
Two radiocarbon dates from Stratum 3 in Sector B fall
fully within the third millennium B.C.
These dates are 4120
± 130 B.P.: 2170 b.c. and 4280 ± 110 B.P.: 2330 b.c. from
Unit 31, Levels 25/26 and Level 30 respectively.
The
associated deposits include abundant lithic material with
several tool types (e.g. projectile points, scrapers,
270
burins, etc.).
The Preceramic dates from Sector B do not
overlap with the Lavasen Phase dates from Sector A which
represent only the last few centuries of the Preceramic
Period.
The paucity of Sector A's (Floors BB through FF)
lithic remains confounds comparisons of the shelter interior
and exterior artifactual components.
For these reasons, it
may be imprudent to assign both to a single Preceramic
Period phase.
In the following chapter, the probability of
a transitional stage separating the Preceramic and Initial
Periods will be addressed.
A date from Feature R-4, the single Sector B hearth in
Unit 31's Level 16, of 2810±100 B.P.: 860 b.c.
(I-17,320)
serves to correlate Stratum 2's rock layer with Manachaqui
Phase Floors
z
and AA in Sector A.
All three Manachaqui
Phase dates fall within the fifty year (uncorrected) span
between 850 and 900 b.c.
A stratigraphic association is
suggested by Floor AA's depth of 61 em below the surface (in
the northeast corner of Unit 17), Feature R-4's depth of 69
em below the surface, and the projected slope of M-lA's
strata from Sector A to Sector B.
The approximate
contemporaneity of the three contexts is corroborated by
ceramic evidence to be described in the next chapter.
Two dates from Unit 22 Level 11 and Unit 6 Level 25
directly date the Suitacocha Phase ceramic component.
These
dates are 2630 ± 100 B.P.: 680 b.c. and 2740 ± 90 B.P.: 790
b.c. respectively, and both of these Sector B contexts may
271
be regarded as roughly coeval with Sector A's Floors X and
Y.
At 41 em below ground surface, the Suitacocha Phase
sherds comprising Feature R-3 were probably deposited during
activities that contributed to the formation of Floor Y at
44 em below the surface (in the northwest corner of Unit
17).
Radiocarbon dates to correlate Sector B's Early
Intermediate Period levels with those identified in Sector A
have not yet been processed.
Based solely upon the "negative" evidence provided by
Manachaqui Cave's series of calibrated radiocarbon dates, it
may be argued that gaps in the occupational sequence lasting
three hundred years or more occurred during the mid-Early
Horizon, the middle Early Intermediate Period and the Middle
Horizon.
Of these three, it seems that the hiatus during
the middle Early Intermediate Period may be more apparent
than real.
Additional radiocarbon dates may fill the
possible gap between the Early Intermediate Period Colpar
and Empedrada Phases, and also provide evidence for Middle
Horizon occupation apparently lacking in Manachaqui Cave's
ceramic assemblage.
Some of the best evidence that occupational hiatuses
occurred within Manachaqui Cave's occupational sequence is
the erosion of sherd surfaces presumably resulting from
extended periods of exposure on the site surface.
In Sector
B, all of the phase assemblages have suffered severe erosion
except for that belonging to the Manachaqui Phase.
This
272
condition suggests that the Suitacocha, Ernpedrada and
Poblano phases were each followed by decades or centuries of
Manachaqui Cave's comparatively light usage or abandonment.
For example, radiocarbon dates and sherd erosion patterning
suggest that a gap in the shelter's sequence corresponds to
the mid-Early Horizon {coeval with the Chavin horizon) .
Manachaqui Cave's occupational hiatuses require explanation
as reflections of regional and pan-regional social,
political and economic events.
Materials Analysis
The analyses of ceramic, lithic and other cultural
remains from Manachaqui Cave are designed primarily to
extract chronological information.
Thus, emphasis is placed
upon relating artifacts and their salient attributes to the
stratigraphy and associated radiocarbon dates,
thereby
establishing relative and absolute sequences for the
shelter's occupation.
Functional consideration of pottery
vessels, stone tools and other artifacts receives sufficient
attention only to address questions of Manachaqui Cave's
changing functions and to assess the comparability of
Manachaqui's assemblages to others in neighboring regions.
Clearly there is a need for raw material sourcing studies,
use-wear analyses and additional specialized investigations,
but these await future opportunities.
273
Ceramic Analysis
The ceramic study presented in subsequent chapters is a
simple attribute analysis much like those performed by
Patterson (1966) on the Central Coast of Peru, and Burger at
Chavin de Huantar (1984b) and Huaricoto (1985b).
The
attributes of primary importance relate to ceramic
technology, vessel shape and decoration.
A total of 72,310
sherds were recovered during the 1988 and 1990 excavations
and subsequently examined for the analysis of Manachaqui's
occupational sequence.
This thesis will examine the ceramic
sequence only through the Early Intermediate Period.
However, the absence of visible, discrete stratigraphic
correlates for proposed phase components combined with the
lack of clear differentiation between Early Intermediate
Period and later wares preclude estimates of the total
sherds corresponding to each of the four established phases.
Of the 72,310 total sherds, approximately 7,500 sherds
exhibiting selected diagnostic morphological, decorative and
technological features were culled and sorted by established
categories.
Many sherds deemed too fragmentary or eroded to
supply reliable information were separated out, leaving
6,889 sherds that supplied the data for the following
analysis.
The study of the Manachaqui artifact assemblages
differs from other analyses because there are few useful
stratigraphic indices to isolate occupations and
independently support proposed phase designations.
274
Necessity requires the utilization of the artifacts and
their horizontal and vertical relationships to one another
to generate a sequence based solely on internal evidence.
At Lathrap's Ucayali sites and at Manachaqui Cave, ceramics
provide the most plentiful and informative material from
which to derive the sequence.
The archaeological phases
presented within the next three chapters represent the
product of macrochronological analysis of ceramic attributes
(what Lathrap termed "ceramic features").
The analysis
focused mostly on the units excavated in Sector B judged
through the course of study to retain the most intact
depositional sequence.
Those units cluster around the crest
of the berm.
Attempts at establishing ceramic chronologies usually
begin by distinguishing between wares presumed to vary in
frequency through time.
The majority of Manachaqui's
pottery sample from all levels consists of plain brown
wares.
Angular grains of igneous rock comprise the
predominant non-plastic inclusions within ceramic pastes
throughout the occupational sequence, and the pastes lack
other useful diagnostic characteristics with which to
consistently distinguish wares.
Hence, wares are regarded
as insensitive analytical units for the purpose of
evaluating change in the bulk of the pottery sample.
Nevertheless, it becomes apparent to the analyst that the
variety of wares increases in later levels of Manachaqui
275
Cave's deposits.
The creation of large "paste groups" based on easily
recognizable gross features constitutes a response to this
problem.
Wherever possible, distinctions are made between
the predominant brown ware and other wares with clearly
observable distinguishing characteristics.
Among these
others are wares with abundant quartz crystal and micaceous
inclusions, or white, pink and gray wares made with kaolinerich clays.
Within the following chapters, these three sets
of wares will be termed Paste Group A (predominant brown
wares), Paste Group B (deviant, often micaceous wares) and
Paste Group C (kaolin wares).
Group A wares were likely
produced at settlements close to Manachaqui Cave by the
populations that most frequented the road and shelter.
Group B and C wares, which appear in relatively small
quantities, represent pottery originating at greater
distances.
They cannot be termed "exotic" because all
pottery was, in a sense, exotic to Manachaqui Cave.
Nor can
they be labelled "trade wares" because they may have arrived
by means other than trade.
The erosion of sherd surfaces precluded classification
of pottery remains by surface treatments (e.g. polished,
burnished, matte, etc.).
However, clearly identifiable
surface treatments and wares can be useful attributes for
analyses provided they are weighted equally with the
presence of other attributes like fine-line incision, red-
276
painting, rim-beveling, etc.
The problem of erosion also
hinders efforts to quantify the frequency of certain
superficial decorations, especially slipping and painting.
Due to these analytical methodologies, and of course
the stratigraphic conditions, sherd totals pertaining to
every phase can only be estimated by an assortment of
techniques, each one frought with difficulties.
Some
estimates are necessary in order to determine proportions of
paste classes within each phase assemblage.
Resulting paste
group ratios provide one index indicating the relative
intensity of
in~eraction
from phase to phase, and these
ratios are obtained using simple rim counts.
The assumption
that rim-type frequency ratios provide approximations of
vessel-type frequency ratios is based upon quantitative
experimentation by DeBoer and Carroll (1992).
Three aids which readers can utilize to visualize the
vertical distributions of ceramic attributes are included in
the appendices.
Appendix A includes a list of the
illustrated artifacts showing the provenience of each.
Appendix D shows the macrochronology which supplied the
initial basis for the creation of phases.
Appendix E
presents the distributions of vessel shapes and rim forms by
provenience, and Tables 5 through 19 summarize vessel/rim
frequencies and dimensions and other ceramic data.
Computer
data base and spreadsheet programs such as Paradox 5.0 and
Quattro Pro 5.0 aided the compilation and sorting of
277
ceramics by provenience and attributes.
Lithic Analysis
The collection of chipped and ground-stone artifacts
from Manachaqui Cave's ceramic period deposits (including
Late Horizon deposits) consists of an estimated 9,500
pieces.
In the absence of use-wear studies, the chipped
stone lithic analysis is admittedly cursory and strictly
technological.
Refuse from a simple core-flake industry,
with the addition of a few unifacially flaked tools and
projectile points, marks the time periods in question.
Functional interpretations are more easily rendered for the
limited yet familiar varieties of ground-stone implements
unearthed at Manachaqui.
Of these, the ground-slate points
constitute a set of tools for which a more penetrating
technological analysis is warranted, especially given the
presence of slate blanks, partially worked preforms and
slate detritus.
The provenience of each stone artifact
category is matched against ceramic macrochronological
distributions in order to determine its pertinent
archaeological phase.
Analyses of Organic Remains
Organic remains recovered during excavations at
Manachaqui Cave are either charred botanical remains, or
burned and unburned bones.
These provide direct evidence of
plant and animal products consumed at the site, and indirect
278
evidence of the site's functions
(see the previous chapter).
In the case of the botanical remains, only the material
accidently charred is preserved.
Nearly all of the
unburned faunal remains from the shelter interior survived,
although trampling and weathering throughout the site
contributed to a generally poor state of preservation.
All
samples had to endure the rigors of transportation to Pataz
by muleback, and to Trujillo by truck.
Botanical Analysis
Through a two-stage process of water and chemical
flotation employing the IDOT technique (Pearsall 1989:3552), virtually all botanical remains were extracted from the
soil and hearth fill samples recovered from Manachaqui.
Sources of sampling error include the loss of minute remains
from floor samples wet screened with 1/16th inch mesh at the
site, and the destruction of remains too fragile to
withstand the wet screening and subsequent rigors of
transport.
Samples from Sector A Unit 15's floors and the
hearths were sent to Dr. Deborah Pearsall at the University
of Missouri for analysis.
After agreeing to emphasize the
earlier components, Pearsall selected a total of 42 samples
for sorting,
remains.
floors,
identification and quantification of botanical
Analyzed were 14 samples from all Lavasen Phase
4 samples from both Manachaqui Phase floors, 4
samples from both Suitacocha Phase floors, 18 samples from
17 Empedrada Phase floors, and two samples from late
279
prehistoric floors.
Some of the larger samples were sub-
sampled, and frequencies for all taxa identified were
tabulated on a computer spreadsheet program.
These are
tabulated within Pearsall's report included here as Appendix
F, and in Table 18.
Pearsall then calculated a series of
ratios to aid interpretation of the rockshelter's changing
functions.
The bulk of the botanical remains consisted of
wood charcoal, which was weighed but not identified.
Faunal Analysis
All of the bones and bone fragments recovered during
1988 and 1990 excavations at Manachaqui were sent to Dr.
Jonathan Kent at the Salango Laboratory in Salango, Ecuador
for analysis (Kent 1994).
At Salango, Kent and student
assistants weighed and counted the faunal materials.
The
bone sample totaled 3,673 bones of which 2,400 (65.34%) come
from Sector A, and 1,273 (34.66%) come from Sector B.
The
remains were sorted in terms of "number of identified
specimens"
(NISP).
These specimens were derived from single
excavation samples (i.e. they share common provenience), and
may be comprised of from one to 85 fragments of a single
bone type and animal size.
Kent's bone types are:
Long bones: humerus, radius, ulna, femur, tibia, fibula,
fragments of phalanges and metapodia that were not
recognized as such due to fragmentation;
and
Irregular bones: recognized phalanges and metapodia,
recognized vertebral parts, carpals, tarsals and hyoid;
Flat bones: scapula, ribs, cranial portions, and vertebral
spines and processes not recognized as such;
280
Teeth: crowns, roots and fragments of enamel;
Scrap: any bone not recognized as any of the other
categories.
Scrap specimens may contain up to 200 tiny bone fragments.
Animal size was broken down into categories of small, medium
and large, and into intermediate categories when necessary.
According to Kent, these are:
Small: rodents such as mice and guinea pigs, and most birds.
Medium: most carnivores, pacas (agouti), a few birds and
pudus, the smallest of the deer (cervidsl.
Large: carnelids and other cervids.
After subtracting bones of the scrap category, the remaining
2,319 bones (63.14 of the sample total) served to analyze:
1) bone types from Sector A vs. Sector B; 2) bone types from
the east vs. the west side of Sector A; and 3) bone types
and animal sizes by period.
Specimens identified and
classified into the above categories were then tabulated by
unit and level or floor, and entered into a spreadsheet
computer program.
Results of the faunal analysis are
summarized in Tables 19-28.
The high total of bones is misleading as the entire
collection weighs 450.8 grams (less than 1/2 kilogram).
Nineteen bones found outside of M-1A comprise 20 percent of
the total weight.
Kent observes that the high degree of
fragmentation can be observed within Unit 15 where 1,850
bones weigh 165.3 grams together, and average 0.089 grams
each.
Kent's taphonomic study observes that bird
activities, "trampling, digestion, weathering, human
281
smashing, and both rodent and carnivore gnawing are all
evident."
Scrap bone tallies that are higher outside than
inside the shelter (44.32% compared to 31.59%) support the
conclusion that Sector B's deposits were exposed to greater
attrition.
Kent concludes that the depositional patterns in
evidence "seem largely the results of human activities with
minor contributions made by birds and mammalian carnivores."
Again, ceramic and radiocarbon evidence provided the
chronological framework for assigning the faunal remains to
particular periods.
Results from analyses of organic
remains for each phase are summarized within each of the
following chapters.
CHAPTER 6
THE MANACHAQUI PHASE
In this and the following two chapters, the material
remains representing each archaeological phase are described
in detail.
As noted in Chapter 5, the Manachaqui Phase
corresponds to the Initial Period in Central Andean
prehistory, when the production and use of pottery first
becomes important in Andean daily life.
Not surprisingly,
pottery constitutes the most abundant material remains
recovered from Manachaqui Phase strata.
Ceramic and stone
artifacts will be described in turn below, followed by
sections reporting non-artifact remains such as floral and
faunal material.
Manachaqui Phase Antecedents
While the Preceramic-Initial Period transition on the
coast coincides with documented shifts in subsistence and
settlement patterns {Lanning 1967), economic stability may
be more characteristic of coeval highland development
{Burger 1992:104).
However, our view of economic change in
the highlands may be constricted by preservation problems as
well as the archaeological focus on monumental architecture
and on rockshelters of the central Peruvian puna.
282
As the
283
only north highland rockshelter occupied during the
Preceramic-Initial Period transition studied to date,
Manachaqui Cave potentially supplies new and valuable
information.
The transition at Manachaqui warrants analysis
beyond the purview of this thesis, but a few preliminary
observations can be offered here.
At Manachaqui Cave the Preceramic-Initial Period shift
apparently included a transitional stage that might be
thought of as the Terminal Preceramic Period.
In Sector A's
Stratum 2B, Lavasen Phase Floors BB through FF represent
this intermediate stage.
The shift from rock-filled to
simple basin-shaped hearths marks the end of the stage.
Underlying Sector B's Manachaqui Phase levels, Lavasen Phase
remains are contained within Stratum 2B's lowermost 10 to 15
centimeters.
These layers yield lesser quantities of
chipped stone artifacts than the Preceramic Period levels of
Stratum 3, and none of Stratum 3's projectile points nor
finely worked tools.
The Lavasen Phase raw stone materials
in both sectors are coarse grained, and sufficiently
isotropic only for a simple core-flake industry.
The early
Initial Period Manachaqui Phase is distinguished by a
remarkable dearth of any chipped-stone tools (see Appendix
D) •
This proposed stage was also identified during
Kaulicke's excavations in highland Cajamarca where "Final
Preceramic Period" material evidence underlay Initial Period
284
Pandanche Phase A deposits:
In the Final Preceramic that forms the basal part of the
cultural sequence we have a poorly documented phase whose
definition is based on the absence of ceramics, and welldefined layers with clear cultural associations such as
hearths ... There are few lithic artifacts and they are not
clearly distinguishable from others found in overlaying
layers with ceramics. One can even speak of an
impoverishment of the industry as raw materials such as
chert [silex] do not appear ... (Kaulicke 1975:40).
A single charcoal sample from two hearths (one simple and
another delineated by stones)
that apparently belongs to
Kaulicke's Terminal Preceramic Period provided two dates:
2010 b.c. and 2125 b.c.
(Kaulicke 1981:388).
The earliest
radiocarbon dates from the overlaying Initial Period
Pandanche Phase A deposits are 1395 b.c. and 1490 b.c.
(Ibid.).
At Manachaqui Cave,
the Lavasen Phase dates of
1720 b.c., 1880 b.c. and 1570 b.c. from Floors BB, EE and FF
represent this Terminal Preceramic stage.
If we are dealing
with the same developmental phenomenon at both Pandanche and
Manachaqui Cave, then we might hypothesize that a Terminal
Preceramic in the northern Peruvian highlands began between
2,200 and 2,100 b.c.
(recall Sector B's Late Preceramic
Period date of 2170 b.c.) and ended around 1500 b.c.
Intensive analysis of Preceramic Period levels and
their lithic contents will be crucial to better understand
the Preceramic-Initial Period transition at Manachaqui.
Analysis of changing frequencies of raw lithic materials and
tool types through the transition provides one avenue for
continued research.
Changes in material culture during
these shifts reflect alterations in the rockshelter's
285
function which in turn indicate modifications in the local
and regional economies.
Manachaqui Phase Ceramics
The Manachaqui Phase deposit is approximately 35 em
thick in Sector B's excavation units at the edge of the
berm, and contitutes 35 percent of the berm's meter-thick
ceramic period deposits.
Above Sector B's Manachaqui Phase
layers, the Suitacocha, Colpar, Empedrada and Poblano Phase
deposits are an estimated 20, 10, 20 and 15 em thick
respectively.
Examples of Manachaqui Phase levels from the
berm are Unit 5, Levels 26 through 33; Unit 28, Levels 14
through 22; Unit 30, Levels 13 through 20; and Unit 32,
Levels 13 through 20.
Of the 6,889 diagnostic sherds culled
(from the total 72,310 sherds) for this study, 3697
(53.7%)
belong to the Manachaqui Phase.
Ceramic Paste A
During the Manachaqui Phase, Paste Group A is composed
of a thin brown ware represented by 3,588 diagnostic sherds.
These constitute 97 percent of the total 3,697 diagnostic
Manachaqui Phase sherds.
The precise geographic origins of
the Paste A ware will remain unknown until archaeological
work on a regional scale reveals more about Initial Period
population and ceramic distributions.
At present there is
no reason to suspect that the large quantity of Manachaqui
Phase Paste A pottery was transported from distances farther
286
than the intermontane valleys flanking both sides of the
Marafion-Huallaga divide.
Manachaqui Phase Paste A is compact, rather than
porous.
Vein quartz comprises the predominant non-plastic
constituent, and angular grains reach a maximum 1.5 mm in
size.
Analysis with the scanning electron microprobe at
Yale's Geophysics Laboratory permitted the identification of
hornblende, ilmenite and numerous igneous rocks of
heterogenous mineral composition.
Unlike the quartz, these
grains are more homogeneously sized, and seem to have been
carefully sorted and added as crushed rock and sand temper.
Like other Initial Period pottery of the Peruvian north
highlands, Manachaqui Phase vessels exhibit thin walls
between 2 and 4 mm in thickness, a characteristic that
doubtless contributed to the severe fragmentation of the
collection.
construction.
A few sherds bear evidence of coil
Restricted vessels with unfinished interior
surfaces occasionally show evidence of wiping with vertical
strokes.
Potters burnished vessel surfaces with a smooth
implement, perhaps a pebble, leaving horizontal tracks one
to two centimeters wide.
The firing was well-controlled.
The pottery is hard, and over 90 percent of the sherds show
complete oxidation to a brown, reddish brown or dark brown
color.
Unintended firing smudges, or smoke clouds, are
present but not common.
described below.
Other technological details will be
287
Morphology
A total of 3,387 sherds were utilized to interpret and
classify Paste A vessel shapes of the Manachaqui Phase.
These were either pieces of vessel rims
(N=2,129), or angled
and decorated sherds from vessel mid-sections (N=l,258).
Represented by 2,017 rim sherds that constitute 94.6 percent
of the Manachaqui Phase Paste A rim sherd collection,
restricted vessel shapes predominate.
rounded, or nearly rounded bases.
All vessels had
Despite this apparent
conformity, there is a considerable range of variation in
formal details especially evident within the collection of
rims.
The Manachaqui Phase style might be considered highly
"sculpted."
sections.
Clay is added to rims, shoulders and midRims are typically reinforced by folding, and
then beveled, notched or deeply incised.
Applique ribs,
flanges and adornos attached to vessel mid-sections serve to
alter body profiles.
Such details blur taxonomic
distinctions between shape and decorative characteristics,
presenting challenges for establishing consistent
classifications and terminologies for ceramic attributes.
These same troublesome qualities give the phase assemblage
its stylistic unity and unique character.
Several other aspects of the Manachaqui Phase Paste A
assemblage hinder the application of schemes traditionally
employed to classify vessel shapes in the Andes.
The first
is the severe fragmentation of the collection which hampers
288
the reconstruction of vessel shapes (sherds are
characteristically less than five em in diameter) .
Trampling, vessel breakage patterns and other factors
frustrate attempts to associate rim shapes with body
profiles.
Consequently, this analysis focuses on rim form
as the attribute exhibiting greatest variability and most
suggestive of each vessel's gross morphological
characteristics.
Through the painstaking exercise of retro-
fitting sherds, the general tendencies of Manachaqui Phase
pottery shapes became evident.
A further hindrance to classification relates to how
the ancient inhabitants conceived of vessel shapes and
functional categories.
Of course the problem of vessel
functions is embedded within the larger problem of site
function addressed in the final chapter of this thesis.
The
Paste A assemblage lacks consistently occuring sets of
attributes that normally form the bases for establishing
analytical dichotomies between coarse vs. fine wares and
cooking vs. serving wares.
Native, or ernie,
formal and
functional categories lack rigid standardization, and some
variables of crucial importance for interpreting formal and
functional classes are virtually continuous.
The most
nettlesome include rim angles, rim thickening, wall
curvature and degree of interior surface finishing.
In
short, there are tendencies, but few inviolate rules, that
governed construction and decoration of pottery utilized
289
during the Manachaqui Phase.
In an attempt to approximate functional categories as
closely as possible, this ceramic classification prioritizes
rim morphology and interior surface treatment as independent
variables.
Given the limited evidence provided by the rim
sherds, five vessel shape categories were delineated
employing strictly objective criteria.
Each shape category
is interpreted as having functional significance.
The
presence or absence of some attempt to smooth vessel
interior surfaces provides a basic dichotomy between vessels
interpreted as "jars" utilized for food preparation, storage
and perhaps transportation, and "bowls" utilized for
serving, eating and perhaps transportation.
While the
interior surfaces of jars remain rough and unfinished, bowl
interiors show evidence of smoothing and/or burnishing.
Functional implications of each of the vessel categories
will be further considered in later discussions.
Rim Forms
The following three-tiered hierarchical classification
is based on 2,129 classifiable rim sherds grouped into the
five vessel shape categories.
The categories are Shape A:
neckless jars and jars with incipient necks; Shape B: shortnecked jars; Shape C: restricted bowls; Shape D:
unrestricted bowls with concave walls; and Shape E:
unrestricted convex bowls.
A finer level of classification
separates rims of differing thickness, orientation, aperture
290
and general treatment.
Thus, each rim form is given a
number designation (e.g. Rim 1, Rim 2, etc.).
A third level
occasionally highlights relatively subtle distinctions
deemed significant, and these carry letter designations.
This heavy emphasis on rim morphology is appropriately
employed here because the Manachaqui Phase rims are so
distinctive.
With few exceptions, they can be distinguished
from rims of later phases, and, hence,
temporal indices.
they do serve as
The rims were shaped by careful trimming
and beveling, and often finished by polishing.
Manachaqui
Phase vessel necks are invariably short where present.
The
degree to which particular rim forms consistently correlate
with specific body shapes and dimensions remains to be
determined with a less fragmented study sample.
Other
singular features of Manachaqui Phase rims will be described
below.
Body Shapes
For the purposes of this analysis, Manachaqui Phase jar
and restricted bowl shapes (Shapes A-E) are considered
either 1) globular, 2) semi-carinated or 3) carinated.
The
bodies of Shape E unrestricted bowls are, of course, semihemispherical.
The globular vessel bodies of jars and
restricted bowls are either spherical, ellipsoid or ovaloid.
Semi-carinated and carinated shapes may feature convex or
concave upper walls.
True carinated vessels exhibit a sharp
corner point or angle separating the shoulder and base,
291
thereby creating a composite silhouette.
The corner point
dividing upper and lower sections of carinated vessels is
often referred to in the literature a shoulder or basal
angle (e.g. Meggers et al. 1965; Lathrap 1962).
Perhaps the
latter term (basal angle) better describes the Manachaqui
Phase assemblage where most joints or points of maximum
width occur comparatively low on the vessel profile (Fig.
34a-h) . ·
Because the term "medial" seems to best characterize
the location of carination angles at, or just below the
vessel's mid-point, it will be utilized in this study.
The
term "shoulder" will signify carination angles above the
mid-point.
Vessel mid-sections may be embellished with a
medial applique rib or flange, lending a carinated
appearance to an otherwise globular or semi-carinated
profile (Fig. 34i-p).
The angularity of the vessel mid-
section is a continuous variable, and Figs. 34a-h illustrate
the graded series from globular to semi-carinated to
carinated profiles.
An additional complex body shape
(Fig.
34c) remains poorly understood, but probably resembles twotiered Machalilla vessel shape 12: Shouldered-neck Jars from
coastal Ecuador (Meggers et al. 1965: Fig. 90 and Table F)
Little is known of Manachaqui Phase vessel heights.
Not one vessel from Paste Group A could be sufficiently
reconstructed to permit measurement.
However, vessel
diameters were calculated from pieces of medial carinations
292
and sherds with medial applique ribs and flanges using a
formagauge and a rim diameter template.
1,097 such medial
sherds were separated, and 208 (19%) of these provided
reliable measurements.
Diameters (always measured on the
sherd interior) range from 12 to 24 em.
is 16 em (N=38).
The mode diameter
These statistics conflate formal and
functional vessel categories, and are not particularly
useful.
However, they do show an absence of large vessels
with maximum diameters of 30 em or greater.
In order to view statistical breakdowns by inferred
functional categories, an attempt was made to examine metric
differences between medial carination sherds with unfinished
vs. finished interior surfaces.
The former are most often
associated with the two previously mentioned jar Shapes A
and B, while the latter typically belong to bowl Shapes C
and D.
The mode diameter provided by 120 examples of the
sherds with unfinished interior surfaces is again 16 em.
A
small sample of 33 sherds with finished interior surfaces
inhibited comparability of distributions, but there are
indications that the bowl category has multiple modes, a
greater average diameter (19.6 compared to 16.9 em) and a
larger standard deviation (3.0 compared to 2.65 em).
These
figures only verify with certainty that, on the average,
bowls are wider and more variable in size than jars.
It should be reiterated that secure associations
between Manachaqui Phase rim forms and vessel body shapes
293
are few.
Generalizing statements such as,
"all short-necked
jars featuring Rim Form 2 had carinated bodies," are rarely
possible, although some possibilities are noted.
The
separate discussions of rims and body shapes just presented
are a response to this dilemma.
Shape A: Neckless Jars and Jars with Incipient Necks
A total of 175 rim sherds grouped into 13 distinctive
rim forms can be attributed to neckless jars, and jars with
incipient necks.
These constitute 8.2 percent of the
Manachaqui Phase rim collection.
Jars with incipient necks
feature inflected, or concave, upper body contours that
approach or end at a point of vertical tangency near the
vessel orifice.
The interior surfaces of these vessels have
been left scraped or wiped in cursory manner.
Exteriors
show careful smoothing and often burnishing.
Of the five
Manachaqui Phase shape categories, Shape A encompasses the
most variability in rim form.
The majority (65 percent} of
vessels exhibit everted or direct rims.
The exceptions are
the semi-spherical neckless jars with inverted rims
represented by Rims 2, 4 and 5.
Within the Shape A class,
functional variability is suggested by the presence of two
principal size categories.
A small category is represented
by vessels with Rims 1, 3, 4 and 7-13 (N=71, x=12.17 em,
:=1.98 em}, and a large category by vessels with Rim 6
(N=lO,
x=23.7 ern}.
Adding the Rim 2 and Rim 5 vessels with
orifice diameters greater 19 ern to the large category
294
renders an average of 24.07 em with standard deviation of
2.35 em (N=15).
The Shape A rims are described below, and
within Table 5, in order of decreasing frequency.
R±m 1 (Figs. 35a-h).
jars with incipient necks.
Rim 1 (48 examples) belongs to
The point of vertical tangency
is reached at the lip (or end point) and some rims flare
slightly outward.
The rim interiors have been gradually
thickened to a maximum thickness of 4 mm near the lip.
While the lip interior edge is rounded, the exterior edge
ends in a dull point.
The abrupt curvature below the vessel
orifice may indicate predominantly globular body shapes
(Fig. 35i).
The exterior surfaces of rims were burnished to
a low luster.
R±m 2 (Figs. 35j-p, 36a-g).
Rim 2 (39 examples)
exhibits convex walls and an inverted rim with interior
thickening.
Maximum thickness occurs at the lip which is
squared or semi-squared with slightly rounded edges.
The
maximum thickness ranges from 5 to 12 mm and tends to
increase with rim diameter.
2 to 3 mm thick.
Vessel walls below the rim are
These rims suggest globular bodies of
variable height (Fig. 36h).
The bi-modal distribution of
measured rim diameters (N=29) correspondes to the two size
categories described above.
Rim diameters of the smaller
class (N=25) range from 10 to 19 em, average 14.56 em and
yield a standard deviation of 2.42 em.
Eight of these rims
295
(e.g. Fig. 36a,b) are decorated with a notched applique
fillet of clay adhered to the lip exterior (the decorative
technique is described later within this chapter).
One
displays notches cut directly into the exterior edge of the
rim (Fig. 36g).
Four undecorated rim sherds (Fig. 36c-f)
represent the larger vessel size category.
Rim 3 (Figs. 37a-c).
Rim 3 (15 examples) pertains to
neckless jars with walls 3 to 5 mm thick.
The rims are
direct, and lip exteriors have been thickened to a maximum
of 11 mm by applying a clay fillet which is subsequently
decorated by notching in a manner similar to the decorated
variants of Rim 2.
Only one sherd is undecorated.
The
vessel body may exhibit any of the three possible profiles.
Rim 4 (Figs. 37d-f).
to globular jars.
The 12 examples of Rim 4 belong
Their characteristic orientation is
sharply inverted and approaches the horizontal plane.
They
show gradual, often uneven, thickening on the interior
surface.
The end of the rim has been "pulled" inward to
leave a 1 to 2 mm overhang or clay cornice, and the lip
exterior edge has been squared.
applique decorations.
Six of the rims bear
The total rim thickness reaches a
maximum of 6. 5 mm.
Rim 5 (Fig. 37g-i).
The 12 examples of Rim 5 are
direct or slightly inverted.
Clay has been applied to
296
reinforce the lip exterior, and the top of the rim has been
flattened.
The diameters from five measured rims indicate
sample partition into the small and large size categories.
Maximum thickness at the lip is 10 mm, and vessel walls vary
between 2 and 3 mm in thickness.
Like Rim 2 vessels, these
may have had simple globular body shapes.
Rim 6 (Fig. 38a-e). Ten rim sherds pertain to neckless
jars with slightly concave upper walls.
The rim form and
treatment are similar to those of Rim 2 although the
exterior edge of the lip terminates in a dull point.
Maximum thickness ranges from 6 to 10 mm.
Rim diameters are
among the largest of the Manachaqui Phase assemblage ranging
from 22 to 29 em and averaging 24.3 em (N=8) in diameter.
Two rims represent particularly deep vessels, and body
profiles were likely globular rather than carinated.
Rim 7 (Fig. 38f-j).
Eight sherds comprise Rim 7.
These rims end just short of reaching the point of vertical
tangency at the orifice.
They are gradually thickened to a
maximum of 6 mm just under the lip.
with rounded edges.
The lip is semi-squared
Several retro-fitted sherds reveal a
semi-carinated body profile (Fig. 38j).
Vessel walls are 2
to 2.5 mm thick.
Rim 8 (Fig. 38k-m).
Eight rims are unthickened or
slightly thickened to 3 or 4 mm.
The lip was trimmed flat
297
while the clay was quite dry, leaving small cracks still
visible on the surface.
Rim 9 (Fig. 38n-p).
Rim 4 has seven examples.
They
are thickened at the lip which is rounded and 6 mm thick.
An overhang of clay is folded over the interior edge of the
lip.
The vessels' upper walls are 3 to 3.5 rnm thick.
Rim 10 (Fig. 39-e).
Six rim sherds represent Rim 10
which features gradual curvature below the point of vertical
tangency at the lip.
The rims are unthickened and tapered
toward the lip which has been lightly flattened.
The
vessels' upper walls are 4 mm thick.
Rim 11 (Fig. 39f).
Five rim sherds may belong to a
single neckless jar exhibiting a direct rim.
The lip shows
exterior thickening and has been beveled flat on the top and
exterior surface.
Its orifice diameter is 11 em.
The upper
walls are 5 em thick.
Rim 12 (Fig. 39g).
Five rim sherds with gradual
exterior thickening show a short, squared protrusion and a
flattened lip.
The maximum thickness is 8 rnm.
The wall
below the rim is 2 rnm thick.
Rim 13 (Fig. 39h).
Two rim sherds show exterior
thickening in the form of a short, beveled flange.
Including the flange,
the rim's thickness is 7 rnm.
The
298
vessel's upper walls are 3 mm thick.
Shape B: Short-necked Jars
Represented by 1,728 rim sherds, short-necked jars are
the most common Paste A vessel type, comprising 81.1 percent
of the Manachaqui Phase rim collection.
The few partially
reconstructed short-necked jars show semi-carinated and
carinated body shapes.
The occurrence of simple globular
shapes is probable but remains undemonstrated.
Statistics
relating to diameters of rim forms with a sample size of 10
or greater are presented in Table 6.
Orifice diameters
range from 9 to 24 em, and most rim forms yield average
diameters between 13 and 14 em.
Interior surfaces remain
unfinished while exterior surfaces were smoothed to a matte
finish.
Potters nearly always burnished the tops and
interior surfaces of rims.
This seems to be an important
feature of Manachaqui Phase short-necked jars, perhaps
related to the vessels' functioning.
Two groups of 298 decorated rims are presented
separately at the end of the Shape B rim form descriptions
as Rims 9 and 10.
They were not included within the
existing morphological categories because decorative
treatment frequently obscured classificatory criteria.
Their larger average rim diameters may indicate different
vessel size and shape tendencies.
R~s
la-c {Fig. 39i-aa).
With 503 examples, Rim 1 is
299
the most numerous of the Manachaqui Phase Paste A rim forms.
Examination of broken sherd edges reveals that the potters
simulated vessel necks by reinforcing the exteriors of
otherwise direct rims.
The technique involved either
bending the upper wall outward, then curling and tucking the
end underneath the bent clay, or simply doubling the end
back over the vessel exterior 5 to 7 mm.
The folded clay
was then joined to the vessel wall exterior, often
haphazardly, with the fingers.
Most surface traces of the
"fold-over" technique were removed by beveling the rim
surfaces flat,
and then wiping the wet clay smooth.
Wiping
the interior surface of the rim often left a tiny cornice of
clay protruding over the top of the rim which potters did
not remove while burnishing the rim surfaces.
The total
thickness of the wall and folded portion ranges from 7 to 16
mm.
While Rim la (250 examples) shows a beveled end and a
rounded exterior (Fig. 39i-p), Rim lb (245 examples) was
beveled flat on the exterior surface as well (Fig. 39q-x).
Rim lc (eight examples) was beveled to create more than two
surfaces, rendering a "prismatic" effect (Fig. 39y-aa).
The
larger sherds show that interior surfaces were invariably
left rough and unfinished during manufacture.
The rim
illustrated in Fig. 40a shows the only handle found in the
Manachaqui Phase assemblage.
A hypothetical Shape B Rim lb
vessel with a globular body is depicted in Fig. 40b.
300
Rims 2a, b
(Figs. 40c-p, 4la-i).
Rim 2 and the
remainder of the Manachaqui Phase jar rims described below
have true curved, albeit short, necks.
Rim 2a (454
examples) is characterized by exterior thickening, often
achieved by the fold-over technique just described, and a
semi-squared lip (Figs. 40c-p, 4la, b).
These rims
typically lack the sharp edges left by beveling.
An
unusually large sherd with Rim 2a has a semi-carinated
profile and an applique flange has been added to the vessel
mid-section (Fig. 4la, b).
The maximum thickness of Rim 2a
varies between 5 to 10 mm.
Rim 2b (41 examples) is left
unthickened, but the lip is similarly semi-squared (Fig.
4lc-i) .
It is 4 to 5 mm thick.
Rims 3a, b
(Fig. 4lj-w).
Rim 3a (224 examples) is
thickened, often by fold-over, and rounding of the lip
frequently renders a bulbous appearance (Fig. 4lj-p).
thickness varies from 5 to 10 mm.
Its
Rim 3b (N=67) is
unthickened with a similarly rounded lip (Fig. 4lq-w).
It
is typically 4 mm thick.
Rim 4 (Fig. 42a-h).
Rim 5 (51 examples) is gradually
thickened, reaching maximum thickness at the lip which has
been trimmed flat.
A highly variable neck curvature
suggests that some of the corresponding vessel shapes were
more "open" than others, perhaps functioning as "bowls."
Fig. 42h shows a large rim sherd with applique decoration.
301
The maximum thickness of these rims varies from 5 to 8 mm.
Rim 5 (Fig. 42i-p, 43a-c, 44a-d).
These flanged rims
(37 examples) project horizontally or slightly upward 8 or 9
mm from a vertical or insloping wall.
The undersides of the
flanges often present smooth curves from the pointed tips of
the rims to the vessel walls.
lightly flattened.
The tips are pointed or
Some examples bear applique decoration
(Fig. 43a, b), while another has been decorated by incising
large squared notches (Fig. 44a).
Still others show deep
incisions, occasionally combined with applique (Figs. 44bd) .
Evidence suggesting that at least some of these vessel
shapes were semi-carinated (Fig. 43c) consists of matching
rim and medial flange decorative motifs.
Decorative motifs
and medial flanges are described in the section on
decoration.
Rim 6 (Fig. 45a-g).
Rim 6 (25 examples) has been bent
outward and its interior surface has been flattened.
Most
are lightly thickened, but taper toward the lip which may be
rounded or pointed as the result of the interior flattening.
At the neck angle, the rims are approximately 6 rnm thick,
and they extend 10 to 18 mm.
Rim 7 (Fig. 45h-l).
Rim 7 (19 examples) has been
thickened on the exterior surface or underside.
The lip
comes to a point where a tiny clay cornuice often protrudes
302
as a result of smoothing the interior surface.
The maximum
thickness ranges from 6 to 8 mm.
Rim 8 (Fig. 45m-o, 46a).
Rim 10 (nine examples) is
distinguished by thickening, a beveled, horizontal upper
surface and a pointed or lightly flattened tip which
approximates a flange (like Rim 5).
A partially
reconstructed vessel shows a semi-carinated profile (Fig.
46a) and bears applique decoration.
The walls are 3 mm
thick and the maximum thickness reaches 7 mm.
Rim 9 {with notched decoration)
48a-c).
(Fig. 46b,c, 47a-f,
Decorated Rim 9 (189 examples)
includes profiles
similar to those of Rims 2 through 4, 7 and 8.
There are
only two examples of Rim 1 vessels decorated by notching.
Three rim sherds with particularly large diameters (21, 22
and 24 em) may represent a special size class of this
vessel.
Rim 10 {with incised decoration)
{Figs. 48d,e, 49a-c).
Rim 10 (109 examples) includes rim forms similar in profile
to Rim 4, but unthickened variants are best compared to Rim
2b.
Where clay was applied to thicken the rim exterior, it
was usually haphazardly smoothed to the vessel wall.
The
process of cutting incisions deep into the lip displaced
clay outward further "thickening" the rim.
thickness varies from 5 to 9 mm.
The resulting
Rim interiors were always
303
burnished down to the point of maximum neck constriction,
and a tiny clay cornice left during the wiping stage extends
from the interior surface 1 to 2 mrn beyond the end of the
rim.
Fig. 49a shows a carinated body.
Some rim and wall
angles suggest that Rim 10 includes relatively open shape
variants (e.g. Fig. 49c).
Shape C: Restricted Bowls
Six restricted bowl shape variants are suggested on the
basis of rim morphology.
The Shape C restricted bowl
category consists of 113 rim sherds representing 5.3 percent
of the Manachaqui Phase rim collection.
The majority of
these bowls (Rims 1 and 2) are believed to exhibit globular,
semi-carinated or carinated body profiles although attempts
to retrofit entire body profiles were again unsuccessful.
Rim diameters normally range from 7 to 19 em, while the two
most popular classes show diameters averaging approximately
14 and 14.5 em.
Interior and exterior surfaces are
carefully burnished.
Statistical data are presented in
Table 7.
R~s
la, b (Fig. 50a-q).
A total of 68 rim sherds
representing Rim 1 probably belong to convex bowls with
globular body profiles.
Both interior and exterior surfaces
of these rims have been burnished to a low luster.
walls range from 2 to 4 mrn in thickness.
Vessel
Typical examples
of Rim la gradually thicken to a maximum 6 mrn (Fig. SOa-k).
304
Lips are rounded or show rounded interior edges and pointed
exterior edges.
The tip of Rim 1b's (11 examples) lip has
been trimmed flat (Fig. 501-p).
Fig. 50q offers an
hypothetical Shape C Rim 1a bowl with a globular body
profile.
Rims 2a, b (Fig. 51a-o).
The concave upper walls of
Rims 2a and 2b (27 examples) suggest that these sherds
belong to semi-carinated or carinated bowls.
Like Rim 1
sherds, they show careful polishing, similar angles of
orientation, occasional light thickening of the interior
surface toward the lip and similar rim and wall thicknesses.
The lip of Rim 2a (Fig. 51a-i) exhibits a sharp exterior
edge and rounded interior edge.
Rim 2b's lip (Fig. 51j-n)
has been lightly flattened by trimming.
Fig. 51o shows an
hypothetical Shape C Rim 2a vessel with a semi-carinated
body profile.
Rim 3 (Fig. 52a-b).
Eight examples of Rim 3 belong to
small, shallow convex bowls with well-smoothed interior and
exterior surfaces.
The rims have been gradually thickened
to 5 mm, and the lips show semi-squared profiles.
Rim 4 {Figs. 52c-e).
Seven rim sherds from small
globular bowls are strongly inverted like those of Shape A
Rim 4 vessels.
The thickness of these rims ranges from 4 to
8 mm, and lips are rounded or semi-squared.
Surfaces are
305
burnished and these bowls frequently carry applique
decorations on their rims.
Rim 5 (Fig. 52f). Two unthickened, direct rims with
squared lips also belong to bowls.
They are smooth, yet
unburnished, and the interiors show remains of a thick white
paint.
They are 3 mm thick.
R~
6 (Fig. 52g).
The sole example of Rim 6 belongs to
a large globular bowl with well-burnished surfaces.
It
shows slight, gradual thickening to a maximum of 6 mm and
the lip is semi-squared with rounded edges.
Shape D: Unrestricted Bowls with Concave Walls
There are 87 examples of unrestricted bowls with
concave walls representing four percent of the Manachaqui
Phase rim collection (Table 8).
This shape class
accommodates rim forms that apparently pertained to open
vessels with concave upper walls and semi-carinated or
carinated profiles.
Again, none of the vessels could be
adequately reconstructed to confirm suspected lower body
contours.
Rim diameters vary between 11 and 26 em, and are
greater on average than diameters of Shape B short-necked
jars.
Both the rim and upper wall interior and exterior
surfaces were usually burnished to a low luster.
Because of
the severe fragmentation of the collection, it could not be
confirmed whether or not a small number of Shape B short-
306
necked jar variants belonging to Rims 4 and 10 might fit
within this category.
It should be recalled that, for the
Manachaqui Phase shape inventory as a whole, the degree of
neck constriction may be a continuous variable.
The
following rim forms were confidently assigned to this
category.
Rim 1 (Fig. 53a-h).
These 32 everted rims are
unthickened with squared, and occasionally semi-squared,
lips.
The rim orientations range from nearly vertical to
approximately 45 degrees, and rim diameters are likewise
highly variable.
Sherd thickness varies from 4 to 5 mm.
Rim 2 (Fig. 53i-l).
Rim 2 (14 examples) is slightly
everted and lightly thickened.
The interior and end
surfaces of the lip have been flattened to create a dull
point.
Their maximum thickness reaches 8 mm.
Rim 3 (Fig. 54a-d).
Rim 3
slightly everted and thickened.
(13 examples) is also
The lip exterior has been
rounded rendering a dull point on the lip interior.
thickness varies between 6 and 7 mm.
Its
Fig. 53m shows a
hypothetical reconstruction of a whole vessel.
Rim 4
(Fig. 54e-g).
Rim 3
(13 examples) is also
slightly everted and thickened, but the lip interior has
been rounded leaving a dull point on the lip exterior.
is 6 to 7 mm thick.
It
307
R~
5 (Fig. 54h-j).
Rim 5 (13 examples) shows flange-
like thickening on the lip exterior.
The lips have been
flattened on their upper and exterior surfaces.
They are
approximately 7 mm thick.
R~
6 (Fig. 54k,l).
Two sherds belong to a single
small bowl with a globular or semi-hemispherical profile and
walls as thin as 1 mm.
mid-section.
An applique rib surrounds the vessel
The rim interior is thickened to 5 mm, and the
lip is pointed.
Shape E: Unrestricted Convex Bowls
The 26 rim sherds comprising this shape category
represent one percent of the Manachaqui Phase rim collection
and suggest a wide variety of vessel shapes and treatments
(Table 9).
Seven distinct rim profiles can be
distinguished, and meager sample sizes for each range from
one to five sherds.
There remains some question whether Rim
3 represents an intrusive example of later non-Manachaqui
Phase bowl rims.
R~
1 (Fig. 55a, b).
Rim 1 (6 examples) belongs to a
convex bowl with direct, vertical walls and rim.
thickness varies from 2.5 to 5 mm.
Wall
The lip exterior has
been thickened to a maximum 8 mrn, and the top of the lip has
been flattened.
R~
2 (Fig. 55c, d).
The six examples of Rim 2 pertain
308
to bowls with straight or slightly curved walls.
The walls'
angle of orientation varies from slightly insloping to
slightly outsloping.
The rim shows gradual thickening and
the lip exterior has been thickened by applying a fillet of
clay which was subsequently incised in a manner similar to
Shape B, Rim 10.
Maximum thickness at the lip ranges from 9
to 13 mm.
Rim 3 (Fig. SSe-h).
The five Rim 3 examples are
direct, nearly vertical, and belong to simple convex bowls.
They show gradual but slight interior thickening below an
in-sloping,
Rim 4
tapering lip.
Thickness is 3 to 4 mm.
(Fig. SSi, j).
Three examples of Rim 4 have
slightly concave walls and the interior surface has been
thickened toward the lip.
The lip is flat but slightly in-
sloping, has rounded edges and is 7 mm thick.
The lower
walls are 3.S mm thick.
Rim 5 (Fig. SSk, S6a).
These three rims show thickened
interior surfaces and rounded lips.
well-burnished.
Their surfaces are
All three examples carry applique button
decorations on the lip.
Rim 6 (Fig. S6b).
or two large open bowls.
The maximum rim thickness is 8 mm.
Two examples of Rim 6 belong to one
The walls have been gradually
thickened from 3 to 8 mm at the semi-squared lip.
309
R~
7
(Fig. 56c).
A single rim representing Rim 7
appears to belong to an large open bowl with polished
surfaces.
The rim is gradually thickened and ends at a
squared lip.
A small decorated ridge of clay protrudes from
the interior surface beyond the lip.
Miscellaneous Shapes
Of the three sherds (Fig. 57a-c) from Manachaqui Phase
strata that do not pertain to vessels, only the spoon handle
depicted in Fig. 57a could be identified with respect to
function.
The dish-shaped end of the spoon is broken and
does not show evidence of polishing or any special attention
to surface finish.
Such spoons occur with considerable
frequency in Early Intermediate Period contexts.
This
example may be the earliest yet recovered in the Central
Andes.
It is doubtful that the decorated sherd shown in
Fig. 57b belongs to a vessel because of its irregular shape.
The interior surface is rough and unfinished.
The third
piece (Fig. 57c) has a smooth tabular shape with rounded
edges backed by an irregular broken fragment.
R~
Decoration
Vessel decoration during the Manachaqui Phase is based
upon plastic modeling techniques.
It is additive in the
sense that applique provides the basis for most decorative
modifications.
Potters applied ribs, buttons and adornos
(defined here as representational protruberances with
310
zoomorphic or anthropomorphic features) to vessel rims,
shoulders and mid-sections.
Applique ribs and buttons were
usually embellished by notching or deep incision.
Fine-line
incised designs and painting occur only rarely in the
sample, and pattern burnishing appears on only one sherd.
Weathering may have affected the quantification of sherds
with these kinds of superficial decoration, but even the
best-preserved Manachaqui Phase sherds from Sector A rarely
show such techniques.
A total of 1,816 rim and non-rim sherds showing
decorative modification can be attributed to the Manachaqui
Phase, and each one of the five shape categories described
above is represented.
It is tempting to assume that vessels
with notched rims carried notched applique ribs on their
mid-sections, and that incised rims and incised ribs were
similarly associated.
However, the evidence at hand shows
plain rims with incised ribs and flanges (Figs. 38j, 41a)
and notched rims with incised medial ribs (Fig. 48c; Plate
VII) .
It still may be reasonable to deduce that rims and
flanges both bearing rare, identical adornos belong to the
same vessel.
Only a less fragmented sample of Manachaqui
pottery can serve to disclose the rules or tendencies
governing possible combinations of decorative elements.
The
following is a description of the techniques utilized to
decorate Manachaqui Phase vessel rims and bodies.
Confirmed
associations between techniques and vessel shapes will be
311
offered wherever possible.
Of the 2,129 Manachaqui Phase
rims, 360 (16.9%) exhibit some kind of decoration.
Decorations are notching, incision, applique buttons,
adornos and red paint.
Rim Notching
The most common technique employed to embellish vessel
rims during the Manachaqui Phase (216 examples) is notching
of the lip.
The inner and outer edges of squared lips, and
the ends of rounded or pointed lips, are notched at
intervals (3 to 20 mm) with a flat or rounded section of a
stick or cana to produce 2 to 3 rnm wide U or V-shaped
notches (Figs. 46b,c, 47a-f, 48a-c).
Often, the
aforementioned clay cornices left during manufacture were
notched by the same technique (e.g. Figs. 46b, 47a and f).
Where notches occur close together,
"serrated" appearance.
the rim presents a
One Shape B Rim 5 sherd (Fig. 44a)
shows unusually wide notches, while unusually thin notches
appear on Shape D Rim 5 (Fig. 54j) and Shape BRim 8 (Fig.
45n) .
The technique of directly notching the lip occurs on
Shape B Rims 2, 4, 7 and 8 necked jars, and on Shape D Rims
2 and 5 and Shape E Rim 7 unrestricted bowls.
On Shape A
Rims 2 and 3 neckless jars, the notching is executed on an
applique band attached to the lip exterior (Figs. 36a, b,
37a-c).
312
Rim Incision
The second most common technique with 133 examples
involves jabbing large punctations directly into the end of
a thickened lip with the end of a stick or cana (Figs. 44bd, 48d,e, 49a-c).
The gouging tool repeatedly struck the
surface at an angle of approximately 45 degrees as the
potter rotated the vessel.
The jabbing motion either
originat·ed directly over the rim, or angled toward or away
from the vessel interior.
The resulting punctations vary in
depth from 2 to 5 mm, reach 9 mm long and 4 mm wide and
range in distance from one another from 1 to 9 mm.
Gouging
of large punctations frequently gave the rim exterior an
uneven, swollen appearance.
However,
the interior of the
rim and neck was always maintained even and well-smoothed.
Incised rim embellishment occurs on Shape B necked jars
(Rims 5 and 10) and Shape E unrestricted bowls (Rim 2).
Occasionally the technique occurs in combination with
applique appendages like buttons and adornos.
Applique Rim Appendages
Eleven Manachaqui Phase vessel rims were also
embellished with buttons and adornos.
Seven sherds bear
buttons that appear as small disks approximately one
centimeter in diameter.
They may be notched or incised to
render a "slashed" appearance.
A button on the tip of
neckless jar Shape A Rim 6 (Fig. 38a) was grooved by
impression with a thin implement like a stick, while a
313
button perched on the edge of restricted bowl Shape C Rim 4
(Fig. 52d) shows two incisions cut with a sharp point.
Buttons on the flat top of necked jar Shape B Rim 5 (Fig.
44c and a similar example not illustrated) were fashioned
into doughnut-like circles before fastening to incised rims.
The buttons on the tips of three Shape E Rim 5 unrestricted
bowls (two illustrated in Fig. 56a,b) appear to have been
altered by pressing a small, nearly square-shaped implement
into their centers.
Four rim sherds are festooned with adornos.
In
addition to buttons, Shape B Rim 5 exhibits thin, undulating
rolls of clay suggestive of serpents (Fig. 43a, b).
A Shape
B Rim 5 sherd with incision carries a broken zoomorphic or
anthropomorphic adorno with only a preserved set of arms and
hands (Fig. 44d).
Fingers were rendered by incision, and
the arms show circular punctations.
off.
The head has broken
The small modelled face attached to the lip of Shape D
Rim 1 probably represents a snake (Fig. 53a).
Vessel Body Decoration
The Manachaqui Phase ceramic assemblage includes 1,456
decorated body sherds.
For the purposes of description, the
1,418 sherds with applique are divided into categories of:
1) medial and shoulder ribs and flanges, 2) notched and
incised bands and 3) other applique such as buttons, adornos
and lugs.
Described last are 31 sherds with fine-line
incised decoration.
The 1,258 examples of sherds with
314
applique medial and shoulder ribs and flanges are either
unembellished, notched or incised.
Rolls of clay modeled
into ridges with triangular cross-sections between 4 and 9
rnm wide were applied horizontally to encircle the exteriors
of vessels with globular, semi-carinated and carinated
profiles.
Potters fastened ribs by smoothing the upper edge
to the vessel walls.
edge and the wall.
Less care was taken to join the bottom
Sherds exhibiting rib ends indicate that
medial and shoulder ribs did not always completely surround
the vessel.
Appendages identified as flanges are
discontinuous, short in length and protrude higher from
vessel surfaces than ribs.
The final burnishing of vessels
followed the application of ribs and flanges.
Unembellished Ribs
Unembellished applique ribs are simply affixed coils of
clay that rise 2 to 7 rom above the vessel wall surface (Fig.
58a-d}.
There are 146 examples.
The high percentage of
sherds exhibiting both unembellished ribs and unsmoothed
interior surfaces (Table 10} suggests a close association
with Shape A and/or B jar shapes.
Only rarely do
unembellished ribs appear on vessel shoulders above the midsection (Fig. 58e-g}.
Notched Ribs
Notched medial and shoulder ribs are modified with the
basic technique of impression utilized on notched rims
315
(Shape B Rim 9).
Of the 350 notched ribs, the majority
(196) exhibit notches impressed directly and vertically into
the clay (Fig. 58h, i).
Oblique, vertical strokes produced
characteristic notches on a total of 55 sherds with ribs
(Fig. 58j, k), direct and diagonal on 24 (Fig. 581, m), and
oblique and diagonal on 37 (Fig. 59a, b).
One of two
notched rib variants, Notched A refers to ribs averaging
approximately 4 mm wide and 2.5 mm high which invariably
occur in the medial position.
Three out of four Notched A
sherds show smoothed or burnished interior surfaces,
suggesting association with bowl shapes C and/or D.
Thirty-
eight sherds with Notched B feature relatively small and
shallow notches on higher ribs protruding an average of 5.5
mm above the surface (Fig. 59c, d).
Even higher percentages
of Notched B sherds exhibit smoothed or polished interior
surfaces, also suggesting association with bowl shapes.
Because most Notched B ribs occur on the vessel shoulder
rather than in the medial position a specific association
with Shape C bowls is likely.
Examples of Notched B ribs in
both medial and shoulder positions can be observed on Shape
C Rim 4 and Rim 6 restricted bowls (Fig. 52e and 52g
respectively) .
Incised Ribs and Flanges
Incised applique ribs and flanges on a total of 617
sherds were decorated by two techniques.
The most common
with 532 examples and termed Incised A in Table 6, involved
316
cutting large gashes in the upper surface of medial ribs
with a long, thin blunt or hollow-ended implement (Fig. 59eg) .
The potter rotated the vessel (usually clock-wise)
while rhythmically cutting into the rib surface with a
jabbing motion.
Drag marks left on the vessel walls by the
jabbing implement are still visible where not erased by
subsequent burnishing.
curling tail (Fig. 60a).
One Incised A rib terminates in a
Incised A ribs occur with Shape A
Rim 7 neckless jars (Fig. 38j) and Shape B Rim 9 (Fig. 48c;
Plate VII) and Rim 10 (Fig. 49a) necked jars with notched
rims.
Eighty-five examples of the Incised B technique feature
deep parallel incisions cut into flanges up to 8mm high
(Fig. 60b-e).
The incisions were produced either by cutting
down and away from the vessel body from above with a
slashing motion, or by cutting across the upper surface of
the flange towards the vessel wall with a stabbing motion.
This latter motion often left round impressions where the
implement penetrated the vessel wall.
Most, or perhaps all,
Incised B flanges are discontinuous, and some resemble bird
wings or fish fins.
An Incised B flange decorates a Shape B
Rim 2a jar in Fig. 41a and b.
Applique Bands
A total of 149 sherds show that Manachaqui Phase
potters also applied bands of clay vertically and diagonally
to vessel shoulders, often in conjunction with medial ribs
317
(e.g. Figs 48c; 49a).
Here the term "bands" distinguishes
these decorative elements from the horizontal medial and
shoulder ribs and flanges just described.
Of these, 118
(84%) were embellished with the same Notched A technique
observed on medial ribs.
Seventy of these sherds (59%)
exhibit smoothed or polished interior surfaces (Fig. 60f-i),
while surfaces of the remaining 48 (41%) were left
unfinished (Fig. 60j, k; 61a, b).
Rim sherds and partially
reconstructed vessels show associations with Shape B Rim 8
(Fig. 45m, n; 46a), Rim 9 (Fig. 48c; Plate VII) and Rim 10
(Fig. 49a) short-necked jars, and Shape C Rim 1a (Fig. 50ik), Rim 2a (Fig. 51i), Rim 2b (Fig. 51n) and Rim 4 (Fig.
52c) restricted bowls.
Fourteen sherds bear applique bands with circular
punctation (Fig. 61c, d).
These bands protrude
approximately 6 mm from the vessel surfaces.
They appear on
the shoulders of Shape A Rim 4 neckless jars, a Shape C Rim
1b restricted bowl, as the tail of an Incised A medial rib
from a vessel with a burnished interior (Fig. 61d),
associated with an Incised A medial rib on a Shape D Rim 6
unrestricted bowl (Fig. 54k) and as the "arms" of the broken
adorno of Shape B Rim 5 (Fig. 44d).
There are six examples of applique bands with ovoid
incisions.
surfaces.
These bands protrude 5-6 mm from vessel
Two sherds, one from a rim (Fig. 42h) and another
bearing an otherwise unembellished shoulder rib (Fig. 61e),
318
may belong to the same Shape B Rim 4 short-necked jar.
One
of two nearly identical sherds with burnished interior
surfaces shows association between bands (perhaps parallel)
with ovoid punctation, buttons with ovoid punctations and
Incised A medial ribs (Fig. 61f).
Nine examples of applique bands from Manachaqui Phase
stratigraphic contexts are too eroded to further identify.
A sherd on which the applique band has broken off features
what appears to be a serpent head (Fig. 61g).
A single
applique band which might be a "renegade" Incised A medial
rib meanders serpent-like across the shoulder of a vessel
with an unfinished interior surface (Fig. 62a).
Other Applique Appendages
Other applique appendages that Manachaqui Phase potters
affixed to vessel bodies include buttons, adornos and a
single lug.
The single lug appears as a knob where a
vertical notched band meets a Notched A medial rib on a
sherd with a burnished interior (Fig. 62b).
Of four body
sherds with buttons, two were described above in association
with applique bands (Fig.
61f).
Another body sherd with an
unfinished interior surface features a button with two
parallel slashes resting atop an unembellished medial rib
(Fig.
62c).
The fourth sherd has a finished interior, a
button with four parallel slashes and an Incised A medial
rib (Fig. 62d) .
Of eight adornos from vessel bodies, six are broken
319
from mid-sections of vessels of unknown shape, while the
seventh adorned the shoulder of a Shape A Rim 4 neckless jar
(Fig. 37c).
This seventh adorno appears as a protruding
snout with applique eyes affixed to either side.
Like other
Manachaqui Phase adornos, it is zoomorphic, rendered in
abstract fashion, and cannot be easily identified as to
species.
Three sherds with snake adornos on unembellished
medial ribs (two are shown in Fig. 62e, f) probably belong
to the Shape B Rim 5 necked jars with identical adornos atop
their rims (Fig. 43a-c).
Fig. 63a illustrates the modeled
head of an amphibian or fish.
Like the Shape A Rim 4 adorno
just described, it was formed using small bits of shaped and
incised clay to render facial features.
This adorno, and
another which may represent a bird wing or fish fin (Fig.
63b), come from vessels with unfinished interior surfaces.
The remaining two adornos are knob-like protrusions,
apparently representing heads, one with only a mouth (Fig.
63c) and another with only eyes (Fig. 63d).
Lack of detail
inhibits speculation with regard to species represented.
Incised Decoration
A total of 31 sherds from three Manachaqui Phase
vessels exhibit incised decoration on their exterior
surfaces.
All of the sherds show burnished interior
surfaces.
Incisions were cut while the clay was still moist
after smoothing or burnishing the entire exterior.
None of
these were included in Appendix D totals for incised sherds.
320
Xncised Vessel 1 (Fig. 64a-c).
Twenty sherds with
zoned punctation pertain to a single vessel.
Despite the
burnished interior surfaces of this "bowl," the rim form is
most comparable to Shape B Rim 4 necked jars.
This
inconsistency demonstrates the potential pitfalls of
applying a strict "bowl" vs. "jar" dichotomy to the
Manachaqui Phase collection.
The design exhibits parallel
vertical lines 16 mm apart that serve to create zoned fields
alternately filled with triangular or dashed punctations.
The entire exterior of this vessel was burnished prior to
incision.
The motif begins just under the rim and ends at
the basal angle or corner point.
Xncised Vessel 2 (Fig. 64d, e).
Ten sherds probably
pertain to a single restricted vessel with a carinated
profile and unknown rim form.
Two partial reconstructions
show zoned triangular and step-shaped fields filled with
thin, slash-like punctations.
The exterior surface was
smoothed to a matte finish prior to incision.
Lines
discernible on one of the reconstructions (Fig. 64e) suggest
that the triangular and step-shaped fields may alternate
around the exterior of the vessel.
Finally, a sherd from
another vessel displays deep and relatively wide incisions
(Fig. 64f).
Two vertical parallel lines border a series of
short horizontal, parallel lines.
321
Pattern Burnishing and Painting
Pattern Burnishing and painting are two techniques
which may be under-represented in the Manachaqui Phase
collection because of its weathered condition.
A single
sherd (not illustrated) exhibits exterior surface burnishing
with a spider web-like pattern similar to that on a
Pacopampa bowl illustrated by Rosas
(1976:572, Lamina Sa).
Six sherds exhibit remains of red paint.
One is a Shape B
Rim 1b necked jar rim sherd with red paint or slip on its
exterior.
Four articulating sherds from a Shape D Rim 5
unrestricted bowl with concave walls (Fig. 45h) show red
paint on both surfaces, except on the rim proper.
One sherd
found in Sector A and pertaining to a Shape D Rim 1
unrestricted bowl bears traces of post-firing red paint on
both surfaces.
The sherd's relatively dry Sector A context
allowed preservation.
Ceramic Paste Group B
The Manachaqui Phase ceramic assemblage includes 109
sherds assigned to Paste Group B.
Twenty-one Paste Group B
rims comprise one percent (0.98%) of the total 2150
Manachaqui Phase rims.
As a group the Paste B pottery is
distinguished by glittery surfaces, yet as many as three
pastes or wares may be represented.
Manachaqui Phase Paste
B pottery differs from Paste A not only in the pastes
employed, but in details of vessel shape and decorative
techniques.
These Paste B vessels may have been traded to
322
the Paste A-producing population, or they may have been
carried in by persons that only rarely frequented the
shelter.
Possible sources will be discussed later.
In general these pastes are even more compact and finegrained than Paste A.
A cursory microprobe analysis
identified quartz, alkali feldspar and ilmenite mineral
constituents.
Clear crystalline quartz and mica produce the
glittery effect.
Non-plastic inclusions are well-sorted
except for occasional large grains of vein quartz.
The
pottery colors range from a light tan to a dark grayish
brown.
Many sherds show incomplete oxidation and, because
the fabric is softer than Paste A pottery, sherd surfaces
tend to be more eroded.
Nevertheless, a glossy finish
observed on well-preserved specimens attests to burnishing.
Paste 8 1
Fifty-six sherds from a minimum of five vessels share
the same paste referred to as Paste B1 •
Surface colors are
7.5YR5/6 (strong brown) and 10YR7/3 (very pale brown).
Vessel 1 (Fig. 65a) is a short-necked jar with an everted
rim, mouth diameter of 13 em, and body (exterior) diameter
of 16 em.
The rim is lightly thickened and the body profile
is carinated.
Vessel 2 (Fig. 65b) is a neckless jar with a
mouth diameter of 14 em and an unknown body profile.
The
rim exterior has been thickened, beveled and grooved below
the lip.
Vessel 3 (Fig. 65c) is a carinated, restricted
bowl with a direct, unthickened rim and a mouth diameter of
323
13 em.
Red, post-firing paint (10R4/4 weak red to 10R4/6
red) covers the exterior but, like the Shape D Rim 5 bowl
described earlier (Fig. 45h), the end of the rim was left
unpainted.
Vessel 4 (Fig. 66a) is a restricted bowl represented by
small rim sherds with an unusual profile.
The lip exterior
has been thickened and brought to a dull point.
interior shows a shallow groove under the lip.
The rim
The five
sherds that constitute Vessel 5 (Fig. 66b) probably belong
to a double-spout-and-bridge bottle, although they may also
represent an asymmetrical bottle with a single spout and
strap handle (DeBoer n.d.).
The hypothetical reconstruction
illustrated here is based upon a double-spout-and-bridge
bottle illustrated by Lathrap (1962: 935, Fig. 49e; 1970:91,
Fig. 9j), and assigned to the Late Tutishcainyo Phase.
Paste B2
The 31 sherds consisting of Paste B2 belong to a single
carinated vessel (Fig. 66c; Plate VIII) that was almost
entirely reconstructed.
The paste is distinguished from
Paste B1 by larger non-plastic grains and a less compact
fabric.
The rim is unthickened and sharply everted, or
flanged, with a diameter of 9 em.
The lip has been squared.
Beneath the lip exterior, two notched applique bands run
vertically down the upper vessel wall to end at the corner
point of the basal angle.
The basal angle has been
decorated with small nicks encircling the entire vessel.
324
Paste
~
Paste B3 contains less mica and larger quantities of
crystalline quartz.
Its color is brownish gray.
These
sherds are more difficult to distinguish from Paste A, but
sixteen were isolated to reconstruct a jar with an angled
neck and an everted rim 13 em in diameter (Fig. 67a).
The
body profile is globular although it carries a notched
medial rib.
The notching is deeper and spaced at wider
intervals than Notched A decoration observed on Paste A
vessels.
Arch motifs rise from the medial rib at opposite
sides of the vessel (only one of the two is depicted in Fig.
67a) .
Lithic Remains
Because of the stratigraphic conditions at Manachaqui
Cave, the assignment of stone artifacts to particular phases
presents some difficulties.
As previously noted, the
ceramic macrochronology serves as a reference for situating
lithic remains within the established chronological
framework.
Examination of the macrochronology reveals that
chipped-stone tools and debitage virtually disappear from
the archaeological record after the Terminal Preceramic
Period, and do not reappear in substantial numbers until the
Early Intermediate Period.
Of the approximately 9,500
lithic artifacts from the ceramic-bearing deposits, only an
estimated 50 (or 0.5%) may truly pertain to the Manachaqui
Phase.
None of the retouched chipped stone tools from the
325
rockshelter's ceramic-bearing deposits can be confidently
assigned to the Manachaqui Phase.
In contrast, nine ground
stone artifacts and cobble tools were retrieved from unmixed
Manachaqui Phase levels.
Ground and polished points of shale and slate
constitute the most numerous class of ground stone artifact
during the Manachaqui Phase.
Although 72 such points were
recovered from the excavations (Plate IX), not one remains
intact.
Like the pottery, these brittle points probably
suffered breakage from trampling.
Four point mid-sections
and a tip were unearthed from Manachaqui Phase deposits.
Mixed Manachaqui and Suitacocha Phase strata yielded seven
more point fragments.
Evidence that points were
manufactured at Manachaqui Cave consists of small stone
chips and preforms.
Twelve chips from Lavasen Phase strata,
and four preforms or blanks from mixed Lavasen and
Manachaqui Phase strata provide evidence of experimental
shale and slate working at the end of the Precerarnic Period.
Other slate artifacts include a disk with two drilled
perforations and incised lines on both sides (Fig. 67b), and
a tiny tubular bead recovered while screening Floor Z soil
samples.
Of six ground stone tools and tool fragments,
three
were recovered from unmixed Manachaqui Phase contexts in
Sector B.
These include a porous sandstone cobble (from
Unit 24, Level 15) faceted and grooved by abrasion that
326
would have been an appropriate implement for shaping the
shale and slate points.
It was found close to a disk-shaped
pebble (from Unit 24, Level 16) with a concave surface that
might have served for polishing or sharpening the points.
The third piece is a granite cobble fragment that lacks
clear signs of wear, but may have been shaped by pecking.
The three remaining pieces from mixed layers were unearthed
in Sector A.
These include a hand-size cobble of felsic
igneous rock with wear suggesting use as a milling stone and
hammer (Unit 11, Level 7).
The edges and one surface show
polish, and the ends show chipping from blows.
Two flakes
from a finely ground and polished tool like an axe bit may
have served as cutting tools.
Rocks and Minerals
Numerous rocks and minerals brought from outside the
Manachaqui Valley were recovered during excavation of
Manachaqui Cave.
Seven chunks of white, angular vein quartz
appear in Manachaqui Phase deposits and the layers
immediately above and below.
Uses for these minerals remain
unclear, but quartz deposits in and around Pataz frequently
bear sulfide ores such as gold and silver.
No such ores
were observed, and the chunks may simply have been
appreciated as hard stone with sharp edges for cutting and
chopping.
A chunk of pegmatite, a small fragment of
prismatic quartz crystal and two platy pieces of mica were
also recovered from Manachaqui Phase levels.
327
Botanical Remains
Because only the samples extracted from Sector A were
analyzed by Dr. Pearsall, ethnobotanical information predating the Terminal Preceramic Period is lacking.
The
botanical taxa identified from Manachaqui Phase Floors AA
and
z
are summarized in Table 18.
These include fruit rinds
of the Sapotaceae family similar to lucuma (Pouteria spp.)
These and other unidentified fruit rinds may represent
"exotic" foods brought from the lower elevations of the
montane forest to the east.
Foods that probably originated
locally in the Tropical Alpine Zone near Manachaqui include
tuber and root fragments that likely belong to sedge
(Cyperus or Scirpus).
Seeds identified are Festuca and
either Chenopodium or Amaranthus, all native to the paramo
grassland formation.
All of these taxa, plus Lupinus;
Ribes; Rubus; and Polygonum/Rumex seeds occur in samples
from Terminal Preceramic Lavasen Phase Floors, which
suggests continuity in the rockshelter's economy into the
Initial Period.
Faunal Remains
A total of 72 faunal specimens were recovered from
unmixed Manachaqui Phase levels in both sectors, and 52 were
identified as Mammalia by Kent (Table 21).
Represented are
guinea pigs, mountain paca, white-tailed deer and armadillo.
That only two of 72 specimens could be positively identified
to the species level is a function of the faunal material's
328
•
severe fragmentation.
Comparison of the Manachaqui Phase inventory with the
Terminal Precerarnic Period inventory is problematic because
Sector B's unmixed Terminal Preceramic Period levels are
difficult to isolate without further artifact analyses.
Sector A's Lavasen Phase strata yielded only seven of the 97
identified taxa recovered from all of Site M-1A's Preceramic
Period deposits (Table 19).
All seven specimens are
Mmmnalia, two are identified as Artiodactyla, and one as
Cavia.
Of the additional 90 Preceramic Period specimens
from Sector B, 88 are Mammalia and two are Aves.
One
Mammalia specimen belongs to the Caviidae Family, while one
Ave specimen is of the Strigiformes Family.
Specimens from
mixed Lavasen and Manachaqui Phase strata were also
tabulated (Table 20).
Manachaqui Phase Chronology
Unfortunately, radiocarbon evidence processed so far
does not offer a clear idea of the Manachaqui Phase's
beginnings.
The three radiocarbon dates of 850, 860 and 900
b.c. date only the upper portion of the Manachaqui Phase
deposit which, on the whole, is thicker and richer in
artifact density than deposits of any subsequent phase.
Radiocarbon dates for the Lavasen Phase (1880, 1720 and 1570
b.c.) and the Suitacocha Phase (790,
680, 610, 500 b.c.)
bracket the Manachaqui Phase occupation approximately
between 1500 and 800 b.c.
Also, future research may allow
329
the subdivision of the Manachaqui Phase into early and late
sub-phases.
Potentially late in the phase are the incised
vessels of Fig. 64, and most of the Paste Group B vessels
(Figs. 65-67a).
Although present evidence is suggestive, my
efforts to sub-divide the phase have not succeeded.
The dates of 1500 to 800 b.c. for the Manachaqui Phase
generally agree with radiocarbon evidence for the
chronological placement of other early Initial Period phases
such as Kotosh Wairajirca (Izumi and Terada 1972:308),
Yesopampa (Terada 1979:173-174), Early Huacaloma (Terada and
Onuki 1985:267-268; Onuki 1993:Fig. 15, p.91) and Pandanche
A (Kaulicke 1981:388).
Comparisons to be offered in the
following sections suggest that pottery styles associated
with these north highland phases are coeval.
The sets of
attributes shared with southern Ecuadorian Formative
Machalilla and Chorrera Phase styles, during the Manachaqui
and Suitacocha Phases respectively, suggest a later
beginning date for the Manachaqui Phase around 1200 b.c.
The shared attributes will be described below and in Chapter
7.
Manachaqui Phase Ceramic Relationships
This examination of relationships during the Manachaqui
Phase will focus on the shelter's abundant pottery.
The
evaluation of migration hypotheses requires detailed
comparative analyses and the identification of specific
shared attributes.
At issue is whether the Manachaqui Phase
330
Paste A assemblage can or should be regarded as a derivative
of ceramic traditions in neighboring regions to the west,
south, east or north.
This comparative analysis of
Manachaqui and other early Initial Period pottery styles
also seeks to elucidate aspects of cultural geography,
spheres of interaction and sources of innovation during the
Preceramic-Initial Period transition.
The Manachaqui Phase
Paste A style should contain evidence for: 1) the cultural
affiliation of Manachaqui Cave's users (e.g. Andean or
Amazonian) and 2) cultural interaction on a variety of
scales.
Of course Manachaqui Phase Paste B pottery provides
the most direct evidence for interaction, and its abundance
provides one gross measure of the interaction's intensity.
The interpretation of the Manachaqui Paste A style's
relations requires assumptions regarding the relative
significance of certain attributes.
One reason that Lathrap
and Meggers can derive conflicting interpretations of
historical relationships from the same data stems from their
disagreements regarding which shape or decorative attributes
are diagnostic of migrations or diffusion (Meggers and Evans
1983:330-331}.
Simple guiding principles utilized in the
interpretations offered here and in subsequent chapters are
that: 1} a relatively high number of shared attributes is
most likely to indicate particularly intense interaction,
and 2} statements regarding directionality are credible only
where temporal priority has been demonstrated for one of the
331
ceramic complexes in question.
As a general rule for the purposes of this study, the
co-occurrences of single attributes, unless they are complex
(e.g. double-spout-and-bridge bottle shapes), will be
accorded little significance.
Co-occurrences of multiple
attributes will be granted greater validity, especially when
similarities in both vessel shape and decorative attributes
converge.
It must also be remembered that, because of
Manachaqui Cave's presumed special functions, its pottery
sample most likely represents a subset of a more diverse
"local" tradition.
Therefore, the absence of ceramic
attributes at Manachaqui (negative evidence) must be
carefully considered before it can be accorded significance.
The final chapter of this thesis will synthesize the
evidence for interregional interaction during this and
subsequent phases in order to illuminate trends that
accompany the emergence of complex Central Andean societies.
Paste A Relationships:
the Central Andes
The Manachaqui Phase Paste A assemblage has
surprisingly little in common with the relatively simple
pottery styles found in the heart of the Central Andes
immediately west and south of Manachaqui Cave.
Early
Guafiape (Strong and Evans 1952), Las Haldas Phase 1 (Fung
1972), Yesopampa (Terada 1979), Toril (Burger 1985b) and
other Central Andean assemblages feature neckless ollas and
332
simple convex bowl and bottle shapes.
The globular variants
of Manachaqui neckless jars (e.g. Shape A Rims 2 and 6) do
resemble the ubiquitous Central Andean neckless ollas, but
these constitute less than eight percent of the Manachaqui
Phase Paste A assemblage.
Manachaqui Cave's special
function may account for the scarcity and relatively small
size of these vessels.
However, Manachaqui neckless jar
shapes also tend to be shallower and more open than their
Central Andean counterparts (further discussion of this
shape will follow below) .
Convex open bowls typical of the
Central Andes are rare at Manachaqui where gourds could have
been used as dishes for food or drink.
Perhaps the
dissimilarity between the Manachaqui Style and styles to the
west and south is most clearly pronounced in the absence of
necked jar shapes, reinforced and beveled rims, and
carinated body profiles in early Initial Period Central
Andean assemblages.
Most early Central Andean pottery styles feature simple
decorations rendered by rudimentary incision, punctation and
applique techniques.
In contrast, Manachaqui Style
decoration is based almost exclusively on relatively
elaborate applique and modeling techniques.
Many of the
Manachaqui Phase techniques can be found in isolation in the
Central Andes.
For example, rim notching occurs at Caballo
Muerto in the Meche Valley (T. Pozorski 1983: Figs. 13, 17).
However, it appears on neckless olla rims and the
333
technique's frequency and temporal position remain to be
assessed.
Also, gouged incisions and rim notching can be
observed on purportedly "early" neckless ollas and convex
bowls in Santiago de Chuco (Perez C. 1988: Lam. 1).
Manachaqui Phase pottery blends these and other singular
decorative features to render a unique and highly
distinctive style.
One·ceramic assemblage in the heart of the Central
Andes to the southwest deserves special note because of a
number of attributes shared with the Manachaqui Style.
La
Galgada in the middle Santa Valley yunga yielded a small
sample of Initial Period ceramics including vessels with
carinated profiles (Grieder 1988: Fig. 156gg), medial
applique ribs (Ibid.: Fig. 156aa), incipient necks (Ibid.:
Figs. 156ee, ff, gg and ii) and flattened and beveled rims.
These La Galgada pieces are strikingly similar to the
earliest Manachaqui ceramics.
Grieder (Ibid.:185) reports
that pottery appears at La Galgada between 1790 b.c. and
1565 b.c.
Nearer to Manachaqui Cave and directly west across the
Marafion Canyon, several investigations in the Huarnachuco
area have failed to isolate early Initial Period
occupations.
Surface finds assigned to the Initial Period
Marnorco Phase by Thatcher (1979: Figs. 5-8) include sherds
from necked vessels with thickened rims resembling
Manachaqui Phase Shape B jars.
However, recent efforts by
334
Topic and Topic failed to corroborate Thatcher's Mamorco
Phase, even at Cerro Mamorco (Theresa Topic personal
communication 1987).
Neckless ollas and brushed pottery
recovered during the Topics' surface reconnaissance at Cerro
Huachac may date to the early Initial Period, but later
materials predominate (T. Topic personal communication
1994).
None of these authors report early carinated vessels
or other design features diagnostic of the Manachaqui Style.
The Manachaqui style most closely resembles the
earliest pottery at intermontane valley and western slope
sites to the northeast in Cajamarca Department.
The best-
documented assemblages come from Huacaloma (Terada and Onuki
1982, 1985), Cerro Blanco (Terada and Onuki 1988) and
Pandanche (Kaulicke 1981) in the intermontane highland
valleys, and Montegrande in the middle Jequetepeque valley
yunga
(Ravines 1982; Ulbert 1994).
Typical Central Andean
neckless ollas predominate in these north-Central Andean
assemblages, yet the presence of necked vessels with
carinated and semi-carinated body profiles distinguishes
them from more southerly Central Andean assemblages.
Early
Huacaloma Phase neckless olla, carinated jar, concave wallbowl, spherical bowl and open bowl shape categories loosely
correspond to Manachaqui Shapes A through E.
Cerro Blanco's
La Conga Phase shape inventory is more closely analogous to
Manachaqui's because of the predominance of carinated
vessels and the dearth of open bowls.
Open bowls are
335
entirely absent during Phase A at Pandanche where restricted
vessel shapes were likewise preferred.
In general, the shape proportions of Manachaqui Phase
vessels differ from those found in the Cajamarca region.
Manachaqui vessels tend to be narrower, necks are
consistently shorter and carination angles occur one or two
centimeters lower on vessel bodies.
True carination angles
are far more common in Cajamarca than at Manachaqui where
potters mimicked them by attaching applique ribs to the midsections of semi-carinated and globular vessel bodies.
At
this early stage of research in northern Peru, it is
difficult to evaluate the effect of site function on shape
inventories and design features.
Ceramic assemblages at
Huacaloma, Cerro Blanco, Montegrande, Pandanche and
Manachaqui ostensibly reflect ceremonial, domestic and
traveling activities.
Cajamarca's repertoire of decorative techniques is
larger and more diverse that Manachaqui's, yet many of the
applique and modeling techniques common to both areas differ
in details.
Montegrande potters habitually notched clay
strips applied to vessel rims in addition to notching the
rim directly.
Some decorative techniques such as grooving
and cane-stamping (Terada and Onuki 1982: Pl. 73-3, 10;
Ulbert 1994: Lam. 36-1, 2) are not found at Manachaqui Cave
and may be unique to the Cajamarca sites.
Among other
decorative techniques rare or entirely absent at Manachaqui
336
are the simple incised line motifs, zoned punctation and
reddish slip at Montegrande and Huacaloma, and the brushed
decoration at Pandanche.
Because the rarity or dearth of some of the Cajamarca
techniques at Manachaqui may be due to the rockshelter's
special functions,
the Manachaqui Phase assemblage's
singular qualities are more significant.
Unique to
Manachaqui are techniques of rim treatment and lip incision,
while rim embellishment is more common.
Also characteristic
of Manachaqui are the inordinate emphasis on "false"
carination angles rendered with medial ribs, and the varied
incision techniques employed on rims and applique ribs.
Modeled adornos occur at Montegrande, but apparently not in
the Cajamarca Basin.
Curiously, Manachaqui also shares
distinctive rim thickening techniques and varieties of
carinated bowl shapes (Ibid.: Lams. 9-13), two-tiered
shoulders (Ibid.: Lam. 1) and applique flanges and notched
strips (Ibid.: Lams. 33-35, 17) with Montegrande that it
does not share with the closer Cajamarca highlands.
Some of
Montegrande's zoned-incised decorative motifs (Ibid.: Lams.
14-3760, 14-3863, Taf. 37-7) are virtually identical to
those of the Manachaqui Phase.
Unlike Manachaqui's, the Montegrande shape inventory is
still dominated by neckless ollas (Ibid.:58).
Jar shapes
typically show longer necks and all are assumed to have
globular bodies.
Montegrande pottery's resemblance to the
337
Manachaqui assemblage is difficult to evaluate because
assemblages belonging to more than one phase appear to be
conflated.
Ulbert points out (Ibid.:149) that all of the
Cajamarca sites share carinated vessel shapes not found
farther south.
South of Manachaqui Cave lie eastern Andean sites in
similar geographic and ecological circumstances that have
also produced early assemblages.
At 3,800 m elevation on
the Marafion-Huallaga divide in Huanuco Department, Piruro
(Bonnier 1988) occupies a location analogous to Manachaqui's
near the montane forest edge.
Illustrations (Rozenberg
1982: Planche 6) show that the classic Central Andean
neckless olla again constitutes the predominant vessel
shape, and Rozenberg (Ibid.:135) expresses puzzlement at not
encountering similarities to the coeval Kotosh Wairajirca
Style nearby, nor any clear traces of Amazonian influence.
Approximately 90 km farther south in Huanuco
Department, local topography creates an arid rain-shadow,
and highland-like environmental conditions extending down
the eastern slopes below 2,000 m.
At the site of Kotosh,
necked jars with thickened and beveled rims comprise a
sizable portion of the Wairajirca assemblage.
Wairajirca
Red Plain Form 6 (Izumi and Terada 1972: Plate 119, Figs. 9-
13) constitutes the second most popular plain ware vessel
shape after the neckless ella.
Necked jars also occur in
significant numbers as Wairajirca Red Line-burnished Form 3
338
(Ibid.: Plate 125, Figs. 11 and 12), Wairajirca Zoned
Hachure Form 1 (Ibid.: Plate 122, Figs. 1-4) and Wairajirca
Black Line-burnished Form 2 (Ibid.: Plate 124, Fig. 6).
These shapes have identical counterparts in Manachaqui's
Shape B Rim 1 and Rim 2 globular jars.
Wairajirca spherical
bowls match Manachaqui's Shape C Rim 1 bowls, although this
somewhat generic shape is less useful for comparative
purposes.
Despite the Kotosh potters' familiarity with carinated
vessel shapes, Wairajirca short-necked jars reportedly have
only globular or "boat-shaped" bodies.
Eccentric vessel
shapes lend Wairajirca a slightly more varied shape
repertoire than Manachaqui's.
The most conspicuous
differences between Wairajirca and Manachaqui ceramics are
found in modes of surface treatment.
Wairajirca potters
emphasized incised and post-firing painted decoration, and
eschewed applique.
However, they did affix zoomorphic
adornos to vessels (Kano 1979).
features are unique to Kotosh.
Numerous other design
These include the eccentric
shapes, and modelled zoomorphic and anthropomorphic faces
(Kano 1979) .
Continuing southward along the eastern highland edge,
earliest ceramics from the upper Mantaro Valley puna of
Junin Department have been only superficially studied.
However, Silva (1988:Figs. 13-21) describes a collection
from Pachamachay that includes short-necked jars with
339
thickened, and often beveled, rims.
He regards the
collection as no earlier than 1200 b.c., and likens the jar
rims to Kotosh Plain jar Forms 5, 6 and 7 {Izumi and Terada
1972: Pl. 114:20-30).
The Pachamachay jar rims, and the few
Initial Period jar rims from nearby Telarmachay {in the
upper Tarma River drainage) illustrated by Lavallee {1977:
Lam. 6b, d), are comparable to the previously cited
Wairajirca Red Plain Form 6 and Manachaqui Shape B rim forms
(e.g. Figs 42a-g this thesis).
Apparently, neither the
carinated vessel shape, nor the diversity of decorative
modes that characterize Manachaqui and Kotosh appear in the
earliest Junin assemblages.
At Waywaka in the upper Apurimac River drainage
radiocarbon dates ranging from 1600 to 710 b.c. justify
comparisons between the Manachaqui Phase assemblage and the
Initial Period Muyu Moqo Style {Grossman 1972, 1985).
The
Muyu Moqo A and subsequent Muyu Moqo B style are based on
simple neckless olla and convex bowl shapes typical of the
Central Andes.
Only notching and incising on occasional
rims {and one possible Phase B sherd from a carination
angle) are techniques reminiscent of the Manachaqui Style
(1972: Figs. 37, 41, 45, 46, 52, 145).
During Muyu Moqo
Phase C-D, which Grossman (1985:59) dates to the second half
of the second millennium B.C., a series of new formal and
decorative features appears.
Notched carination angles
characterize Muyu Moqo C-D, and notched "ridges" of clay
340
decorate the upper shoulders of neckless ollas approximately
two ern below the rim (Grossman 1972: Figs. 79-81, 89-96).
Muyu Moqo C-D necked jars with thickened and beveled rims
(Ibid.: Figs. 131-135) are termed "New, Non-Traditional
vessel forms."
Their consistent association with novel
decorative features including incised and punctate applique
fillets and buttons (Ibid.: Figs. 126, 127, 136, 171, 173,
174) leads Grossman to regard the decorated jars as "the
result of ideas being introduced from outside the area
rather than the result of local experimentation"
(Ibid.:120).
The attributes that constitute Grossman's
evidence for an intrusive complex are among the diagnostic
features of the Manachaqui Style.
The necked jar appears again as the predominant vessel
shape at Initial Period Marcavalle (Mohr-Chavez 1980, 198la,
198lb) at 3314 rn elevation in Cuzco's upper Urubamba River
drainage.
Mohr-Chavez documents Marcavalle's changing
interregional relationships throughout its four or five
century occupation between 1100 and 600 b.c.
Ceramic
evidence suggesting relationships between the Marcavalle
Phase A and B assemblages and the Manachaqui style includes
the aforementioned necked jars and necked ollas, some of
which bear folded-over, flanged and beveled rims (MohrChavez 198la:128), and carinated bowls.
Marcavalle necked
jars (Ibid.: Figs. 27b, c, r) and carinated bowls (Ibid.:
Figs. 331, rn) both show the ridged shoulder or two-tiered
341
profile as well.
One Marcavalle Phase A jar with "IV.A.
stepped element" rendered in red paint and zoned by incision
(Ibid.: Fig. 32k; 198lb: Fig. 36) bears a remarkably close
resemblance to Manachaqui Phase Incised Vessel No. 2.
Mohr-Chavez (1981b:327-328) likens the abundant incised
and punctate applique flanges and lugs (1981a: Fig. 35) to
similar features observed at Kotosh and Tutishcainyo, and
many also have Manachaqui counterparts.
"fillets" occur rarely at Marcavalle.
Otherwise, applique
The use of red
specular hematite paint appears as early as Phase A.
Mohr-
Chavez (1981b:331) notes that design features including
carinated vessel shapes, applique decoration and doublespout-and-bridge bottles with short spouts (Ibid.: Fig. 33a)
link Marcavalle to Muyu Moqo C-D and the south coastal sites
of Hacha and Erizo.
Recent work on the eastern slopes far
south of Marcavalle in Bolivian Cochabamba Department
(Brockington et al. 1995) has yielded short-necked globular
jars with everted, thickened and occasionally notched rims
(Ibid.: Figs 7a, 7b [top], 18a, 18b) contemporaneous with
the Manachaqui Phase and comparable to Shape B.
Paste A Relationships: the
Amazonian Lowlands
In the lowlands directly east of Manachaqui Cave, the
limited archaeological investigation performed within the
lower Hual1abamba and Abiseo, and central Huallaga valleys
has failed to locate evidence for such early human
342
occupation.
One fragment of a "nicked shoulder flange"
(DeBoer 1984: Fig.6h) found at Site HUA-3 and grouped with
relatively late surface finds recalls the Manachaqui Phase
notched applique ribs, but little can be inferred from a
single sherd.
With the Tutishcainyo styles in the Ucayali basin east
of Huanuco, Manachaqui Cave shares the carinated vessel
profile and concave upper walls that Lathrap considers
diagnostic of Tropical Forest Culture (Lathrap 1970:110112).
Late Tutishcainyo Phase pottery in particular bears a
number of specific attributes in common with Manachaqui such
as rim reinforcement, notching (referred to as "nicking") of
carination angles, and the attachment of similarly incised
flanges on basal angles (Lathrap 1962: Figs. 41i; 42d, e;
43a, c) .
Less common at Tutishcainyo are notched applique
bands (Ibid.: Figs. 44c, e) and notched rims (Ibid.: Figs.
47p, 48h).
Morales (1992:156) considers the notched
applique technique indigenous to Peruvian Amazonia, but it
appears in coastal Ecuador's Valdivia Phase VI {Hill 197274:16) prior to its occurrence in Late Tutishcainyo.
The
more unrestricted Manachaqui Shape B carinated jars bear a
general resernbance to common Early (Lathrap 1962: Fig. 20
and Late Tutishcainyo (Ibid.: Fig. 40) vessel shapes.
Overall, the Tutishcainyo styles' emphasis on open vessel
shapes and complex incised decoration contrasts with
Manachaqui's repertoire of restricted shapes and modelled
343
decorations.
Other early Amazonian assemblages from the base of the
eastern Andes exhibit some formal attributes characteristic
of the Manachaqui style.
Near Tingo Maria in the Upper
Huallaga premontane forest, Lathrap and Roys (1963) have
documented pottery from the Cave of the Owls (660 m)
that
might be considered coeval with and stylistically
intermediate between Late Tutishcainyo and Kotosh
Wairajirca.
Only surface decorations link this assemblage
to highland Wairajirca, while the most common shape, a
vessel with globular or carinated body profile, concave
upper wall or neck, and thickened rim resembles rare Late
Tutishcainyo shapes (Ibid.:15; Lathrap 1970:103).
The Cave
of the Owls profiles illustrated in Lathrap and Roys' Fig.
5a-k can also be likened to those of Manachaqui Phase Shape
B Rims 4 and 10 jars, and Shape D unrestricted bowls with
concave walls.
The Cobichaniqui and Pangotsi Phases in the Upper
Pachitea Basin (elevation 300 m) span the centuries from
1600 to 800 b.c.
(Allen 1968:347, 351), and their
assemblages include carinated bowls and globular neckless
jars and bowls.
A Pangotsi Phase Form 3 semi-carinated
"bowl" with a weak neck and thickened, everted rim (Allen
1968: Fig. 7; Lathrap 1970: Fig. 14e) is comparable to
Manachaqui Shape D bowls, while Form 6 is a neckless jar
(Allen 1968: Fig. 8).
The Pangotsi assemblage features
344
numerous restricted vessel shapes and a tendency to thicken
vessel rims.
Allen describes the style as an Andean-
Amazonian composit (1968:350).
Below the northeastern fringe of the Central Andes,
archaeological sites of lowland Bagua investigated by Shady
(1987b) occupy tropical thorn forest between 600 and 800 m.
Shady (Ibid.:464) considers Morerilla Phase necked jars with
applique decoration contemporary with, and closely akin to,
Pandanche A, Early Huacaloma and Montegrande styles.
Morerilla's similarity to the Manachaqui Phase assemblage is
difficult to access because the former is represented by
only 20 sherds (Ibid.: Fig. 2a-e).
Like Montegrande,
Morerilla is a low-elevation site assemblage displaying
features of highland Cajamarca styles.
Northeast of Bagua and Manachaqui in the Ecuadorian
Oriente, the shapes and proportions (especially the low
positioning of basal angles) of Yasuni Phase vessel Forms 7
and 8 (Evans and Meggers 1968: Fig. 8) closely parallel some
Manachaqui Shape B and Shape D variants.
Evans and Meggers
find the single radiocarbon date of 50±90 b.c. for Yasuni
acceptable.
Lumbreras (1981:11), however, believes that the
Yasuni materials are coeval with the oldest South American
ceramic assemblages.
Yasuni Form 7 and 8 flanged rims
resemble those of Manachaqui Shape B Rim 5 and especially
Shape D Rim 5, while Yasuni Form 6 rim profiles recall those
of Manachaqui Shape B Rim 10.
Yasuni rims and body
345
protrusions also show notching embellishment.
The step
motif is shared between Manachaqui Incised Vessel 2 (Fig.
31e) and Yasuni Incised and Punctate sherds (Ibid.: Figs.
11f, 12b), but most incised-line decorations on Yasuni
vessel walls have no equivalents at Manachaqui.
Although
other Yasuni Phase vessel shapes (all open bowls) have no
Manachaqui counterparts, the correspondences constitute
compelling evidence for a close relationship between Yasuni
and Initial Period Manachaqui.
Some early Amazonian assemblages like Chambira (Morales
1992, 1993:636-641), Jauari (Hilbert 1968) and Taperinha
(Roosevelt et al. 1991; Roosevelt 1995) show infrequent and
scattered resemblances to the Manachaqui Phase assemblage.
The co-occurrence of globular bowls and neckless jars at
these sites (and rim notching at Taperinha) may be
historically significant or fortuitous.
Like Manachaqui's
neckless jars, the Amazonian vessels (including Upper
Pachitea Pangotsi Phase Forms 5 and 6) are shallower and
less restricted than Central Andean neckless ollas, with
more nearly vertical walls.
The carinated vessel form constitutes a more useful
indicator of stylistic relationships.
Vessels with
embellished basal angles are usually viewed as diagnostic of
Amazonian ceramic traditions (Lanning 1967:85-88; Lathrap
1963, 1970, 1971, 1974; Burger 1985b:528).
Vessels with
prominent and highly embellished carination angles are
346
widely distributed across the northern continental lowlands
of Ecuador, Colombia and Venezuela by approximately 1000
b.c., and they persist throughout widely separated Amazonian
occupational sequences from the upper, middle and lower
Amazon (Evans and Meggers 1968: Fig. 79; Hilbert 1968).
Nevertheless, and despite Lathrap's judgement in favor of
Amazonian temporal priority, i t remains uncertain whether or
not Amazonian carinated vessels like those of Early
Tutishcainyo predate those from late Valdivia Phase VI
through VIII contexts of coastal Ecuador.
The earliest
Amazonian pottery from downriver at Taparinha (Roosevelt
1995) apparently lacks carinated profiles, but large spatial
and temporal lacunas separate Taparinha and the Central
Ucayali.
Paste A Relationships:
the Northern Andes
Ceramic remains from sites in extreme northwestern Peru
and southern Ecuador are closely allied with the ValdiviaMachalilla necked jar tradition of the Northern Andes
defined by Lanning (1967:85).
In Piura, necked jars,
carinated and semi-carinated body profiles, rim notching,
and notched applique ribs on carination angles occur in an
"extremely small" Paita Phase A sample, a "poorly
documented" Paita B sample (Lanning 1963:156-157, Pl. 1a,
g), and a supplemented Paita C and D sample (Ravines 198687: Lams.1 and 2).
Radiocarbon dates of 1660 and 1440 b.c.
347
associated with Paita ceramics have been reported by
Richardson (1973:203), but Lanning's early phase subdivisions require further corroboration.
Lanning (1963:
Table 22) cross-dates Paita A and B with terminal Valdivia
phases, Paita C with Machalilla and early Chorrera, and
Paita D with late Chorrera.
In the semi-arid southern Ecuadorian interior between
1200 and 1350 m elevation, some Catamayo Phase B (1300 to
900 b.c.) globular, necked jars from Loja carry beveled rims
(Guffroy et al. 1987: Fig. 12a-c) virtually identical to
Manachaqui Shape BRims 1 through 7.
Some rims are lightly
thickened, but apparently not by the fold-over technique.
Bowls are absent from Guffroy's Phase B sample.
Aside from
similar rim burnishing techniques, the austere Catamayo
Phase B pottery shows few additional resemblances to the
Manachaqui assemblage.
The Valdivia, Machalilla and Cerro Narrio assemblages
from the Ecuadorian coastal lowlands and southern highlands
best exemplify Northern Andean styles that feature necked
jars and carinated vessel shapes.
The Valdivia and
Machalilla styles in combination (Meggers et al. 1965) share
the greatest number of parallels in both shape categories
and decorative techniques with the Manachaqui collection.
Despite initial arguments for a Machalilla site unit
intrusion and partial contemporaneity with Valdivia
(Ibid.:171; Lathrap 1963), recent investigations have
348
concluded that Machalilla developed gradually in situ from
Valdivia antecedents (Lathrap 1971:85; Hill 1972-74:24-26).
Hill's (1972-74:21-24) and Lippi's
(1983:351-355) analyses
of available radiocarbon evidence concur with Lanning's
(1967:85) estimate of 1500 to 800 b.c. for Machalilla's
temporal placement.
Thus, Machalilla is coeval with the
Manachaqui Phase.
Manachaqui vessel shapes with counterparts in both
Valdivia and Machalilla assemblages (Meggers et al. 1965:
Figs. 54 and 89, Tables A and F)
include Shape A Rim 2
globular neckless jars (cf. Machalilla Form 15), Shape B
necked jars (cf. Machalilla Forms 11, 12 and 13; Valdivia
Form 19) and Shape D unrestricted bowls with concave walls
(cf. Machalilla Form 9 and Valdivia Form 10).
Manachaqui
Shape A Rim 1 jars with incipient necks closely resemble
Valdivia Form 18 (Ibid.: Fig. 54 and Table A), and
Manachaqui Shape C Rim 1 restricted bowls parallel Valdivia
Forms 4 and 5.
Also, most of the Machalilla and Valdivia
vessel shapes are restricted like Manachaqui's.
Machalilla necked jars occur in both globular (Meggers
et al. 1965: Fig. 83-5),
two-tiered (Ibid.: Fig. 85-12) and
semi-carinated variants (Ibid.: Fig. 86-6).
However, the
proportions of the semi-carinated Machalilla variants differ
from Manachaqui's which have lower basal angles more
comparable to Amazonian vessel proportions (e.g. Yasuni
Phase Shape 8 in Evans and Meggers 1968: Fig. 8).
In
349
contrast with heavily thickened and folded Manachaqui vessel
rims, Machalilla's rims are unthickened or only lightly
thickened.
A fold-over thickening technique frequently
observed on Valdivia jars (e.g. Meggers et al.: Fig. 35-1)
was abandoned prior to the onset of the Machalilla Phase.
Finally, like Manachaqui neckless jars, the Machalilla
variety more closely resembles shallower and less restricted
Amazonian neckless jars than Central Andean neckless ollas.
Along with remarkably precise vessel shape parallels,
the Manachaqui Phase assemblage shares some decorative
techniques with the Machalilla assemblage.
Recent
investigations at La Ponga have documented rim embellishment
by notching (Lippi 1983: Fig. 87: Nos. 875, 1018), incision
(Ibid.:
buttons.
Fig. 87: No. 1612) and the addition of applique
The co-occurence of rim incision at La Ponga and
Manachaqui is especially remarkable because this particular
technique rarely appears elsewhere.
The wide-mouthed
variant of the necked jar (Machalilla Form 13) exhibits a
notched (or "nicked") carination angle {Meggers et al.: Fig.
78-7) .
Despite such similarities in modeling techniques,
line-
incision and painting become the preferred decorative
techniques by Machalilla times.
The rim notching observed
during early Valdivia phases (Ibid.: Plate 95) may be rare
in Machalilla.
In fact, Manachaqui's most common decorative
techniques are more characteristic of Valdivia.
Applique
350
bands and "fillets" occur frequently in late Valdivia phases
(Ibid.: Plates 27-29).
However, neither Valdivia nor
Machalilla potters embellished vessel profiles by the
Manachaqui technique of affixing applique medial and
shoulder ribs.
In the southern Ecuadorian highlands, pottery from
Cerro Narrio and other sites includes necked jars and
notched applique band decorations (Collier and Murra 1943:
Plate 18, Figs. 3, 5, 7-12; Uhle 1922a: Fig. 12; Bennett
1946).
The full range of vessel shape details is not
clearly conveyed by the available illustrations, but some
Red-on-Buff Type jars feature carination angles decorated
with notched applique ribs (Collier and Murra 1943: Plate
18, Figs. 1, 2, 4 and 6).
Provenience information at Cerro
Narrio suggests their association with the early portion of
the sequence.
Uhle (1922a: Fig. 22 middle) illustrates a notched rim
comparable to Manachaqui Phase Shape B Rim 9, and also notes
the presence of applique serpent motifs (Ibid.: 211, Fig.
58A), and a "non-functional" handle (Bennett 1946: 55 and
Fig. 8L; Uhle 1922a: 210, Fig. 6a top).
Manachaqui, Early
Cerro Narrio and Huancarcuchu vessels bear zoomorphic
adornos (e.g. Collier and Murra 1943: Plate 24, Figs. 1-6;
Bennett 1946:55).
The most obvious difference between the
assemblages is that, as on the
~eighboring
coast, south
highlanders emphasized painted and line-incised decoration.
351
A sober analysis of the data fails to support Braun's (1982)
long Cerro Narrio sequence (Bruhns 1989:57).
Early Cerro
Narrio and the early phase at nearby Pirincay (Bruhns et al.
1990) are probably coeval with the Manachaqui Phase.
To
date, only a few illustrations of early Pirincay pottery
have been published.
The more securely dated early Cotocollao assemblage
from the northern Ecuadorian highlands also shows
considerable similarity to Manachaqui Phase ceramics.
Globular neckless jars (Villalba 1988: Figs. 106, 123) and
necked jars with carinated profiles (Ibid.: Fig. 109; Porras
1982: Lam. 20, Form 14) resemble Manachaqui Shape A and
Shape B vessels respectively.
Short-necked jars illustrated
by Porras (1982: Lam 20, Form 17) show thickened and beveled
rims much like Manachaqui Shape B's.
Porras'
(Ibid.: Lam.
18) vessel Forms 5, 7A and 7B match Manachaqui Shape D.
Shared decorative attributes include notched rims and
carination angles (Villalba 1988: Figs. 102, 107; Porras
1982: Lam. 3, Nos. 7-9)
~~d
notched applique (Villalba 1988:
Fig. 94, 99, 100; Porras 1982: Lam. 1).
Rim and carination
embellishment at Cotocollao was executed by incision
directly into the vessel's surface rather than into applique
fillets.
With the early Ecuadorian highland styles, then,
Manachaqui shares numerous attributes among which zoomorphic
adornos, notched rims and notched carination angles are
salient.
352
In summary, we may conclude that Manachaqui Phase Paste
A ceramics again share most design attributes with northCentral Andean Cajamarca styles.
Together, these
assemblages merge the basic Central Andean ceramic industry
with Northern Andean expanded repertoires of vessel shapes
and decorative techniques.
Perhaps the northern aspect is
not surprising since Formative Period Ecuadorian styles have
been favorite workhorses for diffusionist explanation in
mid-twentieth century New World Prehistory (e.g. Meggers et
al. 1965; Ford 1969; Lathrap et al. 1977; Paulsen 1977;
Porter Weaver 1981:493-501).
The Manachaqui style in
particular combines general shape attributes of Machalilla,
Cotocollao and Yasuni with decorative attributes typical of
Machalilla, Valdivia and Early Cerro Narrio.
affinities are particularly strong.
The Machalilla
Low carinated vessel
profiles typical of Amazonian vessels distinguish the
Manachaqui Phase Paste A assemblage from the Cajamarca
styles.
Rare at Manachaqui are the elaborate incised and
painted decorations of Amazonia and the Formative Period
Northern Andes, although some negative evidence may be a
product of the rockshelter's special function.
Yet despite
its far-ranging and far-flung affinities, the Manachaqui
style is unified, coherent and unique.
Paste B Origins
Manachaqui Phase Paste B pottery is most productively
compared to Amazonian assemblages to the east.
Manachaqui
353
Paste B1 Vessel 1 (Fig. 63a) and Vessel 3 shapes (Fig. 63c)
resemble Pangotsi Phase Form 3 (Allen 1968: Fig. 7; Lathrap
1970: Fig. 14e) and Cobichaniqui Phase Form 1 vessels (Allen
1968: Fig. 5; Lathrap 1970: Fig. 13b) from the Upper
Pachitea Basin at the foot of the Central Andes.
The post-
firing paint on Vessel 3 likewise suggests an eastern
origin.
Both Allen and Lathrap view Upper Pachitea
populations as Amazonian representatives of early Tropical
Forest Culture.
However, because Paste B's non-plastic
inclusions are primarily igneous rocks and minerals much
like Paste A's, it is not unreasonable to suggest that the
Paste B pottery was produced in the eastern Andean
foothills, perhaps in the Premontane Forest Life Zone
bordering the Central and Upper Huallaga valley.
The medial
applique rib on the Paste B3 vessel suggests a close
relationship to the Paste A style and origins relatively
close by.
Most extraordinary is the resemblance between the
Manachaqui Paste B1 Vessel 5 bottle spout (Fig. 64a) and
Lathrap's Late Tutishcainyo spout fragment (1962: Fig. 49e)
and hypothetical double-spout-and-bridge bottle
reconstruction (1970: Fig. 9j).
Equally striking is the
similarity between the Manachaqui Paste B2 vessel shape
(Fig. 64b) and Yasuni Phase Form 8 (Evans and Meggers 1968:
Fig. 8) .
The Paste B2 vessel's "nicked" embellishment of
the carination angle is uncharacteristic of the Manachaqui
354
Paste A style, but typical of Late Tutishcainyo pottery as
well as Ecuadorian Machalilla Phase ceramics.
These two
examples provide the clearest evidence for interaction with
Amazonian societies.
The Manachaqui Phase Paste B1 double-
spout-and-bridge bottle provides support to postulate Paste
B origins in the premontane foothills and lowlands of the
Amazon Basin east, or perhaps northeast, of Manachaqui Cave.
CHAPTER 7
THE SUITACOCHA PHASE
The Suitacocha Phase coincides with the first centuries
of Rowe's Early Horizon (800 B.C. - A.D. 1).
Analysis of
Manachaqui Cave's ceramic macrochronology, primarily the
units in Sector B, provided a basis for the assignment of
1,831 diagnostic sherds, as well as associated lithic,
ethnobotanical and faunal remains, to the Suitacocha Phase.
Examples of Suitacocha Phase-bearing levels from the berm
area include Unit 5, Levels 22 through 25; Unit 22, Levels 8
through 12; Unit 28, Levels 9 through 13; Unit 30, Levels 10
through 13; and Unit 32, Levels 9 through 12.
The
Suitacocha Phase deposit in Sector A appears thin in
comparison to the Manachaqui Phase deposit below, and
especially the Empedrada Phase deposit above.
As discussed
in Chapter 5, this may be attributed to alterations of the
shelter interior space during later occupations.
Suitacocha Phase Ceramics
As noted in previous chapters, the Suitacocha Phase
artifact sample is smaller and more weathered than its
underlying Manachaqui Phase counterpart.
Analyses of the
stratigraphy, macrochronology and associated radiocarbon
355
356
dates suggest that cultural deposition was continuous during
a seemingly rapid stylistic transition from the Manachaqui
Phase to the Suitacocha Phase.
Suitacocha Phase vessels,
like their Manachaqui Phase predecessors are predominantly
restricted.
Vessel shape categories A through E remain
basically the same, except that Shape D (unrestricted bowls
with concave walls} drops out.
The popular Manachaqui Phase
carinated and semi-carinated jar shapes are replaced by
globular shapes, and only a small percentage of Suitacocha
Phase bowls have carinated profiles.
Again, virtually all
of the vessels had rounded bases.
Shape A neckless jars dwindle in importance during the
Suitacocha Phase, while Shape B short-necked jars continue
to dominate the vessel shape inventory.
Like their
Manachaqui Phase antecedents, the Shape B jar rims were
often heavily thickened or reinforced by folding the ends
out and down.
A groove left by the process of smoothing the
folded end to the exterior surface is a diagnostic attribute
during the Suitacocha Phase.
Shape
c
restricted bowls and
Shape E open bowls likewise exhibit reinforced rims.
especially open bowls, are still rare.
Bowls,
Shape F is a new jar
shape distinguished from jar Shapes A and B by long, direct
and usually vertical rim forms.
The Shape F rims might have
been lumped into the Shape B category, but the lack of
morphological overlap between the two categories suggests
functional differentiation.
The Shape X category was
357
conceived to contain rim forms that cannot yet be
confidently assigned to any of the established shape
categories.
Suitacocha Phase decorations include a combination of
Manachaqui Phase holdovers and new techniques.
Present, but
diminished in importance, is notched applique and appendages
like flanges, buttons and adornos.
Design emphasis shifts
from applique modelling to geometric renderings utilizing
incision, punctation and a few texturing techniques either
individually or together.
During the Suitacocha Phase, potters frequently
embellished vessel rims by burnishing, or painting and
burnishing them prior to firing, or by adding incised
decorations.
The slips or paints vary in color from maroon
to blood red to pale iridescent red and pink.
Red slip is
also used to create contrasts with unpainted fields filled
with incised decoration.
On Shape B and F jars, the rim and
bottom portion of the vessel body was frequently painted and
burnished.
The unpainted shoulder then served as a field
for incised decorations.
Weathering of the sherd surfaces
becomes a serious impediment to the identification and
analysis of decorative techniques, and especially to the
quantification of painted sherds.
Ceramic Paste Group A
While there is clearly a single ware that predominates
during the Suitacocha Phase, other similar wares
358
insufficiently distinctive to be consistently differentiated
with the unaided eye are also present.
Thus, Paste Group A
consists of more than one brown ware less red in hue, and
darker in color value than Manachaqui Phase Paste A.
The
Paste Group A sherds utilized to characterize the Suitacocha
Phase total 1,792, representing 97.9 percent of the whole
phase diagnostic sherd collection of 1,831 sherds.
That
Suitacocha Paste A wares are not easily distinguished from
the predominant Manachaqui Phase ware reflects the minimal
change in basic ceramic technology.
In general, the Suitacocha Paste Group A contains
greater quantities of vein quartz, but lesser quantities of
quartz crystal, than Manachaqui Paste A.
Microscopic and
microprobe analyses confirm that grain sizes and size
variability differ little from those of the Manachaqui Paste
A.
The microprobe identified only quartz and volcanic glass
among the angular grains of complex igneous rocks.
Surface
finishing techniques utilized also differed little from
those employed earlier.
One sherd shows cord impressions on
its interior surface (Fig. 67c).
The impressions were
likely produced by the technique termed "netting" observed
on Urabarriu Phase sherds from Chavin de Huantar and
illustrated by Burger (1984b: Figs. 143, 144).
Apparently
comprised of twined fibers, the 3-4 rnrn thick cords probably
wrapped cloth pads that were pressed against vessel
interiors during the burnishing process.
Vessel walls are
359
slightly thicker and more variable in thickness than those
of the preceding phase, ranging from 3 to 6 mrn.
rims tend to be thinner and less massive.
However,
A slip with red
pigment was often applied to vessels, or only to their rims
before burnishing.
Well-controlled firing left nearly all
of the pottery completely oxidized.
Morphology
A total of 979 rim sherds, 19 sherds from vessel basal
portions and nine miscellaneous body parts supplied
information regarding Suitacocha Phase vessel shapes.
The
Suitacocha Phase sample is just as fragmented, and much more
eroded than the Manachaqui Phase sample.
Yet despite a
smaller sample, greater consistency in technological
treatment of more standardized vessel shapes permits clearer
analytical distinctions between so-called jar and bowl
shapes.
In other words, the vessels classified as bowls
consistently exhibit the smoothed interiors that jars
invariably lack.
This shift suggests that vessel shape
classes were now more functionally specialized.
Thus, the
shape characterizations offered in the following pages
contain fewer caveats.
The most troublesome feature of the
collection is the similarity shared between some jar and
bowl rim forms that necessitated the creation of a "sink
trap" category, Shape X, described below.
Rim forms again constitute the focus of the following
classification.
The same three-tier scheme developed for
360
Manachaqui Phase rims is extended to accommodate Suitacocha
Phase rim forms and interpreted body shapes (Appendix E) .
Suitacocha Phase rims can almost always be distinguished
from Manachaqui Phase rims based solely on morphology.
Differences in paste qualities, the presence or absence of
red paint, and the degree of erosion provide additional
evidence to confirm or question initial determinations.
Shape A: Neckless Jars
A total of 30 sherds grouped into four rim forms belong
to neckless jars (Table 11}.
Together, these comprise only
three percent of the Suitacocha Phase rim collection.
The
corresponding vessels were apparently small in size, and all
but one example exhibits either exterior thickening, or an
incised groove below the lip exterior.
Rim 14 (Fig. 68a-c).
Rim 14 (14 examples) thickens
gradually to reach maximum thickness (between 6 and 8 rnrn) at
a flattened and squared lip.
All but one slightly upturned
rim (Fig. 68c} are direct or lightly incurving.
An incised
groove encircles the rim exterior one centimeter below the
lip.
Vessel walls approximate 4 mrn in thickness.
These
rims exhibit red paint on the lip and exterior surface.
Rim 15 (Fig. 68d-f).
Rim 15 (8 examples) belongs to
incurving rims that have been thickened on their exterior
surfaces to a maximum of 6 to 9 mrn, and grooved below the
361
thickened portion.
Their lips are rounded, and vessel walls
range from 4 to 6 mrn thick.
R~
16 (Fig. 68g, h).
The seven examples of Rim 16 are
similar to Rim 15, but they lack the groove below the
thickened exterior.
Vessel walls show a thickness of 5 to 6
mrn.
R~·
17 (Fig. 68i).
This single sherd has a thickened
(7 rnrn) interior surface.
point.
The lip exterior ends in a dull
It belongs to the largest of the neckless jars, and
its walls are 4 mm thick.
Shape B: Short-necked Jars
The 612 short-necked jar rims constitute 62.5 percent
of the Suitacocha Phase rim collection (Table 12).
Like
their Manachaqui Phase antecedents, these jar rims are
typically thickened on their exterior surfaces, and lips are
often beveled or flattened.
They range from 7 to 18 ern in
diameter, averaging between 11.5 and 12 em.
Rim
orientations vary from sharply everted with an angular neck,
to nearly vertical with a negligible neck.
Lip exteriors
were usually reinforced by folding the end of the rims over.
Reinforced rims may feature a thin (1 rnm) groove directly
beneath the thickened portion.
Also, some Suitacocha Phase
Shape B rims show a light outward inflection where the neck
meets the shoulder 1 to 1.5 em below the everted rim.
362
Rims lla-g (Figs. 68j-s, 69, 70).
The 585 examples of
Rim 11 are divided into seven variants.
Each of the
variants represents only a morphological tendency.
they constitute a continuous series.
Together
The reader will
observe that the average rim diameter for each variant is
nearly identical, at least among the larger samples.
Corresponding vessel walls average 4 or 5 mm in thickness.
Rim 11a (Figs. 68j-s, 69a-c) exhibits a thickened exterior
(maximum 7 mm) and a flattened lip.
Grooves below the
reinforced lip exteriors lend these rims a "squared"
profile.
Three examples (e.g. Fig. 68s) show thickened
protrusions, or lip flanges, to securely grasp the vessel by
the rim.
Rim 11b (Fig. 69d-k) examples are everted, with rounded
lips and exteriors thickened to a maximum of 6 mm.
Rim 11c
(Fig. 691-r) differs little from 11b, but the thickened
exterior appears more bulbous reaching a maximum of 7mm, and
the lip ends at a dull point.
Rim 11d (Fig. 70a-h) is
vertically oriented, but otherwise it resembles Rim 11a
because of its flattened lip and the groove frequently found
below its thickened exterior (maximum thickness 4 to 7 mm) .
The sherd illustrated in Fig. 70h bears iridescent pink
paint.
Rim 11e (Fig. 70i-l) is only slightly everted and
nearly vertical and direct.
The lip has been flattened and
slopes toward the vessel exterior.
6 to 7 mm.
Its maximum thickness is
One sherd (Fig. 70i) exhibits pink iridescent
363
paint.
Rim 11f (Fig. 70m-s) reaches a maximum thickness near 7
mm at the lip which has a dull point at its interior edge.
The rim's thickened exterior surface forms a right angle
with the neck, which also tends to be inflected again where
it merges with the vessel shoulder (e.g. Fig. 71g, s).
Rim
11g is essentially the same as Rim 11e, except that its
flattened lip slopes toward the vessel interior.
Rims 11e
and 11g might show the same shoulder inflection as Rim 11f,
but the fragments are too small to verify the possibility.
One Rim 11g rim (Fig. 70v) shows traces of pale red
iridescent paint.
Rim 12 (Fig. 71a-f).
These 16 examples are unthickened
everted rims with rounded or semi-rounded lips.
Curvature
at the neck is smooth and nearly uninterrupted.
The rim,
neck and vessel walls are only around 3 rnm thick.
In some
instances, these rims might be mistaken for Manachaqui Phase
Shape B Rim 3b, and their pastes best serve to distinguish
them.
Rim 13 (Fig. 71g-l).
The 11 sherds representing Rim 13
are lightly thickened and slightly everted.
The maximum
thickness at the rounded lips reaches 7 or 8 rnm.
Shape C: Restricted Bowls
The 54 rim sherds from restricted bowls represent 5.5
364
percent of the Suitacocha Phase rim collection (Table 13).
There are four variants of which all but one have thickened
lip exteriors.
Rim diameters range from 9 to 20 em.
Despite weathering, some sherds show traces of burnishing on
exterior surfaces.
R~
7 (Figs. 71m-72a).
There are 37 examples of Rim 7.
The top of the rim is thickened (7 or 8 mm maximum) and
lightly grooved below the flattened, reinforced lip to
render a "squared" profile reminiscent of Shape B Rim lla.
Maximum thickness reaches between 7 and 9 mm at the lip, but
vessel walls are thin, between 2 and 3 mm.
The rim
illustrated in Fig. 71u exhibits a pale red iridescent
paint, while the bowl depicted in Fig. 72a shows only traces
of red paint on the exterior surface.
R~
8 (Fig. 72b-d).
Nine Rim 8 sherds pertain to three
bowls that may be considered morphologically related to the
Rim 7 bowls.
Rim thickening is less pronounced, and all are
decorated with a band of zoned punctation.
approximately 3.5 mm thick.
Vessel walls are
Two of the three are entirely
painted red (Fig. 72b and c), and one shows a lug handle on
the rim (Fig. 72b) analogous to the lugs observed on Shape B
Rim lla jar rims (see Fig. 68s).
The bowl's rim lug has
been flattened like the rest of the rim.
Twenty-four sherds
belonging to this particular bowl were recovered from
Feature R-3 and surrounding unit levels.
The lip of the Rim
365
8 example viewed in Fig. 72c is rounded rather than
flattened.
R~
9 {Fig. 73a-c}.
Seven rim sherds from round bowls
with rounded, lightly thickened lips represent Rim 9.
Maximum thickness at the lip varies from 7 to 9 mrn.
Vessel
walls are 5 rnrn thick.
R~
10.
10 {Fig. 73d}.
A single rimsherd constitutes Rim
The rim is gradually, and only slightly thickened from
2 to 3 mrn.
The lip has been flattened and a tiny clay
"cornice" protrudes from the lip's interior edge.
Shape E: Unrestricted Convex Bowls
Only 15 sherds belong to the Shape E category, which
constitutes 1.5 percent of the Suitacocha Phase rim
collection {Table 14}.
There are four rim forms,
which are represented by a single sherd.
two of
Like other
Suitacocha Phase vessel rims, Shape E bowl rims typically
show some sort of thickening treatment.
R~
members.
8 {Fig. 73e-h}.
The Rim 8 category has seven
The rims vary in overall thickness, but all share
a thickened and rounded lip that protrudes on the vessel
exterior.
Maximum thickness at the lip ranges from 3 to 6
mm.
R~
9 {Fig. 73i}.
Six examples of Rim 9 are assigned
366
to the Suitacocha Phase.
As with the Manachaqui Phase
examples, they may have intruded from later deposits above.
Rim 9 represents a simple, convex bowl with a rounded lip.
Rim 10 (Fig. 73j).
The single example of Rim 10
belongs to a large bowl with a thickened rim that has been
rendered by folding 2 em of the rim end down over the vessel
exterior.
The maximum thickness attained is 10 rnm.
Rim 11 (Fig. 73k, 1).
Rim 11 is one rim sherd that
shows gradual light thickening and a rounded lip.
thickness reaches 5 mm.
Maximum
A sherd with matching paste, color
and decorative technique probably corresponds to the base of
the same vessel.
Rim 11 is unique in representing the only
Paste Group A vessel with a flat base.
Shape F: Jars with Vertical Rims
A total of 272 sherds represent nearly 28 percent of
the Suitacocha Phase rim collection (Table 15).
If combined
with the 612 examples of Suitacocha Phase Shape B shortnecked jars, the total number of necked jars increases to
884, or 91 percent of the phase collection.
The Shape F
rims range from 4 mm to 6 em in height, averaging
approximately 2.5 em.
The majority are between 2 and 3 em
high and oriented vertically, although inverted and everted
varieties exist.
A diagnostic feature of Shape F rims is an
angular junction where the base of the rim/neck meets the
367
vessel shoulder.
angle.
Rim fragments nearly always broke at this
Rim diameters range from 5 to 21 em, and average
between 11 and 14 em.
Vessel bodies were globular, often
spherical, and walls varied from 3 to 4 mm in thickness.
Virtually all of the Shape F rims were burnished or slipped
red and burnished.
Rim la, b
(Figs. 74, 75a-v).
Rim 1 sherds total 138.
They may be gradually thickened toward the rim end, or
unthickened.
Lips of variant Rim 1a (Fig. 75a-n) are
rounded, or end in a dull point.
slightly everted.
They may be vertical or
Rim 1b (Fig. 75o-v) may exhibit an
inverted, vertical or everted orientation.
The lips are
squared, or "semi-squared" with rounded edges.
Together,
the Rim 1 variants range from 1.5 to 3.5 em in height, and 4
to 6 mm in maximum thickness.
Some Rim 1 sherds feature a
narrow, incised groove at the union with the vessel body.
Rim 2 (Fig. 76a-j).
Rim 2 is represented by 79 sherds.
The lip exterior has been thickened at the top to a maximum
of 7 mm.
The rims range from 2 to 4 em high and are often
slightly inverted.
incision.
Occasionally these are decorated by
The sherd illustrated in Fig. 74b bears
iridescent paint.
The incised sherds illustrated in Fig.
76j and k may belong to the same vessel.
Rim 3 (Figs. 761-n, 77a-f, 78a).
There are 26 examples
368
of Rim 3.
These resemble Rim 2, except that the tops of the
rims are bent outward and their undersides are thickened.
The lips may be rounded or squared.
The average height of
Rim 3 cannot be determined because of breakage.
Most of the
Rim 3 sherds show embellishment by painting and/or incised
decoration, and the average rim diameter is greater than
those of other Shape F rim variants.
Rims 4a, 4b (Fig. 78b-f).
divided into two variants.
Ten examples of Rim 4 are
These rims are invariably short
and vertically oriented, ranging from 4 to 7 mm high.
Rim
4a (Fig. 78b-d) has been thickened to 6 mm and has been
grooved at the rim's base.
The lip has been squared.
example illustrated in Fig. 78b bears iridescent paint.
The
Rim
4b's (Fig. 78e, f) neck is negligible, and its lip has been
slightly thickened before rounding.
Rim 5 (Fig. 78g, h).
Six examples of Rim 5 have been
thickened by folding 1 to 2 em of the rim down over the rim
exterior.
The vertically-oriented rim ranges from 2 to 3 em
high, and the lip has been rounded.
The folded over portion
may extend over halfway down the height of the rim, and has
not been smoothed to the rim exterior wall.
Rim 6 (Fig. 78i,
j).
The six sherds representing Rim 6
are bent at an angle approximately half way below the lip.
They are 2 em high, unthickened and the lips have been
369
squared.
Rim 7
(Fig. 78k).
comprise Rim 7.
Four rim sherds from the same vessel
It is the largest of the Shape F rims, and
represents the largest Suitacocha Phase Shape F vessel.
It
is 6 em high, slightly convex and thickened to 1 em at its
rounded lip.
The mouth diameter is 21 em.
Rim 8 (Fig. 79a-c).
The three sherds representing Rim
8 are all vertically-oriented and slightly everted.
ends have been thickened to 8 mm and squared.
The
Rims
illustrated in Fig. 79b and c have shallow grooves beneath
the thickened portions of the lip exteriors.
All three are
painted red.
Shape X: Undetermined Vessel Shapes
The rim forms grouped under the Shape X category cannot
be confidently assigned to any of the previously described
vessel shapes, although informed guesses may be offered in
some cases.
They are a diverse lot, and are described below
in order of decreasing abundance.
Rim 1 (Fig. 79d-j).
Rim 1 belongs either to a
restricted bowl similar to Shape C Rim 7, or to jars with
insloping rims like Shape B Rims 11g or 1le, or Shape F Rim
2.
The Rim 1 problem typifies the dilemma that frustrates
attempts to determine the vessel shapes for all of the rims
grouped under Shape X.
Shape X rims often feature carefully
370
burnished and painted exterior and interior surfaces.
Lacking evidence for a neck base or shoulder beneath the
rim, it is virtually impossible to confidently group the
rims in question with either bowls or jars.
The 25 examples
of Rim 1 are all thickened on the lip exterior to a maximum
of 9 mm.
The lip interior edge ends in a dull point.
Thirteen rim sherds measured range from 9 to 14 em in
diameter, averaging 11.54 em (S.D.=1.28).
is 12 em.
The mode diameter
These dimensions are more typical of the smaller
Shape B and F orifices than those of Shape C.
R~
2
(Fig. 79k-n).
The eight rims are similar to Rim
1, but the thickened portion of the lip is smaller,
attaining a maximum thickness of 7 mm.
range from 11 to 20 em in diameter.
Six sherds measured
Three of them, probably
from bowls, are decorated by incision.
R~
3 (Fig. 79o-r).
These seven rim sherds all thicken
gradually to a maximum 9 mm at the flattened lip which
slopes down toward the interior edge.
Five sherds measured
range from 11 to 16 em in diameter.
R~
4 (Fig. 80a).
These three rim sherds belong to one
large Shape C or Shape F vessel.
to 8 mm below a rounded lip.
R~
5 (Fig. 80b).
The rim gradually thickens
The diameter is 20 em.
Two Rim 5 sherds also belong to
either Shape C or Shape F.
They are insloping and thickened
371
to 1 ern at the flattened lip.
Rim 6 (Fig. 80c).
The rim diameter is 12 ern.
The one Rim 6 sherd probably
pertains to a Shape E vessel.
It is vertical and gradually
thickened to a maximum of 8 mm below a flattened lip which
slopes slightly toward the vessel exterior.
Its diameter is
19 ern.
Basal Sherds
Of the 19 sherds from the basa: portions of vessels,
three sherds might belong to a recessed jar or bowl base
with raised painted and polished walls (Fig. 80d).
vessel shape is not well understood.
The
The remaining 16
sherds pertain to at least four carinated bowls with
protruding basal angles (Fig. 80e-h).
Two of the bowls
exhibit red slip on both interior and exterior surfaces.
Diameters measured from the interiors of Figs. 80e and 80g
are 17 em and 22 em respectively.
The basal portions
protrude 3 to 7 mm from the upper walls.
The angles of
orientation illustrated represent best estimations.
Miscellaneous Shapes
Nine sherds (Fig. 8la-i) pertain to ceramic vessels or
objects of unknown shape.
Three of these (Fig. 8la-c) are
interpreted as pieces of "mammiform" vessel legs.
Five
(Fig. 8ld-h) seem to belong to handles, and a single sherd
with incised decoration and tubular cross-section (Fig. 8li)
372
remains unidentified.
Decoration
A total of 785 sherds serve to evaluate Suitacocha
Phase decorative techniques.
The principal techniques may
be grouped under the headings: painting, applique and
incision.
these.
All of the shape categories show at least one of
Apparently the Shape A neckless jars lack decorative
treatment other than red paint and incised grooves on rim
exteriors.
The Shape F jars, and probably the Shape B jars,
show all three of the major techniques.
The Shape F Rim 1
jar illustrated in Fig. 74 represents a conventional
Suitacocha Phase design layout in which the unpainted upperthird of the vessel body provides a field for applique and
incised decorations, while the rim and lower two thirds are
red slipped and burnished.
Incised lines and applique ribs
or flanges may separate the decorated jar shoulder from the
painted lower hemisphere.
Incised lines also delineate
zones filled with red paint or punctation.
punctation occur individually on rims.
Incision and
Additional
observations regarding jar decorations will be offered
below.
The small sample of Suitacocha Phase bowls also
exhibits
frequent decoration.
Shape C restricted bowls may
be painted and incised while, of the Shape E open bowls,
only the incised Rim 10 is decorated.
Examination of body
sherds confirms that applique occurs on bowls, but precise
373
shape and rim associations remain uncertain.
As with the
Manachaqui Phase collection, severe fragmentation limits the
capacity to deduce tendencies governing associations between
decorations
~~d
vessel shapes; however,
the overwhelming
majority of decorated sherds pertain to jars (89%).
It
should also be borne in mind that this particular Paste
Group A sample may represent more than one style or set of
stylistic norms.
Conversely, it probably fails to include
techniques, motifs and combinations that might be observed
within an expanded Suitacocha Phase sample.
Painting
Pre-firing red paint or slip, and traces of these, were
observed on 461 sherds assigned to the Suitacocha Phase.
Painted body sherds total 225, while 236 rim sherds (24% of
the phase total) exhibit remains of red paint.
The
proportion of unpainted to painted sherds is, of course,
distorted by the severe weathering of the Sector B sample.
For example, none of the rim sherds from Units 7, 8 and 39
at the base of the berm show paint.
Within the sheltered
Sector A, proportions average close to 2 to 1 which suggests
a similar proportion for the entire Suitacocha Phase rim
collection, and a substantially higher total for painted
body sherds as well.
Of the 225 painted body sherds, 106
display red paint with no other techniques except polishing,
and 16 of these carry red paint on interior surfaces.
The
remaining 90 sherds show red painted exterior surfaces, and
374
the majority belong to polished and painted jar bases like
that of the reconstructed Shape F Rim 1 example (Fig. 74).
Red paint is associated with applique and incision on 119
sherds.
On many Suitacocha Phase vessels polished red paint
serves to contrast with, and therefore highlight, unpainted
decorative panels or fields filled with incised and punctate
designs.
The red paint utilized for vessel decoration varies
from dark red (10R3/6) to a pale red or pink (10R5/3).
It
is actually a slip composed of water, clay and iron oxide
which achieves a metallic luster and iridescent effect when
the slip is especially dilute (Lathrap et al. 1975:53-55;
Sonin 1977).
The iridescence is best observed by splashing
water onto a painted surface.
Burnishing tracks left on
rims and jar bases particularly stand out with the addition
of water.
The slip's use on vessel rims and jar bottom
exteriors strongly suggests that the iridescent effect was
incidental rather than intentionally produced.
Faint
remains of white paint appear on the exterior surfaces of
two sherds from Sector A's Floor Y, and one sherd to be
described bears two thin black-painted lines.
Applique
A total of 171 sherds with applique decorations,
including three flanges, can be assigned to the Suitacocha
Phase.
Applique ribs and bands appear independently, as
well as in combination with the other techniques of painting
375
and incision.
Applique may be used to frame fields with
painted or incised decoration.
Techniques of applique
embellishment include notching and punctation.
Notched applique.
A total of 94 sherds bear applique
ribs and bands embellished by notching (Figs. 82a-g}.
These
can be divided into sherds with 1} single, isolated notched
ribs or bands, 2} parallel notched ribs and 3} high relief
notched ribs and bands.
The Suitacocha Phase notching
technique involves cutting or incising with a sharp pointed
tool, thereby differing from the Manachaqui Phase technique.
Only the high relief notched ribs and bands are still
notched by impression, probably with sprigs of cana.
Sherds
with single notched ribs or bands (Figs. 82a-d} from sound
Suitacocha Phase contexts total 52.
Notched applique
decorations also appear in later chronological components.
An additional 15 sherds with notched applique were recovered
from mixed Suitacocha Phase and Empedrada Phase levels.
Suitacocha Phase sherds isolated and parallel notched
ribs and bands are invariably straight.
Bands are
vertically oriented while framing design panels on jars
(e.g. Fig. 74}, while ribs run horizontally around vessel
mid-sections.
12 mm.
These applique strips vary in width from 6 to
On 20 sherds, two and three notched ribs run
horizontally and parallel to one another (Fig. 82e}.
pertain to one or more bowls with red-slipped interior
All 20
376
surfaces.
These parallel ribs are 5 to 7 mm wide, and
separated from each other by 2 to 4 rnrn.
The 22 examples of "high relief notched ribs and bands"
may be straight or curving, and all belong to jars (Figs.
82f and g).
rnrn high.
These strips are typically 4 mm wide and 3 to 5
Notches are spaced 2 to 5 rnrn apart.
High relief
ribs may separate the unpainted upper from the painted lower
hemispheres of jars (Fig. 82f), or bands may undulate
sinuously around vessel shoulders (Fig. 82g).
Additional
examples of applique demonstrating associations with incised
decorations are illustrated later within this chapter.
Punctate applique.
A total of 71 sherds bear applique
ribs and bands embellished with either ovoid or round
punctation.
The ovoid punctation is rare, appearing on only
4 sherds from jar shoulders or mid-sections (Fig. 82h and
i), and one sherd from a vessel leg (Fig. 81a).
The 66
sherds bearing bands with round punctation are 3 to 4 rnrn
wide, and the punctations are usually 2 mm apart (Figs. 82j
and 83a).
3 neck.
In Fig. 78a, an arching band adorns a Shape F Rim
Punctate applique buttons appear only on one sherd
(Fig. 82i).
The motif depicted in Fig. 82j deviates from
the majority of these bands which usually run sinuously
around jar shoulders (e.g. Fig. 83a).
They are frequently
associated with incised designs and red paint as later
illustrations show.
377
Unembellished applique.
There are only three eroded
examples of applique bands with no embellishment (Fig. 83b
and c).
All three pertain to jars.
Flanges.
Three sherds represent flanges (Fig. 83d-f).
A short, curving,
"shelf-like" flange has three small
incised nicks on one side (Fig. 83d).
A flange fragment
from a jar mid-section protrudes with a triangular crosssection (Fig. 83e).
It is grooved at both top and bottom
junctures with the vessel wall.
The upper surface shows
short incisions, while the bottom is covered with red paint.
The third flange is an incised tab projecting out from the
upper shoulder of a jar, and sloping downward (Fig. 83f).
Incision and Surface Texturing
A total of 525 sherds bearing incision and/or surface
texturing are assigned to the Suitacocha Phase.
The
decorative techniques grouped under this category include:
cross-hatching, stamped circles, punctation, combing,
rouletting, wet incision or inciso cortante, engraving,
rocker-stamping and fabric-impression.
Design motifs
comprised of simple rectilinear patterns appear on jar rims,
necks and shoulders, or on bowl exteriors.
As previously
noted, zones delineated by incision and/or applique may be
filled with red paint, punctation, stamped circles, or
incised line motifs.
Patterns outlined with incised lines
include triangles, arches and inverted rectilinear U-shapes.
378
A total of 174 sherds show straight or curved incised
lines appearing alone, in sets or as parts of simple line
motifs.
Lines are narrow, and were created by cutting into
leather-hard clay with a sharp-tipped implement.
never more than 1 mm wide or 1 mm deep.
They are
Incised lines are
found on Shape B jar necks (e.g. Figs. 69c, 83g), Shape F
jar rims (Figs. 76i-l, 77a, e), Shape C bowl exteriors
(Figs. 72a, 73d, 83e) and jar shoulders (Fig. 83h, i).
Divergent arrays of parallel lines commonly occur on both
bowls
(Fig. 72a) and jars (Figs. 77e, 83j), while cross-
hatching ornaments at least one jar represented by seven
sherds (Fig. 83k and 1).
Stamped circles.
Stamped circles, concentric circles
and circle-dots were observed on 76 fragments of jar rims
(Figs. 69a, 77b-d), necks (Fig. 84a) and shoulders (Figs.
84b-l, 85a-f).
They appear in association with red, and in
one case black, paint (Fig. 84b), with divergent arrays of
parallel lines (Fig. 84d-f), within incised zones (Fig.
84i), with zoned punctation (Fig. 84j, k), with applique
ribs with round punctation (Fig. 85a-c) and high-relief
notched ribs (Fig. 85d, e).
Two sherds offer fragmentary
glimpses of designs featuring stamped semi-circles
intersected by incised lines (Fig. 841), and zoned by
unusual applique ribs embellished with crowded rows of small
punctations (Fig. 85f).
379
Punctation.
A total of 226 sherds show punctation.
In
the following paragraphs, the punctation shapes
differentiated include: round, ovoid, oblique, dash,
triangular and crescent.
Punctation may be zoned by
incision or applique, but also occurs unzoned.
Unzoned rows
of punctations may run independently or parallel to incised
lines.
Infrequently they sketch linear patterns.
Thirty-
seven of these sherds belong to Shape C Rim 2 restricted
bowls, while 188 pertain to Shape B and F jars, and one is
assigned to Shape X.
The thirty-seven Shape C sherds with zoned punctation
derive from three bowls.
The first features a single row of
oblique incisions zoned by incised lines comprising a band
one em wide under the rim exterior (Fig. 72b).
A second
bowl has a similar band filled with two offset rows of ovoid
punctations (Fig. 72c).
On both bowls, the exterior
surfaces above and below the bands have been painted red and
polished.
The third bowl features a band with irregular
rows of crescent-shaped punctations and, except for the
band, has been polished inside and out.
Shape B and Shape F jars feature punctation on necks
and shoulders, and some sherds are sufficiently large to
reconstruct incised and punctate motifs that likely repeat
around the vessel circumference.
For example, Fig. 85g
illustrates a Shape F jar with a design field on the
shoulder zoned on the top by red paint and on the bottom by
380
incised lines and red paint.
Parallel incised lines outline
triangles separated by fields with minute dash punctations.
The punctations form diagonal rows running parallel to the
incised lines.
A Shape B jar (Fig. 85h) neck features
oblique punctations within pendant triangular fields zoned
by an incised line above, and incised lines and red paint
below.
Another sherd from a jar shoulder (Fig. 86a) depicts
four parallel arching incised lines bordered above by a
field with large triangular punctations and below by
incision and polished red paint.
A fourth jar shoulder
fragment (Fig. 86b) shows oblique punctations within a zone
(perhaps triangular)
formed by a single curving incised
line.
The remaining illustrations of sherds with punctation
decorations portray the variety of designs and design
elements rendered on jar necks and shoulders.
Many of these
incorporate rows of punctations zoned by parallel incised
lines.
Fig. 86c exhibits a row of oblique punctations zoned
by parallel incised lines which in turn delimits the bottom
of a larger field filled with oblique punctations.
Single
rows of ovoid and widely spaced punctations may occupy large
zones created by incision (Fig. 86d).
Incised lines may be
marked (unintentionally?) by rows of perpendicular dash
punctations (Fig. 86e).
Three sherds (Fig. 86f-h) show
small, oblique punctations in arching rows zoned by single
or parallel incised lines.
A similar design element
381
incorporates parallel diagonal and horizontal rows of
punctations (Figs. 86i-l, 87a, b).
An unusual sherd (Fig.
87c) bears punctations that may trace an undulating pattern
around the vessel circumference.
A miscellany of sherds demonstrate other decorative
uses of punctation.
Punctations surrounding jar necks
frequently appear independent of incision (Figs. 69b, k;
77f; 78f; 79n; 87d-h).
Fig. 87h shows rare round
punctations, while another single sherd demonstrates that
punctation can be zoned or bordered by an applique band
(Fig. 87i).
Finally, the most complex motifs are rendered
by combinations of incision, zoned punctation and notched
applique bands on panels framed by red paint and repeating
around jar shoulders (Figs. 87j; 88a, b).
Combing and Brushing.
Sets of very closely spaced
incised lines were created either by combing with an
implement with rigid bristles, or by brushing with an
implement with flexible bristles.
The techniques, which are
not easily distinguished from one another, were observed on
14 sherds.
One rim assigned to Shape X (Fig. 80a), and
perhaps belonging to a large bowl, shows vertical sets of
six very fine incised lines that on some sherds can only be
observed under certain lighting conditions.
A coarse
combing technique was applied to a vessel neck depicted in
Fig. 78c., and to a jar mid-section shown in Fig. 88d).
382
Fig. 88e illustrates extremely fine lines created by
brushing, but utilized to render design elements typically
executed with conventional incised lines.
Four thin, eroded sherds from a jar show
Rouletting.
rouletting beneath parallel, curving incised lines (Fig.
88f) .
The lines are wide and shallow, and were produced
with a blunt-tipped tool.
polished.
The gray exterior surface is
Although classed within Paste Group A,
this
pottery may have originated in more distant regions.
Incised boss.
A boss from a jar has a conical shape
and exhibits the end portions of two parallel incised lines
(Fig. 88g).
Traces of red slip and polish remain visible.
Fine-line Scratched.
A single sherd (Fig. 88h) shows
an incised line and light scratches made by apparently
haphazard strokes with a fine-pointed implement.
Squiggles.
Eight sherds exhibit sets of shallow, wavy
lines and could all belong to the Shape X Rim 2 vessel
illustrated, probably a restricted bowl (Fig. 79k).
Some of
them (e.g. the sherd at left) may have been rendered by a
combing technique.
Inciso cortante.
Seven tan sherds show inciso
cortante, a technique of incising deep lines into wet clay
with a sharp pointed implement (Fig. 73j, k).
The technique
383
leaves wider (1.5 mm}, deeper lines than is usual for
Suitacocha Phase pottery.
Small ridges on both sides of
each incision are left by the implement as it pushes aside
the wet clay.
All of these sherds probably pertain to the
same flat-bottomed open bowl designated Shape E Rim 10.
Engraving.
Five sherds exhibit engraving, or incising
the vessel after the clay is dry and hard.
leaves grooves with rough, chipped edges.
The technique
All of the sherds
seem to come from the same Shape C Rim 7 vessel (Fig. 71t}
which also shows red paint below the incisions.
Rocker-stamping.
Rocker-stamping can barely be
discerned on three very eroded sherds which could not be
usefully illustrated.
Isolated non-dentate rocker-stamping
lines zig-zag vertically down the vessel shoulder.
They
belong to a jar with a red-slipped exterior.
Mat and/or fabric impressed.
Three sherds show traces
of having been decorated by pressing or adhering a fabric or
woven mat to their exterior surfaces.
Two of the these are
unusually thick and show small, chevron-like marks occurring
in parallel, vertical rows (Fig. 89a}.
A third has faint,
white-stained marks, apparently left by a gauze or net-like
fabric (Fig. 89b}.
Adornos
There are four adornos in the Suitacocha Phase
384
collection {Figs. 89c-e, 90a).
Three are hollow, while one
is solid; all are zoomorphic and anthropomorphic heads that
probably adorned jar shoulders or mid-sections.
Fig. 89c
depicts a bird, most likely a parrot, with incised eyes and
mouth, and a distinctive hooked beak.
It was painted red.
Fig. 89d shows an unidentified animal with inset, applique
eyes and prominent snout reminiscent of the opposurn's.
small appendage under its cheek may be a paw.
A
Fig. 89e, a
lug-like appendage with deep incisions and a notched
applique rib vaguely resembles a bat's head and face.
Fig.
90a is an anthropomorphic head which originally projected
from the vessel at an angle.
Ceramic Paste Group B
The Suitacocha Phase ceramic assemblage includes 39
sherds assigned to Paste Group B.
These represent 2.13
percent of the total phase sherd collection.
The three
Paste Group B rim sherds represent only 0.305 percent of the
Suitacocha Phase collection of 982 rims.
This figure under-
represents the quantity of Suitacocha Phase pottery from
relatively distant sources.
Most likely, some Paste Group A
sherds with rare decorative techniques {e.g. rouletting,
inciso cortante) originate from sources other than that of
the bulk of the assemblage.
Paste B4
Thirty-eight sherds assigned to the Paste B4 category
385
are light tan or beige in color (7.5YR6/6 reddish yellow),
and contain a sandy temper.
Electron microprobe analysis
identified grains of quartz, alkali feldspar, chlorite,
hornblende and iron oxides.
be crumbly.
The semi-compact paste tends to
A darker core indicates that the pottery was
incompletely oxidized during firing.
At least two necked
jars are represented by rims and decorated sherds.
One rim
(Fig. 90c) from a short-necked jar was thickened by the
fold-over technique and grooved where the folded end meets
the neck exterior.
Were it not for its distinctive paste
and Suitacocha Phase context, this rim would have been
classified as Manachaqui Phase Shape B Rim 2a.
The other
rim is slightly thinned toward the lip and sits atop a
longer neck (Fig. 90d).
Shape BRim 12.
It somewhat resembles Paste Group A
The decorated sherds show a thick, notched
flange (Fig. 90b), horizontal rows of round punctations
zoned by parallel incised lines, and unzoned vertical rows
of punctations (Fig. 90e-g).
Paste B5
Paste B5 is represented by a single sherd from the
basal angle of a flat-bottomed bowl (Fig. 90h).
Its Sector
A context (Unit 11 Level 8) is mixed, and assignment to the
Suitacocha Phase is based on regional comparisons described
at the end of this chapter.
The paste is very compact and
microprobe analysis identified angular grains of plagioclase
feldspar, apatite, quartz and an unidentified glass.
The
386
core of the sherd is dark brownish gray, and the surface is
black (2.5Y3/0 very dark gray).
The interior of the bowl is
highly polished to a metallic luster.
The uneven exterior
surface has been scored by short strokes with a sharp
implement.
Lithic Remains
As noted in the previous chapter, stone tools and
debitage do not appear in abundance at Manachaqui Cave until
the Early Intermediate Period.
An estimated 100 stone
artifacts, representing one percent of Manachaqui Cave's
total lithic remains, belong to the Suitacocha Phase.
It is
unlikely that all of the chipped stone flakes from
Suitacocha Phase layers intruded from earlier and later
deposits.
However, only ground and polished shale and slate
artifacts are confidently assigned to this phase, while
three chipped stone artifacts are tentatively included.
Of the ground and polished stone artifacts, one
preform, five flakes and ten broken segments of ground shale
and slate points were retrieved from Suitacocha Phase
levels.
Another preform, seven flakes and ten point
fragments are from mixed Suitacocha, Colpar and Empedrada
Phase levels.
The three chipped-stone artifacts include two
from Unit 4 level 21.
One is a tabular core of felsic
igneous rock from which flakes were removed around the edges
leaving some of the cortex.
The other is a similar tabular
piece of felsic rock, but flakes have been stricken from
387
many directions.
One short side has been retouched to
create a concave working edge.
The third chipped-stone
artifact is a gray chert biface with ovoid cross-section.
Lustrous surfaces indicate that the stone was thermally
altered, perhaps intentionally, prior to flaking.
Numerous
large hinge and step fractures suggest that the chert
continued to present insuperable flaws despite heating.
Rocks and Minerals
Rocks and minerals from Suitacocha Phase levels include
one chunk of unworked vein quartz, and a small piece of
mica.
Another piece of mica and a broken piece of a
prismatic quartz crystal were recovered from mixed
Suitacocha and later levels (37-7 and 18-6 respectively).
Botanical Remains
The reader may refer to Table 18 for the list of
botanical taxa identified by Pearsall in Suitacocha Phase
Floors Y and X in Sector A.
Again, Festuca and
Chenopodium/Amaranthus represent potential food sources
obtainable within the immediate site environs.
The
Sapotaceae and other unidentified fruit rinds recovered may
have been brought in from lower elevations.
The most
significant new food counted among those exotic to the
Tropical Alpine zone are maize (Zea mays) and beans
(Phaseolus).
These both appear in Floor Y, the earliest of
the two Suitacocha Phase floors.
Maize also occurs in Floor
388
X although beans do not.
Faunal Remains
A total of 41 faunal specimens were retrieved from
unmixed Suitacocha Phase levels {Table 23), and 32 of these
could be identified by Class {Kent 1994).
In addition to
Mammalia, the inventory includes one Ave {bird), one
Reptilia {reptile) and two Osteichthyes {fish).
Seventeen
of the 28 Mammalia specimens could be identified at finer
taxonomic levels.
Of eight Rodentia, one belongs to the
Muridae Family, and three Caviidae were identified to the
Genus Cavia.
None of the Suitacocha Phase remains could be
positively identified to the species level, although one
White-tailed Deer (Odocoileus virginianus) specimen was
retrieved from Unit 25 Level 7, a mixed ManachaquiSuitacocha Phase context {Table 22).
One Artiodactyl
specimen recovered in deposits between Suitacocha Phase
Floor X and Colpar Phase Floor W {Unit 14, Level 34) is
tentatively classified as Camelid {Table 24).
This specimen
would be the earliest occurrence of Camelidae excavated from
Manachaqui Cave.
Suitacocha Phase Chronology
Given the acceptable radiocarbon dates of 850, 860 and
900 b.c. for the uppermost layers of the Manachaqui Phase
deposits, and 500, 610, 680 and 790 b.c. for the Suitacocha
Phase deposits in both Sectors A and B, 800 b.c. presents
389
itself as an appropriate beginning date for the Suitacocha
Phase.
Some of the vessel forms and virtually all of the
decorative techniques that characterize the Suitacocha Style
are likewise found in late Initial Period/Early Horizon
occupations at sites like Ancon, Kotosh, Huacaloma, Cerro
Blanco, Kuntur Wasi, Kotosh and Chavin de Huantar.
Radiocarbon evidence from these sites (Burger 1981:599,
1992:230-233) suggests that mid-Initial Period stylistic
shifts at Manachaqui Cave lagged one or two centuries behind
major restyling events across the north-Central Andes.
However, some of these sites suffer dating problems, and
only additional well-dated cultural sequences will modify or
reify the current panorama of uneven patterning in Central
Andean cultural development.
Discerning the end of the Suitacocha Phase is more
problematic.
The next ceramic component clearly isolated
utilizing Manachaqui's macrochronology features kaolin ware
pottery, and can be confidently cross-dated to the Early
Intermediate Period.
In Chapter 5, an occupational hiatus
from 450 b.c. to 200 b.c.,
hypothesized.
evidence.
(during the Chavin horizon) was
This hypothesis is sustained only by negative
There are no clear indications of the Chavin
cult's stylistic influence (as defined by Burger 1984b,
1988, 1992) on Manachaqui Cave's pottery, nor is there
ceramic evidence of any style change prior to the mid-Early
Intermediate Period.
The centuries between 500 and 200 b.c.
390
are not represented within Sector A's floor sequence, nor in
Manachaqui Cave's radiocarbon sequence.
Most likely, the
shelter continued to serve local populations infrequently
during the mid-Early Horizon, but played no significant role
within Chavin interaction spheres.
The problem of isolating
evidence for a ceramic component corresponding to the first
centuries of the Early Intermediate Period will be addressed
in Chapter 8.
Paste A Relationships:
the Central Andes
Early Horizon Central Andean pottery styles are
characterized by expanding inventories of vessel shapes and
decorative techniques.
Many Central Andean pottery
traditions now feature polished wares, either red-slipped or
smudged black, on which incised, punctate, engraved,
stamped, line-burnished and post-fire painted decorations
were executed.
While the principal vessel shapes remained
neckless ollas and open bowls, short-necked jars with
everted and sometimes beveled rims like Manachaqui and
Suitacocha Phase Shape B appear in variable amounts along
the eastern edge of the Central Andean cordillera at Piruru
(Rozenberg and Picon 1990: Figs. 2f,
j), Kotosh (Izumi and
Soto 1972:203), Chavin de Huantar, Pachamachay, Telarmachay,
Waywaka and Marcavalle.
Rim lobes are common features on
such eastern highland jars.
Rarely are the rims reinforced,
however, and none exhibit profiles like the distinctive
391
Suitacocha Phase Shape B Rim 11 variants.
Jars with tall, direct rims identical to Suitacocha
Phase Shape F begin to appear in the Early Horizon Central
Andean heartland in relatively small percentages.
Examples
are Middle Guafiape Form 4 (Strong and Evans 1952: Fig. 51),
La Pampa Brown Form 3 (Terada 1979: Pl. 103:18-21) and
perhaps Chavin de Huantar Urabarriu Phase Jars 2 through 5
(Burger 1984b: Figs. 48-51).
In the Chicama Valley,
globular jars with vertical rims and decorated shoulders
constitute minority elements of the Cupisnique Culture as
defined by Larco (1941).
An illustrated example (Ibid.:
Fig. 78 top, far right) presents an zoned-punctate motif of
pendant triangles resembling Suitacocha decorated sherds
(Fig. 85h).
Neither systematic ceramic studies nor analyses
of absolute chronology have been effected for Cupisnique to
date.
At Chavin de Huantar, jars with vertical rims
constitute only five percent of the Urabarriu Phase shape
inventory (Burger 1984b:44), and they apparently lack Shape
F's characteristic right angle joint at the base of the
neck.
Jars identical to Suitacocha Shape F first appear
during the Chakinani Phase (500 to 400 B.C.) as Jars 3B, 6
and 7 (Ibid. :Figs. 168-172).
Shape F-like jars are entirely
absent from Early Horizon Kotosh (Izumi and Sono 1972),
Huaricoto (Burger 1985b) and the central coast.
Thus,
the
shape's distribution is uneven, but largely restricted to
392
the north-Central Andes.
Most importantly, Central Andean
necked jars rarely, if ever, carry decoration other than
polish or slip.
Central Andean potters mostly embellished
bottles, cups and open bowls, shapes absent or rare in the
Suitacocha assemblage.
Decorative techniques like zoned punctation and redslip can be found throughout the Early Horizon Central
Andes.
The manner in which slip is utilized at Huaricoto
for creating contrasts between the painted vessel body and
the unpainted, zoned decoration on the vessel shoulders
(Burger 1985b) resembles the Suitacocha Phase use of slip
painting.
At Chavin de Huantar however, slipping rarely
serves as more than a finishing technique (Burger 1984b:51)
Suitacocha incised motifs found elsewhere in the Central
Andean heartland include pendant triangles filled with
punctation, and zoned rows of punctations.
Kotosh seems to
share the greatest number of motifs with Suitacocha,
including arches and rectilinear U-shapes (Izumi and Sono
1963, 1972).
The latter motif is more typical of the
earlier Kotosh Wairajirca Phase, and it is also found in an
early Nepefia Valley coastal context (Proulx 1985: Pl. 3c).
Closer to Manachaqui Cave, the Huamachuco area shows
evidence of Early Horizon occupation at several sites
(Thatcher 1979; T. Topic and J. Topic 1987), but the
collection of diagnostic material assigned to the Colpa
Phase is small (T. Topic personal communication 1994).
Some
393
"ollas with straight outflared rims"
(T. Topic and J. Topic
1987:14) from Site 102 apparently confirm the presence of
necked jars.
Thatcher's Shape 3B (1979: Figs. 18, 19) may
represent Suitacocha Shape F's Huamachuco counterpart.
His
bowl Shape 16 resembles Suitacocha Shape E Rim 8.
Nevertheless, neckless ollas and open bowls with red slip
comprise the bulk of the Colpa Phase assemblage.
Thatcher
reports short-necked jars during his Mamorco Phase (1972:
Fig. 4n-r) but, as previously noted,
the existence of the
Mamorco Phase has been placed in doubt (and the pertinent
rim sherds thus consigned to temporal limbo).
Incised
decorations on Colpa Phase bowls show little resemblance to
Suitacocha Phase examples.
With the exception of neckless
ollas with grooved rims like Shape A Rim 14, the unusual
Cerro Pelon style (Zaki 1983) displays few resemblances to
Suitacocha.
Again during the Suitacocha Phase, the most salient
similarities are shared with the northernmost Central Andean
sites in Cajamarca.
Suitacocha Phase Shape F jars are
common, although still outnumbered both by neckless ollas
and open bowls at Huacaloma, Cerro Blanco and perhaps
Pacopampa.
Some Huacaloma and Cerro Blanco examples are
identical to Shape F Rims 1 and 4a, even showing the
characteristic red slip and polish (e.g. Terada and Onuki
1982: Pl. 79:4, 1985: Pl. 60:11, 1988: Fig. 28:7-8).
A few
Huacaloma Line Burnished Type Form 3 vessels are identical
394
to Shape B Rim 11f, including the distinctive inflection at
the base of the neck (Terada and Onuki 1980: Pl. 80:5, 1985:
Pl. 58:6).
Such jar rim forms occur in minor proportions at
the Cajamarca sites.
Pacopampa and Pandanche jar shapes also have Suitacocha
Style counterparts.
Pacopampa Group II rims,
some of which
bear red slip, closely resemble Suitacocha Shape F Rim 2
(Fung 1975: Lams. 9:3, 10:23-25, 11:26-37), Shape BRim 11b
(Ibid.: Lam. 10:12, 13) and Shape BRim 11e (Ibid.: Lam.
10:15).
Likewise, Pandanche Phase B1 rims parallel those of
Suitacocha Shape F (Kaulicke 1975: Lam. XII: second row from
top, five at left), and Shape B (Ibid.: top row,
left).
five at
Kaulicke (Ibid.:45) reports that some short-necked
jars were decorated.
The popularity of Inciso cortante like
that observed on the solitary Suitacocha Shape E Rim 10
eclipses other decorative techniques of the Pacopampa
Pacopampa Phase.
Burger suspects an Urabarriu Phase bowl
sherd with inciso cortante of being foreign to Chavin de
Huantar (Burger 1984b:67).
The Suitacocha Phase inciso
cortante example, and the Shape X Rim 4 sherd with combed
decoration (cf. Rosas and Shady 1970: Fig. Sb) could be
"imports" from Cajamarca.
Kaulicke (1975:45,
53) postulates
that both non-local influences and non-local pottery in the
Phase B1 assemblage derive from the lowland tropical forests
and Ecuador.
Cajamarca decorative techniques also display close
395
Suitacocha Phase parallels.
Among these are the use of red
slip to contrast with zoned incised and punctate fields
(Terada and Onuki 1988: Figs. 27:2, 28:13, 14, 16),
restricted application of red slip to the lip and upper
portions of jar rims (Ibid.: Figs. 27:18, 22; 28:7, 8), and
parallel notched applique ribs on red-slipped bowls (Ibid.:
Figs. 30:9, 12).
The red-slipped, polished and incised
bowls illustrated in the Huacaloma (Terada and Onuki 1985:
Pl.56:10) and Cerro Blanco (Terada and Onuki 1988: Fig.
27:2) monographs provide the closest matches for comparison
to distinctive Suitacocha Shape C Rim 8 decorated bowls.
the other hand,
On
there are no polychrome bowls or jars (e.g.
Terada and Onuki 1988: Fig. 32:4) at Manachaqui Cave, and
white slip is likewise absent.
In the western slope yunga of Cajamarca, middle
Jequetepeque valley sites have also yielded Shape F type
rims
(Ravines 1982: Fig. 110:7; Alva 1986: Figs. 407 and
411,
413, 414; Ulbert 1994: Lam. 29) identical to the
previously cited Cupisnique example (Larco 1941: Fig. 78).
The Montegrande assemblage described by Ulbert continues to
display stylistic correspondences to Manachaqui' Cave's
early pottery assemblages.
Some jars from private
collections illustrated by Alva, especially Fig. 411, might
be almost indistinguishable from Suitacocha Phase Shape F
Rim 1b examples.
The middle Jequetepeque publications
illustrate no Shape B analogues, but in the upper Zafia
396
valley montane forest, Alva recovered a small sample of
"short-necked jars with reinforced rims" identical to Shape
BRims 11c (Alva 1988a: Figs. 37:7 and 38:19 and 25).
Red
slip, zoned red painting, geometric zoned punctation and
incised line geometric motifs common to the Jequetepeque and
Zafia valleys (Ulbert 1994: Lam. 14; Alva 1988a:347) are
among diagnostic Suitacocha attributes.
Montegrande basal angles from red-slipped carinated
bowls (Ulbert 1994: Lam. 15: Bowl A7) provide potential
analogies for the Suitacocha Phase carinated vessels.
Kaulicke (1975:45) unearthed one decorated sherd from a
similar vessel (Ibid.: Lams. XII bottom right, XIIIb) in a
Pandanche B1 context.
The Early Horizon Jequetepeque, Zafia
and Pandanche materials all lack associated radiocarbon
dates.
While the Cajamarca assemblages are clearly the closest
stylistic peers of the Suitacocha Style, some significant
differences should not be overlooked.
striking, yet they are sporadic.
Affinities may be
At Huacaloma and Cerro
Blanco, like elsewhere in the Central Andes, neckless ollas
are still far more important than necked jars.
Decorated,
often polychrome, open bowls outnumber both olla and jar
shapes at these Cajamarca sites.
In contrast, the
Suitacocha neckless jar is the least popular of three jar
shapes.
Shape BRim 11 short-necked jars with distinctive
folded and beveled rims constitute 60 percent of the
397
Suitacocha Phase rim collection, yet similar rims are rare
or absent elsewhere in the Central Andes.
Open bowls
comprise only one percent of the Suitacocha Phase
collection, while open and restricted bowls together
encompass only 6.6 percent.
Undoubtedly, Manachaqui Cave's prehistoric functions
were largely responsible for the site's vessel shape
inventory.
Nevertheless, neckless jar shapes, which
differed morphologically from the Central Andean variety
during the Manachaqui Phase (see Chapter 6), may never have
been very important in the northeastern montane forests.
Kaulicke neither mentions nor illustrates neckless
ollas/jars in his discussion of Pandanche Phase Bl vessel
shapes (Kaulicke 1975:45)
One might also consider that
only its smaller size and the addition of its short, everted
rim distinguishes Manachaqui Cave's Initial Period shortnecked jars (especially Shape B Rims 1 and 2)
Central Andean neckless olla.
from the
The neck addition may be
deemed a stylistic detail, and the two vessel shapes'
respective functions may consequently be regarded as
interchangeable if not redundant.
The rarity of the ubiquitous Early Horizon Central
Andean decorated bowls also likely reflects Manachaqui
Cave's specialized functions.
Such bowls are typically
associated with ritual contexts such as Chavin de Huantar's
Ofrendas Gallery (Lumbreras 1993).
They are thought of as
398
having functioned, at least on occasion, to convey ritual
offerings (Mohr-Chavez 1981b:343).
Ulbert (1994:142)
finds
that red-slipped and incised bowls are closely associated
with the monumental constructions at Montegrande.
They also
frequently accompanied the Early Horizon Central Andean dead
into the after-world as grave goods (e.g. Larco 1941, Alva
1986) .
The Suitacocha Shape C Rims 7 and 8 vessels must
represent a larger, more diverse decorated bowl tradition on
the Marafion-Huallaga divide.
Also, decorated Suitacocha
Phase jars may have functioned in some of the same social
and religious contexts as Central Andean bowls.
Regarding
regional comparisons, however, it is most significant that
there is no evidence for the Cajamarca post-firing or
polychrome painting techniques on Suitacocha Phase bowls or
jars.
Manachaqui Cave's divergence from eastern slope
styles farther south may be deduced from the lack of shape
analogues, and painted and applique decorations at Kotosh.
Paste A Relationships: the
Amazonian Lowlands
East of Manachaqui Cave in the Central Huallaga valley
the first documented human occupation is coeval with the
Suitacocha Phase.
At S-Huay-5 on the lower Huallabamba,
Ravines excavated red-slipped and excised pottery that he
assigns to the Shakimu Style first identified on the Central
Ucayali River course by Lathrap (1962, 1971).
Lathrap
offers the radiocarbon date of 650 ± 100 b.c. to which
399
DeBoer (1972-74:95) adds a date of 830 ± 135 b.c. for the
closely related Bin6 Style of the upper Ucayali.
DeBoer's
illustrations of rim treatment on Bin6 cups and jars
(Ibid.:Figs. 12-17; also Figs. 46 and 50) show close
resemblances to Suitacocha Shape BRim 11 and Shape C Rim 7.
At variance with the Suitacocha Style, however, large open
vessels appear to be the most important Shakimu shape class.
All of the common Shakimu vessel shapes, with the exception
of the convex excised bowls, have sharp basal angles and
carinated profiles.
Shakimu carinated bowl shapes (Ibid.:
Figs. 5-8; Lathrap 1970: Fig. lOa-d, k) more closely
resemble the earlier Manachaqui Phase Shape D unrestricted
bowls with concave walls, but temporal disparity makes such
similarities difficult to evaluate.
Red slip seems to be
the only decorative feature shared between the Shakimu and
Suitacocha assemblages.
Although larger ceramic samples and
radiocarbon dates obtained from the Central Huallaga may
prove otherwise,
the two styles do not appear to be closely
related.
Following the Pangotsi Phase in the Alto Pachitea
drainage, the Nazaratequi Phase may be coeval with, or
slightly later than the Suitacocha Phase (Lathrap 1970:98).
From the Pangotsi Phase, Nazaratequi retains a large, necked
jar shape with an unthickened rim (Allen 1968: Fig. 15,
Forms 12 and 13), but carinated bowls dominate the phase
assemblage.
The presence of these necked jars distinguishes
400
Nazaratequi from the neighboring Shakimu Style and loosely
allies it with styles to the west and north (i.e. the
Central and Northern Andes).
Allen (Ibid.:352) again notes
the composite nature of the assemblage.
Nevertheless, there
is little, especially in the way of decorative attributes,
to suggest a specific relationship to the Suitacocha Style.
Far more arresting are Suitacocha's formal and
decorative similarities to Bagua I and II assemblages from
the Marafion and Utcubamba Valley lowlands to the north.
Neckless ollas are also rare at Bagua (Shady and Rosas
1979:112), but a polychrome bowl tradition links the lower
Utcubarnba region to highland Cajarnarca.
Bagua's vessel
shapes include necked jars with folded rims (Shady 1987b:
Fig. 2f-i) like Suitacocha Shape B Rim 11a, and necked jars
with vertical rims (Ibid.: Fig. 4a-d, Sg-i, 6a-c) like
Suitacocha Shape F Rims la, 1b and 3.
Bagua Style jar rims
are much taller on average, but they do commonly bear red
slip.
The jars also carry simple fine-line incised
decorations on the rims (Ibid:. Fig. Sg, h) and upper
shoulders (Shady and Rosas 1979: Fig. 130).
Zoned
punctation diagnostic of the Suitacocha Phase only becomes
important during Bagua II.
Carinated bowl shapes of Shady's
Alenya Complex (Shady and Rosas 1982: Fig. 4b) provide
analogies for the Suitacocha basal angles from carinated
bowls.
Alenya rims match Suitacocha Shape X Rim 3.
Intrusive trade sherds at Bagua indicate long-distance
401
exchange networks, and support the Bagua Phase's temporal
alignment with the Kotosh Kotosh Phase in Huanuco, the
Urabarriu Phase at Chavin de Huantar and the Pacopampa
Pacopampa Phase in Cajamarca (Shady 1987b:481}.
Limited reconnaissance and test excavations in the
lower Utcubamba and nearby Chinchipe areas have yielded
early examples of necked jars (e.g. Rojas 1969: Lam. VI top
a-c, e), and jars or bowls with reinforced rims (Ibid.:
Lams. VI top d, g and bottom a).
The basal angle from a
carinated bowl illustrated by Bushnell (1966: Fig. 1 top) is
similar to the Suitacocha examples, while the distinctive
rim from a "massive vessel"
(Ibid.:
Fig. 2 bottom) is
virtually identical to Shape E Rim 9.
These parallels are
striking given the small size of the Utcubamba-Chinchipe
samples.
So far this lowland region lacks radiocarbon
evidence for absolute chronologies.
Farther afield, on the eastern slopes of the Ecuadorian
cordillera near Macas, early necked jars are found at the
Sangay mound complex (Porras 1987, 1989}, and at the Yaunchu
site originally excavated by Harner (Rostoker 1988).
Despite numerous radiocarbon measurements (Porras 1987:296},
the absolute dating of these Upano Valley finds remains
problematic.
Based on Porras' dates, Athens (1986:112)
estimates a maximum antiquity for the Sangay Upano Tradition
of 400 b.c.
Rostoker (1988:75-76) employs several lines of
evidence to suggest greater antiquity for the stylistically
402
related Yaunchu remains.
Although these ceramic samples
might post-date the Suitacocha Phase by several centuries,
some similarities are too specific to be overlooked.
Interpreted Upano Tradition jar shapes from Sangay
closely resemble Suitacocha Shape F Rims 1, 2 and 3 {Porras
1987: Figs. 76a-ch, 77a; see also Ramp6n 1959: Fig. 2).
Many Sangay jars are polished and bear incised-line and redbanded-incised decorations on the rims and vessel shoulders
{Porras 1987:199-202).
Yaunchu also yielded necked {or
"collared") jars {Rostoker 1988: Figs: 65-70), one of which
has a distinctive lip treatment identical to Suitacocha
Shape F Rim 8 (Ibid.: Fig. 68).
These Yaunchu rims carry
red paint or incised-line decorations, and the upper
shoulders may be incised as well (Ibid.: Figs. 69, 71).
From a private collection reportedly from Macas, Bushnell
{1946: Fig. 2f) illustrates a necked jar with vertical rim
and incised decoration on the shoulder.
These Upano rims
are on average about one or two em. higher than their
Suitacocha Phase counterparts.
Porras (1989: 378) notes that 89 percent of the Upano
Tradition pottery at the Sangay mounds consists of shallow
bowls and plates.
The only attributes paralleled in the
Suitacocha bowl assemblage might be the basal angle on these
Upano style vessels {Porras 1987: Figs. 70ch third from
left, and 73c far right; Ramp6n 1959: Fig. 3, Nos. 15, 16),
and mamiform legs {Porras 1987: Fig. 99).
Rostoker also
403
illustrates carinated bowls (1988: Figs. 25-29, 37).
His
Fig. 28, 29 and 37 bowls carry rims identical to Suitacocha
Shape X Rim 3.
Another unusual bowl rim with a raised
surface (Ibid.: Fig. 54) may provide an analogy for the
sherd illustrated in Fig. 80d (this thesis).
Upano decorative motifs most similar to Suitacocha's
include simple incised lines, arrays of divergent lines
(Porras 1987: Figs. 47c, 48b, 54a), incised arches (Ibid.:
Fig. 51i) and zoned punctation (Ibid.: Figs. 58 and 59).
None of these motif correspondences is very precise.
More
intriguing are parallels in subject matter of Upano
Tradition and Suitacocha Phase adornos.
These are
zoomorphic and anthropomorphic heads that seem to have been
attached to the shoulders of jars.
Porras describes three
zoomorphic adornos as an anteater or opposum (Ibid.: Fig.
116b), a bat (Ibid.: 116c) and a bird (Ibid.: Fig. 116d).
The similarities in detail shared by the actual pieces
cannot be evaluated using Porras' illustrations.
One
anthropomorphic adorno not discussed by the author (Ibid.:
Fig. 107e) is a human head bent sideways at the neck much
like its Suitacocha Phase analogue.
A sketch of an
additional zoomorphic adorno representing an unidentified
animal head is presented by Bushnell (1946: Fig. 2g).
Many resemblances shared between Suitacocha and the
distant Ecuadorian Oriente styles are precise and,
therefore, difficult to dismiss regardless of what criteria
404
archaeologists believe paramount for comparative analyses.
Necked jars, especially jars with vertical rims and
decorated shoulders, are diagnostic elements in early
eastern Ecuadorian pottery styles.
Necked jars illustrated
for the Pastaza Phase (Porras 1975b: Fig. 8) dated somewhere
between 100 b.c. and a.d. 700 (Athens 1986), and Tivacundo
Phase jars (Evans and Meggers 1968: Fig. 19 Shape 5)
radiocarbon dated at a.d. 510 apparently represent the
shape's continued popularity in the Oriente.
Morphologically analogous vessels do not appear in
southwestern Amazonia until approximately a.d. 800 when
Cumancaya and related styles spread throughout the Central
Ucayali (Raymond et al. 1975) and Central Huallaga valleys
east of Manachaqui Cave (DeBoer 1984:114-115; n.d.),
puportedly carried in by Lathrap's Quechua migrants.
Paste A Relationships:
the Northern Andes
In the coastal lowlands separating the Central Andes of
Peru and the Northern Andes of Ecuador, coeval pottery
styles exhibit features diagnostic of both areas.
Necked
jars and elaborate incised bowls dominate Formative Period
assemblages in which neckless ollas are rare or absent.
Necked jars with direct rims have been recovered from Batan
Grande (Shimada et al. 1982: Fig. 39; 1990: Fig. 13b-e),
Lanning's (1963: Fig. 1f) and Ravines'
(1986-87: Lam. 1:3)
Paita Phase C and D sites and Nafiafiique's Panecillo Phase
405
contexts (Guffroy 1989: Fig. 8a-c).
These typically
resemble the tall and slightly flaring Bagua and Pacopampa
rims that rise to between 4 and 8 em, much higher than the
vertical Suitacocha Shape F rims.
Paita Phase D, Panecillo Phase and Pechiche Phase jars
mostly continue Valdivia and Machalilla tendencies toward
gently curving, everted necks and unthickened rims (e.g.
Lanning 1963: Figs. 7a-c, 8; Ravines 1986-87: Lam. 1;
Guffroy 1989: Fig. 8; Izumi and Terada 1966: Pl. 32:1-13).
One rim assigned to Paita Phase B is identical to Suitacocha
Shape BRim 11a (Lanning 1963: Fig. 21b).
The distinctive
inflected upper shoulder diagnostic of Suitacocha Shape B
Rim 11 jars is illustrated by Lanning (Ibid.: Fig. 4b) and
Guffroy (1989: Fig. 8e).
Incised zoned punctation and/or
zoned red slip and painting appears on all of the phase
assemblages coeval with Suitacocha.
Some jars exhibit the
tendency to locate design motifs on the upper shoulders
(e.g. Lanning 1963: Fig. 7; Izumi and Terada 1966: Pl.
32:15; Guffroy 1989: Fig. 8d. c).
Farther inland in the Loja Province of southernmost
Ecuador, the transition from Catamayo B short-necked jars
with everted rims and beveled lips (Guffroy et al. 1987:
Fig. 12) to Catamayo C necked jars with vertical rims
(Ibid.: Fig. 13) mirrors an identical shift in jar rim
morphology at Manachaqui Cave.
Catamayo C rims closely
parallel Shape F Rim 1 (Ibid.: Fig. 14f, i), Rim 2 (Ibid.:
406
Figs. 13a-c, 14a-c) and Rim 5 (Ibid.: Fig. 14d) in details
of thickening and average height.
Loja jar bodies and lips
are frequently slipped red (Ibid.:90).
Simple incised line
decorations ornament rim exteriors (Ibid.: Figs. 13a, b,
14a, g, k), and notched applique, anthropomorphic and
zoomorphic adornos are also associated with the jar shapes
(Ibid.: Figs. 16d, e).
Other decorative techniques such as
zoned punctation are apparently absent.
Nonetheless, the
Catarnayo C Style's correspondences to Suitacocha Phase
ceramic attributes are perhaps more precise than the
parallels shared between Suitacocha and Bagua.
Archaeologists working in the frontier between the
Central and Northern Andes often note similarities between
their materials and the widespread Chorrera style (800 - 300
b.c.) of coastal Ecuador (e.g. Izumi and Terada 1966:88;
Shimada et al. 1982:146-147; Guffroy et al. 1987:113).
Information regarding Chorrera material culture is found
scattered in numerous publications (e.g. Bushnell 1951;
Estrada 1958; Meggers et al. 1965; Meggers 1966; Lathrap et
al. 1975; Evans and Meggers 1982 and Bischof 1982).
The
unambiguous resemblances between Chorrera and the Suitacocha
style are only surprising if one overlooks the numerous
design features shared by the Valdivia/Machalilla and
Initial Period Manachaqui assemblages.
Bushnell (1951: Figs. 37a-e) illustrates two Engoroy
(the Santa Elena Peninsula Chorrera variant) classes of
407
necked jars with vertical rims.
The wide-mouth class in
particular (Ibid.: Figs. 37c, e) mirrors Suitacocha Shape F.
Oddly, Bischof's (1982) Engoroy jars do not include this
shape.
Estrada illustrates jars of the same narrow and
wide-mouth classes from the coastal site of Ayangue (1958:
Fig. 48:1 and Fig. 47:4 respectively).
The coastal
Formative Period sequence established by Meggers et al.
(1965) not only demonstrates the great antiquity of
decorated necked jars in the Northern Andes, but documents
their persistence into the Chorrera Phase.
Red-slipped and
polished vessels date to the beginning of the Valdivia
sequence, as does the tendency to decorate jar shoulders.
The Chorrera counterpart of Suitacocha Shape F is
"Machalilla Polished Red" Vessel Form 6 (Ibid.: Fig. 84:6).
The Form 6 rim grades into a gently curved everted form, but
an angular junction of the neck and body visible on the
interior surface is a diagnostic attribute.
Red slip covers
the jar exterior and the rim interior down to the angle.
Machalilla Striated Polished Plain Vessel Form 9 is the same
shape without red slip.
Both of these "Machalilla" pottery
types figure prominently during the subsequent Chorrera
Phase (Ibid.: Fig. 89).
At Pefion del Rio, a Late Formative
Period habitation site in the Guayas Basin, vessel Forms
15A-C mirror Suitacocha Shape F.
The sharply angled joint
at the base of the rim is diagnostic of the Chorrera jars.
However, like other western Ecuadorian jars, the Pefion del
408
Rio varieties carry strictly everted rims of varying lengths
and thicknesses.
Suitacocha Shape F rims show more variable
rim orientation angles and rim thickening techniques, thus
they more closely resemble the Upano Tradition jars from the
Oriente.
Machalilla Polished Plain Jar Form 4, Machalilla
Striated Polished Plain Jar Form 10 and Machalilla DoubleLine Incised Jar Form 6 with reinforced lips vaguely recall
Suitacocha Shape B Rims 11d-g.
The Suitacocha Shape B
short-necked jars really have no precise Chorrera match.
Chorrera shallow carinated bowls from La Ponga illustrated
by Lippi (1983: Fig. 53b [bottom row, second from right])
exhibit basal angles analogous to the Oriente and Suitacocha
examples.
Even more precise resemblances between the Chorrera and
Suitacocha styles are found in shared decorative techniques.
A Suitacocha Shape C Rim 7 sherd with engraved decoration
(Fig. 71t this thesis) displays a faithful duplication of
the Double-line Incised technique and motif illustrated by
Meggers et al.
(1965: Pl. 137a).
Lippi (1983: Fig. 63,
#1506) illustrates the same engraving technique and motif.
The Punctate, and Incised and Punctate techniques (Meggers
et al. 1965: Pl. 144w-z, 147b-i) mirror the Suitacocha
examples depicted in Figs. 86c-k, 87a-h.
Red slip creates
contrasting painted and punctate fields separated or "zoned"
by incision (Lippi 1983: Fig. 73).
Bosses embellish
409
shoulders of other Machalilla/Chorrera vessels (Meggers et
al. 1965: Pl. 143), and Suitacocha incised motifs common in
coastal Ecuador include simple arrays of divergent parallel
lines (Ibid.: Pl. 144h), zoned superimposed X motifs (Ibid.:
Pl. 133n; Lippi 1983: Figs. 64, 85, 88), cross-hatching
(Meggers et al. 1965: Pls. 131, 133), arches (Ibid.: Pls.
145e, 148; Lippi 1983: Figs. 71, 72) and pendant triangles
(Meggers et al. 1965: Pl. 132h).
Delineating, or "zoning" red-painted fields with
incision gained popularity during the Machalilla Phase
(Ibid.: Pl. 145a-i) and became a diagnostic attribute during
Chorrera (Ibid.: 139 and Pl. 157b, c; Meggers 1966:58; Evans
and Meggers 1982), as do polished red rims,
incised rims,
iridescent paint, applique rib, plain rocker stamping, and
some techniques unknown for Suitacocha.
Interpretations of
Suitacocha's external relations must take into account that
many of the shared Northern Andean incised motifs: 1) are
engraved, 2) occur on bowls as a preferred medium rather
than jars, 3) are distributed widely over the Northern Andes
and the Oriente, and 4) may appear at different prehistoric
moments in different regions (they rarely appear in welldated stratigraphic contexts).
Furthermore,
the temporal
and stylistic relationships between Machalilla and Chorrera
remain to be clarified with stratigraphic evidence.
In the Ecuadorian highlands, necked jars with redslipped and polished vertical rims and incised shoulders
410
appear with the earliest documented pottery styles at
Monjashuaico (Bennett 1946: Figs. 10, 11), Huancarcuchu
(Ibid.: Figs. 6, 7), Cerro Narrio (Collier and Murra 1943:
Figs, 10 top two rows left, 12 top row) and Cotocollao
(Villalba 1988: Fig. 128).
Some Huancarcuchu jars with
fields of punctation and red paint zoned by incision
(Bennett 1946: Fig. 71, 9o) closely resemble decorated
Suitacocha jars in general design layout.
Unlike
Suitacocha, however, Huancarcuchu red paint was habitually
applied in parallel bands.
The manner in which jar
shoulders from Cafiar Province carried zoomorphic and
anthropomorphic adornos (Uhle 1922: Lams. X-XII) provides an
analogy for the placement of the Sangay and Suitacocha
adornos.
Incised line motifs common to Suitacocha and south
highland Ecuador (Ibid.: Figs. 9, 11, 13; Collier and Murra
1943: Pls. 19: 12, 13) likewise match motifs on the
Machalilla/Chorrera pottery types just discussed (Meggers
1966:64).
Similar motifs, occasionally combined with
punctation, are common at Cotocollao (Villalba 1988: Fig.
88, 90, 92, 96) and in the northern Ecuadorian province of
Irnbabura (Myers 1976: Figs. 2, 3).
Porras graphically
demonstrates these pan-Ecuadorian resemblances with
illustrations (Porras 1982: Lams. 38-53).
Observe the
likeness shared by incised sherds from Machalilla (Meggers
et al. 1965: Pl. 144h), Sangay (Porras 1987:54a), Charnbira
411
in the Peruvian Amazon (Morales 1992: Fig. 4a) and Cerro
Narrio (Collier and Murra 1946: Pl. 19, Figs. 12).
Clearly
there is a need to sort out the temporal and stylistic
relationships among these assemblages.
The reader will recall that Early Cerro Narrio shares
affinities with the preceding Manachaqui Phase style as well
as Suitacocha.
Meggers (1966:62-64) observes that early
south highland pottery combines Machalilla incised motifs
and Chorrera vessel shapes.
assemblage.
So does the Suitacocha Phase
Unfortunatley, reliable temporal information
regarding Northern Andean ceramic attributes most pertinent
for a comparative analysis of the Suitacocha Phase
assemblage is not yet available.
Nonetheless, it should be
apparent to the reader, especially in light of the
relatively solid data base for Formative Period south
coastal Ecuador, that the Suitacocha Style is in large
measure comprised of diagnostic attributes that possess
great antiquity and wide geographic distribution in preHispanic Ecuador.
Paste B Origins
The Suitacocha Phase Paste B sample offers additional
evidence for interaction over very long distances.
Paste B4
pottery likely originated far to the north in the Ecuadorian
Oriente.
The sources for Paste B5 probably lay on the
Central Andean north coast.
These identifications have not
been substantiated by hands-on comparisons with sherds from
412
the suspected source areas, although such comparisons are
warranted.
Paste B4
The Paste B4 sample of 38 sherds shows similarity to
several coeval styles with respect to decorative attributes.
Rows of large punctations arrayed in zoned-incised fields
occur on Kotosh Phase bowls at Kotosh (Izumi and Terada
1972: Pl. 116), and Urabarriu Phase neckless ollas, bowls
and cups at Chavin de Huantar (Burger 1984b: Figs. 65-69,
21, 99B).
Sherds with these kinds of motifs also appear
intrusively in a Machalilla context at Salango (Norton et
al. 1983: Fig. 12a-e).
Because the Paste B4 sherds belong
to decorated necked jars, origins farther north than Kotosh
or Chavin are indicated.
The Paste B4 sherds approximate descriptions and
illustrations of Upano Punteado (Upano Punctate) Type
pottery recovered from excavations at Sangay (Porras
1987:153-158).
Porras describes non-plastic inclusions in
the paste as "river sand, feldspar and quartz."
Porras does
not specify associated vessel shapes for his small sample of
49 decorated sherds, but necked jars are well-represented at
Sangay.
The B4 decoration matches his description of Motif
3: "Circular punctations inscribed in double horizontal
line, double curved line in some cases, and double oblique
line in others"
(Ibid. :154).
Porras' Fig. 58fi provides a
basis for comparing Upano zoned punctation with Suitacocha
413
Paste B4 sherds, and the B4 applique flange may be compared
with Upano flanges illustrated in his Figs. 65e and f.
Origins in the Amazonian montane or lowland forests of
southern Ecuador or northern Peru are most likely.
Paste B5
The single black polished sherd of Paste B5 displays a
texturing technique observed at Chavin de Huantar, and
termed "rough scratch," by Bennett (1944:87).
Bennett
recovered sherds with these scored applique nubbins from pit
CH-10 (Ibid.: Fig. 30c) which also yielded numerous
Chakinani Phase (500 to 400 B.C.) sherds (Burger 1984b:167).
Similar surface scoring appears in Qotopukyo Style pottery
of Lumbreras' Ofrendas Phase (800 to 700 B.C.} at Chavin
(Lumbreras 1993: Lam. 45).
vessels are bowls.
Apparently, none of the Chavin
It is also possible that Paste B5
originated in a Cupisnique context on the Pacific coast
directly west of Manachaqui.
In either case, the sherd is
typical of polished black wares of the Early Horizon Central
Andes.
Prehistoric Cultural Development and Interregional
Interaction in the Tropical Montane Forests of Peru
Volume 2
A Dissertation
Presented to the Faculty of the Graduate School
of
Yale University
in Candidacy for the Degree of
Doctor of Philosophy
by
Warren Brooks Church
Dissertation Director: Richard L. Burger
December, 1996
CHAPTER 8
THE COLPAR PHASE
The Colpar Phase was established during the final
laboratory analyses of Manachaqui Cave's radiocarbon dates
and ceramic macrochronology.
Chapter 7 presented
radiocarbon evidence from Sector A (the shelter interior)
suggesting at least two distinct occupations (or series of
occupations) during the Early Intermediate Period.
Colpar Phase represents the earlier of the two.
The
Analysis of
stratified remains from Sector B resulted in the isolation
of several vessel shapes and rim forms which did not fit
comfortably into either the Suitacocha nor the ostensibly
subsequent Empedrada Phase.
These shapes and forms
constitute a weakly represented, yet nonetheless
discernible, component labeled the Colpar Phase.
Future
radiocarbon analyses will attempt to correlate Sector B's
evidence for Early Intermediate Period occupations with
Sector A's radiocarbon and ceramic sequences.
An additional indication that a temporal component
between the Suitacocha and Empedrada Phases can be isolated
consists of several "intermediate" levels containing neither
incised decoration nor kaolin ware sherds.
particularly evident in Units 28 and 30.
414
These levels are
The levels that
415
provide most of the Colpar Phase remains include Unit 28
Level 9, Unit 30 Level 9 and Unit 31 Level 9, but it cannot
be demonstrated that any of these levels are completely
unmixed.
The assignment of vessel shapes and rim forms to
the Colpar Phase required careful consideration of pertinent
sherd proveniences.
There is considerable room for error,
especially because a small sample is involved.
Most
difficulties were encountered when attempting to separate
shapes/rims belonging to the Suitacocha and Colpar phase
assemblages.
Colpar Phase ceramics show closer stylistic
affinity to Suitacocha Phase pottery than to Empedrada Phase
pottery.
Finally, the difficulty in isolating unmixed Colpar
levels hampers the assignment of lithic, rock and mineral,
botanical and faunal remains to the Colpar Phase, and
frustrates attempts at even rough quantitative estimates for
some of these assemblages.
Hopefully, continued
archaeological investigations in this region will locate
unmixed deposits with which to define more clearly the
Colpar Phase.
Colpar Phase Ceramics
The Colpar ceramic collection is the smallest of
Manachaqui Cave's diagnostic phase assemblages, represented
by only 134 diagnostic sherds.
In most respects, the Colpar
Phase represents a continuation of stylistic norms
established during the preceding Suitacocha Phase.
Most
416
notably, restricted vessels with reinforced rims, globular
bodies and round bases continue to dominate the shape
inventory.
Notched applique and red slip painting are the
only decorative techniques confidently assigned to the
Colpar Phase.
Because of the small sample and the
analytical problems just cited, the apparent absence of
attributes such as line-incised and textured decoration,
punctate applique, flanges and adornos during the Colpar
Phase may be apparent rather than real.
Shape A neckless
jars, Shape F jars with vertical rims and carinated body
profiles also drop out of the shape inventory.
The highly
weathered condition of the small collection further reduces
the number of rims that can be securely classified and
accurately illustrated.
It also hinders evaluation of the
role of painted decoration.
During the Colpar Phase, ceramics of Paste Group B
increase in number and diversity, especially in proportion
to the Paste Group A ceramics.
appear for the first time.
Paste Group C kaolin wares
Both paste groups can be
considered constituents of a nascent category of "fine
wares" contrasting with "coarse wares" of Paste Group A.
As
in the previous phases, these assemblages are assumed to
consist of pottery produced at greater distance from
Manachaqui Cave than Paste Group A pottery.
It is primarily
the stylistic diversity that strongly indicates that both
Paste Groups B and C are "intrusive" relative to Paste Group
417
A which, in contrast, exhibits stylistic unity,.
The so-
called "fine wares" of the pre-Abiseo Phase assemblage at
Gran Pajaten show the same lack of stylistic coherence
(Church 1994:288).
Ceramic Paste Group A
The Paste Group A category is comprised of 93 sherds
representing 69.4 percent of the total Colpar Phase
assemblage.
Like its Suitacocha Paste A predecessor, it may
consist of more than one ware.
Because the Colpar Phase was
first distinguished and defined after the collections were
no longer available for study, pertinent Paste Group A
sherds were not purposefully included in the hands-on
analyses.
That the overall similarity of Colpar Phase
sherds to Suitacocha Phase sherds obstructed recognition of
the phase assemblage during the hands-on pottery analysis in
Trujillo demonstrates that the basic ceramic technology
remained virtually unchanged (at least from a macroscopic
perspective) .
Morphol.ogy
Because of the difficulties in isolating the Colpar
Phase ceramic assemblage, it should be kept in mind that the
totals (e.g. of rims and decorated sherds) are understated
and, consequently,
the percentages representing attribute
frequencies are distorted approximations.
Only 71 rim
sherds are assigned to the Colpar Paste Group A assemblage.
418
Some vessel shape and rim form categories seem to have
carried over from the Suitacocha Phase.
B Rim 11b and Shape X Rim 1 may persist.
persistence are less certain.
For example,
Shape
Other cases of
The contours and proportions
of Colpar vessels remain unknown, although comparison with
earlier Suitacocha Phase vessels and coeval vessels from
neighboring areas may provide analogies.
Shape B: Short-necked jars
The 22 rims identified as Colpar Phase short-necked
jars represent 31 percent of the Paste Group A rim
collection.
Like earlier Shape B rims, these are beveled
and reinforced by adding clay to the lip exterior, or by
folding over the rim end and smoothing it to the vessel
exterior.
Gone is the narrow groove under the thickened lip
that characterized many Suitacocha jar rims.
Jar mouth
diameters range from 10 to 16 em in diameter, but only ten
shape B rim diameters were measured.
Rims llh-j
(Fig. 91a-j).
into three variants.
These 22 rims are divided
The 13 examples of Rim 11h (Fig. 91a-
e) resemble Suitacocha Phase Shape B Rims 11b or 11c, but
the pointed lip tends to be bent further outward, and the
thickened underside is less pronounced and joins the vessel
neck in an unbroken curve.
mm.
Its thickness ranges from 5 to 9
Five examples of Rim 11i (Fig. 91f-h) resemble
Suitacocha Phase Shape B Rim 11d, but the flattened lip is
419
thicker (7 to 10 mmj, and the rim may be more vertically
oriented like Shape F rims.
The three measured specimens
range in diameter from 10 to 12 em.
The four examples of
Rim 11j (Fig. 91i, j) are vertically oriented, lightly
thickened to 7 or 8 mm and the lip is beveled to a dull
point.
Both of the examples measured showed diameters of 16
em.
Shape C: Restricted bowls
Colpar Phase restricted bowl rims number 26,
representing 37 percent of the Colpar Paste A rim
assemblage.
All but one of the rim variants are thickened,
and all likely pertain to low, rounded vessels.
As
expected, rim diameters are more variable than those of
jars, measuring fr,om 12 to 19 em (N=12).
R~
11 (Fig. 91k-o).
The 16 examples of Rim 11 show
rounded lips reinforced by rolling the rim end down over the
exterior.
Thickness at the lip ranges from 5 to 8 mm.
Diameters range from 12 to 19 em (N=6).
between 3 and 4 mm in thickness.
Vessel walls vary
The only unweathered sherd
from Sector A shows highly burnished interior and exterior
surfaces, and clearly visible tracks approximately 1 mm wide
left by a blunt-edged burnishing tool.
R~
12 (Fig. 91p-s).
The six examples of Rim 12 are
lightly thickened to 6 or 7 mm and have flattened lips.
420
Diameters range from 12 to 16 em (N=4).
4 to 5 mm thick.
The walls vary from
An eroded rim (Fig. 91p) exhibits applique
nubbins.
Rim 13 (Fig. 91t).
Two examples of Rim 13 are
unthickened or lightly thickened and show a shallow groove
running around the lip exterior.
Both fragments were too
eroded to obtain a diameter measurement.
The walls are
approximately 3 mm thick.
Rim 14 (Fig. 91u).
The single example of Rim 14 is
thickened to 8 mm, has a flattened but slightly concave lip,
and diameter of approximately 18 em.
Its walls are 4 mm
thick.
Rim 15 (Fig. 9lv).
The lone Rim 15 sherd is
unthickened with a dull point at the lip exterior.
It shows
a diameter of 12 em and the walls are 3 mm thick.
Shape E: Unrestricted convex bowls
The 23 Shape E rims constitute 32 percent of the Colpar
Paste A rim collection.
the lip.
Again, most exhibit thickening near
Two of the five rim forms (Rim 3 and Rim 9) occur
in the earlier phases (e.g. Fig. 92a-c).
The other three
thickened and beveled rims are unique to the Colpar Phase.
Together the Shape E rims range from 11 to 20 em (N=l3)
diameter.
in
421
Rim 12 (Fig. 92d-j).
The 13 examples of Rim 12 are
direct rims with a reinforced exterior extending
approximately one em below the rounded or pointed lip, and
ranging from 4 to 7 mm in thickness.
measure only 2 mm in thickness.
Unusually thin walls
The rim depicted in Fig.
92j has a small notched lobe projecting about 3 mm above the
lip.
Rim 13 (Fig. 92k, 1).
Two sherds represent Rim 13.
They thicken gradually to 6 mm toward the lip which is
beveled to leave a dull exterior point and a flat sloping
interior surface.
A shallow groove runs around the lip
exterior 5 mm below the tip.
The bowls' walls are 3 to 4 mm
thick.
Rim 14 (Fig. 92m).
The single Rim 14 sherd resembles
Rim 13, but the lip is beveled to create two flat surfaces
culminating in a point.
There is no exterior groove.
The
maximum thickness reaches 7 mm near the lip, and the walls
below are 4 mm thick.
Decoration
As previously mentioned, the difficulty encountered in
stratigraphically isolating the Colpar Phase assemblage
coupled with its severe erosion results in the underrepresentation of decorative techniques and inexact sherd
counts.
At least 22 decorated sherds belong to the Colpar
422
Phase assemblage.
The collection's extreme fragmentation
hinders the identification of associations between shapes
and rims and particular techniques.
The following
observations are offered with these limitations in mind.
Painting
Pre-firing paint or slip is usually observed only on
sherds from Sector A's sheltered, but often mixed deposits.
Based upon these observations, however, it would seem that
all of the Colpar Phase vessel shapes and most of the rim
forms may bear paint.
Red is clearly the most common paint
color, although its frequency cannot be accurately assessed.
Five of the Shape B Rim 11h and 11i examples (including Fig.
91a, c-e) show traces of red paint.
Two of these also show
traces of black paint on the lip (Fig. 9la, c).
One Shape C
Rim 12 sherd (Fig. 91r) exhibits red paint on both surfaces.
The surfaces of the single Shape C Rim 14 (Fig. 9lu) example
have traces of white slip.
Applique
Colpar Phase applique may occur as small nubbins as
seen in Fig. 91p.
The notched applique lobe previously
described adorns one Shape E Rim 13 example (Fig. 92j)
I
and
the tradition of decorating jar bodies with notched applique
strips apparently continues.
Fifteen sherds bearing wide (8
to 12 mrn) strips with cut or incised (not impressed) notches
were recovered from suspected Colpar contexts (Fig. 92n-p)
I
423
and mixed Suitacocha, Colpar and Empedrada contexts.
Probably only a fraction of these sherds truly belonged to
Colpar Phase vessels.
Their specific location on vessel
bodies remains unknown although they were probably medial
ribs surrounding jar mid-sections.
Ceramic Paste Group B
Paste Group B sherds are more numerous and diverse in
design details than those of the two preceding phases.
Again, because Manachaqui Cave's ceramic macrochronology
provides only a small window on the Colpar Phase, it is
difficult to assign small samples of Paste B and C wares
(especially those from Sector A) to their proper phases.
Wares were assigned to phases based primarily upon macrochronological position.
However, information regarding
regional stylistic developments was an additional
consideration for temporal placement.
Some 38 sherds
representing five sub-groups (or wares) made conspicuous by
atypical pastes have been assigned to Colpar Phase Ceramic
Paste Group B.
In some cases, sherd surface erosion
impaired the identification of attributes judged diagnostic
of each paste sub-group during the ceramic analysis.
Therefore, these totals are also conservative.
This may be
the case for Paste B10 •
The 38 Paste Group B sherds represent 28.4 percent of
the Colpar Phase diagnostic sherd total.
Of the 86 Colpar
Phase rim sherds, 15 or 17.4 percent belong to Groups Band
424
C.
This percentage is artificially high because of the
difficulties in separating the Colpar Phase assemblages from
the preceding Suitacocha Phase and the following
Phase assemblages.
R~pedrada
Recall that some Colpar Phase Paste A
Shape B Rim 11 sherds were counted among the Suitacocha
Phase rims.
One or more of the Paste B wares listed below
conceivably belong in the Empedrada Phase Paste B
assemblage.
Paste B6
Fourteen highly eroded sherds of Paste B6 pottery
probably pertain to a single bowl with flat base, direct
divergent walls and unknown rim diameter (Fig. 93a, b).
The
rim shows an unusual grooved or recessed exterior surface.
The soft paste is completely oxidized a brick red color (SYR
5/6 yellowish red).
Quartz, mica and angular grains of
igneous rock comprising non-plastic inclusions vary greatly
in size.
Although the paste matrix is very fine, some
grains reach up to 4 mm in diameter.
Paste B7
The thirteen sherds representing this pottery exhibit a
dark brown paste with a dark gray core left by incomplete
oxidation during firing.
They are covered by a brownish
black veneer (5YR3/2 dark reddish brown) created by smudging
the fired pot in a carbon-rich atmosphere.
The dark surface
glitters because of abundant mica and smaller amounts of
425
quartz.
The sherds pertain to open convex bowls (Fig. 93c-
f) with slightly thickened rims and flattened, in-sloping
lips closely resembling Shape E Rim 13.
A deep incised
groove runs around the vessel circumference just
lip exterior.
~~der
the
One bowl interior (Fig. 93f) exhibits
decoration with a wide curving line of pink iridescent paint
(lORS/3 weak red) that may have been applied with the
finger.
A small amount of paint is also visible within the
groove on the rim exterior.
Paste B8
Five hard sherds from three bowls show light gray
surfaces (10YR6/2) with smoke clouds and black, unoxidized
cores (Figs. 93g; 94a, b).
The paste is micaceous with
angular bits of quartz and sand.
Three articulating sherds
(Fig. 93g) comprise a fragment from a simple convex bowl
covered with pink slip (lORS/6 red) and painted with three
thin, curving white painted lines.
The other two bowl
fragments are undecorated, but one displays a carinated
profile (Fig. 94a) while the shape of the third (Fig. 94b)
resembles that of the slipped and painted bowl.
Paste B9
Four rim sherds belong to one or two bowls or jars with
vertical rims (Fig. 94c-e, Plate XI).
The incompletely
oxidized paste is fine, light gray (10YR7/2 to 10YR7/4} on
the surfaces and dark gray within.
Non-plastic inclusions
426
are mostly angular grains of igneous rock.
A red (10R4/4
weak red) painted design repeating around the rim consists
of Xs and pendant triangles on the interior, and Xs, pendant
triangles, lines and dots on the exterior.
Paste
B~
The two Paste B10 sherds belong to a single open bowl
with a thickened rim interior (Fig. 94f).
compact and fired hard.
The paste is
A dark gray core and brown surfaces
(5YR3/4 dark reddish brown) indicate incomplete oxidation,
and a small smoke cloud covers part of the rim exterior.
The non-plastic inclusions are angular grains of igneous
rock.
Orangish-red paint (2.5YR5/8 red) covers much of the
exterior surface, and the small amount of paint visible on
the interior surface shades into brownish tan with a
metallic luster.
The entire bowl is well-polished.
Ceramic Paste Group C
The Colpar Phase marks the first appearance of fine,
well-made pottery rich in kaolin clay at Manachaqui Cave.
The wares united under the category of Ceramic Paste Group C
in this and the following Empedrada Phase assemblage vary
greatly in color and texture.
Undecorated or eroded kaolin
sherds pertaining to the Colpar Phase could not be isolated
for the reasons previously cited.
Only three Paste Group C
sherds (2.2% of the phase collection of diagnostic sherds)
are assigned to the Colpar Phase.
427
Paste C1
Three sherds of extremely fine,
fully oxidized (7.5YR7/
light gray) pottery probably belong to an open bowl with a
lightly thickened, flaring rim (Fig. 94g).
inclusions are invisible to the unaided eye.
Non-plastic
Two of the
three sherds are rims, while the third is a body sherd.
The
lip exterior is rounded and the interior edge forms a point.
Vessel walls are 3 mm thick.
The lip exterior shows traces
of red paint, and the wall exteriors exhibit a motif
composed of thin continuous and broken painted lines.
Lithic Remains
By viewing the frequency of lithic remains per level
within the Colpar Phase macro-chronological contexts,
especially those in Units 28 and 30, a pattern of increasing
utilization of stone tools during the Early Intermediate
Period can be discerned.
The elevating numbers mostly
reflect the increasing deposition of chipped-stone cores and
manufacturing detritus left by a simple core-flake industry.
Some flakes of gray and dark green igneous rock and chert
were retouched with percussion blows to render steep working
edges but, again, no specific totals for tools, cores or
waste flakes can be cited for the Colpar Phase.
Balancing the suspected increased rate of lithic debris
deposition with the overall paucity of artifacts within this
rather effemeral phase assemblage, a total of 500 lithic
428
artifacts (about five percent of the Manachaqui Cave total)
can be offered as a reasonable estimate.
Of the ground-
stone tools, only polished slate and shale points occur in
Colpar Phase contexts.
These points occur in abundance
during both the preceding Suitacocha Phase and the
subsequent Empedrada Phase.
Excavations in Sector B
recovered 10 point fragments, one possible preform and seven
laminar pieces of slate or shale debitage from levels
approximating the Suitacocha-Empedrada phase interface.
Rocks and Minerals
One piece of mica (Unit 37 Level 7) and one fragment of
a prismatic quartz crystal (Unit 18 Level 6) may have been
deposited during the Colpar Phase, but could also belong to
the earlier or later deposits.
Botanical Remains
The only floor radiocarbon-dated to the Colpar Phase is
Floor W, although Floors U and V might also constitute
Colpar Phase deposits.
Flotation of Floor Wand V samples
produced food remains, while the Floor U sample yielded only
charred wood (Table 18).
kernels and cob fragments,
Floor W contains abundant maize
cf. Sapotaceae fruit,
fruit
rinds, tuber and root fragments, Lupinus, Festuca and
Chenopodium and/or Amaranthus.
fruit rinds and cf. Ribes.
Floor V contains maize,
429
Faunal Remains
Only the Sector A specimens recovered from Floors W, V
and U are described here (Table 25).
Thirty Floor W
specimens are classed as Mammalia and one as Osteichthyes
(fish).
Three of the Mammalia belong to the order
Artiodactyla of which two are Cervidae cf. Mazama, and and
one is cf. Camelidae.
Eleven Mammalia are Rodentia or cf.
Rodentia, mostly Caviidae.
Two Dasypodidae specimens are
probably remains of armadillo.
One Osteichthyes from Floor V (probably Colpar Phase
but tabulated in Table 26 as mixed deposit) belongs to the
ct. Eleotridae family of sleepers, a fresh water fish found
in warm estuarian river environments.
Nineteen Mammalia
specimens were identified from Floor V, among which six
Cavia were identified.
Floor U (probably Ernpedrada Phase)
yielded one specimen identified only as Ave and nine
Mammalia, one of which is assigned to the genus Cavia.
Radiocarbon Evidence for Colpar
Phase Chronology
The Colpar Phase is established on the basis of: 1) the
two radiocarbon dates from Sector A representing the first
centuries of the Early Intermediate Period, 2) small, but
perceptible gaps separating the two macrochronological
components bearing incised-line decoration (Suitacocha) and
abundant kaolin fine wares (Ernpedrada), and 3) an ephemeral,
yet distinguishable, ceramic stylistic component straddling
430
the macro-chronological interface between the Suitacocha and
Empedrada Phases.
In the following sections, cross-dates to
this same time period will be intimated by the resemblances
shared between Colpar ceramics and pottery styles in
neighboring regions.
The Colpar Phase awaits further
definition through additional work in this area.
The radiocarbon dates of 160 b.c. from Unit 15's Floor
W, and a.d. 110 from Unit 14's Floor P constitute evidence
for Manachaqui Cave's utilization immediately following the
Chavin horizon.
The temporal lacuna of 340 radiocarbon
years (359 calibrated radiocarbon years) separating the
latest of the Suitacocha Phase dates from Floors X and Y
(500 b.c.) and the Floor W date of 160 b.c. is interpreted
as a period of the shelter's relative disuse.
Manachaqui
Cave probably sheltered travelers infrequently during the
Chavin horizon.
The Sector A radiocarbon evidence supports
a beginning date of approximately 200 b.c. for the Colpar
Phase.
The determination of an ending date is more
problematic.
Ceramic evidence for Colpar Phase chronology
will be discussed at the end of this chapter.
Paste A Relationships:
the Central Andes
Some scholars have characterized the first centuries of
the Central Andean Early Intermediate Period as a stylistic
horizon marked by pre-firing "White-on-Red" painted
decoration (e.g. Willey 1971; Lumbreras 1974).
The
431
proliferation of this decorative technique follows on the
heels of the Chavin horizon at many sites previously
incorporated within Chavin's sphere of influence, especially
in the Callej6n de Huaylas and on the north-central coast.
Coastal assemblages such as Bafios de Boza, Puerto Moorin and
Salinar are relatively well-dated and described, while
coeval highland styles generally lack detailed descriptions
and/or absolute dates.
At the onset of the Early Intermediate Period,
jar
shapes most often feature short necks, everted rims and,
occasionally, tall spouts.
first time on the coast.
Carinated jars appear for the
Convex bowls exhibit both
restricted and open shapes, and both jars and bowls commonly
feature vertical and horizontal strap handles.
The stirrup-
spout bottle shape is most common in the north-coastal
valleys directly west of Manachaqui Cave, while doublespout-and-bridge bottles predominate in south-coastal
valleys like Ica and Nazca.
In the Central Andes, the
white-on-red technique is relatively short-lived and
eventually eclipsed by negative resist painting.
Jar shapes
increasingly feature long, flaring collars with direct and
tapered rims.
Details of design, especially surface
decoration, vary from assemblage to assemblage during these
centuries, although pre-firing slips and painted geometricline motifs are characteristic.
Ceramic variability at this
time in the Central Andes is difficult to evaluate given the
432
current lack of knowledge.
The clearest similarities to the Colpar Phase
assemblage are found one long day's walk eastward to the
montane forest site of Gran Pajaten.
All three Colpar Shape
B rim forms have precise counterparts in collared bowl rims
of the Pre-Abiseo Phase pottery assemblage {Church 1994:
Fig. 12, 13a-g).
(Ibid.: Fig. 12t).
Some Gran Pajaten examples show red slip
These parallels raise the issue of
whether some of the eroded Colpar Phase Shape B vessels
should be considered bowls rather than jars.
Only
comparative analyses of surface treatments on better
preserved samples from this period will settle the question.
It should be kept in mind, however, that the dichotomy
between so-called jar and bowl morphologies is never very
rigid during Manachaqui Cave's early ceramic phases.
Gran Pajaten's pre-Abiseo Phase assemblage includes
cultural remains left by occupations between 420 b.c. and
a.d. 460 (Ibid.: Table 1).
Based upon the new chronological
information from Manachaqui, it is reasonable to suppose
that Gran Pajaten's collared bowls are temporally associated
with the set of four early dates {420 b.c., 250 b.c., a.d.
20, a.d. 40) recovered within Building No. 1's construction
fill.
Colpar Phase Shape E Rim 3 open bowls and Gran
Pajaten's simple bowls (Ibid.: Fig. 131, u) are also alike,
but such common bowl shapes are less useful for comparisons.
More significant are resemblances between Colpar Shape E Rim
433
13 and Gran Pajaten beveled-rim bowls (Ibid.: Fig. 40b).
The Colpar assemblage apparently contains numerous vessel
shapes that are absent at Gran Pajaten, and Gran Pajaten
yielded no evidence of notched applique decoration.
Nevertheless, the similarity between the Colpar and preAbiseo Phase assemblages is close enough to consider them
local variants of a single stylistic tradition.
The montane forest site of Cuelap farther north may
have been occupied this early, but so far lacks absolute
dates for the Early Intermediate Period Cancharin Phase
(Ruiz 1972).
Cancharin deposits predating Cuelap's
impressive architecture include open bowls with beveled rims
similar to Colpar Shape E Rim 13 (Ibid.: Lam. XXXIVa-d).
These rims and some bowl fragments with white-on-red and
negative resist decoration may indicate that at least some
of the Cancharin Phase collection dates to the first half of
the Early Intermediate Period.
However, most of the
Cancharin Phase materials, especially the jar shapes, are
more akin to later Early Intermediate Period pottery
assigned to Manachaqui's Empedrada Phase.
Future
investigations at Cuelap may permit sub-division of the
Cancharin Phase.
Looking west from Manachaqui to the Central Andean
coast, the Puerto Moorin and Gallinazo Periods in the Viru
valley should be roughly coeval with the Colpar Phase
(Strong and Evans 1952; Fogel 1992).
Huacapongo Polished
434
Plain Shapes 3/4 and 5 represent jars with reinforced rims
somewhat analogous to the Colpar variety.
Shape 4 persists
into the subsequent Gallinazo Period (Strong and Evans 1952:
Figs. 37, 40).
Thus, coeval north coast assemblages do
share reinforced rim forms with the Colpar Style.
However,
the associated vessels are very large and ovoid in shape
with mouth diameters ranging from 20 to 40 em during Puerto
Moorin, and 15 to 40 em during Gallinazo.
Unlike Colpar,
the Virli reinforced rims only occur on a limited range of
vessel shapes during each period.
Like Puerto Moorin pottery, Salinar Phase ceramics from
the site of Cerro Arena in the Moche Valley are
characterized by the white-on-red technique.
The inventory
of vessel shapes illustrated by Brennan (1978: Fig. B-1)
includes neckless ollas and jars with both short and long
necks.
While the short-necked variety might be considered
analogous to the Colpar jars, their everted rims are
unthickened.
In sum, most of the features that typify Early
Intermediate Period coastal assemblages are absent in
Manachaqui's Colpar Phase Paste A assemblage.
These include
the white-on-red technique, negative resist decoration
typical of Gallinazo, strap handles, neckless jars and jars
with long, flaring rims.
Thus, the overall degree of
similarity between the Colpar and coastal assemblages is
negligible.
In the highlands separating Manachaqui Cave and the
435
coast, the precise definition and dating of early sites and
ceramic assemblages in Huamachuco, Santiago de Chuco and the
upper Chicama valley remains problematic due to the dearth
of excavated and radiocarbon-dated stratigraphic sequences.
The recent Huamachuco investigations by T. Topic and J.
Topic (1987:23) substantiate Thatcher's (1979) Sausagocha
Phase which opens the Early Intermediate Period.
The
corresponding ceramic assemblage is dominated by neckless
ollas and open bowls
(T. Topic and J. Topic 1987:23).
Thatcher (1979: Figs. 33-35) illustrates jars with long,
flaring rims.
Colpar-like short-necked jars appear to be uncommon or
altogether missing in Huamachuco at this time.
Short-neck
jar rims from McCown's (1944: Fig. 20q, r) excavations at
the Sausagocha Phase settlement of Cerro
Ca~pana
East, and
from Thatcher's Mamorco Phase collections (1972-74: Fig. 5oq) are heavily reinforced and more massive than Colpar Phase
examples.
Only during the Sausagocha Phase does rim
notching become important in Huamachuco.
The Colpar style
notched applique decoration seems to be entirely absent.
Red slip is common to both Colpar and Sausagocha
assemblages, yet neither exhibit incised decoration nor
white-on-red decoration.
Nevertheless, it would seem that
the only significant parallel between the two assemblages is
found in the co-occurrence of beveled-rim bowls.
Rims from
Cerro Campana East parallel Shape C Rim 12 bowls (cf. McCown
436
1945: Fig. 20qq), Shape E Rim 13 (Ibid.: Fig. 2lf, t) and
Shape E Rim 15 (Ibid.: Fig. 21d).
In each case the
similarity is precise and, therefore, significant despite
the many differences between the Sausagocha and Colpar
assemblages.
In the neighboring highlands of the upper Chicama
valley, Krzanowski (1986: Table 11.1, 1986: Table 1)
characterizes the Totorapamba Phase as a local variant of
the Sausagocha Phase, and the final phase of the Early
Horizon Pelon Tradition.
The Totorapamba pottery has not
yet been fully reported, but Krzanowski (1986:245) remarks
that it "is almost identical to that of the Sausagocha
Phase" as Thatcher defined it.
Farther south in the upper Santa drainage, the Quinu
Period at Pashash (Grieder 1978) has not been directly
dated, but it is probably coeval with the later portion of
the Colpar Phase.
Vessel shapes include the neckless olla
and some necked jars devoid of similarity to Colpar Phase
shapes.
An isolated, yet arresting parallel is again the
presence of beveled-rim bowls including two forms which
resemble Colpar Shape E Rim 13 (Ibid.: Fig. 341) and Shape C
Rim 14 (Ibid.: Fig. 34m).
These are the same rim forms
reported from the eastern montane forest sites, and
Huamachuco collections just described.
Beveled-rim bowls resembling Colpar Shape E Rim 13
appear a short distance south of Pashash at La Pampa.
437
Terada (1979:179) proposes a sub-division of the undated
Tornapampa Phase into early and late components.
The
beveled-rim bowls (Ibid.: Pl. 98-6, 14; 99-6, 7, 9; 101-5)
belong to the earlier component which also features whiteon-red decoration recalling the Huaras style at Chavin de
Huantar.
Bowls and cups with beveled rims date to the late
Initial Period Urabarriu Phase at Chavin de Huantar.
None
of these, not even the later Janabarriu Phase open bowls and
cups, exhibit rim profiles like Shape E Rim 13.
However,
Janabarriu Jar 10 rim profiles (Burger 1984b: Fig. 293-295)
are identical to Suitacocha Shape B Rim llc, and Colpar
Shape B Rims llh and llj necked jar rims.
Some other
Janabarriu bowl rims with beveled lips somewhat resemble
Colpar Shape C Rim 12.
None of the Chavin decorative
features appear in the Colpar assemblage.
The Colpar Phase's closest overall stylistic
affiliations with Central Andean assemblages found to the
west are once again in Cajamarca.
At a glance these
similarities are not readily apparent as Layzon Phase
ceramics are best known for distinctive red-on-white (as
opposed to the more widespread white-on-red) painted
decorations that appear abruptly, yet remain absent in
Colpar Paste A pottery.
Also unlike the Colpar Style,
Layzon retains notched rims and incised decoration from
earlier phases.
In fact,
the only decorative techniques
438
shared by the two styles are the application of red slip to
vessel rims and the limited use of notched applique (Terada
and Onuki 1985:92).
Despite the contrasting decorative traditions, a
comparison of vessel shapes reveals remarkably close
correspondences.
Layz6n neckless ollas occur in greatly
diminished numbers, and the shape inventory now features
globular jars with long, flaring rims, short-necked jars
with short rims, wide-mouthed jars and open, semispherical
bowls (Ibid.).
There are clear morphological parallels
between Colpar short-necked jars and Layz6n wide-mouth jars.
In particular, Colpar Shape B Rim 11h jars find counterparts
in Layz6n Red-on-White and Layz6n Red Slipped Form 4 jars
(Ibid.: Pl. 67-5 and Pl. 68-17).
Colpar Shape BRim 11i
mirrors Layz6n Red-on-White Form 4 jars (Ibid.: Pl. 67-9),
while Shape B Rim 11j jars resemble Layz6n Brushed Form 2
wide-mouth jars (Ibid.: Pl. 70-3).
Also indicative of stylistic relationships are the
Huacaloma Black Polished Form 10, Huacalorna Red Polished
Form 7 and Huacaloma Brown Polished Form 3 open bowl
variants with reinforced lips (Terada and Onuki 1985: Pl.
59-15, 59-28, 60-5) identical to Colpar Shape E Rim 13.
According to Terada and Onuki (Ibid.: Tables 9-11, 28-30,
53-55, etc.), these shapes occur during the Late Huacaloma
Phase and persist through the EL Phase.
The EL Phase is
coeval with the Chavin horizon (Onuki 1993: Fig. 15).
The
439
common Layz6n convex bowl shapes resemble Colpar Shape C Rim
15 and Shape E Rim 9, but again, these shapes are not
particularly diagnostic.
Pottery of the following Initial Cajamarca Phase is
said to retain much of Layz6n's stylistic character with the
addition of fine, slipped and painted kaolin ware bowls.
However,
the Colpar-like short-necked and wide-mouthed jars
drop out of the sequence.
Initial Cajamarca non-kaolin
ceramics feature neckless ollas and globular jars with
either "swollen" rims (Terada and Onuki 1982:122-123, Pl.9411-14) or long, flaring rims (Ibid.: Pl. 95-1-9).
These two
jar shapes persist into subsequent phases of what Matsumoto
(1993) terms the "Cajamarca Tradition."
It is Layz6n that
lays a foundation for this tradition characterized by a
distinctive set of pottery attributes that persist
throughout Cajamarca's Early Intermediate Period.
The only
Cajamarca Tradition non-kaolin shape with a Colpar analogue
is the open or slightly restricted bowl with beveled rim
(Terada and Onuki 1982: Pl. 91-1, 2).
Also in Cajamarca and showing some remarkably similar
features with Colpar is the Pandanche Phase CII ceramic
assemblage from the highlands farther north.
In contrast to
Layz6n, earlier incised-line techniques disappear after
Phase CI and the Chavin horizon (Kaulicke 1975:47-48).
It
is in Pandanche Phase CI that bowls with beveled rims like
Colpar Shape E Rim 13 and others previously cited occur
440
(Ibid.: Lam. XIX: top row, fourth and fifth from left).
These do not persist into Pandanche Phase CII, and they are
conspicuously missing from the Layz6n assemblages described
above.
Wide incisions and red lines rendered in pre-firing
paint decorate Phase CII vessels, most of which are
reportedly globular jars with vertical necks.
Like Colpar's
however, Phase CII rims are reinforced on the lip exterior,
and occasionally beveled.
Some short-necked jar rims have
clear Colpar Phase counterparts (e.g. Ibid.: Lam. XXI top
row third from left, bottom row third from left).
neither mentions nor illustrates neckless ollas.
Kaulicke
That he
does not differentiate between jars and bowls in his
descriptions may indicate the same ambiguities present at
Manachaqui Cave and Gran Pajaten.
Phase CII is the final
phase in Kaulicke's published sequence.
Finally, the Colpar
Phase Shape E Rim 13 beveled-rim bowl also finds an analogue
at the western montane forest site of Poro-Poro (Alva 1988a:
Figs. 38-17).
On the eastern slopes of Huanuco farther south, the
Kotosh Sajara-patac Phase pottery displays few resemblances
to Colpar Phase ceramics.
Sajara-patac vessel shapes
include neckless ollas and open bowls.
vertical or flaring rims.
Jars have short
Izumi and Terada (1972: Pl. 102)
illustrate a Sajara-patac jar strikingly similar to
Suitacocha Phase Shape F jars with incised decorations laid
441
out in panels on the shoulders.
However, this shape has
already come and gone at Manachaqui Cave.
Sajara-patac
"short-necked jars" exhibit longer necks than their Colpar
counterparts, and they lack the reinforced lip.
Sajara-patac broad-line incision, punctation and
applique techniques are absent in the Colpar assemblage,
while Kotosh potters employed no painted decoration other
than an occasional red slip.
The continuing tendency to
bevel vessel rims remains one of the few pottery attributes
common to both styles.
In fact, beveled-rim bowls are not
only ubiquitous during the Kotosh Sajara-patac Phase (e.g.
Ibid.: Pl. 100-5, 105-16-24), but they have a long history
that may pre-date the Chavin horizon (e.g.
Ibid.: Pl. 118-
11-17).
The following Kotosh Higueras Phase at Kotosh shows
even less similarity to Colpar pottery than Kotosh Sajara
Patac.
Onuki (1993: Fig. 15) dates the beginning of the
phase at A.D. 1, and the argument that the phase ushers in
new, different and perhaps "foreign" cultural patterns at
Kotosh is well-known (Izumi and Terada 1972:205, 311;
Lathrap 1970:173; Isbell 1974).
The beginning of the Kotosh
Higueras Phase signals the introduction of new pastes,
vessel shapes and decorative techniques.
Globular jars with
long flaring rims (Izumi and Terada 1972: Pl. 98, 99-1-5),
strap handles, notched applique and white-on-red and
negative resist decorations co-occur.
Of these attributes,
442
only applique characterizes Manachaqui's Colpar Paste A
assemblage.
To recapitulate, the greatest degree of overall
similarity is found between the Colpar Phase assemblage and
the earliest ceramics from nearby Gran Pajaten.
Colpar's
relationship with Layzon is difficult to evaluate given that
the dissimilarities between their respective repertoires of
decorative techniques are as striking as the similarities
between their respective inventories of vessel shapes.
It
is more difficult to lend substantial weight to the
parallels in jar rim morphology shared with Janabarriu Phase
Chavin de Huantar.
On the other hand,
they are difficult to
dismiss given the absence of parallels between Colpar Shape
B rims and jar rims at any other Central Andean site south
of Cajamarca.
During the Colpar Phase another indication of
linkage between Manachaqui Cave and neighboring highland
groups to the west is the presence of the distinctive
beveled-rim bowls.
The appearance of these bowls within
north highland sequences serves as a marker indicating
cross-dates, or temporal overlap, between the pre-Abiseo
Phase assemblage, Cancharin Phase, Sausagocha Phase, Quinu
Period, Tornapampa Phase, Early Cajamarca Phase, Pandanche
Phase CI and Kotosh Sajara-patac Phase as each is defined by
its respective author.
Again it is also difficult to
evaluate the significance of these parallels, since this rim
profile cross-cuts otherwise disparate stylistic traditions.
443
Paste A Relationships: the
Amazonian Lowlands
In the Amazonian lowlands to the east, Barrancoid
pottery has been found over a wide geographic area extending
from the lower Orinoco in the north to the Central Ucayali
in the south.
Lathrap dates the arrival of populations
bearing the Barrancoid Style to 200 b.c. on the Central
Ucayali {1970:117}.
Lathrap's Barrancoid Hupa-iya Phase
assemblage includes wide-mouthed vessels with flaring,
unthickened collars {1962: Figs. 84, 85}.
dominate the vessel shape inventory.
Open bowls
Neither shapes, nor
the typical incised and modelled Barrancoid decorations
suggest a relationship to Manachaqui's Colpar Style.
Lathrap reports that, by a.d. 100 the Central Ucayali
Barrancoid populations populations were replaced by
producers of the Yarinacocha Style.
The predominantly open,
neckless vessels of the subsequent Yarinacocha Phase
likewise show no relationship to Colpar pottery.
Morales
{1992} excavated Barrancoid pottery in northeastern Peru's
Chambira River valley but his Siamba Phase assemblage has
not yet been fully described or dated.
Ravines'
{1978} brief field campaign in the Central
Huallaga valley directly east of Manachaqui Cave did not
uncover any traces of a Barrancoid presence.
His
stratigraphic sequence includes two related ceramic
complexes that overlay the Shakimu remains, but still lack
absolute dates.
The Huayabamba Complexes 1 and 2 include
444
short-necked "ollas" with everted, reinforced rims (e.g.
Ibid.: Lam. 4b).
However, a cross-date with the later
Empedrada Phase to be described in Chapter 11 appears more
likely.
At the foot of the Andes to the south, globular,
necked jars of the Nazaratequi phase (Allen 1968: Fig. 16,
Forms 12 and 13) cited in the previous chapter may be
contemporaneous with the Colpar Phase and suggest continued
ties to the west and /or north.
In the Amazonian lowlands north of Manachaqui Cave,
Shady's (1987b) Bagua sequence concludes with the El Salado
Phase that she dates to the end of the Early Horizon.
She
reports that El Salado vessels frequently bear exterior
reinforced rims painted red and occasionally exhibiting an
incised groove beneath the lip (Ibid.:476).
In profile, the
two rims illustrated in the upper right corner of Fig. 10
are not unlike Colpar Phase Shape BRim 15 examples.
Shady
(1987b:474) notes that such rims are also popular in the
forested Tabaconas valley of Jaen where they appear to have
been reinforced by the fold-over technique (e.g. Miasta
1979: Lams. 9a, 12a, 24c, 31b, 45f, SOb).
Some of the
Tabaconas rims (e.g. Ibid.: Lams. 28b, 57a, 61d) resemble
those Colpar restricted bowl Shape C Rim 11.
Miasta's
undated assemblages also contain vessel shapes comparable to
earlier Suitacocha Phase Shape F jars with vertical necks
(e.g.
Ibid.: Lams. 39a, 49b, 59e).
The wide diversity of decorative techniques utilized in
445
the Bagua-Jaen region and elsewhere across the Amazon Basin
might support arguments against relationships to the
seemingly austere Colpar Style.
Farther north in the
Ecuador Oriente, none of the styles potentially coeval with
the Colpar Phase (e.g. Sangay, Pastaza, Kamihun, Tivacundo)
show similar decorative or formal attributes.
More detailed
information regarding vessel shapes is not yet available for
the ceramic complexes of the Bagua-Jaen region, and the lack
of absolute dates likewise hinders efforts to relate them to
Colpar remains.
However, comparisons utilizing available
descriptions and illustrations indicate that Colpar Phase
ceramics maintained a stylistic relationship with styles at
the base of the eastern Andes directly to the northeast that
is primarily reflected in distinctive rim treatments.
Paste A Relationships: the
Northern Andes
In the lowland regions of extreme northern Peru,
archaeological assemblages coeval with the Colpar Phase
continue to display Northern Andean features.
In the Piura
River valley, assemblages of the coastal Sechura Phases A-D
and mid-valley Chapica, Tamarindo A and Tamarindo B Phases
are now dominated by jars with flaring rims, and vertical
rims like the earlier Suitacocha Shape F examples (Lanning
1963: Fig. 23a-c, f, g, t; Ravines 1986-87: Lam. 5; Guffroy
1989: Fig. 7e, f; Kaulicke 1991: Figs. 14, 15).
exhibit angled joints at the base of the neck.
Many
Lanning
446
(1963:209-210) regards white-on-red and negative resist
decorative techniques in Sechura A through D as
introductions from southern Ecuador.
These same techniques
characterize assemblages from the Tumbes area where Izumi
and Terada (1966:81, 88) conclude that the closest stylistic
affiliations of the Garbanzal style are found in the
Ecuadorian Jambeli and Guangala styles.
Precise similarities between the Colpar Phase Paste A
assemblage and the Piura and Tumbes collections are few.
In
Piura, notched and plain applique strips can be observed on
body sherds until the end of Sechura B (Lanning 1963: Fig.
22:i-m; also Guffroy 1992:110).
Perhaps most significant is
the presence of bowls with beveled rims during Sechura
Phases C and D (Lanning 1963: Figs. 11i, 12a-b), and perhaps
as early as Sechura A (Ravines 1986-87: Lam. 3-16),
identical to those described from the Peruvian north
highlands and montane forest earlier in this chapter.
They
also occur during the mid-valley Tamarindo Phase B (Kaulicke
1991: Fig. 15 lower right) which Kaulicke likens to Sechura
PhaseD (Ibid.:417).
From Loja in the southern Ecuadorian interior, the
reinforced rims on necked vessels (Lecoq 1987) bear a
general resemblance to those of the Colpar Phase, but the
necks tend to be longer and Loja decorative techniques
continue to emphasize incision.
Lecoq points to Late Cerro
Narrio and Jambeli as Loja's closest stylistic affiliates.
447
The Late Cerro Narrio assemblage from the southern
Ecuadorian highlands is likely coeval with the Colpar Phase.
It includes necked jars, bowls and compoteras.
Most of the
jar shapes feature unthickened short and long everted rims
(Collier and Murra 1943: Fig. 10 bottom row, Fig. 14).
One
of the fifteen illustrated jar profiles closely matches
Colpar Shape BRim 11i or 11j
third from left).
(Ibid.: Fig. 10 bottom row,
Otherwise, the Late Cerro Narrio
assemblage shows little resemblance to the Colpar pottery.
Guangala and Jambeli are North Andean archaeological
cultures assigned to the Regional Developmental Period in
Meggers'
(1966) schematization of Ecuadorian prehistory.
She perceives these assemblages as outgrowths of a Chorrera
stylistic heritage.
Necked jars, convex bowls, compoteras
and vessels with multiple leg supports typify many
collections (Ibid.: Fig. 18).
White-on-red and negative
resist decorative techniques are characteristic of Guangala
(Bushnell 1951) and Jambeli (Estrada et al. 1964) pottery.
The only close morphological analogies to Colpar shapes
might be found in the Early Guangala beveled-rim bowls'
likeness to Colpar Shape E Rim 13 (Bischof 1982: Fig. 6d;
Stothert 1993: Fig. 55e).
These bowls with beveled lips and
grooved exteriors are especially numerous in the
southwestern Ecuadorian lowlands where early Engoroy
(Chorrera) contexts indicate temporal priority (cf. Bischof
1982: Fig. 2a, f,
3b, d).
448
Of all of the reported northern assemblages reviewed
for this study, Porras'
(1975a) Cosanga collections from the
montane forest east of modern-day Quito bear the strongest
overall resemblances to the Colpar Phase assemblage.
Porras
(Ibid.:189) dates the Cosanga "Phase" to the centuries
between 500 b.c. and a.d. 800.
Although there are no
indications of Cosanga-type compoteras at Manachaqui,
the
rim profiles illustrated for Shapes 3, 5 and 9 (Ibid.: Figs.
19 and 20) closely parallel Colpar Shape C Rims 12, 15 and
11 respectively.
Although the Colpar Shape B rims find no
precise matches among the Cosanga jar rims, the fold-over
technique of rim reinforcement is a particularly diagnostic
shared element.
This technique is so distinctive that,
coupled with the bowl rim resemblances, a stylistic
relationship between the Colpar and Cosanga assemblages is
clearly indicated.
Paste B Origins
The suspected sources of Colpar Phase Paste Group B
ceramics are not so far-flung as Suitacocha Phase sources,
but when considered as a group, they suggest emerging
patterns of interaction that foreshadow the intensive
interaction of the subsequent Empedrada Phase.
Again,
confirmation of suspected sources awaits substantiation
through additional analyses.
449
Paste B6
Open bowls with modeled rims are found throughout the
north-Central Andean highlands, in far northern coastal Peru
and in southern Ecuador during this time period.
McCown
{1945: 21g) illustrates a rim form similar to Paste B 6 's
{Fig. 93a) among a collection of modeled rims from Cerro
Campana East, but offers no specifics regarding its paste
characteristics.
During his surface survey, Thatcher
recovered several sherds of a distinctive pottery from Cerro
Campana East and other sites with Mamorco and Sausagocha
Phase remains:
The paste cross section is a brick-red color with inclusions
that look like gravel, including a noticeably large
percentage of mica. The inclusions range up to more than 2
mm. in size and make up 40-50% of the rather soft paste.
The surface is the same color as the cross section and the
inclusions are quite noticeable.
The surface is grainy and
matte. No decoration is present (Thatcher 1979:92-93).
That these sherds are intrusive at Thatcher's sites is
implied by their scarcity.
Paste B6 's place of origin
remains unknown, but a Peruvian north highland source seems
likely.
Paste B7
Paste B7 sherds {especially Fig. 93f) unite features
diagnostic of a classic Chorrera or Guangala bowl decorated
with iridescent paint.
The profiles of these southern
Ecuadorian beveled rims are comparable to Colpar Shape E Rim
13 and, by extension, to the northern examples cited within
the above discussion of Colpar Paste A relations.
A group
450
of four non-local sherds of a fine, black micaceous pottery
from Gran Pajaten's pre-Abiseo Phase assemblage also consist
of Paste B7
(Church 1988:216-217).
A single rim sherd
represents a simple, convex bowl with unthickened rim and
rounded lip (Ibid.: Fig. 51b).
An Ecuadorian source is suggested by the presence of
iridescent paint and the practice of smudging, both made
popular during Chorrera.
Bowls with beveled rims and
iridescent paint persist during Guangala (Bischof 1982;
Stothert 1993).
Vessels with micaceous paste at Early
Guangala Phase sites have been considered intrusive (Paulsen
1970:72), or fabricated with imported micaceous tempering
material (Masucci 1992).
It is difficult, therefore,
to
offer a more precise source location than "southwestern
Ecuador."
Paste B8
The Paste B8 sherds include the only example of whiteon-red decoration at Manachaqui Cave (Fig. 93g), although
nothing truly identical could be located among other
published collections of white-on-red style pottery.
The
motif of thin, roughly parallel lines looping below the rims
of simple convex bowls is common in Cajamarca Layz6n Phase
collections
However,
(Terada and Onuki 1985: Pl. 67-11, 67-12).
in Layz6n the motif is executed in red paint on a
white slip.
From Huamachuco, Thatcher (1979: Fig. 38)
illustrates a Sausagocha Phase bowl with the same motif
451
rendered in red paint on orange slip.
The carinated bowl
shape (Fig. 94a) also occurs in Huamachuco (McCown 1945:
Fig. 19e).
These Paste B8 sherds probably all originated in
the neighboring highlands.
Paste B9
The literature search failed to turn up a match for the
rims with red painted geometric designs representing Paste
B9 (Fig. 94c-e).
Similar motifs rendered in white diagonal
and crossing painted lines and dots appear on red Garbanzal
bowl rims (Izumi and Terada 1966: Pl. 13a-15 and 16) and jar
rims (Ibid.: Pl. 13b-8) from the far north coast of Peru.
Perhaps these sherds originated in extreme northern Peru.
Paste B10
Both McCown (1945: Fig. 17h, j) and Thatcher (1979:
Fig. 42) illustrate rims with profiles resembling the
distinctive Paste B10 rim (Fig. 94f), but in their
Sausagocha Phase assemblages the decoration is executed with
red paint on a light orange paste.
The Paste B10 sherd
probably represents pottery brought from neighboring
highlands.
Paste C1 Origins
The Colpar Phase Paste C, rim sherd (Fig. 94g) is
identical in both form and decoration to Layzon Red-on-White
pottery from Cajamarca (cf. Terada and Onuki 1982: Pl. 8912, 13 and 14).
However, neither the Layzon Phase wares,
452
nor similar red-on-white pottery at Cerro Arena (Brennan
1978:609; Mujica 1984:12), are fashioned in pure kaolin
paste although exotic kaolin sherds do occur at Cerro Arena
(Brennan 1978:599).
Because of their isolated occurrence on
the coast at Cerro Arena, Mujica argues that the red-onwhite style ceramics are imports from Cajarnarca.
Despite
the difference in clays employed for Layz6n Red-on-white and
the C1 sherds, it still appears likely that the Colpar Phase
examples were brought from Cajarnarca.
Ceramic Evidence for Colpar
Phase Chronology
At present, the Colpar Phase is envisioned as the span
of centuries between 400 b.c. and a.d. 200 during which
Manachaqui Cave was only sporadically used.
Manachaqui
Cave's artifact assemblages lack attributes suggestive of
participation in the Chavin cult, and the isolated match
between Colpar and Janabarriu jar rims appears to be weak
evidence of stylistic relationship.
However, this popular
rim form "with no antecedent for its short everted neck" is
either "innovative"
(Burger 1984b: 127), or intrusive,
perhaps as a form borrowed from eastern montane forest
pottery styles like Suitacocha or Colpar.
Note that this
co-occurrence does not necessarily imply a Chavin horizon
cross-date for the Colpar Phase.
On the other hand, a
cross-date suggested by Colpar Shape E Rim 12's EL Phase
counterpart in Cajamarca does suggest Manachaqui Cave's use
453
during the Chavin horizon.
Shape E Rim 13 beveled-rim bowls
also appear in the Central Andes during the Chavin horizon,
but because the shape persists into the Early Intermediate
Period, this evidence for Manachaqui Cave's use during the
Chavin horizon remains less conclusive.
Colpar Phase ceramic evidence, especially the close
formal correspondences to post-Chavin horizon Cajamarca
assemblages, most clearly indicates a beginning date of 200
b.c.
It appears likely that the shelter was also
infrequently utilized during the Chavin horizon,
hypothetically advancing the beginning of the Colpar Phase
to 400 b.c.
Estimating the end of the Colpar Phase with
ceramic evidence is also problematic, mostly because the
nature and timing of style changes in the neighboring
highlands during the Early Intermediate Period remain poorly
understood.
Because a paucity of fine kaolin ware pottery
seems to characterize this phase, it is unlikely that it
extends beyond a.d. 200.
Radiocarbon dates for the
subsequent Empedrada Phase range from a.d. 450 to 570.
Whether or not the Colpar and Empedrada Phases are separated
by an occupational hiatus will be considered in the next
chapter.
CHAPTER 9
THE EMPEDRADA PHASE
Cultural occupations at Manachaqui Cave during the midEarly Intermediate Period are assigned to the Empedrada
Phase.
As described in Chapter 5, the phase is
distinguished in the ceramic macrochronology by the abundant
remains of fine, painted bowls fashioned from kaolin-rich
clays.
Again, levels from the deep Sector B berm deposits
comprise the most intact stratigraphic contexts with which
to isolate the Empedrada Phase materials.
Representative
and relatively unmixed samples come from Levels 4 through 7
in Units 27 and 28; and Levels 5 through 8 in Units 30, 31
and 32.
Empedrada Phase deposits occur virtually at the
surface near the mouth of the shelter (e.g. Units 20 and 21}
where the 1986 "cleaning" activities removed the uppermost
soil layers.
In Sector A, numerous compact, superimposed floors rich
in organic contents (Table 1, Stratum 2A} attest to the
intensive utilization of Manachaqui Cave at this time.
These floors contain the best-preserved samples of ceramics
and organic remains, and charcoal from Sector A's hearths
rendered all of the Empedrada Phase radiocarbon dates.
Future radiocarbon analyses will target the kaolin ware454
455
bearing layers in Sector B in order to correlate the
stratigraphic sequences from the two sectors, thereby
confirming or modifying the absolute dates proposed for this
and other Manachaqui Cave phases.
An estimated total of 5,977 artifacts constitute the
Ernpedrada Phase collection.
The inventory of non-ceramic
remains is more diverse than those of previous phases at the
shelter.
The large sample of lithic remains now includes an
assortment of tool types.
The full range of Manachaqui's
faunal and botanical remains are found in Ernpedrada Phase
strata.
Empedrada Phase Ceramics
A total of 1227 diagnostic sherds represents the
Ernpedrada Phase.
The transition from the Colpar to the
Ernpedrada Phase marks the most substantial stylistic shift
in Manachaqui's ceramic sequence.
A nearly two thousand
year-old pottery tradition characterized by short-necked
jars and restricted bowl shapes ends abruptly, and is
replaced by new vessel shapes such as globular jars with
long,
flaring rims and compound jars with bowl-shaped rims.
The restricted bowl shape drops out entirely, and bowl rims
rise to nearly 50 percent of the rim assemblage.
The
Ernpedrada Phase ceramic assemblage shows fewer shape
categories than any previous phase assemblage at Manachaqui.
Empedrada Phase jars with flaring rims are
morphologically, and perhaps functionally, analogous to the
456
Suitacocha Shape F jar shape.
Shape F vanishes from the
Manachaqui Cave shape inventory by the beginning of the
Colpar Phase, but its disappearance might be attributed to
modifications in the shelter's functions over time.
It
should be kept in mind that the Colpar Phase sample is
small, and Shape F constituted only a fraction of the jar
shapes during the earlier Suitacocha Phase.
In other words,
Empedrada Phase jars might be functional and stylistic
outgrowths of Shape F jars.
The ancient tradition of fold-
over rim reinforcement persists in variants of the Empedrada
Phase jars.
Also enduring virtually unchanged are
decorative techniques featuring only red slip or paint, and
notched applique strips.
Again, severe weathering of the
ceramic sample, especially in Sector B, frustrated aspects
of the analysis.
The assemblages of Paste Group B and Paste Group C
pottery expand to unprecedented size and diversity during
the Empedrada Phase.
In sherd frequencies the Paste Group C
now eclipses Paste Group B.
The identification and
chronological placement of these sherds met with the same
problems previously described for the eroded assemblages of
the previous phases.
The most intact sherd surfaces are
found in Sector A, while the largest samples and most intact
stratigraphy are found in Sector B.
The weathered condition
of the Sector B sample frustrates endeavors to subdivide the
Empedrada Phase.
457
Ceramic Paste Group A
The Empedrada Phase Paste Group A assemblage consists
of 359 diagnostic sherds, representing 29.3 percent of the
phase diagnostic sherd collection.
During this phase, Paste
Group A encompasses an unprecedented diversity of pastes or
wares.
Again, the basic ceramic technology changes little.
Microprobe analysis of one Empedrada Phase rim sherd enabled
identification of angular grains of quartz, alkali feldspar,
hornblende and ilmenite.
It would seem that, although the
selection and processing of clays and non-plastic additives
changes from phase to phase, the criteria for raw material
selection remains the same.
The impression gleaned from
handling the entire Manachaqui Cave ceramic collection is
that the number of distinct pastes that constitute Paste
Group A increases with each phase.
The Empedrada Phase
represents a burst of paste diversity second only to the
extraordinary profusion contained in the Late Horizon
Poblano Phase Paste A assemblage.
Morphology
A total of 342 sherds rendered information with which
to reconstruct Empedrada Phase Paste Group A pottery shapes,
338 of which are rims.
The shapes of bowls are typically
semi-spherical and jar bodies are globular.
There is still
no evidence for vessel bases that are not rounded.
458
Shape E: Unrestricted convex bowls
A total of 163 Shape E rim sherds represent 48.2
percent of the Empedrada Phase Paste A rim collection (Table
16).
Most are unthickened or slightly thickened.
None of
the bowl rims are reinforced in the manner of the previous
phases, and beveled varieties now constitute a small
minority.
Rim 3 and Rim 9 persist from the Colpar Phase
(Figs. 95a-c and 95d-k respectively).
Together the
Empedrada Phase bowls range in diameter from 6 to 23 em
(N=79), although one bowl diameter reaches 30 em.
The
diameters average 14.22 em with a standard deviation of
3.68.
R~
15 (Fig. 951-s).
The 26 examples of Rim 15 belong
to semi-spherical bowls with slightly thickened, unthickened
and tapering rims that end in rounded and semi-flattened
lips.
The form is identical to Rim 9 except that it is
slightly in-curving.
Walls range from 4 to 7 mm in
thickness.
R~
16 (Fig. 95t-w).
The 14 examples of Rim 16 belong
to semi-spherical bowls with unthickened rims and flattened
lips.
Wall thickness ranges from 4 to 5 mm.
R~
17 (Fig. 96a-d).
The 10 examples of Rim 17 pertain
to carinated bowls with vertical or nearly vertical rims and
beveled, in-sloping lips.
The walls are 4 to 5 mm thick.
459
R~
18 (Fig. 96e-h).
Nine examples of Rim 18 represent
bowls with diverging walls and unthickened, direct rims.
The lips are rounded.
The vessel walls range from 4 to 8 mm
thick.
R~
19 {Fig. 96i-k).
The six examples of Rim 19 belong
to relatively shallow open bowls with unthickened rims
flattened at the lip.
R~
Wall thickness ranges from 3 to 6 mm.
20 {Fig. 961-n).
Four Rim 20 sherds are from open
bowls with unthickened, sharply inverted rims.
The lips are
rounded and wall thickness varies from 3.5 to 4.5 mm.
R~
21 {Fig. 96o).
The four Rim 21 examples all
probably belong to the same large bowl with a 30 em
diameter.
The rims are vertical and unthickened with
flattened lips.
R~
The maximum wall thickness is 8 mm.
22 {Fig. 96p, q).
Three Rim 22 examples represent
relatively shallow bowls with divergent walls and thickened,
bulbous rims.
Wall thickness is 3 rnm, but maximum thickness
at the rounded lips reaches 7mm.
R~
23 {Fig. 96r).
Two eroded rims may belong to the
same bowl of unknown diameter.
They are thickened to 4 mm
on the exterior surface and near the lip which ends in a
dull point.
R~
24 {Fig. 96s).
These two eroded rims may also
460
belong to a single bowl with unknown diameter.
The walls
reach 10 mm in thickness and the unthickened rim ends at a
rounded lip.
Shape G: Jars with Long, Flaring Rims
A total of 152 Shape G rims (45 percent of the
Empedrada Phase rim collection) pertain to globular jars
with short, constricted necks and long, flaring rims.
The
jar rims are separated into 11 categories based upon
morphological characteristics (Table 17).
Many are
thickened, and some even exhibit the folded over, reinforced
lip so characteristic of Paste Group A pottery of previous
phases.
Overall, the vessel orifices vary from 7 to 24 em
in diameter, and show great size variability.
Collectively
they (N=69) average 15.32 em in diameter with a standard
deviation of 3.80 em
Rim 1 (Fig. 97a-m).
The 56 examples of Rim 1 belong to
jars with long, unthickened rims.
Thickness varies from 4
to 9 mm. The lips are either rounded or form a dull point.
Some of the rims bear a flanged lobe (Fig. 97i,
Rim 2 (Fig. 98a-h).
j).
These 43 examples of Rim 2
represent jars with long, flaring rims that have been
reinforced by folding over the end.
The union of the folded
end and the rim exterior is typically left uneven and
unsmoothed.
Apparently the length of the rim from neck to
461
lip is highly variable.
Rim thickness below the folded
portion varies from 3 to 7 mm, and the reinforced end ranges
from 7 to 12 mm thick.
Rim 3 (Fig. 98i-k).
The 21 Rim 3 examples belong to
jars with gradually thickened, bulbous r1ms.
invariably rounded.
The lips are
Below the thickened portion, the rim
measures 3 to 4 mm in thickness, and the thickened end
measures 7 to 9 mm.
Rim 4 (Fig. 981-n).
Twelve sherds constitute Rim 4
which shows a thickened, rounded interior surface and a
nearly flat underside.
The lip ends in a dull point.
Maximum thickness reaches 8 mm.
Rim 5 (Fig. 98o).
Eight Rim 5 sherds pertain to jars
with thick, direct rims.
The illustrated rim tapers from 14
mm thick at its sharply angled neck to 6 mm at its semisquared lip.
Rim 6 (Fig. 98p, q).
Four rims represent one or more
jars with unthickened rims exhibiting slightly concave
interior surfaces.
The lip ends in a dull point.
Thickness
ranges from 4 to 5 mm.
Rim 7 (Fig. 98r).
The three Rim 7 sherds represent one
or more jars with unthickened rims.
The necks are sharply
angled and the lip has been beveled flat on top leaving a
462
dull point at the exterior edge.
Rim 8 (Fig. 98s, t).
Two rims, apparently from the
same jar are nearly identical to Rim 7, except that the end
has been reinforced and "squared" by beveling.
The rims are
4 mm thick and the reinforced ends measure 6 mm thick.
Rim 10 (Fig. 98u).
One Rim 10 sherd represents a jar
with an unthickened, direct rim.
The lip is rounded and rim
thickness is 8 mm.
Rim 11 (Fig. 98v).
One Rim 11 sherd belongs to a jar
with a straight neck approximately 2.5 em high and an
everted, unthickened, direct rim 1.5 em long.
The neck and
rim are both 5 mm thick.
Shape H: Jars with Bowl-shaped Rims
A total of 22 Shape H rims represent 6.5 percent of the
Empedrada Phase rim collection.
The Shape H category was
established to distinguish these compound vessels with bowlshaped rims (also commonly referred to as S-shaped) ,
although there is no evidence that body shapes differed from
those of Shape G jars.
Four rim form categories represent
variability in rim profiles and sizes.
Mouth diameters
range from 10 to 16 em, although one is 25 em.
Only nine
rim diameters were measured, and the mode and mean are both
approximately 14 em.
463
R~
1 (Fig. 99a-d).
Nineteen sherds represent Rim 1.
The carefully modeled S-shaped profiles display vertical or
slightly inverted upper walls and everted ends.
The lip of
one rim (Fig. 99d) is slightly thickened or "beaded."
The
others are generally unthickened, ranging from 3 to 7 rnm in
total thickness.
Six measured orifices exhibit diameters
from 10 to 16 em.
R~
2 (Fig. 99e).
The single Rim 2 sherd is identical
to Rim 1, but a mouth diameter of 25 em indicates a very
large vessel.
R~
The rim is 7 mm thick.
3 (Fig. 99f).
bowl-shaped profile.
rounded lip.
One Rim 3 sherd shows a simple,
The rim is unthickened and ends in a
The mouth diameter is 11 em, and the rim walls
are 5 mm thick.
R~
4 (Fig. 99g).
Rim 4 is represented by a single
bowl-shaped rim that has been slightly thickened to 4 rnm
approximately
1 ern below the lip.
The orifice measures 10
em in diameter.
Miscellaneous Shapes
Two of five sherds are believed to be broken spoon
handles (Fig. 100a, b).
Fig. 100a is 3.5 ern long,
The handle fragment illustrated in
1.1 ern thick and tipped with a
zoomorphic head featuring incised eyes, mouth and nostrils.
It probably represents a llama.
The other handle fragment
464
(Fig. lOOb) is 4.5 em long, 1.5 em at its thickest point,
and has a long, conical shape.
Three articulating sherds
represent an unknown object (Fig. lOOc).
It has a concave,
scoop-like shape.
Decoration
Only 23 sherds supply information on Empedrada Phase
decorative techniques.
Again, we can be certain that
painted decoration is under-represented due to the weathered
state of the collection.
Eight of nine body sherds with red
paint or traces of paint on their interior surfaces
apparently belong to bowls.
The illustrated example (Fig.
96h) shows that potters still preferred to apply the paint
to limited areas of the rim.
One Shape E Rim 16 bowl (Fig.
95t) exhibits red paint on the interior and white paint on
the exterior surface.
A single sherd bears traces of white
paint on the exterior surface.
Fourteen sherds from Empedrada Phase levels display
plastic decorative techniques.
Nine bear notched applique
strips identical to those of the Colpar Phase (Fig. lOOd,
e).
The surface of the sherd depicted in Fig. 100e shows
shallow incisions where the potter scratched the vessel
surface while "nicking" a discontinuous applique fillet,
probably with long, downward strokes.
Again, the applique
strips seem to have been affixed to jar mid-sections.
Shape GRim 1 jar rims carry notched lobes (Fig. 97i,
Two
j).
Three Shape G Rim 1 jar rims feature rows of small round
465
punctations 1 to 2 rom in diameter on their interior surfaces
(Fig. 97k-m) .
Ceramic Paste Group B
The Paste Group B category consists of 153 sherds
representing 12.5 percent of the Empedrada Phase collection
of diagnostic sherd collection.
paste sub-groups.
They are divided into eight
Eighty percent of the Paste Group B sherd
collection belongs to two sub-groups of particularly
micaceous pastes.
The remaining 20 percent consists of sub-
groups with between one and eight sherds.
Eleven (2.7%) of
the total 410 Empedrada Phase rims belong to Paste Group B.
Paste B 11
Ninety-one sherds representing Paste B11 may all belong
to a single, large necked jar.
Firing left the paste
incompletely oxidized as indicated by the dark grey core
sandwiched by light brown interior and exterior surfaces.
Sand was added as temper and the abundant mica could be a
natural constituent of the clay.
The thickened,
flaring rim
of the jar is nearly 7.5 em long, 11 rom thick near the lip,
and shows an orifice diameter of 30 em (Fig. 100f).
constricted neck is sharply angled.
exhibit traces of red paint.
Its
Unweathered sherds
The sherd interiors show
striations left by a bristled tool utilized to finish the
surface.
466
Paste Bu
Twenty-three heavily eroded sherds may have all
belonged to a single vessel, perhaps an open bowl with
everted rim (Fig. 100g).
The paste exterior is 5YR5/6
yellowish red, and a dark grey core shows incomplete
oxidation.
Mica and quartz are principle constituents of
the non-plastic inclusions that comprise more than 50
percent ·of the paste.
The great quantity of mica renders
the sherds very brittle, but when intact it must have lent
this vessel a spectacular glittery appearance.
The vessel
has a tapered, flaring rim, and the orifice measures 15 em
in diameter.
The lip is rounded.
Paste Bu
Eight sherds come from a bowl fashioned from a fine
paste with well-sorted particles of mica, pyrite and igneous
rocks.
It was fully oxidized to an orange color (5YR5/6
yellowish red) .
The lightly thickened rim corresponds to a
simple open bowl with an annular base (Fig. lOla-c).
A
geometric motif now partially eroded was rendered in prefiring red paint (10R4/6 yellowish red) on the bowl
exterior.
Paste Bu
Eight sherds comprised of a distinctive coarse, sandy
paste belong to two bowls with red (10R4/4 weak red) and
brown (7.5YR5/6 strong brown) painted lines.
The red paint
467
approaches purple in hue.
Also distinctive is the exterior
surface which displays wide tracks left by the burnishing
tool.
One large rim exhibits brown lines criss-crossed by
short, red strokes on the interior surface (Fig. lOld).
brown paint also covers the lip exterior.
The
Two sherds from a
second bowl (Fig. lOle, f) show dots and alternating brown
and red lines on the interior surface.
Paste B 15
Four sherds grouped as Paste B15 also pertain to a bowl
(Fig. 102a, b).
The paste is coarse and sandy with a light
brown (10YR7/4 very pale brown) color.
smooth, but undulating and unburnished.
The surface 1s
A reconstructed rim
indicates that the bowl has a carinated shape with slightly
inverted walls and unthickened rim ending in a dull point at
the lip exterior.
It apparently rested on an annular base.
The rim measures 20 em in diameter and the base, 13 em.
The
bowl interior shows a brown (10R3/6 dark red) painted design
comprised of wide uneven lines.
The exterior was slipped
with a pale orange (5YR6/6 reddish yellow) wash, and then
painted with two parallel lines, each approximately 3 mm
wide.
Paste B 16
Three sherds belong to a bowl with fine, sandy, dark
brown (5YR4/2 reddish gray) paste and evidence of three
drilled repair holes (Fig. 102c).
Two of the sherds are
468
lightly thickened rims with flattened lips, indicating a
simple semi-spherical shape.
diameter.
They measure 14 em in
On one rim sherd a painted decoration rendered in
iridescent specular hematite paint is barely visible on the
exterior surface.
The design beneath the lip consists of a
stepped motif underlined by two parallel horizontal lines.
Below these, the motif is probably a reclining "scroll-S"
with another horizontal line beneath it.
The second rim
shows only traces of paint.
Paste B 17
Two body sherds from jars with similar decorative
treatments (Fig. 102d, e) may represent two distinct wares.
The pastes are very fine with fine sandy inclusions.
One
sherd is fully oxidized to a light tan color (Fig. 102d),
while the other shows a gray core indicating incomplete
oxidation (Fig. 102e).
thickness.
They range from 3 to 4 rnm in
The exterior surfaces of both sherds are covered
with a dark brown slip (7.5YR4/4 brown and 5YR4/4 reddish
brown respectively), wide orange painted lines (2.5YR5/8 red
and 5YR6/6 reddish yellow) and a thick white paint.
Paste B 18
A single rim sherd from a bowl consists of a hard,
relatively coarse pinkish gray (10YR7/4 very pale brown)
paste with well-sorted igneous rock temper (Fig. 102f).
The
walls are 5 mm thick, while the rim is gradually thickened
469
to 7 mm ending at a rounded lip.
The polished bowl exterior
displays an orange slip (2.5YR5/8 red), and the potter
employed the negative resist technique to produce a pattern
of circles surrounded by smudged black (perhaps triangular)
fields.
Ceramic Paste Group C
The Empedrada Paste C category consists of 715 sherds
representing 58.3 percent of the Empedrada Phase collection
of diagnostics.
Sixty-one (14.9%) of the 410 Empedrada
Phase rim sherds belong to Paste Group C.
If combined with
Group B's 11 rims, the total 72 rims comprise 17.6 percent
of the Empedrada Phase rim collection.
Empedrada Phase
Paste Group C characteristics are essentially the same as
those noted for the Colpar Phase.
Because of their
unusually high diversity and poor preservation, wares (or
individual pastes) represented within Ceramic Paste Group C
cannot be consistently distinguished from one another with
the unaided eye.
Where the fired clay remains unoxidized,
all kaolin-rich pastes may appear dark grey.
The fully
oxidized pastes show a continuum of colors and hues ranging
from snow white to cream to beige to bright orange,
depending upon the impurities in the clay.
Therefore, the
entire collection of 715 Empedrada Paste Group C pottery was
lumped into Paste Group C2 of necessity.
470
Paste C2 a-h
Empedrada Group C2 pastes are usually fine-grained,
compact and hard, without visible non-plastic inclusions
(Plate XII) .
Finely sorted sand was added to some of these
pastes as temper.
The pottery was often slipped, polished
and painted before firing to produce decorated thin-walled
bowls, many with annular bases.
The average rim diameter is
14.68 em (N=59) with a standard deviation of 2.69 em and
mode of 14 em (N=29). Wall thickness ranges from 2 to 6 mm.
Distinctions within the Paste C2 category are made according
to salient technological, formal and/or decorative
attributes, or combinations of attributes.
Some of these
most basic categories (indicated by lower case letters, e.g.
Paste C2 a, C2 b, C2 c ... etc.) may contain more than one ware.
It should be kept in mind that the sole purpose behind this
categorical scheme is to distinguish groups of sherds by
easily recognizable features,
and comparative analysis.
thus facilitating description
Analysis of the fine wares
recovered from excavations at Gran Pajaten was handled in
the same manner (Church 1988, 1994).
A total of 553 Paste C2 sherds show no signs of painted
decoration, although some off-white sherds carry white slip
or traces of white slip.
An unknown number of these are
highly eroded sherds from painted vessels.
From this
collection, 14 rims are illustrated (Fig. 103a-m, p), along
with two sherds from annular bases (Fig. 103n, o).
These
471
portray a representative variety of characteristic vessel
shapes and rim treatments.
A total of 170 Paste C2 sherds
are distinguished by clearly recognizable (usually
decorative) attributes or clusters of attributes and these
are described in the following paragraphs.
Paste C2 a
(Red Painted)
A total of 124 Paste C2a sherds show decoration, or
traces of decoration, with red paint.
with a white slip prior to painting.
Some were covered
Profiles of seven of
the 12 rims with traces of red paint are presented in Fig.
103q-w.
Figs. 103x, 104a-e and 105a-e show painted designs
on rims sufficiently preserved for illustration.
Fig. 106a-
e shows five of a total 23 body sherds with painted red
designs.
Six sherds (including Fig. 106b-e) bear
decorations on bowl interior surfaces.
Most painted motifs
consist of straight and curving lines tracing simple
geometric patterns.
Line width is highly variable.
Twenty-
three sherds are from vessels covered with an orange (5YR7/6
reddish yellow) or reddish orange (2.5YR5/8 red) slip.
Both
bowl rims (Fig. 104g-j, m, n) and annular bases (Fig. 106f,
k, 1) are represented.
Paste C"b (Black Painted)
Twenty-two sherds grouped as Paste C2b exhibit dark
brown or black painted designs.
The paint color is
typically 5YR3/2 and SYR 3/3 (dark reddish brown).
Six
472
sherds consist of white (oxidized) and gray (unoxidized)
kaolin paste.
One rim with black paint applied to the lip
area (Fig. 107a) represents a rare jar shape in Manachaqui's
Paste Group C collections.
Three rims (Fig. 107b-d) belong
to bowls with simple linear painted designs.
A body sherd
shows a design painted on the exterior with both thin and
wide lines (Fig. 107e).
Sixteen of the Paste C2 b sherds consist of a fully
oxidized orange paste.
One rim exterior displays a painted
row of pendant triangles above a set of two parallel,
A sherd from an annular base
horizontal lines (Fig. 108a).
exhibits crossed lines on the bottom interior (Fig. 108b).
The profiles of two eroded rims with indistinguishable
designs are also illustrated (Fig. 108c, d).
Paste C2 c
(Red and Black Painted)
Seven sherds exhibit exterior surfaces decorated with
both red and black paint.
Two rims bear simple red and
black painted lines (Fig. 108e,
f).
Two sherds show that
black lines may delineate red painted zones (Figs. 108g,
109a) .
Others show black and red alternating with the
natural paste color (Fig. 109b, c).
Paste C"d (Tan on Buff)
Five sherds with a tan colored paint (7.5YR6/6 reddish
yellow) constitute Paste C2 d.
Two similar bowls with in-
curving rims may be represented (Fig. 109d-f).
The motif
473
consists of simple geometric lines and a step pattern.
The
rims portrayed in Figs. 109d and f probably belong to the
same bowl.
Paste C2 e (Black on Orange Slip)
Two sherds, both rims from separate bowls, have
decorated exteriors showing black painted lines (2.5YR3/
very dark gray) on an orange slip (SYRS/6 yellowish red).
One bowl rim (Fig. 109g) has parallel sets of divergent
diagonal lines, while the other (Fig. 109h) exhibits
parallel horizontal lines.
Paste C2 f
One body sherd with a gray core left incompletely
oxidized during firing belongs to a decorated jar (Fig.
llOa) .
The sherd is 5 mm thick.
Over the natural polished
light brown surface, a brown slip (2.5YR3/6 dark red) was
applied leaving an unslipped horizontal band 3 em wide.
Black pigment appearing in the upper part of the band was
probably produced by resist smudging.
A row of concentric
circles and circle-dots rendered in thick white paint repeat
within the band which likely surrounded the jar's shoulder
or mid-section.
Smears of white paint obscure the motifs,
especially the circle-dot pattern at right.
Paste C2 g
The sherd with a gray core depicted in Fig. llOb
pertains to a restricted vessel with thin walls left
474
incompletely oxidized during firing.
On its light brown,
highly-polished exterior surface, a design rendered in red
(2.5YR4/8 red} paint is barely visible.
Paste C2 h
One eroded sherd of grayish white kaolin paste has
cracked and eroded surfaces (Fig. llOc}.
It belongs to a
jar neck and rim closely resembling Shape GRim 2.
The end
of the rim has been reinforced by folding it over.
Maximum
thickness is 8 mm, while the neck walls are 4 rnm thick.
Lithic remains
An estimated 50 percent of the Manachaqui Cave ceramic
era lithic remains (approximately 4,750 artifacts} belongs
to the Empedrada Phase.
The 43.5 percent of the collection
remaining after accounting for the four phases under study
pertains to the Late Horizon Poblano Phase.
Of the 4,750
Empedrada lithic artifacts, 4,706 (99%} are chipped-stone
products of the same simple core-flake industry that
characterized the Colpar phase.
Thirty-four chipped-stone
artifacts exhibit indications of retouch, leaving an
estimated 4,672 non-retouched cores, flakes and blocky
shatter.
Of seven cores, six consist of mafic igneous rocks
including basalt, and the seventh is made of gray chert from
the outcrop crossing the ancient road to the east (Site M11).
Four of the cores have battered surfaces.
More frequent than in any of the previous phase
475
assemblages are retouched flakes.
Twenty-nine exhibit
somewhat haphazard unifacial retouch on one or two edges.
These same "scraper" edges show tiny step fractures and
crushing from heavy use.
Fourteen consist of igneous rock,
nine pieces are of the local gray chert and six consist of
chert from unknown sources.
Three obsidian flakes were
recovered from Empedrada Phase contexts.
Two of these are
small pressure flakes picked out of the fine screens used to
sort the Ernpedrada Phase floors (Floors I and K} in Sector
A.
A third is the proximal end of a thick, broken blade
with a prominent dorsal spine and parallel edges showing
evidence of heavy use.
A piece of blocky obsidian shatter
was collected from a mixed Empedrada and Poblano level (Unit
31 Level 5} .
Two narrow biface fragments appear to be the basal
portions of broken un-stemrned projectile points made from
non-local chert (Plate XIII}.
Both points were long, narrow
(1.8 and 2.4 em} and relatively thick (7 and 11 rnrn} with
rhomboidal cross-sections.
The point from Unit 23 Level 8
exhibits a crudely-shaped tip, while the other from Unit 23
Level 10 shows a semi-squared base.
In size and basic
outline they are not unlike the ground slate points, and
they may have served a similar function.
The 44 ground stone artifacts from Ernpedrada Phase
levels are pieces of ground and polished slate and shale.
One oblong pebble (5.1 x 2.4 x 2.1 em} shows no signs of
476
use, but was certainly carried to the shelter from
elsewhere.
The ground and polished point fragments from
Empedrada Phase levels number 24.
Five fragments from mixed
Empedrada and Poblano Phase levels are assigned to the
Empedrada Phase because there are no such points in unmixed
Poblano Phase levels.
Thus, the Empedrada Phase ground and
polished points total 29.
An additional piece of slate with
a convex, polished surface apparently split off a larger
ground and polished slate artifact.
An ovoid slate disk ground smooth, but otherwise
unworked,
x 0.3 ern.
from Sector A Unit 16 Level 8* measures 2.7 x 2.1
Two similar ovoid disks from Unit 10 Levels 2 and
3 at the rear of the shelter may date to the Empedrada Phase
or to the Poblano Phase.
x 2.2 x 0.4 em.
The disk from Level 3 measures 3.2
The disk from Level 2 measures 2 x 1.5 x
0.25 em, but a perforation 3 mm in diameter has been drilled
into its center, and the wide end is notched or serrated
{Fig. 108d).
The presence of two blanks and 11 flakes of
the same slate or shale provides evidence for the
manufacture of pendants, projectile points and perhaps other
items at the shelter during the Empedrada Phase.
Rocks and Minerals
A small quantity of rocks and minerals brought to the
site during the Empedrada Phase include two small pieces of
mica, a chunk of quartz combined with galena and one
complete prismatic quartz crystal.
477
Botanical Remains
Unit 15's Floors G through T (but excepting Floors LL
and 0) provided samples of the Empedrada Phase carbonized
botanical remains examined by Pearsall.
Maize kernels and
cob fragments continue in abundance, while beans occur
relatively rarely.
Fruit rinds, tuber/root fragments and
all of the taxa identified in the earlier phase assemblages
appear in varying quantities during the Empedrada Phase
(Table 18).
Faunal Remains
With the Empedrada Phase, the variety of taxa, most of
which is presumed to represent food remains, continues to
expand (Table 27).
Bird bones, some identified as owl, are
more frequently represented than in previous phases, and
reptiles appear for the first time.
Finer taxonomic
resolution reveals the presence of the llama, white-tailed
deer, Andean fox, squirrel and guinea pig.
A Homo sapiens
premolar from Floor 0 in the shelter interior directly
represents the human protagonists under study.
Radiocarbon Evidence for Empedrada
Phase Chronology
Three radiocarbon dates from Sector A represent the midEarly Intermediate Period.
Dates from Unit 14 Floors L and
H are a.d. 570 ± 80 and 450 ± 80 respectively, while a date
from Unit 15 Floor T is a.d. 490 ± 80.
The Unit 15 date is
associated with levels bearing the kaolin sherds diagnostic
478
of the Empedrada Phase in Manachaqui's macrochronology.
However, as with the Colpar Phase, future radiocarbon dating
should seek to correlate Sector B's stratigraphic chronology
with Sector A's.
At present, the radiocarbon dates do not
provide clear resolution of the phase's beginning or ending
dates.
In the previous chapter it was suggested that a hiatus
may separate the Colpar and Empedrada Phase utilization of
Manachaqui Cave.
Two undated occupation floors separate the
radiocarbon dates of 160 b.c. and a.d. 490 in Unit 15, and
three undated floors separate the dates of a.d. 110 and a.d.
450 in Unit 14.
Additional radiocarbon dating may close
this 340 year gap (only a 180 year gap if the sigmas of
Floor P and Floor H dates are factored in) .
On the other
hand, evidence for a similar break exists in the radiocarbon
sequence from construction fill at nearby Gran Pajaten where
more than four centuries separate the uncalibrated dates of
a.d. 40 and a.d. 460.
At present, the radiocarbon evidence
indicates a dating of a.d. 400 to 600 for the Empedrada
Phase, but ceramic cross-dates providing additional
chronological information will be discussed at the end of
this chapter.
Paste A Relationships:
the Central Andes
For the first time in the Manachaqui Cave sequence,
Paste A pottery shows virtually all of the characteristic
479
features of Central Andean assemblages to the west and
south.
Large, globular jars with long, flaring rims and
semi-spherical bowls constitute the two most basic vessel
shapes in the Central Andes from the mid-Early Intermediate
Period until the Spanish invasion.
so generic and widespread,
Because these shapes are
they are poor indicators of
specific relationships by themselves.
However, a brief
review of specific parallels between the Empedrada
assemblage and some of the better-known assemblages from the
neighboring highlands enables sufficient resolution for
determining which of the Central Andean archaeological
cultures show clearest stylistic affinities to Manachaqui
Cave's Early Intermediate Period assemblages.
Similarities between Empedrada Phase ceramics and Gran
Pajaten's pre-Abiseo Phase assemblage are predictable given
the ancient settlement's proximity to Manachaqui and its
radiocarbon date of a.d. 460.
Shape G Rims 1, 3, 4 and 6
are identical to jar rims encountered in Building No. 1's
construction fill
(Church 1988: Fig. 36h-m; 1994: Fig. 11f-
k, Fig. 11b, c; 1988: Fig. 40g respectively).
The simple
bowls are likewise indistinguishable (Church 1994: Fig. 13hu), although some of the more restricted Gran Pajaten
examples recall Colpar Phase bowl shapes.
Red slip is
common to both sites, but rim punctation and notched
applique decorations have not been observed at Gran Pajaten.
Gran Pajaten's pre-Abiseo Phase rim notching (Church
480
1988:168) is absent from the Empedrada assemblage.
Empedrada Phase pottery also finds analogous
counterparts in the Cancharin Phase assemblages from the
vicinity of Cuelap.
Profiles of Shape G Rims 1, 2 and 6
resemble those depicted by Ruiz {1972) in Lams. VIII:1-31,
X:a-q and IX:c-g respectively.
Bowl rim profiles are also
similar, and carinated varieties occur at both sites.
Cuelap's·bowls with everted rims and thickened interiors do
not occur at Gran Pajaten nor at Manachaqui Cave.
Directly west in Huamachuco, the dates A.D. 400 to 600
coincide with the Early Huamachuco Phase {T. Topic and J.
Topic 1987: Fig. 13).
Only the most generic similarities
between the profiles of flaring jar rims suggest some
relationship to Empedrada Phase ceramics.
Unlike the
Empedrada Phase, the Early Huamachuco Phase is characterized
by the presence of fine wares with cursive-style decoration
{Thatcher 1972-74: Figs. 23-36), both imported from
Cajamarca to the north and imitated locally {T. Topic and J.
Topic 1987:23).
In Cajamarca, the production of cursive-
style pottery marks the beginning of the Middle Cajamarca
Phase {Matsumoto 1993:188).
Matsumoto's {1993: Tabla 1)
absolute chronology for Early Intermediate Period Cajamarca
is based upon Shimada's {1987:140) suggestion of an a.d. 450
cross-date for Moche Phase IV and the Middle Cajamarca
Period.
The dates for both the Early Huamachuco and Middle
Cajamarca assemblages are ostensibly supported by the
481
association of Early Huamachuco Phase cursive-style pottery
with radiocarbon dates averaging a.d. 405 from the principal
constructions at Marcahuamachuco (J. Topic and T. Topic
1982:21, 1983b: Table 3}.
The points just outlined in the previous paragraph lead
to the suppositions that either: 1} the Empedrada Phase
pottery predates its associated radiocarbon dates as well as
the Early Huamachuco and Middle Cajamarca Phases featuring
cursive-style pottery or 2} Manachaqui Cave's interactions
at this time were focused on Central Andean regions south of
Huamachuco and outside of Cajamarca's interaction sphere.
Evaluation of the first supposition requires an examination
of earlier phase assemblages in Huamachuco and Cajamarca.
These would be the Purpucala Phase in Huamachuco dated from
A.D. 1 to 400 (T. Topic and J. Topic 1987: Fig. 13}, and the
Early Cajamarca Phase dated from A.D. 200 to 450 (Matsumoto
1993: 188} .
Although a few resemblances to Huamachuco's Purpucala
Phase remains have been observed in the pre-Abiseo Phase
assemblage at Gran Pajaten (Church 1994:288-289}, there is
little in the Empedrada Phase assemblage that presents
similar correspondences.
Thatcher's
Empedrada Shape GRim 6 matches
(1972-74: Fig. 19a-h} Jar 8 rim, and red and
white paints were also applied to Purpucala Phase coarse
wares.
In fact, Thatcher (1972-74:112}
regards the
decorative use of red and white paints as the only
482
remarkable similarity between Purpucala and contemporary
Cajamarca styles.
Unlike their Empedrada counterparts,
Early Cajamarca Phase jar rims are often modelled,
especially around the lip (Terada and Onuki 1982: Pls. 90,
92, 94-96).
In terms of shared design features,
the
Empedrada assemblage bears little more than superficial
resemblance to Purpucala and Early Cajamarca Phase ceramics.
The closest stylistic match for the Empedrada Phase
assemblage can be found south of Huamachuco in the upper
Santa River drainage best known for Recuay Style pottery.
Based on his finds at the site of Pashash, Grieder (1978)
has published the only Recuay ceramic sequence.
The
centuries between a.d. 300 and 600 correspond to Grieder's
Quimit, Yaia and Huacohu Phases of the Recuay Period
(Ibid.:63, Table 10).
Empedrada Shape G Rims 5 and 6
resemble Recuay jar shapes B-1 and B-4 (Ibid.: Fig. 35k, s)
of the Quimit Phase, while Empedrada Shape G Rims 1, 2 and 4
belong to jars identical to Recuay shape B-5 (Ibid.: Figs.
38f, l; 36i) of the Huacohu and Yaia Phases.
The
distinctive Empedrada Shape H Rims 1, 2 and 3 find precise
parallels in later Usu Period (a.d. 600-700) vessel shapes
B-10 and B-8 (Ibid.: Figs. 40i, 38j, k).
That these latter
shapes only appear late in the Pashash sequence (Ibid. :62)
supports Thatcher's (1979:99) argument that identical rim
shapes at Cerro Campana East (McCown 1944: Fig. 18a-g)
represent a post-Sausagocha Phase re-occupation of the site
483
(J. Topic and T. Topic 1982:20 present a contrary point of
view).
In the eastern Andes of Huanuco to the south of
Manachaqui Cave, archaeological survey and excavations have
not yet turned up mid-Early Intermediate Period occupations
(Rozenberg 1982:135; Bonnier et al. 1985:96).
Farther south
however, Hastings (1985) has grouped two sets of mid-Early
Intermediate Period assemblages on the Tarma Valley eastern
slopes into the Malambo and Camonal Complexes.
The former
is distributed from the Tropical Alpine and Subalpine Life
Zones down through the eastern Montane Forests (Hasting's
"Lower Tier") and into the Premontane Forest (his "Montana
Uplands").
The Camonal Complex is mostly restricted to the
Montana Uplands and Bottomlands at the foot of the Andes
(Ibid.: Table 11-11).
However, Hastings (Ibid. :590) notes
that shared attributes occasionally render the two complexes
virtually indistinguishable.
Both of Hastings' complexes share design features with
the Empedrada assemblage.
Empedrada's resemblances to the
Malambo Complex are general, as jars with flaring rims,
semi-spherical bowls and notched applique decorations are
typical (Ibid.: Fig. 9-11, 9-13).
Parallels with the
Camonal Complex (Ibid.: Figs. 9-14, 9-15) are more specific
as Empedrada Shape G Rims 1, 2 and 4 mirror the rims
depicted in Hastings' Fig. 9-14t-v, z-ae and k-1
respectively.
Jar rims of both Tarma Valley complexes
484
feature rims with exterior thickening.
The "folded" rim,
however, is almost exclusive to Carnonal (Ibid.:590).
Distinctive bowl-shaped rims like Empedrada Shape H Rims 1
and 2 also occur in the Camonal Complex (Ibid.: Fig. 9-14hk).
Carnonal's rare corrugation and frequent use of strap
handles are not shared by the Empedrada assemblage.
The regional comparisons offered in the above paragraphs
demonstrate that the Empedrada Phase Paste A ceramic
assemblage is closely affiliated with coeval Central Andean
pottery styles, yet maintains specific design features in
common with early assemblages from eastern slope localities
such as Gran Pajaten, Cuelap and the Tarma River valley.
The most compelling resemblances to Empedrada Phase Paste A
pottery are found in the upper Santa valley Recuay region
and, to a far lesser extent, in Huamachuco.
Both Grieder
(1978:75) and Thatcher (1972-74:112) have noted the
similarities shared between their respective regions of
study.
The complete absence of Huamachuco and Cajamarca
cursive-style pottery at Manachaqui Cave draws attention
south to Recuay as the Empedrada assemblage's closest
stylistic companion in the highland Central Andes.
Paste A Relationships: the
Amazonian Lowlands
Below Manachaqui Cave to the east Ravines'
(1978)
Central Huallaga sequence presents close stylistic
affinities to highland pottery traditions for the first
485
time.
The Huayabamba I Complex recovered from three sites
on the lower Huallabamba River below its confluence with the
Abiseo and Jelache Rivers displays several parallels with
the Empedrada assemblage.
Ravines'
(Ibid.: Lam. 3a, b)
Forms A and B are jars with bowl-shaped rims conceptually
similar to distinctive Empedrada Shape H Rim 1.
The
likeness may not be immediately apparent because the end of
the Huayabamba vessel's rim (see Lam. 3a profile) has been
folded over, and the rim's carination angle is sharper.
A
second formal correspondence is evident between Huayabarnba
Form G necked jars (Ibid.: Lam. 4d, e) and Empedrada Shape G
Rim 2 jars with folded-over rims.
Note that these are the
same formal resemblances that characterize the stylistic
relationship between the Empedrada assemblage and Hastings'
Camonal Complex.
The Empedrada Phase pre-dates the appearance of the
Cumancaya Tradition on the river floodplains east of the
Central Andes.
It is presumably coeval with Lathrap's
(1962) Pacacocha Phase on the Central Ucayali, and perhaps
with Allen's (1968) Naneini Phase from the foot of the
eastern Andes.
While these lowland assemblages carry design
features which are unmistakably Amazonian, it may be
significant that the jars with bowl-shaped rims like
Empedrada Shape Hare found in both (Lathrap 1962: 113a, c,
e; Allen 1968: Fig. 25: Form 4).
Lathrap (1962:351)
maintains that they may have early Amazonian antecedents in
486
Hupa-iya (Barrancoid) shapes.
Allen (1968:357) observes that Naneini is the only Alto
Pachitea assemblage that can be "linked" to lowland
Amazonian pottery traditions to the east.
However, Naneini
retains a globular, necked jar shape from earlier phases
(Allen 1968: Fig. 26: Form 6) suggesting continued ties to
the Central Andes.
Hastings (1985:590-591) regards the rim
folding technique as a common thread connecting Malarnbo,
Carnonal and Alto Pachitea styles, and the Ernpedrada and
Cancharin (from Cuelap) assemblages should be added.
Otherwise, coeval lowland Amazonian assemblages, and the
Curnancaya assemblages that succeed them, are stylistically
dissimilar to Ernpedrada and other Central Andean
assemblages, especially with respect to decorative
techniques and motifs.
Paste A Relationships:
the Northern Andes
The Empedrada Phase at Manachaqui Cave corresponds to
the centuries spanning the Regional Developmental PeriodIntegration Period transition in the neighboring Northern
Andes
(Meggers 1966).
Some of the Ernpedrada features do
have northern correlates, yet it is not necessary to range
beyond the Central Andes to document Ernpedrada's closest
resemblances and hypothetical affiliations.
Nevertheless, a
stylistic relationship apparently does exist with ceramics
from the analogous Ecuadorian montane forest setting of
487
Cosanga (Porras 1975a).
Empedrada jar Shape G Rims 1, 2,
and 3 match illustrated rim profiles for Cosanga Shapes 7
and 10 and 11 (Ibid.: Figs. 20, 21), while Cosanga Shape 5
represents less diagnostic semi-spherical bowls much like
Empedrada examples.
As noted in the previous chapter, the co-occurrence of
the distinctive fold-over rim reinforcement technique
constitutes a commonality among styles found along the
eastern slopes of the Central Andes.
Based on her analysis
of Cosanga/Panzaleo pottery from northern Ecuador, Bray
(1995:146) contends that folded rims are the unique
"hallmark" of this eastern slope style.
Additional
Cosanga/Panzaleo features occurring in the Empedrada
assemblage are the parallel rows of punctations on rim
interior surfaces (Porras 1975a: Lams. 29g-l, 30a-m).
At
Cosanga, the punctation occurs more typically on short rims
of wide-mouthed vessels.
Nevertheless, the technique has a
limited distribution in the Northern Andes and has not been
reported this far south.
Bray (1995:140) considers this
punctation a diagnostic feature of Cosanga/Panzaleo pottery
during early centuries of the Integration Period.
Paste B Origins
Paste B,,
Useful clues in determining hypothetical sources for the
micaceous pottery of Paste B11 may be found in Porras'
(1975a) report on Cosanga ceramics, and Bray's (1995)
488
summary of Cosanga/Panzaleo vessel morphology.
All three of
the Cosanga pastes are sandy and micaceous, and Bray
(Ibid.:137) observes that "thin vessel walls, ash-colored
paste, and micaceous inclusions" distinguish Panzaleo
pottery found in archaeological assemblages throughout much
of northern Ecuador in a variety of temporal contexts.
The
single Paste B11 jar rim profile (Fig. 100f) precisely
mirrors Porras'
(1975a: Fig. 21) Shape 10 profile, and
Bray's (1995: Fig. 7g) Form III/7 profile.
Porras
(1975a:115) reports similar surface treatments, and observes
that Shape 10 jars frequently exhibit red slip (Ibid.:127).
The Manachaqui Paste B11 rim diameter of 30 ern exceeds the
Shape 10 diameter range of 10 to 25 measured by Porras, but
is smaller than the Cosanga maximum of 40 ern (Ibid. :107).
Paste B12
This pottery is too eroded and fragmented to determine
the shape of the corresponding vessel or vessels.
The
micaceous paste suggests some relationship to Paste B11
pottery from the eastern slopes of the Northern Andes.
Little more can be inferred without a larger, better
preserved sample.
Paste B13
Again, a micaceous paste suggests a possible
relationship with the Paste B11 and B1 :: pottery just
described.
Porras (1975a:ll3-118) writes that fully
489
oxidized Cosanga pastes often appear orange or yellowish
red.
The Paste B13 bowl shape (although admittedly generic)
matches Porras' Shape 5 (Ibid.: Fig. 20).
Its painted
design (Fig. lOla, b) resembles the motif depicted in
Porras' Lam. 36a, although the red-painted Cosanga design is
rendered on white slip.
A similar triangular scroll motif
can be found closer to Manachaqui on Cajamarca Tradition
kaolin bowls of the Cajamarca Linear-painted Type dating to
the Early Cajamarca Phase (Terada and Onuki 1982: Pls. 44c5, 102-25; Matsumoto 1993).
Taking paste characteristics
into account, it is not unreasonable to suggest origins on
the eastern slopes of the northern Peruvian or Ecuadorian
Andes.
Paste B14
The eight sherds representing Paste B14 pertain to two
bowls that originated far to the south in the highland
Central Andes between Lake Junin and the modern city of
Ayacucho.
The use of purplish red paint is a diagnostic
characteristic of mid-Early Intermediate Period pottery
reported from the Salinas de San Blas (Morales 1978:331) and
the upper Tarma and Mantaro valleys (Hastings 1985:525-528;
Browman 1970; Benavides 1971).
The sherds illustrated by
Morales (1978: Lams. 5, 6) show similar decorative panels
painted on the rim interiors.
The same red and black cross-
hatch motif (Fig. lOld) appears on the bowl depicted in
Morales' Lam. 6.
490
Ceramics from Pachamachay described by Silva (1988:2628
I
Figs. 56-58) may also be related to Paste B14 pottery.
Silva observes that highland populations in Huancayol
Ayacucho
1
Junin~
Huancavelica and Huanuco shared a common
decorative tradition yet maintained regional autonomy during
the Early Intermediate Period.
Huacrapuquio
1
Browman has termed the style
and suggested dates of A.D. 500 to 600/650.
Hastings (1985:592-595) reports that this eastern highland
style penetrates the upper edges of the Tarma canyon montane
forests as an imported ware (Ibid.: Fig. 9-12).
The Junin
sherds described by Morales and Hastings provide
sufficiently close resemblances to conclude that the Paste
B14 sherds likewise represent this south-central highland
stylistic tradition.
Paste B15
Sherds from decorated bowls like the Paste B15 example
have been recovered from Cancharin Phase contexts at the
site of Cuelap (Ruiz 1972).
Ruiz's Cuelap Pintado Tosco
Type (Ibid. :77-79) is likewise unburnished and frequently
shows horizontal striations left after smoothing the
surfaces.
Horizontal red painted lines (Fig. 102a) may
appear on bowl lips and exteriors (Ibid.: Lam. XXXVIIa-d)
I
while vertical lines and haphazard dots commonly decorate
bowl interiors (Ibid.: Lam. XXXVIal c).
The match between
Paste B:s and Ruiz's written descriptions and illustrations
of the Cancharin pottery is remarkably close.
The "finas
491
capas roj izas"
{fine, reddish layers}
{Ibid. : 77} probably
refer to the thin orange wash observed on the exterior
surface of the Empedrada example.
Ruiz does not mention
annular bases, but his sample consists of only 43 sherds,
all of which come from the "fortaleza" or main settlement of
Cuelap.
Paste B16
Decorated bowls such as these (Fig. 102c} have been
recovered during the investigations at Cerro Ochoconday on
the western slopes of the Andes below Huarnachuco {Alfredo
Melly and Theresa Topic personal communications 1990}.
The
Cerro Ochoconday examples are better preserved, and show the
same horizontal lines, step motifs and scroll-S motifs
rendered in iridescent paint on the exterior surfaces of
brown, polished bowls.
The only published illustration of
this ware might be Wilson's {1988} Fig. 212p.
Paste B17
The two sherds with white and orange paint on polished
brown slipped surfaces {Fig. 102d, e)
fit Strong and Evans'
{1952:344-347 and Fig. 80} descriptions of Castillo White,
Red, Orange Type ceramics recovered from Gallinazo contexts
in the lower Viru valley.
Strong and Evans regard this
pottery as intrusive on the coast, and Theresa Topic
{personal communication 1987} has found larger amounts in
the upper Viru drainage.
Topic {personal communication
492
1991} has examined these two particular sherds and believes
that they are not the same wares that she observed during
the upper Virli surveys.
In the highlands above Virli, Perez (1988: Lam. 4)
illustrates similar pottery from Santiago de Chuco.
Grieder's Pashash publication includes a photograph of
Huacohu Phase (late Recuay Period a.d. 500 - 600}
"white and
red on orange" sherds which might be related to the Paste
B'- 7 wares.
Actually, the Paste B17 sherds match Strong and
Evans' descriptions and illustrations so closely that it may
not be imprudent to conclude that they pertain to styles
akin to Castillo White, Red and Orange that originated in
the north highland Recuay area, perhaps near Santiago de
Chuco.
Paste B18
The Paste B18 rim sherd with negative resist decoration
(Fig. 102f) invites comparison to Gallinazo Negative,
Decorative Sub-type VIII from Viru (Strong and Evans
1952:302 and Fig. 59L, M}.
However, Gallinazo Negative
utilizes natural or white-slipped surfaces as a base.
The
presence of an orange-red slip may indicate origins farther
north.
Ruiz (1972: Lam. XXXVc-g} assigns similar red-
slipped and negative resist decorated sherds at Cuelap
and/or Pumahuanchina to his Cuelap Polished pottery type and
the Early Intermediate Period Cancharin Phase.
The quantity
of these negative decorated sherds is unspecified, but the
493
pottery may be exotic at Cuelap as well.
Estrada et al.
(1964: Figs. 30, 31) illustrate similar
negative dot motifs on red slip for the Jambeli Culture of
southwestern Ecuador.
Cafiar Polished and Tuncahuan pottery
from Cerro Narrio (Collier and Murra 1943: Pls. 26, 38-41)
frequently displays similar dot motifs in negative, but the
vessels are typically unslipped.
Izumi and Terada (1966:84)
urge caution in attaching great temporal or spatial
significance to this widely distributed motif.
Nevertheless, we might speculate that the Paste B18 example
originated in extreme northern Peru.
Paste C Origins
Interaction spheres characterized by the circulation of
fine kaolin wares in the north-Central Andes expand during
the Early Intermediate Period.
Cajamarca and Recuay are the
best known archaeological cultures producing this pottery,
but decorated kaolin ware pottery has also been recovered
during excavations at nearby Gran Pajaten.
The Gran Pajaten
sample was judged to be intrusive, non-local pottery because
of its unusual variety of pastes, decorative techniques and
motifs, coupled with its corresponding lack of stylistic
cohesiveness (Church 1994:288).
The Empedrada Phase sample
at Manachaqui Cave should be regarded as intrusive (relative
to Paste A ceramics)
for precisely the same reasons.
It
should be reiterated that the variety of Empedrada kaolin
ware ceramics was subsumed for descriptive purposes within a
494
single category (Paste Group C)
in response to the
multiplicity of separate, yet visually indistinguishable,
kaolin wares.
Sequences from the Cajamarca and Recuay areas have been
developed, although the Cajamarca region has received more
extensive study.
Chronology and intra-regional variability
of the Recuay tradition remain to be fully investigated.
This study is handicapped especially by the lack of
fieldwork in the upper Marafion valley or Callejon de
Conchucos to the southwest of Manachaqui Cave where much of
the Empedrada Phase Paste C pottery may be originating.
The
hypothetical sources for Paste C described below do not
include Cajamarca which, judging by the absence of cursivestyle pottery, does not seem to interact directly with the
study area during the Empedrada Phase.
Paste C2
Virtually all of the Empedrada Phase Paste C2 decorated
kaolin wares can be subsumed within the highland Recuay
pottery style as described by Bennett (1944:99-106}, Grieder
(1978} and Wegner (1981}.
While illustrations of the most
elaborate Recuay pottery featuring negative-resist
decoration have been widely published (e.g. Larco Hoyle
1962; Bankmann 1977; Grieder 1978}, the simpler positivepainted bowls are more often represented as sherds
encountered during surface surveys performed outside of the
upper Santa Valley core area (e.g. Strong and Evans 1952;
495
Thatcher 1972-74; Gambini 1984; Proulx 1985; Wilson 1988).
Hence, it appears that the positive-painted variety was more
widely distributed than the Recuay negative-resist variety.
Unfortunately, the decorative variability of the positivepainted style remains one of Recuay's least-known aspects.
Despite the problems cited above, some illuminating
comparisons between Ernpedrada and Recuay positive-painted
wares can be offered.
The semi-spherical bowl shapes and
simple rim profiles representative of the highland kaolin
ware traditions are not particularly diagnostic.
Decorative
techniques, and especially motifs, constitute the most
useful attributes for revealing stylistic relationships
although some of these too were widely shared.
The
following comparative analysis focuses primarily on
decorative attributes.
Paste C2 a-h
The Paste C2a motif rendered in red paint on white
kaolin, and depicted in Figs. 103x and 104a, also occurs in
Thatcher's Sausagocha Phase sample from Cerro Campana East
which, as previously noted, may contain ceramics from
multiple Early Intermediate Period occupations.
Apparently
depicting interlocking serpent heads, the motif also appears
in Nepefia Valley contexts (Proulx 1985: Pl. 7A: PV 31-59p).
An illustration of a sherd from an Early Intermediate Period
context at Cerro Campana West, and remarkably similar to the
Empedrada example, was shown to me by T. Topic (personal
496
communication 1994).
The Nepefia example appears to be
rendered in red and black on white kaolin, while the Cerro
Campana West sherd exhibits black on buff-colored kaolin.
A motif composed of a red band with zig-zag edge and
thin parallel lines identical to Fig. 104b was unearthed in
the Virli Valley (Strong and Evans 1952: Fig. 81c).
Both
thin and wide lines painted on kaolin pottery resembling
Empedrada Paste C2 (Figs. 104c-e, 105a-e, 106a-e this
thesis) are frequently illustrated in the Recuay-related
publications cited here.
The painted wavy line on the bowl
depicted in Fig. 104e also appears at Chavin de Huantar
(Bennett 1944: Fig. 31a), in the Santa Valley (Gambini
1984:138-139; Wilson 1988: Fig. 212b) and on an illustrated
sherd from the Callej6n de Conchucos site of Cashajirca
shown to me by Richard Burger (personal communication 1993).
Decorative bands of pendant triangles above horizontal
lines like that depicted in Fig. 108a occur at Gran Pajaten
(Church 1994: Fig. 15a-c), at Cerro Campana East in
Huamachuco (Thatcher 1979: Fig. 46) and again in the lower
Santa Valley (Wilson 1988: Fig. 212i-k).
Similar designs
ornament kaolin bowls recovered from late occupations at
Chavin de Huantar (Tello 1960: Fig. 174 upper left).
Black
horizontal stripes on red paint or slip (Fig. 109h) occur in
Thatcher's (1972-74: Fig. 14, 15) Purpucala Phase
collection.
The Paste C2 f sherd (Fig. 110a) shows both the
negative-resist and positive-painting techniques.
From
497
Pashash, Grieder (1987:228, No. 6) illustrates a Recuay jar
with negative resist decoration and an upper border executed
in "cream circle dot on black."
The Empedrada sherd
probably originated in the Recuay core area.
Many less specific correspondences between Empedrada
Phase Paste C sherds and other north-Central Andean ceramics
of the Early Intermediate Period could be listed here in an
extended, but belabored, discussion.
the Recuay area is clearly evident.
Close interaction with
Only one Empedrada
Phase Paste C sherd identified, the single Paste C2 g jar rim
sherd, was probably produced outside of the Recuay and
Cajamarca core areas despite the utilization of kaolin-rich
clay.
The rim resembles Empedrada Shape B Rim 2 and may
represent a vessel fabricated within the study area.
The
clay may have originated to the west, but because of the
lack of accurate geological information from the Huallaga
basin montane and premontane forests, the existence of
kaolin deposits in the study area cannot be ruled out.
Lithic Evidence
Lithic materials from ceramic era sites in the Andes are
often ignored as archaeological evidence.
However, the
Empedrada Phase chipped-stone lithic assemblage contains a
few formal elements such as projectile points that can be
productively compared to the few reported finds in
neighboring areas.
While similar points are found in the
north highlands, none of these date to the mid-Early
498
Intermediate Period.
Possible stylistic analogues have been
recovered in Cajamarca (Terada and Onuki 1982: Pl. 54-6, 7},
but they date to the Layz6n and Initial Cajamarca Phases
(200 b.c. to a.d. 200}.
Similar "willow-leaf-shaped" points
at La Pampa date to the Late Horizon, and possibly to the
early Initial Period (Terada 1979: Pl. 81a}.
Urabarriu
Phase points from Chavin de Huantar (Burger 1984b: Figs.
384-387} also resemble the Empedrada Phase points.
Whether or not projectile points appear in a particular
phase assemblage at a particular archaeological site depends
upon site functions and the kinds of activities that took
place over time.
The regional panorama of seemingly
scattered appearances of similar projectile points during
different time periods at different sites described in the
literature leads one to suspect that these kinds of
projectile points were produced continuously throughout most
of the prehistoric ceramic era.
Where and when they appear
in site reports can be viewed as the result of a series of
factors, sampling and reporting procedures utilized not
being the least important.
Unequivocal evidence for long-distance interaction at
Manachaqui Cave has been obtained by trace element analysis
of the Empedrada Phase obsidian artifacts.
The three
obsidian pieces from Empedrada and mixed Empedrada and
Poblano Phase contexts were subjected to neutron-activation
analysis by Burger et al.
(1995}, and were found to consist
499
of stone from the Quispisisa source in Huancavelica.
This
identification compliments the evidence for linkage between
Manachaqui Cave and interaction spheres far to the south
represented by the Paste B14 Huacrapuquio-like sherds.
Ceramic Evidence for Empedrada
Phase Chronology
In the previous chapter, it was observed that the Colpar
Phase may not have extended beyond a.d. 200.
Ceramic and
stratigraphic evidence attest to a major style shift in the
Manachaqui Cave sequence occurring at a.d. 200 or shortly
thereafter.
The cross-dating of pottery is crucial to
further define the Empedrada Phase in the absence of more
radiocarbon evidence.
Unfortunately, the north highland
ceramic sequences do not offer particularly high
chronological resolution for the early centuries of the
first millennium A.D.
Nevertheless, there are comparative
data with which to estimate a beginning date for the
Empedrada Phase, and to determine whether or not a hiatus
exists between a.d. 200 and 400.
The chronological contexts in which the widely dispersed
Central Andean kaolin wares occur all correspond to the
Early Intermediate Period.
The most solid evidence with
which to temporally situate much of the Empedrada Paste C
pottery comes from the Recuay site of Pashash and the Viru
valley where non-local Recuay sherds are found in Gallinazo
Period contexts (Strong and Evans 1952).
The Gallinazo
500
occupation of the Virli Valley may have lasted until
approximately a.d. 400 (Shimada and Maguifia 1994: Cuadro
1.2).
Based upon his radiocarbon evidence from Pashash,
Grieder (1978:63) posits a.d. 300 to 600 as dates for the
Recuay style.
However, the beginning of the Recuay Period
at Pashash has not yet been directly dated, and Wegner's
(1981:6) suggested a.d. 200 beginning date is more probable
in view of the Gallinazo-Recuay temporal overlap documented
by Strong and Evans in the Virli valley.
Manachaqui's ceramic cross-dates do not provide evidence
of a hiatus between the Colpar and Empedrada Phases.
On the
contrary, the presence of Pastes B17 , B18 and C2 a implies
Gallinazo cross-dates.
These, coupled with Recuay cross-
dates and the radiocarbon evidence,
indicate continuous
occupation at Manachaqui Cave between a.d. 200 and 600.
An
eighth century cross-date to the Usu Period suggested by the
co-occurance of Shape H Rims 1 and 2 at Manachaqui Cave,
Pashash and perhaps Cerro Campana East extends the proposed
ending dates for the Empedrada Phase to a.d. 700.
A
terminal Early Intermediate Period occupation during the
Empedrada Phase is further supported by the presence of
Paste B14 Huacrapuquio-like imported ceramics.
Finally,
absence of cursive, floral or Huari style pottery in
Manachaqui's deposits provides negative (and therefore
inconclusive) evidence that the Empedrada Phase ended by
a.d. 700 and the Middle Horizon.
the
CHAPTER 10
MANACHAQUI CAVE AND MONTANE
FOREST MIGRATIONS
The previous four chapters presented evidence for four
archaeological phases constructed primarily on the basis of
stratigraphic correlations between ceramic attributes and
radiocarbon dates.
This chapter implements Rouse's
(1986)
three-pronged strategy for inferring migrations by surveying
evidence supplied by historical linguistics and physical
anthropology in conjunction with Manachaqui Cave's
archaeological sequence.
Because relatively little
anthropological research has been conducted in this
geographic area, none of these data sets is sufficient, and
sufficiently detailed, to successfully argue for the
postulated prehistoric migrations.
The scant linguistic
evidence is highly equivocal, while there is little evidence
from physical anthropology with which to gain useful
insights into prehistoric biological populations.
As usual,
archaeological evidence is the most abundant, albeit prone
to the interpretive problems discussed in Chapter 1.
Evaluation of the archaeological evidence for
population movements and colonization entails the
investigation of Manachaqui Cave's prehistoric functions,
501
502
and assessment of the wayside station working hypothesis.
Once Manachaqui Cave's functional contexts are understood,
then full attention can be devoted to critical examination
of diagnostic indices of migration such as anomalous
disconformities in assemblage structures or contents.
Gaining an understanding of the rockshelter's functional
contexts is likewise crucial to the formulation of
alternative hypotheses.
Re-interpretation of the "migration
evidence" is the subject of the final chapter.
Discussion of the evidence for and against migration
from the rockshelter and from the study area will proceed
phase by phase.
Unfortunately, the utilization of
diagnostic indices requiring high stratigraphic and temporal
resolution described in Chapter 1 is infeasible given the
nature of Manachaqui's palimpsest occupation floors.
Discussion of the Manachaqui evidence will be followed by
evaluation of migration hypotheses from other localities in
the eastern montane forest.
First, however, brief summaries
of study area pre-Conquest language and human biological
distributions will be presented.
Evidence from Historical Linguistics
Chapter 3 presented evidence that southern Chachapoyas
societies and their Amazonian Cholon and Hivito trading
partners in the Central Huallaga valley were brought under
the control of the Inca.
During successive periods of Inca
and Spanish domination of the Marafion-Huallaga divide,
the
503
Quechua language was imposed as lingua franca in order to
facilitate administration and control of the local
populations.
This imposition, coupled with the catastrophic
depopulation of both highland and lowland areas, ultimately
left only mestizo, Spanish-speaking populations and a lack
of scholarly consensus regarding the area's pre-Inca, and
perhaps pre-Quechua, languages.
The linguistic problem encountered here is complex
because it is the language "substrate" that is sought rather
than the last native language documented by European
observers.
In the studies described below, place names and
personal names that often survive language replacement
provide scholars with clues, but conclusive evidence has not
been forthcoming.
Tello's and Lathrap's hypothesis would be
served by the identification of an Arawakan sub-stratum, and
Isbell's hypothesis requires the substantiation of a two
thousand year development of Quechua A in the study area.
Verticality scenarios might be served by finding evidence of
language archipelagos to match colony archipelago
distributions, all of the migration hypotheses would be
strengthened if pre-Inca Chachapoyas and Central Huallaga
languages were identical.
A survey of opinions for the
Chachapoyas and Central Huallaga areas is recounted below.
Chachapoyas Languages
While evidence for pre-Inca Quechua dialects would
clearly serve Isbell's and Lathrap's hypothesis of early
504
Quechua expansion from the northeastern Peruvian or
Ecuadorian lowlands, some scholars have considered the
"modern" Chachapoyas dialect of Quechua A (Torero's Quechua
2) only a late prehistoric veneer spread as consequence of
Inca imperial expansion (Rowe 1946:185; Torero 1974; Bird et
al. 1983-84).
As noted in Chapter 2, Isbell challenges
Torero's and Bird et al.'s hypothesized proto-Quechua
coastal Central Andean homeland (Isbell 1974:151, 198384:245), but neither he nor Lathrap provide evidence for
Quechua origins based upon analysis of linguistic data from
Chachapoyas or from the Ecuadorian Oriente.
The question
remains: did Inca Quechua replace other Quechua dialects in
Chachapoyas, or some other unknown non-Quechua language or
languages as Bandelier (1907), Zevallos (1987) and others
have suggested?
Culle (Torero 1974) and the Cajamarcas division of
Puruha-Mochicas (Jijon y Caamafio 1943) are extinct languages
believed to have been spoken in the Huamachuco and Cajamarca
northern Peruvian highlands prior to the expansion of
southern Quechua.
Little is known of Culle.
Jijon y
Caamafio (1943:463, Map III, No. 82) includes Chachapoyas
within the domain of Cajamarcas, which belongs to his
proposed Northern Andean Macro-Chibchan phylum.
(1950:194, Map 18) limits Puruha-Mochicas'
Mason
(he calls it
Yunca-Puruhan) range to the northern coast of Peru and
southwestern Ecuador, and regards affiliations of
505
Chachapoyas languages as undetermined.
Zevallos (1982) contends that Yunca toponyms were
introduced into the Chachapoyas province by Yunca-speaking
mitimae groups from the north-central coast relocated by the
Inca.
Based on an onomastic analysis of 645 names taken
from 16th century Chachapoyas documents, he argues against
their affiliation with any language group in the adjacent
Andes.
He suggests further inquiry into northern and
eastern lowland, especially Carib, relationships (Zevallos
1982:4).
In summary, there is no agreement whether an affiliate
of Quechua, Yunca, Chibchan, Carib or some other language
family (or families) was spoken among the ethnic groups
ultimately subsumed within the Inca province of Chachapoyas.
The adoption of a lingua franca (whether Inca Quechua or any
other) to facilitate interaction probably masks a complex
mosaic of indigenous languages in pre-Incaic Chachapoyas.
Chol6n and Hivito Languages
The extinct Chol6n and Hivito languages remain as
mysterious as other aspects of their pre-Hispanic culture.
Steward (1948:514, 598) reports that much Quechua spoken in
the central Huallaga during the nineteenth century was a
post-Spanish contact or late prehistoric introduction.
The
Amazonian partial assimilation of Andean Quechua supports
suggestions of habitual highland-lowland interaction in the
study area.
Regarding pre-Quechua language distributions,
506
Steward characterizes upper Amazonian language distributions
as extremely complex (Steward 1948:507).
Mason's map of
South American linguistic distributions shows both Hivito
and Chol6n as language isolates (Mason 1950: Map 18).
Chol6n has been more adequately documented (by de la
Mata 1923; Tessmann 1930:546-547) than Hivito, but most
scholars regard them as phylogenically related.
As noted in
Chapter 2, Tello (1942:629) regards both, with Arnuesha and
Muchik (Yunca), as belonging to a larger family of Arawakan
languages representing Central Andean civilization's
demographic sub-stratum.
Jij6n y Caarnafio (1943:465) groups
"Cholona" along with Chachapoyas in his Macro-Chibchan
language phylum, although Mason (1950:192) views Cholona's
relation to Chibchan as "doubtful."
Tessmann (1930:626) finds similarities between Chol6n
and Quechua.
The degree of Quechua infusion remains
unknown, but it may be attributed to commercial and social
modes of interaction dating to the Late Horizon, if not
earlier.
Evidence supporting a hypothesis of substantial
borrowing from Quechua by the Cholones and Hivitos comes
from the foot of the eastern Andes farther south at Cerro de
la Sal where Wise (1976) observes that the Amuesha's
assimilation of Quechua elements has hindered efforts to
classify their indigenous language.
Much of the lexicon
borrowed from the Quechua highlanders is related to "social
and commercial dealings" (Ibid.:358),
including words for
507
buying and selling.
The Inca are presumably responsible for
many intrusive features of Amuesha language and mythology,
and commerce sponsored by Colonial Period missionaries that
employed Quechua trade jargon also exerted a lasting impact.
Most importantly, Wise observes that lexical borrowing from
Quechua has led to transformation, and even partial
replacement of Amuesha Arawakan (Ibid.).
Wise (1985:208)
has recently listed Cholon and Hivito as both "unclassified"
and "extinct."
Other recent classifications group them
(Cholon specifically) with other Andean languages (Landar
1977; Greenberg 1987; Kaufman 1990) including Quechua.
According to Tessmann (1930:626), Hivito may have been
mixed with Pano-Ge.
Pefia Meza (1935:344) reports that
during the 19th century, an isolated band of refugee Hivitos
and Panoan-speaking Conibos from the Ucayali inhabited the
forested slopes between the Huallabamba and Jelache Rivers
(north of the Abiseo) .
Consequent intermarriage may account
for the mix later observed by Tessmann.
The central
Huallaga region's role as a refuge for natives fleeing
Spanish slavers (Reeve 1994:123) and retribution for
rebellious activities (de la Riva Herrera 1907) undoubtedly
contributed to additional linguistic mixing since the
sixteenth century.
Hivito and Cholon apparently absorbed
multiple linguistic influences and we may not attain a
clearer picture of Central Huallaga indigenous languages
unless more illuminating Colonial missionary documents turn
508
up.
The evidence at hand both here and in southern
Chachapoyas, however, cannot be easily utilized to support
of refute any of the postulated migrations.
Evidence from Physical Anthropology
Needed for systematic studies of prehistoric population
distributions are large skeletal samples from welldocumented geographic and temporal contexts.
The semi-arid
micro-climates in which Chachapoyas dead were often interred
favor preservation of skeletal material, and such studies
based upon field investigations are feasible.
For example,
a study of 154 skulls from five funerary sites in
Chuquibarnba by Jakobsen et al.
(1986-87) suggest the
presence of two quantitatively identifiable populations.
The local Late Intermediate Period population tends to be
dolichocephalic and rather typical of Central Andean
highland regions of nearby Cajarnarca, while skulls from the
Late Horizon site of Salsipuedes may pertain to an intrusive
population from the south, perhaps arriving as a result of
Inca-directed population movements (Ibid. :154).
Examination
of stature produced evidence of a similar north-south
dichotomy.
Salsipuedes individuals were on average 4 ern
shorter.
Such data may be useful to evaluate the origins of
Chachapoyas populations (Ibid. :154-155), especially since
some comparative information is available from studies
elsewhere on the continent (Stewart and Newman 1950).
Truly
509
necessary, however, are human remains representing
prehistoric populations in surrounding lowland and highland
regions.
Studies of modern populations in these areas may
offer clues, but these have not been conducted either.
Demographic shuffling under the Inca, followed by chaos and
depopulation under the Spaniards, will have undoubtedly
produced complex distributions and numerous interpretive
problems.
On a larger scale, paleodernographic relationships
between the upper Amazon and Central Andes are impossible to
examine while skeletal samples from the prehistoric humid
lowlands are scant.
Evidence from Archaeology: Manachaqui Cave
and the Pataz-Abiseo Area
In this section, the material evidence from Manachaqui
Cave will be brought to bear on the population movement and
colonization hypotheses for each phase.
A point of
departure for consideration of the postulated migrations
will be provided by the assessment of the rockshelter's
function.
The reader will recall from Chapter 1 that some
functional contexts are better suited than others to
effectively perform the comparative analyses required to
evaluate the postulated migrations.
Specifically, domestic
or quotidian contexts are more desirable than ceremonial
contexts.
Discussion of botanical, faunal and artifactual
evidence for site function from each phase will be followed
by detailed evaluation of pertinent migration hypotheses.
510
Prior to discussing the phase data however, some general
observations can be offered regarding the spatial patterning
of cultural remains throughout the Manachaqui sequence.
Based on her analysis of the botanical remains from
Sector A's sequence of floors,
Pearsall (Appendix F)
concludes that her study "generally support(s) the
hypothesis that Manachaqui Cave was a camp for travelers for
much of its history."
As evidence she notes the low
quantities of food remains relative to wood charcoal and the
presence of "exotic" food items during the entire sequence.
Problems with the interpretations of maize evidence were
cited in Chapter 4.
In the ratio histograms presented
within Appendix F (Pearsall's Figs. 1-7), several floors
show marked deviation from relatively even distributions,
and these will be noted below.
One of the most striking patterns discovered during the
excavation of Manachaqui's Sector A, was the consistency
maintained by the spatial distributions of features
hearths), artifacts
a~d
surface down to bedrock.
(e.g.
organic remains from the site
The ethnographic analogy and
surface distributions appear to be effective tools for
interpretation of subsurface prehistoric patterning and
identification of activity areas.
Virtually all of the
hearths were prepared on the west side of the shelter to the
extent permitted by the shelter's wall and bedrock
protrusions.
The two exceptions are visible in the south
511
profile of Unit 1 (Fig. 18) and the east profile of Unit 3
(Fig. 14).
Evidence that east side surfaces were repeatedly
scooped out to create a concave sleeping surface is seen in
the dipping strata visible in Unit 1 and 2 profiles (Figs.
11 and 14).
The lack of orderly sequences seen in the Unit
1, 2, 11 and 25 macrochronologies
(Appendix D) attest to
heavy disturbance of the east side's stratigraphic
integrity.
Kent's faunal analysis provides additional evidence for
repetitive spatial patterning throughout Manachaqui Cave's
prehistoric occupation.
His tallies of specimens on the
west side (Units 12, 13, 14, 15, 16, 17) and east side
(Units 1, 2, 3, 11, 25) of Sector A reveal that east side
units yielded only 74 specimens while west side units
contained 2,284.
Nearly a third of the west side faunal
material (745 specimens or 32.6%) is classified as
unidentified scrap, while east side excavations produced no
scrap at all.
Kent concludes that most of the bone
deposition, activity and resultant trampling took place on
the west side while the east side was kept relatively clean.
Kent's conclusion ostensibly supports the inference
that the shelter's occupants habitually cooked and ate
around the west side hearths, while the east side
consistently served for sleeping.
However, ceramic and
lithic remains are more concentrated on the east side, away
from the hearths, where they average 79.6 sherds per
512
excavated 5 em level compared to the west side's 60.6.
sherds.
If the samples from Units 16, 17 and 3 at the
shelter's mouth are omitted, the east side average rises to
85.8 sherds per level and the west side average drops to
38.3.
Lithic remains show similar patterning with 12.7
artifacts per level on the east side compared to 3.9 on the
west.
These patterns contrast with artifact distributions
documented at the Initial Period rockshelter herding camp at
Telarmachay where lithic debris and soot-caked potsherds
clustered together with faunal remains around the hearths
(Lavallee 1977:87).
While cleaning activities conceivably account for the
Manachaqui Cave distributions,
the separate distributions of
hearths and food refuse on the west side, and pottery and
lithic refuse on the east side may also indicate that
pottery was not necessarily utilized for cooking and stone
tools were not utilized strictly for food processing.
This
co-variation is consonant with the interpretation that preprepared food was routinely brought into the shelter and
consumed around the hearth where the shelter's occupants
warmed themselves.
In this scenario, personal pots and
stone tools might have been kept close to their owners.
Because of stratigraphic mixing in Sector A's east side, the
spatial distributions of the artifacts cannot be broken down
by phase, but other kinds of archaeological patterning will
be examined phase by phase below.
513
Pre-Lavasen occupations and
the Lavasen Phase
Pre-Lavasen and Lavasen Phase occupations at Manachaqui
Cave await future opportunities for more thorough analysis.
Some general observations are offered here because there are
preliminary indications of major changes in the
rockshelter's functions prior to the Initial Period
Manachaqui Phase.
The reader will recall from Chapter 5
that there are no apparent pre-Lavasen strata in Sector A,
and Sector B's pre-Lavasen Phase strata yielded a variety of
functional stone tool types along with two radiocarbon dates
of 2120 and 2330 b.c.
Associated projectile points,
scrapers and burins indicate that activities at Manachaqui
included hunting and the processing of animal products.
The
importance of hunting and butchering at Manachaqui Cave is
difficult to evaluate given the poor preservation of the
pre-Lavasen faunal sample.
Additional evidence for hunting activities is provided
by the statistical distribution of animal sizes represented
among the pre-Lavasen remains (Fig. 111).
Although poor
preservation may bias the pre-Lavasen distribution to some
extent, a prevalence of medium to large-sized animals
(e.g.
pacas, pudus and cervids) is documented in Kent's analysis.
Despite the occasionally small sample sizes, the changing
distributions of faunal remains from pre-Lavasen times to
the Early Horizon Suitacocha Phase clearly show a
514
diminishing utilization of large animals and a corresponding
increase through time in the utilization or consumption of
small animals (Table 28, Figs. 111-116), perhaps as preprepared travel food like that described by Young and
Morales (see Chapter 4).
Similar emphasis on relatively large game is evident at
Early and Middle Preceramic Northern Andean sites like
Chobshi Cave in Ecuador (Lynch and Pollock 1981: Cuadro 1,
p.98) and Nemocon and Sueva Rockshelters in Colombia
(Correal 1979: Cuadros 1, 4) where large and medium sized
animals (white-tailed deer, pudu and paca) were most
frequently utilized.
Faunal assemblages from Preceramic
Period Central Andean puna rockshelters consist of primarily
cervids and camelids (e.g. Rick 1980: Table 10.1, p.234;
Wheeler 1985: Tableau 2).
and manes
II
and
II
II
Hundreds of ground-s tone slabs
small mammals, rodents and birds
II
dominate
lithic and faunal assemblages from Middle Preceramic house
remains clustering just below the Upper Zafia Valley western
montane forests (Dillehay et al. 1989:749-750), but these
were apparently left by semi-sedentary horticulturalists
rather than by mobile hunters and/or herders.
Stone
grinding tools are rare or absent in Manachaqui Cave strata,
and the few pre-Lavasen Phase macrobotanical samples
recovered from Sector B have not yet been studied.
Nearly
all of the animals represented in Manachaqui's faunal
collection for all phases are native to paramo and forest
515
edge habitats.
During the Lavasen Phase we find the first evidence for
conditioning Sector A, presumably for habitual and/or
prolonged use.
The catch-basin and drainage canal (Feature
R-5) unearthed in Unit 12 would likely have been destroyed
by runoff entering from the rear of the shelter interior
(where Units 9 and 10 are located) had it not been prepared
in combination with some sort of sod or stone wall.
It
remains unknown whether Feature R-5 represents the first
attempt to condition Sector A, or if it is simply the first
to be preserved by site formation processes.
Lavasen Phase artifacts in both sectors include stone
tools produced by a simple core-flake industry, but analyses
of raw materials and tool functions have not yet been
attempted.
Much of the selected stone has poor flaking
qualities and was probably opportunistically gathered from
the surrounding subalpine valleys.
The paucity of faunal
remains suggests that the tools served functions other than
meat processing.
It may be significant that small animals
constitute a larger proportion of the faunal sample (Fig.
112)
Pearsall's botanical evidence for Manachaqui Cave's
functions appears in Appendix F as food:wood ratios with
(Appendix F, Fig. 1) and without Festuca (Appendix F, Fig.
2), and local:exotic+local counting unidentified fruit rinds
as local (Appendix F, Fig. 3), and excluding fruit rinds
516
altogether (Appendix F, Fig. 4).
Charred Festuca seeds may
have been left by burnt straw bedding.
However, an
estimated 344 of the 404 (85%) Festuca seeds from Floor CC
were recovered from its rock-filled hearth.
quantity of 3,132
The inordinate
(estimated) Festuca seeds from Floor BB
were all found concentrated within its associated simplebasin (but perhaps re-utilized, rock-filled) hearth.
This
patterning strongly suggests that these locally gathered
seeds were being cooked as food.
High food:wood ratios in
Floors FF through BB (Appendix F, Fig. 1) support a
habitation scenario rather than the wayside station working
hypothesis.
The local:exotic+local ratio (Appendix F, Fig.
3) also supports hypothetical habitation, but the appearance
of maize drastically lowers the ratios above Floor
z.
The stone-filled hearths in Floors FF through CC
suggest cooking, perhaps by stone-boiling in vessels
fabricated with perishable materials, or roasting seeds
and/or meat among the hot rocks.
Neither possibility seems
more likely than the other given: 1) the difficulty of
stone-boiling at high altitude and 2) the scarcity of animal
bones during this phase (N=21).
The animal sizes
represented in the small Lavasen faunal sample reflect
utilization of the full spectrum of fauna characteristic of
an ecotone setting.
Animal sizes show a distribution
similar to that of the larger pre-Lavasen assemblage, but
the number of small animals increases.
The only other
517
Andean rockshelter to show relatively high proportions of
small animal specimens is Guitarrero Cave (2,580 m), but
Wing (1980:152-153) reports that the profusion of unburned
bones of opossums and "small rodents" is "incidental" and
unrelated to subsistence.
Needed for future reconstruction
of Manachaqui Cave's changing behavioral contexts is an
analysis of changing animal consumption patterns featuring
anatomical breakdowns by species for each phase.
If we also consider Rodbell and Hansen's palynological
evidence for anthropogenic disturbances in the area after
4000 b.c., then it seems likely that the beginning of the
Lavasen Phase at Manachaqui Cave was marked by the local
introduction of crop cultivation, especially of puna tubers
which are seldom preserved in archaeological botanical
assemblages but might have been roasted in the rock-filled
hearths.
Wild Festuca seeds also contributed to the diet.
Thus, evidence for the utilization of mostly Tropical Alpine
Zone resources suggests that Manachaqui Cave likely served
as a short-term habitation for individuals or small families
engaged, at least seasonally, in seed gathering and tending
crops.
The botanical taxa listed are mostly the same as
those utilized at the "hunting base camp" of Pachamachay
(Pearsall 1980).
Manachaqui may have served intermittently
as a wayside station for travelers, possibly carrying preprepared meals that included guinea pigs and small game.
At Manachaqui Cave there is no reason to suspect that
518
any changes at the onset of the Lavasen Phase should be
attributed to migrations.
The Late Preceramic Period
throughout the Central Andes was marked by ongoing economic
and social change (Quilter 1991).
It should be observed,
however, that Preceramic Period human occupation of the ceja
de selva is not anticipated by many migration theories
recounted in Chapter 3.
Colonization scenarios often fail
to address the problem of indigenous eastern montane forest
populations.
Lathrap et al.
(1985) refer to the eastern
montane forest prior to mid-Early Horizon migrations from
the Ecuadorian Oriente as an "empty niche."
The evidence
just recounted suggests utilization of the Tropical Alpine
Zone by groups fully acquainted with and adapted to the
local environment.
The Manachaqui Phase Occupation
The Manachaqui Phase remains provide evidence for the
intensification of the rockshelter's use, perhaps
alternating as a temporary habitation and trailside
waystation.
Suggesting habitual use is the attention given
to refurbishing the shelter with masonry plugs at the rear
(Feature R-7), and walls (R-6 and R-8)
mouth.
flanking the cave
The occupants of Floor AA would have had ample space
to move about beneath a ceiling 1.60 m high.
Although
unequal in height, R-6 and R-8 may have functioned along
with a simple roof of branches or cana and ichu to enclose
an effective sheltered space extending a meter or two from
519
the cave mouth across the berm.
The great fallen stone slab
visible in Unit 26 and the northern edges of Units 18
through 21 (Fig. 28) provided a low wall partially enclosing
the front of the shelter.
In this "enclosure scenario" the leveling and
maintenance of interior floors, combined with movement
within the confined space, may be responsible for the
turbated condition of Manachaqui Phase levels beneath Floor
AA (Levels 43 to 50 in Units 15-17).
Of the two Manachaqui
Phase floors, Floor Z stands out as possessing two hearths,
one of which is the largest unearthed at the rockshelter.
Floor Z's Hearth 1 is also noteworthy because it was
centrally situated at the mouth of the shelter.
Here the
shelter's dripline, coupled with partial exposure to the
elements, would have likely disturbed or obliterated at
least part of the hearth if not for protection by a roof.
The partially scattered charcoal fill on the western side of
Lavasen Phase Floor EE's hearth (Fig. 25b) was probably
dispersed by exactly this kind of exposure to heavy rain and
roofline dripping.
Evidence from Botanical
and Faunal Remains
Floor Z's Hearth 1 is also the only Sector A hearth
containing potsherds.
Although the organic content of Floor
Z's hearths has not yet been examined, the Floor Z sediments
contained high food:wood ratios (Appendix F, Figs. 1 and 2)
520
and a high local:local+exotic food ratio if the problematic
fruit remains are regarded as local (Appendix F, Fig. 3).
Floor AA's two sets of ratios present conflicting evidence.
High local:local+exotic food ratios contrast with very low
food:wood ratios.
Overall, the Manachaqui Phase taxa remain
unchanged from the previous Lavasen Phase, except that
Lupinus and Chenopodium/Amaranthus are now absent.
The Manachaqui Phase faunal remains show the diminished
importance of the larger animals utilized during the
Precerarnic Period and a predominance of medium and smaller
animals
(Fig. 113).
Initial Period travelers may have
consumed guinea pigs as pre-prepared trail meals as modern
villagers do today.
However, Kent (1994:25-26) points out
that guinea pig (Cavia porcellus) remains might be regarded
as evidence of the shelter's function as a habitation since
the animals frequent kitchen areas of Andean households
where they find warmth and food scraps.
They do not survive
left alone at high altitudes for more than a few days.
Hence, guinea pig remains alone do not constitute sufficient
evidence for either wayside station or habitation functions.
Nonetheless, the shift to smaller game is uncharacteristic
of coeval Andean faunal assemblages as is the sudden
disappearance of the chipped-stone tool industry.
The use
of such tools may have been obviated by habitually bringing
pre-prepared meat to Manachaqui Cave.
Pre-prepared meat
could have been carried by travelers sheltering for the
521
night, but it also may have accompanied individuals camped
for a number of days while dedicated to specific activities
such as tending a tuber crop.
In sum, the faunal assemblage
provides only equivocal evidence of site function, perhaps
thereby suggesting multiple functions, including both
habitation and expedient use as a wayside station
Evidence from Artifacts
Although the use of Lavasen Phase rock-filled hearths
remains unknown,
the shift from stone-filled to simple basin
hearths at Manachaqui Cave might be likened to evidence from
Chiapas where Clark and Grosser (1995) describe an inverse
relationship between decreasing quantities of fire-cracked
rock and the increasing frequency of pottery vessels
presumably utilized for cooking.
At Manachaqui Cave
however, it is not clear which of the Manachaqui Phase
vessel types might have been utilized for cooking.
None of
the five shape classes show patterned fire-blackening,
soot
or other organic residues like the Initial Period
Telarmachay pottery assemblage (Lavallee 1977:87), although
food may still have been warmed in pots prior to
consumption.
Additional analysis of Manachaqui Cave's
Preceramic Period remains may help resolve these questions.
The Manachaqui Phase addition of pottery offers another
interpretive tool for evaluating the rockshelter's function
during the Initial Period.
My literature search failed to
turn up archaeological "transport assemblages" other than
522
mass-produced containers for long-distance commercial
shipping of oil or wine described for the Old World.
In
order to formulate a series of expectations with which to
generate reasonable inferences regarding vessel functions,
I
consulted several references, among which Rice's (1987)
sourcebook, and Henrickson and McDonald's ethnographic
survey (1983) were most useful.
In her discussion of vessel
functions, Rice (1987:225) notes that the many decisions
made during pottery production are structured by
considerations of the vessel's projected capacity,
stability, accessibility of contents and transportability.
Using her observations and assumptions, the Manachaqui
Phase's potential utility as a transport assemblage can be
assessed.
According to criteria emphasized by Rice (1987:125141), Manachaqui Paste A ware is ideal for transport.
The
pottery is thin, fired very hard and is therefore
lightweight (although perhaps brittle) .
Restricted shapes
provide adequate containment of vessel contents during
transport.
The round base is appropriate for carrying in a
cloth bag, as are the rough exterior surfaces which prevent
slippage.
Recall that only the rims of Shape B jars are
consistently burnished.
Flanges provide purchases for
handling and also prevent slipping.
Thickened lips
strengthen everted rims and furnish a means of fitting a
skin or lid over the orifice for greater security during
523
transport.
As expected, the assemblage is mostly composed of small
vessels.
With some exceptions, Manachaqui Phase vessel
sizes are uniformly small, while many Central Andean
assemblages feature both small and larger, bulky vessels
intended for use around the home.
Early Guafiape neckless
ollas range from 30 to 50 em in maximum diameter (Strong and
Evans 1952:253).
Wairajirca Phase neckless ollas at Kotosh
are likewise very large.
While Manachaqui Phase Shape A Rim
2 and Rim 5 neckless jar categories include large variants,
there are relatively few of these vessels.
The dual size
modes for the Shape A jars may reflect the site's alternate
functions as habitation and wayside station.
Perhaps vessel size comparisons should center on
Manachaqui Phase Shape B short-necked jars which dominate
the Paste A assemblage.
With a modal rim diameter of 12 ern,
and modal waist diameter of 16 em, Shape B jars are smaller
than their Wairajirca counterparts with rim diameters
between 15 and 19 em.
On average they are smaller than
analogous Huacalorna Coarse Brown jar Forms 4 and 5, and
Brown Smoothed jar Forms 1, 2 and 3 which show larger rim
and waist diameters and greater size variability.
They do
seem to share similar rim diameters with Machalilla Form 11
necked jars, but Form 11's rim and body proportions are
unknown.
Of course the Machalilla assemblage includes a
much wider diversity of vessel shapes.
524
Specific references to vessels designed for transport
include Henrickson and McDonald's {1983:634) description of
a "true canteen" for water transport showing a maximum
diameter between 10 and 20 em, and maximum height of 20 em.
With a modal width of 16 em, Manachaqui Phase Paste A
vessels meet these criteria.
Only thirty-two of 208 sherds
{approx. 15%) from carinated and semi-carinated vessel midsections measured show diameters greater than 20 em.
Although vessel heights for the fragmented Paste A
assemblage could not be measured, none of the hypothetical
reconstructions project maximum heights greater than 15 em.
DeBoer {n.d.) describes historic Amazonian Shipibo vessels
utilized as portable cook pot,
(kenti vacu), traveler's
canteen (chomo vacu), an individual drinking mug (kenpo
vacu) and a small food bowl (kencha vacu), and provides
volume calculations for each.
Manachaqui volummetric data
are unavailable for comparison, but DeBoer's illustrations
of all four Shipibo vessels depict sizes well within
Henrickson and McDonald's range for canteens.
Rice {1987:241) suggests that small orifices are
preferable for transporting liquids and that the addition of
vessel necks facilitates pouring.
She observes that liquid
containers usually show burnished and/or slipped (or glazed)
interior surfaces to minimize absorption into the vessel
walls, but this may not be true in the Andes.
Manachaqui
Shape B vessels exhibit variable attention to interior
525
surface finish,
and vessel proportions may be a convenient
compromise taking both secure containment and easy access to
dry foods into account.
The Manachaqui Phase assemblage's
embellished rims (especially the notched rims) are
impractical for pouring liquids.
Archaeologists have found that vessel sizes correlate
roughly with the size of groups serviced (Turner and Lofgren
1966; DeBoer n.d.).
The previously noted bi-modal
distributions of some Manachaqui Phase vessel types may
prolonged stays by larger groups.
Among the Shape B necked
jars, larger orifice diameters (up to 24 em) lend decorated
Rim 9 and 10 vessels an unusually wide range of sizes.
Numerous individuals, perhaps a small family of semipermanent shelter inhabitants might be serviced by the
larger, more elaborately decorated vessels.
Inordinately
large Shape A neckless jars (e.g. with 19-29 em rim
diameters) would have been awkward for transport on long
journeys, and may instead provide ceramic evidence for the
rockshelter's alternate function as a permanent or semipermanent habitation.
A few ground-stone artifacts may also serve as evidence
for prolonged stays at Manachaqui Cave.
Slate point
fragments and chips in all of the phase assemblages suggest
that points were habitually worked at the site, but
Manachaqui Phase ground-stone tools appropriate for shaping
and polishing the points represent an additional stage in
526
tool manufacture that may indicate extended occupations.
A
rare combination edge-ground cobble and hammer from probable
Manachaqui Phase deposits in Unit 11 may have been utilized
for processing food within the shelter.
These items provide
potential evidence for tool working and food preparation,
activities that might take place in an Andean habitation.
Only the fragment of quartz resembling other crystals found
in Andean ritual contexts (Shady 1983; Burger 1984b) offers
evidence of ritual activities.
In general, the evidence for Manachaqui Cave's Initial
Period functions suggests both habitation and use as an
overnight shelter.
However, contradicting interpretations
that the rockshelter may have housed permanent
agriculturalists, or that it was functionally connected to
intensive agricultural activities at any time, is the
virtual absence of milling stones, hoes, clod-breakers and
other such artifacts.
Consequently, the wayside station
function appears predominant.
The five-fold Manachaqui
Paste A vessel shape inventory resembles the five-part
breakdown of shape categories at Cajamarca sites like Cerro
Blanco.
It is most probable that, like the Shipibo (DeBoer
n.d.), Manachaqui Cave's Initial Period users chose to
travel with small-sized pots representing the same shape
categories available for quotidian use in their domiciles.
The Paste A assemblage should therefore be amenable to the
kinds of comparative analyses required for this study.
527
Manachaqui Phase Evidence
for Migrations
Chapter 3 detailed the various migration theories
postulated to explain eastern montane forest settlement and
utilization.
hypotheses,
According to Meggers' and Kauffmann's
the Manachaqui Phase Paste A ceramic assemblage
should belong to Andean populations moving eastward from the
highlands to the lowlands, while a Central Andean highland
source is also presupposed by Moseley's highland-based
verticality model.
Theories of upslope population movement
postulated by Tello and Lathrap dictate that Manachaqui
Phase Paste A pottery, and other local styles along the
Marafion-Huallaga divide between Bagua and Chavin de Huantar,
should be derivatives of early upper Amazonian styles.
These migration hypotheses can be evaluated empirically
by considering the spatial and temporal patterning of early
highland and lowland artifact styles.
In Chapter 1 it was
noted that the identification of ostensibly supportive
stylistic distributions is fundamental to the plausibility
of migration theories.
Highland to Lowland Migrations
Of the proposed migrations, eastward downslope
incursions into the Pataz-Abiseo study area, whether by
expanding highland populations or by highland colonists, are
the least plausible.
The distinctive Manachaqui Phase Shape
528
B necked jars, carinated body profiles and elaborate
incised-applique decorations are rare or absent from early
assemblages in the adjacent Central Andean highlands to the
west and south.
Initial Period Huamachuco appears to
represent a demographic and developmental lacuna (Burger
1992), hardly the source of surplus populations or
archipelago colonies.
Even Manachaqui Phase Shape A jars or
"neckless ollas" appear more closely allied to northern and
eastern lowland neckless jars than to their Central Andean
counterparts.
While we expect that a special function site
like Manachaqui Cave would contain only a subset of the full
range of vessel shapes and sizes potentially available to
the site's users,
the Paste A assemblage actually shows a
wide assortment of relatively complex vessel shapes and a
rich decorative style with no clear highland antecedents.
As the Manachaqui Phase assemblage finds its closest
resemblances in the coeval Pandanche A, Early Huacaloma, La
Conga and Montegrande styles of Cajarnarca, proponents of
migration scenarios may point to the far northern Central
Andes as a likely wellspring for Pataz-Abiseo area
populations.
However, these styles cannot be regarded as
indigenous to either the highlands or the lowlands.
The
lack of unequivocal evidence that any of these Cajamarca
styles predates the others precludes conjecture on origins.
Also,
the irregular distributions of shared attributes among
these closely-related styles contradict expected wave-like
529
patterning left in the wake of population movements (Rouse
1986:177).
All of the early Cajamarca ceramic assemblages
differ from one another in design details.
Yet even if all
of these conditions were met, migration advocates would be
forced to explain how it behooved Cajarnarca montane forest
populations to colonize similar montane forest environments.
Meggers has ascribed resemblances between the coastal
Ecuadorian Valdivia/Machalilla styles and Initial Period
Kotosh ceramics (e.g. tiered shoulders, zoned punctation,
dot-ended incision, zoned dentate, dot-in-ring and zoned
parallel hachure)
to stylistically and temporally
intermediate "stepping stones" still to be discovered in the
intervening Andean highlands (Meggers et al. 1965:174).
Neither Manachaqui Cave nor other recently investigated
Peruvian highland sites show evidence of the postulated
tell-tale route of descent to the Central Ucayali and
surrounding western Amazonian lowlands.
Unfortunately,
Meggers' postulated migratory forest-free corridors
providing passage to the lowlands cannot be credibly
substantiated or discounted with the new Manachaqui Valley
pollen evidence.
Locating such hypothetical eastern slope
corridors in prehistoric time and space presents a daunting
challenge for paleoenvironmental reconstruction.
Lowland to Highland Migrations
As noted in Chapter 2, Tello, Lathrap, Lanning and
others have emphasized the importance of the Marafion River
530
valley as an Initial Period migratory and/or diffusionary
conduit leading from the Amazon Basin to the Central Andean
highlands, and especially to Chavin de Huantar.
However,
the deeply incised Marafion Canyon has negligible bottomland
for settlement by more than dispersed agricultural
populations.
While it is not impossible that cultigens,
technologies, ideologies and their bearers worked their way
up the valley, the upper elevations of the Marafion-Huallaga
divide (above the Dry Forest Zone) are more likely to have
attracted agriculturalist settlers.
Traditional travel and transport routes in the
precipitous eastern Andes tend to follow ridgetops rather
than the circuitous, and often impassible, river bottoms.
The pre-Hispanic Huanuco to Chachapoyas road described in
Chapter 4 and the Suarez ridge reported by Hastings (1985)
provide examples.
For upslope Initial Period population
movements from lowland Amazonia, the valley slopes and
ridgetops above the Utcubamba river would have provided a
likely entry as it purportedly did for Langlois' Utcubamba
migrants nearly a thousand years later.
It was to the
Chachapoyas province that Tello had gravitated searching for
Chavin civilization's origins when his 1937 expedition to
the Marafion ended prematurely (Tello 1938; Mejia X. 1956).
According to Lathrap (1970:107), archaeological
evidence for a "collateral relative" of the Tutishcainyo and
Wairajirca styles should be found on the narrow Central
531
Huallaga floodplains adjacent to the Pataz-Abiseo area.
He
also indicated the importance of the bend of the Marafion in
the Jaen-Bagua area (Ibid.:107-109) investigated by Rojas
(1969), Miasta (1976) and Shady (1987).
Evidence for both
Tello's and Lathrap's postulated migrations should therefore
be found near the mouth of the Utcubamba River as well as
distributed along the Marafion-Huallaga divide.
At the mouth
of the Utcubamba, Bagua's sequence of early assemblages
consistently reflects stylistic development and patterned
interaction occurring widely across the adjacent Cajamarca
montane forests, rather than derivation from Central Ucayali
or other reportedly early Amazonian styles.
Some recent and
potentially important investigations of early ceramic
occupations in the upper Amazon potentially bearing on this
problem have not yet been fully reported (Morales 1992;
Ledergerber-Crespo 1995).
At Manachaqui Cave, the most conspicuous potential
evidence of upslope population movements from the adjacent
lowlands is the carinated vessel profile, hallmark of the
Central Ucayali Tutishcainyo phase assemblages.
Lathrap's
claim that Ravines had recovered Late Tutishcainyo ceramics
from the Central Huallaga (Lathrap et al. 1984:46) is
apparently a mis-reading or misinterpretation of Ravines'
reports (e.g. Ravines 1978, 1981).
Lathrap does not
elaborate on claims that Pandanche A and Cotocollao are
derivatives of the Late Tutishcainyo style.
Tutishcainyo-
532
like low-positioned carination angles and embellished
flanges do impart an Amazonian aspect to the Manachaqui
Phase Paste A style that is rare or absent in coeval
Cajamarca assemblages more frequently compared to late
Valdivia styles.
However, the principal Manachaqui Phase
vessel, the restricted Shape B necked jar, has no Amazonian
counterparts with the exception of Bagua's montane forest
Morerilla assemblage and perhaps the deep, carinated Yasuni
bowls.
The applique and incised-applique techniques that
typify the Manachaqui style contrast with the complex,
incised-line decorations characteristic of early Amazonian
styles.
The Manachaqui Phase evidence, especially when
combined with aforementioned evidence for the dearth of
lowland impact on the Initial Period pottery further south
at Piruro (Chapter 6), fails to support Tello's and
Lathrap's hypothesized lowland to highland population
movements.
The Manachaqui Phase Paste A assemblage is
stylistically unique, rich in singular details and
internally coherent.
described above,
Regardless of scattered resemblances
the assemblage cannot be considered a
derivative of any known Central Andean or Amazonian
tradition.
Manachaqui Phase Paste A shares more attributes
in common with distant Valdivia-Machalilla and perhaps Early
Cerro Narrio styles to the north than with styles
immediately to the west and east.
Migrations from these
533
Northern Andean areas are improbable given salient
differences between the Manachaqui assemblage and early
ceramic assemblages from intervening areas such as the Loja
and Catamayo valleys.
The Suitacocha Phase Occupation
The Suitacocha Phase occupation at Manachaqui Cave
lacks the range of associated architectural evidence that
was available for evaluating the rockshelter's function
during the previous Manachaqui Phase.
Two Sector A floors,
Floors Y and X, and Feature R-3, a horizontal sherd scatter
in Unit 21 Level 8, represent occupation surfaces, but
information regarding specific activities must be gleaned
from the organic remains and artifact assemblages.
The lack
of evidence for intentional reconditioning of the shelter
interior may be considered evidence for the cave's expedient
utilization strictly as a wayside station, but negative
evidence alone is not convincing.
Evidence from Botanical
and Faunal Remains
Important Suitacocha Phase additions to the inventory
of Manachaqui Cave's organic remains are maize and beans,
both "exotic" to the Tropical Alpine Zone.
Otherwise, the
utilization of local plant resources remains unchanged.
An
examination of Pearsall's food:wood and local:exotic+local
ratios (Appendix F, Figs. 1-4) for Floors X andY shows
relatively low values consistent with the wayside station
534
interpretation.
Suitacocha Phase faunal evidence now includes fish,
birds and reptiles occur, but quantities of these are not
significant and none of the specimens are clearly exotic to
the Manachaqui Valley.
Paca and armadillo specimens in the
Manachaqui Phase assemblage do not occur among the
Suitacocha Phase remains.
Sector A's Manachaqui Phase
deposits are more substantial and richer in faunal remains
than the Suitacocha Phase deposits (Tables 21 and 23), yet
the latter contain more rodent bones than the former.
The
Suitacocha Phase animal size distribution (Fig. 114) is now
clearly weighted toward utilization of small animals.
With
the addition of maize, the Suitacocha Phase organic deposits
may now reflect the patterned consumption of the travel
meals of parched maize and pre-cooked guinea pig for which
Cobo, Young and Morales provide ethnographic evidence.
Evidence from Artifacts
Artifact remains from Suitacocha Phase deposits also
provide support for the wayside station hypothesis.
The
scarcity of lithic remains has already been described as
potentially supportive.
The Paste A pottery continues to
show a preference for small, restricted vessels with round
bottoms, constricted necks and reinforced rims.
A
comparison between Suitacocha Phase and Manachaqui Phase rim
measurements presented in Tables 5 through 15 shows a number
of significant trends.
First, the Suitacocha Phase Shape A,
535
B and F categories generally do not include large variants
with rim diameters over 18 ern.
On average, more constricted
Suitacocha jar orifices are one centimeter smaller than
their Manachaqui Phase counterparts, and this may be equally
true for the smaller sample of Shape C vessels.
The addition of Shape F jars during the Suitacocha
Phase suggests that the two jar Shapes B and F may have
served separate functions.
Shape F's constricted neck and
extended rim or "collar" would be more appropriate for
transporting liquids although occasional similarities in
Shape B's and Shape F's rim morphologies may indicate at
least some functional overlap.
Unfortunately, the
reconstruction illustrated in Fig. 74 with an unusually
large rim diameter of 15 ern and waist diameter of 21 em
offers the only clue for interpreting jar sizes and
proportions.
If jar shape proportions are constant,
however, then vessels with rim diameters of 11 ern (the Shape
F mode) should exhibit modal waist diameters under 16 ern.
Recall that sixteen centimeters is the modal and the average
waist diameter of Manachaqui Phase Shape B carinated and
semi-carinated vessels.
From these calculations, we can
deduce as a working hypothesis that Suitacocha Phase jars
had volurnmetric capacities similar to Manachaqui Phase jars.
Finally, the Suitacocha Shape F rims may be shorter than
Bagua and Loja counterparts simply to minimize breakage
during transport.
536
The Suitacocha Phase remains indicate changes in
Manachaqui Cave's functions, and the trend seems to be
toward increased functional specialization.
Only a fragment
of a prismatic quartz crystal from mixed Suitacocha and
Colpar Phase deposits (Unit 18 Level 6), and a tiny flake of
mica might be interpreted as indicators of ritual
activities.
In the absence of additional evidence of
ceremonial activities, the wayside station working
hypothesis seems the most probable.
Suitacocha Phase Evidence
for Migrations
Migrations postulated for the Early Horizon occupation
of the eastern montane forest are said to have originated
either in the Central Andean highlands, or in the Amazonian
lowlands northeast of Manachaqui Cave.
The Suitacocha Phase
assemblage shows even less stylistic resemblance to
neighboring coeval Central Andean and western Amazonian
assemblages than that of the preceding Manachaqui Phase.
Stylistically, the Suitacocha assemblage might be likened to
a Northern Andean finger extending southward deep into the
Central Andean eastern montane forest.
Populations in
Huamachuco apparently remain dispersed, while nucleated
Cajamarca populations to the northwest continue to produce
large quantities of neckless ollas with the addition of
polychrome bowls.
In sum, hypothesized intrusions from the
adjacent highlands and lowlands find no support from
537
Suitacocha Phase Paste A assemblages at Manachaqui Cave.
The seemingly abrupt onset of the Suitacocha Phase, the
appearance of maize, and the northerly aspect of its pottery
style ostensibly bolster Lathrap and Isbell's postulated
intrusions of Quechua-speaking migrants from the Ecuadorian
Lathrap's (Lathrap et al. 1985) expression of full
Oriente.
support for Isbell's (1974) postulated Quechua expansion
conjoins the two complex hypotheses, although Isbell's
Quechua migrants were presumably entering the Central Andes
at Huanuco nearly a thousand years before Lathrap's migrants
supposedly arrived on the Central Ucayali bearing the Sivia
style.
According to this scenario, Suitacocha Phase Paste A
should represent the CB Series, pottery diagnostic of
Quechua-speaking intruders.
Manachaqui Cave lies directly in the path of Lathrap et
al. 's migration arrow (1985: Fig. 1) connecting Macas to the
Central Ucayali by way of the Marafion-Huallaga divide.
Despite evidence for the rockshelter's functional
specialization, the diagnostic features of Lathrap's and
Isbell's CB Series are sufficiently generalized across an
assortment of functional vessel types that effective
comparative analysis of the Suitacocha Paste A pottery is
feasible.
If the Suitacocha Phase Paste A pottery is a
member of the CB Series, then its stylistic derivatives
would include the Central Andean Kotosh Higueras and
Amazonian Sivia styles among others.
I intend to evaluate
538
the Quechua migration hypothesis empirically utilizing the
archaeological data, but some problems with the assumptions
common to both Lathrap's and Isbell's variants should be
noted first.
While some of his most basic theoretical assumptions
regarding relationships between language and culture were
challenged during Lathrap's lifetime, his convictions
remained mostly unaltered (see Lathrap et al. 1987).
Here
it may be more appropriate to point out problems with
assumptions specific to the Quechua expansion hypothesis.
First, the eastern slopes were not an "unoccupied niche"
(Lathrap et al. 1985:76) as the evidence from Kotosh and
Manachaqui Cave demonstrate.
Scholars such as Murra (1980),
Lathrap (1970) and Bonavia (Bonavia and Ravines 1967;
Bonavia 1978) have argued for associations between maize,
terraced-slope agricultural systems and Quechua-speakers,
yet evidence of ostensibly pre-maize occupations at
Manachaqui Cave suggests that maize (the alleged technoeconomic key to the opening of the new niche) was not the
determining factor that permitted eastern slope settlement.
Furthermore, Schjellerup (1985:119) has described large
terracing systems in Chachapoyas at altitudes between 3,200
and 3,800 m, too high for maize cultivation but ideal for
high altitude tubers and grains.
Admittedly, it is not
impossible that natural or artificially-created microenvironments permitted cultivation of the most frost-
539
resistant varieties of maize.
Regarding the location of the proto-Quechua homeland
there is still no consensus.
Stark (1985) has proposed an
Ecuador Oriente homeland and an initial westward expansion
around 400 b.c.
She links this expansion to Porras'
(1975:154-155) postulated upslope migration of populations
that purportedly left Cosanga-Panzaleo pottery distributed
throughout the highland Northern Andes.
A later split is
said to have sent populations to the north-Central coast of
Peru by a.d. 800, and the final pan-Andean spread of Quechua
occurred with the Inca conquest.
Most importantly, there is
no published evidence that Quechua was spoken in Chachapoyas
prior to Inca conquest.
Recall that non-Quechua language
substrates have been suggested by Bandelier (1907) and
Zevallos (1987).
The ceramic evidence is even more damaging to the
Quechua expansion hypothesis than the weakness of its basic
assumptions.
The Lathrap-Isbell hypothesis dictates that
Kotosh-Higueras should be derived from {or at least
historically related to) Suitacocha, yet the Suitacocha
Phase Paste A assemblage more closely resembles the Kotosh
Sajara-patac assemblage to which, by Lathrap's {1970:173)
own reckoning, Higueras is decidedly not related.
An
assemblage morphologically comparable to Isbell's CB Series
does not appear at Manachaqui Cave until circa a.d. 200-400
and the Empedrada Phase.
Empedrada Phase pottery generally
540
has thicker walls, but Empedrada pastes lack the diagnostic
"very coarse temper"
(Isbell 1974:139).
The purportedly
diagnostic CB Series necked jar shape occurs in Northern
Andean Valdivia, Bolivian Cochabamba's Formative phases and
all around the eastern Central Andes long before the Early
Horizon.
Finally, there is no evidence for CB Series-like
strap handles from the study area until the Late Horizon
Quechua presence (Bonavia 1967; Church 1988).
If Lathrap's hypothesis is correct, then the Sivia
style must count both early Upano valley and Suitacocha
styles as progenitors.
Of the 10 Sivia vessel forms
illustrated by Raymond et al.
(1975: Fig. 53), only Form 8
convex bowls might find a Suitacocha counterpart.
Lathrap's
Quechua-speaking migrants purportedly carried pottery
decorated with red-banded-incised techniques, stepped-fret
and locking scroll motifs and applique face motifs.
None of
these elements occur on Suitacocha Phase Paste A pottery.
To the above observations, DeBoer and Raymond's judgment can
be added: "if Sivia were brought in by a progressive
movement of Quechua-speakers colonizing the eastern slopes
and valleys, one would expect some evidence of a Sivia-like
complex to have turned up in the surveys of Hastings,
Lathrap, Lathrap et al. in the Alto Perene valley, or of
Ravines and DeBoer in the Huallaga valley"
Raymond 1987:127).
(DeBoer and
Based upon their analysis of the
evidence, these authors view Sivia as a lowland riverine
541
intrusion into the eastern slope premontane forests, rather
than an Andean intrusion into the lowlands and the Central
Ucayali.
The Colpar Phase Occupation
As Chapter 8 attempted to convey, the Colpar Phase
occupation at Manachaqui Cave is rather ephemeral, and most
clearly substantiated by Sector A's Floor W, and the hearth
in Unit 14's Floor P.
The paucity of evidence for intensive
use of any kind between 500 b.c. and a.d. 200/400 is not
inconsistent with the working wayside station hypothesis.
By 200 b.c., travelers probably utilized Manachaqui Cave en
route eastward to Gran Pajaten and other Montane Rain Forest
Zone settlements.
Pearsall's food:wood and
local:local+exotic ratios buttress the wayside station
interpretation.
Floor W's soaring kernel:cob ratio stands
in contradiction but, considering potential problems with
this last ratio noted in Chapter 4, agreement between the
other two provides more convincing evidence, in this case,
favoring the wayside station hypothesis.
Following the Chavin horizon, camelid remains
predominate within faunal assemblages at north-Central
Andean sites like Huacaloma where llamas assumed great
importance as cargo animals (Burger 1992: Miller and Burger
1995).
Camelid remains occur in small quantity during the
Colpar Phase, but the presence of camelids alone does not
substantiate either wayside station or habitation
542
interpretation.
The Colpar Phase distribution of animal
sizes (Fig. 115} displays the same preference for small
animals that characterized the Suitacocha Phase.
Colpar
Phase evidence suggests that Manachaqui Cave continued to be
functionally specialized, probably as a wayside station.
Because artifacts of the Colpar Phase are difficult to
isolate in Manachaqui Cave's macrochronology, caution
warrants only the observation that Colpar Phase vessel sizes
are comparable to those of the preceding Suitacocha Phase.
Colpar Phase Evidence
for Migrations
Despite an apparent two or three hundred year gap in
the local sequence corresponding to the Chavin horizon,
the
Colpar Phase assemblages provide clear evidence for cultural
continuity in the Pataz-Abiseo area.
Jar shapes and rims
are altered only slightly from the preceding Suitacocha
Phase.
Only a few new bowl shapes could be assigned to the
phase, and these sustain the long study area tradition of
folding over and beveling rims.
The few Colpar Phase
decorative techniques all have Suitacocha Phase antecedents.
This lack of marked change supports the interpretation that
Pataz-Abiseo societies remained outside of the Chavin
horizon interaction sphere.
Changes in population cannot be
postulated with support of the Colpar assemblages which
instead provide evidence for local cultural stability.
Proponents of verticality colonization scenarios may
543
contend that the ceramic linkages between Manachaqui Cave
and Gran Pajaten indicate a highland-based colonization of
the Montane Rain Forest Zone.
In the Pataz-Abiseo area,
Moseley's (1992:100) suggestion that the domesticated llama
is the harbinger of "true verticality" is undermined by the
observation that Gran Pajaten is poorly situated for
agricultural colonization (Church 1994:292).
Preliminary
investigations indicate that primary economic activities at
Gran Pajaten and Manachaqui Cave involved participation
within interregional interaction networks (Ibid.).
The Empedrada Phase Occupation
Material remains from Manachaqui Cave's Empedrada Phase
occupation suggest that the rockshelter served as a wayside
shelter, and perhaps occasionally as a temporary habitation.
Architectural elements include Sector A's palimpsest Floors
T through G, some of which were apparently prepared with a
light brown sandy silt.
The high organic content and rich
macrobotanical sample in these Sector A strata (Tables 1 and
18) attest to the shelter's intensive utilization during
this period.
Three of six Empedrada Phase hearths are of the
embedded "single-stone" variety.
Near Eastern pastoral
nomads utilize the stones in similar hearths to prop up
cooking pots (Frank Hole, personal communication 1996).
It
may be difficult to establish whether or not the singlestone hearths were used for cooking or warming food in the
544
absence of sooty potsherds or macrobotanical evidence from
the still unstudied hearth fill.
The embedded stone
absorbed heat, perhaps serving to radiate warmth for the
shelter's inhabitants after the fire's coals had cooled.
Evidence from Botanical
and Faunal Remains
If fruit rinds are counted as local foods,
then
Pearsall's food:wood and local:exotic+local ratios may
indicate that Floors T and S were deposited during
habitation.
The kernel:cob ratios stand in contradiction,
especially for Floor S.
Ratios from the remaining Empedrada
Phase floors tend to support the wayside station hypothesis.
Like the faunal assemblages from preceding phases, the
Empedrada Phase assemblage contains a preponderance of small
animal remains (Fig. 116).
Kent observes that a Sciuridae
(squirrel) specimen is "exotic," having been brought from a
forest habitat.
Larger animals representing greater
percentages of the faunal assemblage may indicate increased
large game hunting and the rockshelter's alternate function
as a habitation.
However, the increase in animals
classified as "large" more likely reflects the presence of
domesticated llamas, and it is only surprising that
identified camelid specimens still do not outnumber cervid
specimens (Table 27).
Even while llamas are clearly in
evidence, a herder's camp site function hypothesis is
untenable.
545
Evidence from Artifacts
The large quantity of unretouched and unifacially
retouched chipped-stone artifacts may be interpreted as
evidence of animal processing during extended habitation at
Manachaqui Cave.
The two chipped-stone point fragments were
both recovered in Unit 23 inside the eastern wall (Feature
R-6) extending from the cave mouth.
At least 29 (42%) of
the 69 ground and polished slate points belong to the
Empedrada Phase assemblage.
Worked chunks, disks and flakes
suggest slate-working at the site.
A complete quartz
crystal again would seem to indicate ritual activities at
the site.
Major structural changes in Manachaqui Cave's ceramic
assemblage have implications for interpretation of the
rockshelter's Empedrada Phase functions.
First, the
unprecedented predominance of bowls suggests that these
shapes were more important within Manachaqui's functional
context than jars.
While this may be true, it is also
possible that the jar had been at least partially supplanted
by some other container equally or more practical for
transport by cargo animals, perhaps cloth satchels or
alforjas (saddlebags).
The larger rim diameters of the
Empedrada Phase jars represent proportionately larger jars
(Table 17) less practical for human transport, yet
appropriate for storage use within a habitation, or for
546
securing within cloth or net alforjas as Guaman Poma's
sketch demonstrates (Fig. 117).
Empedrada Phase Evidence
for Migration
The abrupt shift in Manachaqui Cave pottery from a
relatively localized to a Central Andean-affiliated style
may be cited as evidence for mid-Early Intermediate Period
migrations from the highlands.
The inordinate number of
specific similarities between Empedrada Phase and Pashash
assemblages even permits conjecture of an upper Santa valley
or Conchucos source for the invading populations.
Such
migrations ostensibly explain the presence of a developed
Chachapoyas stone sculptural tradition (see Curtin 1951;
Kauffmann 1983:525; Campana 1992) that is related in
technique and motif to the Recuay tradition as defined by
Schaedel (1948).
Can the Empedrada Phase Paste A and Paste
C assemblages be attributed to Recuay colonization of the
study area and perhaps even greater southern Chachapoyas?
Despite the striking qualitative change in the PatazAbiseo Paste A pottery style, a migration hypothesis is
difficult to sustain.
Style attributes that persist from
anterior phases include the use of red paint on vessel rims,
notched applique applied to the vessel waists, folded rims
and rim lobe "handles."
In fact, the eastern montane forest
tradition of rim reinforcement by folding appears to have
been introduced at Pashash at the beginning of the Recuay
547
Period.
Other new Empedrada Phase features are not Central
Andean such as the Panzaleo-like punctations on rim
interiors.
Furthermore, similar episodes of pottery re-
styling were occurring at other locations around the
perimeter of the Central Andes around this time, but I will
return to this subject later.
Given the controversies that always surround migration
hypotheses generated on the basis of "traits," in this case
strictly stylistic evidence (Adams et al. 1973: 488, 503),
it is appropriate to weigh other kinds of evidence against
the expectations of a colonization hypothesis.
Some of the
problems identified below pertain equally to other migration
hypotheses addressed in this thesis.
First, there is no
clear evidence that Grieder's Recuay Period is temporally
prior to the Empedrada Phase, a necessary condition to
sustain migration hypotheses.
Second, potential motivations
for Recuay population movement or colonization must be
considered.
Incursion into a distant, occupied territory
with an unfamiliar ecology is no small undertaking.
Recall
that Adams et al. 's (1973) world survey of migration
theories determined that society-level population movements
across major environmental transitions are rare indeed.
Lack of evidence for highland demographic pressure or any
other motivating condition also renders population movement
hypotheses highly suspect.
Interpretations generated within the verticality or
548
"complementarity" theoretical framework would likely invoke
ecological symbiosis or "diadic interdependence" (cf.
Salomon 1985:515, Fig. 23.1) to explain the EmpedradaPashash relationship.
In this case, strictly ecological
hypotheses are implausible considering geographic
relationships between the upper Santa valley and the study
area.
At Pashash, access to nearby lower valley ecological
zones lies in the surrounding Cabana, Santa and Marafion
drainages.
Proulx (1982) has documented Recuay control of
western slope ecological zones down to 300 m in the coastal
Nepefia Valley.
The Abiseo drainage montane forests across
the Marafion Canyon and over 100 km distant ostensibly offer
few if any advantages for agriculture that cannot be found
closer to home.
The Marafion is known to have coincided with
major cultural boundaries by the Late Intermediate Period
(Garcilazo 1966; Espinoza 1969; C. Julien 1985; D. Julien
1993).
If Gran Pajaten was the site of mid-Early
Intermediate Period verticality colonization as ceramic and
radiocarbon evidence from the study area might suggest, then
we still must confront the aforementioned problem of the
site's unfavorable location for agriculture.
But perhaps the most serious problems with the
colonization hypothesis are exposed by considering expected
archaeological patterning under conditions of archipelago
colonization or "direct control."
Scholars regard economic
self-sufficiency engendered by colonization and kin-based
549
exchange as the raison d'etre and/or the result of "vertical
control of a maximum of ecological tiers"
(Lynch 1981:226;
J. Topic and T. Topic 1985:58; Moseley 1992:44).
Because
colonization establishes "ideal" conditions of direct
control intended to obviate external exchange, we should
therefore expect minimal material evidence for external
exchange relationships.
Contrary to this expectation, both Gran Pajaten's preAbiseo Phase assemblage and Manachaqui Cave's Empedrada
Phase assemblage contain substantial quantities of pottery
(7% and 18% respectively) not only exotic to the study area,
but exotic to the north-Central Andean highlands as well.
Empedrada pottery's contextual associations with styles of
the Junin puna and Northern Andean montane forests undermine
verticality arguments.
These data do not conform to
patterning consistent with the verticality model.
It should
be added, Van Buren's (1996) evidence that the Lupaca colony
of Terata Alta was indeed engaged in external exchange
further undermines Murra's model.
Summary of Evidence for Migration
at Manachaqui Cave
Both the long sequence of southern Chachapoyas cultural
development excavated at Manachaqui Cave, and the shorter
northern Chachapoyas ceramic sequence excavated by Ruiz
(1972) in the upper Utcubamba Valley show remarkable
stylistic continuity through time.
Common to both northern
550
and southern areas
~s
an ancient decorative tradition based
on applique and incised applique frequently rendering
serpent motifs.
Cuelap sequence.
Serpent motifs occur throughout Ruiz's
In the Pataz-Abiseo area, they appear on
the earliest pottery assemblages at Manachaqui Cave and
throughout the rockshelter sequence, as well as on late preColumbian pottery recovered from Gran Pajaten (Bonavia
1968).
The population movement hypothesis first posed by
Langlois, and Moseley's colonization hypothesis are likewise
crippled by archaeological evidence for long-term
continuity.
Taken together, the two Chachapoyas sequences
from the Marafion-Huallaga divide render hypothetical
migrations highly improbable.
The great irony inherent within the migration
hypotheses addressed in this thesis is the heavy burden
placed upon stylistic evidence to support fundamentally
materialist frameworks that typically assign "style" to the
stochastic or epiphenomenal realms of culture.
Meggers'
evolutionary scenario equates ceramic zoned hachure
incision, incised rims and polychrome decoration with
biological populations,
just as Lathrap identifies neolithic
Arawakan populations by their pottery's carinated basal
angles.
A conservative materialist approach might focus
more narrowly on evidence for changes in diet and
subsistence technology potentially indicating
transformations in means of production.
551
Evidence of an intrusive "neolithic" population should
be recognizable, although Manachaqui Cave is probably not
the most appropriate functional context in to find such
evidence.
While root-crop cultigens are always important
seldom preserved subsistence components, it is significant
that, from the Lavasen Phase through the Empedrada Phase,
the inventories of botanical and faunal taxa recovered from
the excavations remain unchanged excepting the predictable
Early Horizon additions of maize and beans (Pearsall 1992),
and Early Intermediate Period inclusion of camelids (Miller
and Burger 1995).
While an argument for the impermeability
of the Andean-Amazonian ecotone cannot be maintained in the
face of contrary historical, ethnographic,
linguistic and
archaeological evidence, the magnitude of the transition
between radically disparate environmental and cultural
worlds cannot be dismissed as historically inconsequential.
Pearsall's (1992, based on Harlan 1975)
Agricultural Complex"
"Lowland
(below 1500 m), and "Andean
Agricultural Complexes"
(1500 - 3500 m) show only four cases
of taxa overlap or co-occurrence.
In other words, a
"neolithic" society migrating across the ecotone would have
to replace its means and modes of production completely.
Archaeological evidence for such fundamental change at the
crucial momencs in prehistoric time has not been
forthcoming.
552
Evidence for Colonization in the Greater
Eastern Montane Forest
In Chapter 1, additional indices of intra-site
functional variability, regional stylistic distributions and
settlement patterning were described as potential predictive
aids with which to identify archipelago colonies implanted
in the eastern montane forest by highland-based, vertically
distributed economies.
Unfortunately, the intra-site
functional homogeneity that Bonavia cites as characteristic
of these "dependent settlements" cannot be fully evaluated
without appropriate sub-surface sampling strategies that
have yet to be implemented in the eastern montane forest.
Bonavia's view that it is impossible "to establish
differences in the social structure by means of samples from
the sites"
(Bonavia 1978:400) is unduly pessimistic in view
of the limited fieldwork undertaken to date.
The Southeastern Montane Forest
Apart from the Late Horizon Inca occupation of the
montane forests east of Cuzco and Lake Titicaca, there is
little published information bearing on indigenous
settlement and cultural development in the southeastern
montane forest.
Scant details of a Middle Horizon Huari or
Tiahuanaco presence are found in Savoy's (1970:99, 117)
report of explorations in the lower montane or premontane
forests of the lower Urubamba Valley.
Also in the lower
Urubamba Valley, Kendall (1976) provides some descriptions
553
and plan maps of Late Intermediate Period settlements
consisting of round and oval constructions and yielding
Killke or Killke-like pottery.
However, the extent of the
lower Urubarnba occupations and their ecological contexts
have not yet been analyzed.
The late prehistoric
settlements within the Inambari Valley east of Puno
cursorily described by Isbell (1968) have yet to receive
more intensive study.
Clearly the southeastern montane
forests were occupied long before the Late Horizon, but the
lack of archaeological data prohibit serious consideration
of hypothetical prehistoric migrations or verticality
colonization.
The Central Montane Forest
The reader will recall from Chapter 1 that Bonavia
(1978) regards absolute functional homogeneity a chief
characteristic of central and northeastern montane forest
site architecture.
In the central subdivision, Hastings'
(1985) more recent work contributes a relatively extensive
description and analysis of prehistoric settlement patterns
in the Tarma Valley.
At the Early Intermediate Period or
Middle Horizon site of Tranca, Hastings (Ibid.:632-633)
identified approximately 71 mostly rectangular stone
buildings, some of which are arrayed in rows or "welldefined patio clusters."
The Late Intermediate Period site
of Paraupunta, which succeeds Tranca at the apex of a
similar hierarchical settlement distribution, contains
554
approximately 150 round buildings showing a bi-modal size
distribution.
Like Tranca buildings, they occasionally
articulate with one another around patios (Parsons and
Hastings 1988: Fig. 7.4).
As Bonavia observes, some aspects of social structure
are apparently difficult to detect from central montane
forest site architecture.
However, such intra-site
distributions of habitations and patio groups are not
restricted to sites like Caballoyuq, Tranca and Paraupunta.
They are also characteristic of the highland Upper Mantaro
Valley Wanka centers of Hatunrnarca and Tunarnarca (Earle et
al. 1978, 1987).
This observation suggests that, based upon
surface architecture, Tranca, Paraupunta, Tunarnarca and
Hatunrnarca all exhibit similar degrees of intra-site
functional heterogeneity which may reflect similar forms of
social organization among the populous.
Apparently, surface
architecture is not a reliable indicator of settlement
dependency or autonomy at central montane forest sites.
In support of his position disputing local autonomy for
Tarrna Valley settlement systems, Hastings argues that: "the
drastic shift in settlement locations ... suggests economic
changes not previously underway during the Tranca Phase"
(1985:731).
However, he describes the transition from the
Tranca to the Paraupunta settlement system as "a gradual
process"
(Ibid. :727).
Most damaging to arguments for
colonization is his observation that the early Tarrna Valley
555
settlement systems have no highland counterparts
(Ibid.:642}.
Hastings' statement that administration and
decision-making were "rendered locally rather than in a
hypothetically higher-order center somewhere in the sierra"
(Ibid.:732} may be interpreted as contradicting his
theoretical position that such systems reflect "externally
introduced change"
(Ibid.: 731}.
In essence, Hastings'
archaeological data equally support interpretations of
fully-autonomous Tarma Valley settlement systems.
Bonavia's
(1972} Mantaro Valley site of Taipi may be functionally
analogous to Hastings' Tranca and Paraupunta type-sites at
the apex of indigenous, two-tiered montane forest settlement
systems.
The postulated Huari affiliation of Taipi and Raymond's
Lower Apurimac Valley colonies can also be called into
question on the basis of scant evidence.
Raymond (1976:206)
judges Bonavia's eastern Ayacucho collection of artifacts
"meager and non-diagnostic," and admits that stronger
correlates of Huari attribution are needed to verify his own
colonization hypothesis (Raymond 1976, 1985}.
The
hypothetical Huari outposts of Vista Alegre and Palestina
remain undated and only superficially described.
importantly,
Most
the Simariba area ruins that Raymond {1992:27)
observed, but could not investigate under the difficult
field conditions, almost certainly pertain to indigenous
montane forest populations.
556
Even granted Huari attribution for the premontane
forest sites of Palestina and Vista Alegre,
it remains to be
demonstrated that models of Huari imperial conquest are not
better suited than colonization to interpret the montane
forest Simariba archaeological data.
In a framework
invoking political hegemony, Palestina and/or Vista Alegre
potentially functioned as administrative centers established
to control local populations as well as to manage
interregional exchange.
Neither of these administrative
functions falls within the purview of verticality colonies
as defined by Murra.
The Inca sites recently reported by
Schjellerup (1992) deep in the lower montane forests above
the Central Huallaga may be functionally analogous to the
lower Apurimac sites.
The Northeastern Montane Forest
As in the central and southeastern subdivisions, the
problem of northeastern montane forest intra-site functional
homogeneity cannot yet be addressed with published
archaeological data from sub-surface contexts.
However,
surface architecture at Cuelap (Narvaez 1988) clearly
exhibits functional heterogeneity.
The site is clearly
divided into upper and lower barrios or sectors, and
constructions include "ceremonial" rectangular buildings, Ushaped structures and the unusual "Tintero."
buildings occur in a wide variety of sizes,
Circular
show a diversity
of stone inlay ornamentation and group themselves in four
557
distinct configurations, two of which articulate with open
patios.
The functions of numerous massive rectangular
structures that dominate the ancient settlement of PirkaPirka above Uchucmarca (Vega 1982} remain to be interpreted
through sub-surface sampling.
At Gran Pajaten (Rojas 1967; Bonavia 1968} the
differential arrangement and decoration of buildings
likewise· strongly suggest functional diversity.
The most
recent excavation data pertinent to the problem of building
functions has not yet been published.
However, Bonavia's
hypothesis that Gran Pajaten's rectilinear slate mosaic
motifs and sandstone curvilinear motifs represent foreign
colonists and local populations respectively cannot be
sustained in view of the motifs rendered in both techniques
illustrated by Pimentel G.
(1967: Fig. 3}.
Stylistic
continuity evident in Gran Pajaten's pre-Abiseo Phase and
Abiseo Phase ceramic remains reflects autochthonous cultural
development since the Early Horizon, and likewise
constitutes evidence contradicting hypothetical site-unit
intrusions.
Forthcoming publications of excavation data
from the nearby site of La Playa will also shed further
light on the problem of intra-site functional heterogeneity
in the northeastern montane forest.
Although little regional sampling has yet been effected
in the northeastern montane forest,
the distributions of
settlements, architectural attributes and ceramic styles
558
exhibit considerable uniformity, rather than archipelagolike heterogeneity.
The northeastern montane forest has
been singled out as a distinctive archaeological region with
unique characteristics that suggest a long period of local
and autonomous cultural development (Church 1994:283).
Among these characteristics are architectural elements such
as cornices, staircases and stone frieze decorations, and
the previously described pottery with applique decoration
featuring serpent motifs.
These same widely distributed
attributes prompted Lumbreras
(1974) to identify the region
as a "new culture area," and motivated Ravines and Brush to
propose state-level socio-political organization for preIncaic Chachapoyas.
If Raymond's set of settlement pattern expectations for
eastern slope verticality colonization is applied to the
Pataz-Abiseo area,
then we should find comparatively light
montane forest settlement corresponding to the highly
compacted ecological zonation.
Yet in Chapter 3,
it was
noted that prehistoric populations in the Pataz-Abiseo study
area were clearly concentrated within the forested Huallaga
watershed, rather than on the unforested slopes of the
Marafion drainage.
Demographic patterns, coupled with the
documentary evidence for autonomous forest populations
subsumed by the encomienda of Sucos y Puymal, cannot be
reconciled with the expectations of a verticality
colonization framework.
Neither the early settlement at
559
Gran Pajaten, nor the early Tarma Valley settlements
correspond to any archaeologically visible highland polities
capable of subsidizing their establishment.
The prehistory of the Central Andean eastern montane
forests suffers from many of the sampling biases that plague
the archaeology of the Amazon lowlands.
Yet while the
scarcity of archaeological data from Amazonia has resulted
in polemic debates, the scant information from the eastern
montane forests has occasioned mostly pessimistic
assessments of their potential productivity and
habitability.
Just like the Amazon Basin and the
Intermediate Area, the montane forests have been offered as
a case study in stunted cultural evolution.
However,
because the eastern montane forests correspond to the
boundary between contrasting, yet juxtaposed culture areas
or co-traditions, they may also be alternatively viewed as
loci of cultural creativity and archaeological
possibilities.
In the following final chapter,
archaeological data from Manachaqui Cave and the Central
Andean montane forests are re-interpreted utilizing a
framework emphasizing local cultural development and
interregional interaction.
CHAPTER 11
MONTANE FOREST CULTURAL DEVELOPMENT
AND INTERREGIONAL INTERACTION
All pristine developments are secondary developments
dependent on outside resources, alliances and events (Clarke
and Blake 1994:20).
In the preceding chapter, I attempted to show that the
archaeological data from Manachaqui Cave fails to support
migration scenarios formulated to account for stylistic
distributions and patterns of cultural development in the
Central Andes.
In the remainder of this thesis, I will
present an alternative interpretation of this material
patterning utilizing evidence from Paste Group A phase
assemblages at Manachaqui Cave, and archaeological daca
reported from surrounding regions,
prehistoric "interaction spheres"
to reconstruct
(Caldwell 1964).
Caldwell's term is implemented here to refer to broad fields
of stylistic similarity among the ceramic assemblages of
interacting societies distributed across the landscape.
The
ceramic interaction spheres are reconstructed on the basis
of shared clusters of pottery design attributes that reflect
limited kinds of interaction in which ceramics presumably
functioned as visible media for symbolic expression.
Thus,
they reveal only one of numerous nested, overlapping and
560
561
complementary fields of interaction that occurred
simultaneously across space.
Early Andean ceramic interaction spheres cross-cut
other interaction spheres defined by co-occurrences of
architectural features like conical adobes, sunken circular
courts and ceremonial hearths.
As heuristic devises, they
are units of analysis that provide one kind of evidence for
frequent,
if not habitual, interaction.
In the Central
Andes, as in most parts of the world, they are small pieces
in very large puzzles.
Ceramic interaction spheres
correlate with archaeological phases because the same
ceramics serve to define each phase.
However, it should
also be emphasized that interaction spheres, like other
aspects of culture, were in constant flux.
Archaeologically
detected interaction spheres may be comprised of smaller
interaction spheres that did not necessarily operate
simultaneously.
The utilization of stylistic similarity to reconstruct
fields of interaction in the Central Andes is rare, but not
unprecedented.
MacNeish et al.
(1975) and Browrnan (1975)
have contributed interpretations of early exchange networks,
but they did not fully implement stylistic similarity as a
criterion.
The studies of Mohr-Chavez (1981b) and Burger
(1984a, 1988, 1993) have made the most explicit use of style
and sets of style attributes to interpret early Andean
interaction within and between regions.
Inherent within
562
their analyses is the view that shared clusters of pottery
design attributes facilitated cultural interchange by
conveying a sense of sameness or common social identity, and
open inter-group boundaries (Mohr-Chavez 1981b:343; Hodder
1982; Burger 1984a:45).
This chapter's point of departure is the assumption
that ethnic, linguistic and genetic boundaries may coincide
with interaction sphere configurations, but such
correspondences cannot be presumed solely on the basis of
archaeological evidence.
Rather, ceramic interaction
spheres merely signify the intent to exchange goods and
information across social, political and/or ecological
boundaries.
Whether they provide a window on ethnic
affiliation, social interchange, ritual interaction and/or
political alliance remains to be determined by the
evaluation of complementary data sets and the contexts of
interaction.
Apparently, some modes of interaction can
produce precisely the same kinds of material patterning that
archaeologists have cited as evidence for migrations and the
second half of this chapter will be devoted to discovering
these.
While Manachaqui Cave's Paste A assemblages provide
indirect evidence enabling the reconstruction of interaction
spheres, Paste Groups B and C provide complementary direct
evidence for exchange networks.
These networks overlay and
cross-cut the aforementioned interaction spheres.
At
563
Manachaqui Cave, small Paste Group B vessels probably mark
the physical presence of travelers from distant localities.
The quantities of Paste Group B and C sherds might
constitute a gross measure of interaction's intensity, but
they might not.
Based upon Burger's (1992, 1993; Miller and
Burger 1995} interpretation of the Chavin horizon as a
widespread "socioeconomic metamorphosis" accompanied by the
establishment of new political boundaries, we might
hypothesize that Paste Groups B and C are more likely to
reflect long-distance travel prior to the Chavin horizon,
and mediated exchange following the Chavin horizon.
Pre-Lavasen and Lavasen Phase
Evidence for Interaction
Finding Preceramic occupations at the edge of the
northeastern montane forest, a purportedly "empty niche,"
should not be surprising as forested terrain did not deter
Central Andean Preceramic Period populations from making a
living within and adjacent to northern (Kaulicke 1981} and
western (Dillehay and Netherly 1983; Dillehay et al. 1989}
montane forest ecological zones.
In the Northern Andes,
Salazar (1983:102-104} argues that El Inga and Chobshi Cave
lay at the paramo-montane forest ecotone during the early
Holocene occupations.
He considers the ecotone a preferred
habitat that permitted access to resources in both
ecological zones, as well as to specialized edge species
(Ibid.:108-110}.
While habitation can be inferred from
564
Lavasen Phase material patterning, there is little evidence
indicating the occupants' degree of permanence.
Paleoenvironmental studies in the north-Central Andean
highlands are needed to determine the past extent of montane
forests before early prehistoric subsistence patterns can be
accurately reconstructed.
One Preceramic Period find at Manachaqui Cave merits
discussion here, because it portends long-distance
interactions characteristic of the rockshelter's ceramic era
occupations.
Burger et al.
(1995) report that a tiny
pressure flake (Unit 2 Level 30) and the distal portion of a
unifacially retouched flake (Unit 6 Level 36) made of
obsidian from the Lavasen Phase deposits constitute unusual
evidence for far-flung exchange linkages.
Not only do these
pieces represent the first documented occurrences of Alca
obsidian (from northern Arequipa) in a Preceramic Period
context, but they also constitute the most northerly
discovery of Precerarnic Period obsidian artifacts in the
Central Andes.
Burger et al. observe that both Alca and Ecuadorian
obsidian sources lie approximately 1000 km from Manachaqui,
and the Cuzco Quispisisa source lies 655 km distant.
They
note that "the natural routes of transport from the north
present less natural obstacles, but this was apparently less
significant than cultural forces that shaped patterns of
interaction."
While the Lavasen assemblages offer few
565
additional clues with which to reconstruct interaction
spheres closer to horne, Shady and Rosas (1979:127} point out
that earliest Initial Period ceramic styles provide a view
of Terminal Precerarnic Period interaction spheres.
Manachaqui Phase Evidence
for Interaction
By 1200 b.c. and the mid-Initial Period, Manachaqui
Cave's users were carrying pottery to the rockshelter.
Although they may have been involved in subsistence-related
activities involving habitation at the rockshelter, and
perhaps limited transhurnance, the broad field of ceramic
similarity shared by western, northern and eastern montane
forest populations indicates that they were simultaneously
engaged in long-distance exchange activities.
The
Manachaqui Phase Paste A assemblage's closest, or primary,
relationship is clearly with the earliest Cajarnarca and
Bagua assemblages.
Shady (1987b} and Onuki (1985} associate
these styles with the north-Central Andean yungas, while
Onuki further identifies them with a hypothetical Initial
Period subsistence strategy termed the "Yunga Tradition."
The Montegrande, La Conga, Early Huacalorna, Pandanche
A, Morerilla and Manachaqui styles are distributed along an
altitudinally-restricted swath
approxirnati~g
past
distributions of northwestern, northern and northeastern
montane forests.
This "Northern Montane Forest Ceramic
Interaction Sphere" describes a wide arc around the northern
566
rim of the Central Andean region and was apparently
positioned along a broad front to facilitate interaction
with Northern Andean societies (Fig. 118).
Like its
companion Central Andean styles of the Northern Montane
Forest Interaction Sphere, the Manachaqui Phase Paste A
assemblage incorporates numerous Northern Andean formal and
decorative attributes seen in the coastal Valdivia and
Machalilla styles, and in the highland Cerro Narrio and
Cotocollao styles.
populations
Unlike styles exhibited by migrating
(Anthony 1990:903), Manachaqui Paste A's
northern influence cannot be attributed to "a highly
restricted point of origin."
The archaeological evidence indicates that montane
forest environments were attractive for early Central Andean
population nucleation.
Rather than from a demographic push,
as Kauffmann postulates, nucleation resulted from the pull
of desirable opportunities for interaction with groups in
neighboring regions, and with other nucleated groups
situated at the Central Andean boundaries.
Despite the
suitability of Huamachuco's intermontane valleys for a
variety of productive subsistence systems, it's populations
apparently remained scant and dispersed within a demographic
"vacuum" left in the wake of attraction to the boundaries.
Based upon their fieldwork in the early 1980s, J. Topic and
T. Topic {1985:60-62) concluded that early Central Andean
north-south interaction routes bypassed Huamachuco by
567
following coastal routes.
La Galgada participated in this
same interaction sphere which in effect surrounded
Huamachuco.
Grieder and Bueno (1985:106), Burger (1992:53)
and Shady (1993:116), describe La Galgada as a node in early
interregional exchange that connected the coast and tropical
lowlands.
The Manachaqui Paste A assemblage shows a weaker, yet
clearly distinguishable relationship to other Initial Period
assemblages recovered from sites in the eastern Central
Andes.
Manachaqui, Wairajirca,
"early formative"
Pachamachay, "formative" Telarmachay, Muyu Moqo C-D and
Marcavalle A and B are connected to early Northern Andean,
especially Valdivia-Machalilla, styles by co-occurrences of
short-necked jar shapes and scattered features such as
carinated body profiles, notched applique, specular hematite
paint, stepped or two-tiered shoulders, rim beveling and
complex bottle spouts.
Lathrap has argued that these
similarities are diagnostic of early migratory (Lathrap
1963) or exchange (1971:95) links between the Northern
Andean Machalilla and Amazonian Late Tutishcainyo styles.
Archaeologists investigating the earliest ceramicbearing occupations on the eastern slopes of the Bolivian
Andes have emphasized similarities between Cochabamba
Formative pottery exhibiting pattern-burnishing and complex
incised decoration and the earliest assemblages from far
northern Peru and southern Ecuador (Ibarra and Querejazu
568
1986:152-163; Brockington and Sanzetenea 1989; Brockington
et al. 1995:164-166).
Evidently, this eastern-slope
interaction sphere extended more than 2,500 km from southern
Ecuador to southern Bolivia, and perhaps beyond.
Some
Northern Andean, Amazonian and south-Central highland
features are found on the south-Central coast of Peru at
Hacha (Robinson 1994: Fig. 14) and Erizo, suggesting that
the interaction sphere penetrated south and westward in a
manner analogous to its inclusion of the inter-Andean
littoral at the modern border of Ecuador and Peru.
Considered together, patterns of stylistic similarity
indicate that prominent Initial Period interregional
interaction networks were longitudinally oriented north and
south along the Central Andean flanks.
Apparently,
principal north and south regional communication and
exchange networks followed montane forest routes on the
slopes of the cordillera rather than descending to the coast
as the Topics propose.
The stylistic discontinuity between
the Jequetepeque and Viru valleys noted by Ulbert (1994:149)
reflects the permeable southwestern boundary of the Northern
Montane Forest Ceramic Interaction Sphere matched on the
southeast by a similar boundary represented by salient
differences between the Initial Period assemblages at
Manachaqui and Kotosh.
Cross-cutting longitudinal montane forest interaction
spheres, more narrowly targeted east-west interaction
569
spheres linked Upper Huallaga Kotosh and upper Urubarnba
Marcavalle populations to Amazonian Central Ucayali
societies.
While the longitudinal sphere features design
attributes found among the Valdivia and Machalilla styles,
the eastern montane forest interaction sphere (or spheres)
features complex line-incised decoration, post-fire paint
and flanges observed in highland Wairajirca and Marcavalle A
ceramics and coeval lowland styles like Cave of the Owls
Fine Ware and Tutishcainyo.
These relatively localized and
intense interaction relationships excluded the nearby
Amazonian Upper Pachitea societies.
In Chapter 6, I suggested that the Paste B vessels,
along with the two Paste A vessels exhibiting incision (Fig.
64), may occur late in the Manachaqui Phase.
They are
stratigraphically associated with the uppermost Manachaqui
Phase deposits corresponding to Floor AA, Floor
z
and
Feature R-4 (radiocarbon dated 900 ± 90, 850 ± 90 and 830 ±
100 b.c. respectively), and seem to reflect developments in
adjacent regions east and north.
Based upon known spatial
and temporal distributions of bottles with asymmetric spouts
in upper Amazonia, DeBoer (n.d.) proposes a "flash horizon
dated to about 800 BC."
He observes that the spatial
distribution of these bottles mirrors later upper Amazonian
interaction spheres that correspond to prehistoric
Curnancaya, historic Shipibo and 17th century Jesuit
missionary travel routes and interaction networks.
The
570
Manachaqui Phase B pottery apparently arrived at the
rockshelter during what appears to represent a brief period
of increased long-distance foot travel.
Suitacocha Phase Evidence
for Interaction
The terminal Initial Period flash horizon can be
conceptualized as a time when Central Andean, Amazonian and
Northern Andean societies became acutely aware of new
possibilities for interaction with one another.
Ultimately,
it foreshadowed abrupt stylistic change in the Pataz-Abiseo
area and the beginning of the Suitacocha Phase around 800
b.c.
During the Early Horizon, it is more difficult to
evaluate Paste Group A's stylistic relations by degrees of
similarity.
Styles are more locally distinctive, and the
spatial distribution of resemblances is uneven, and perhaps
typified by the spotty appearance of Shape F-like rims in
Central Andean assemblages.
The irregular spatial
distributions of many shared attributes not only constitute
evidence against wave-like migrations, but they suggest
specific modes of interaction which target distant
localities while bypassing some that are closer.
The most extraordinary example of this stylistic leap
is the Suitacocha Paste A assemblage itself.
Rather than
displaying closest resemblances to the Kotosh, Huamachuco,
Bagua or Shakimu styles flanking Manachaqui Cave at the four
cardinal points, the Paste A assemblage finds its closest
571
relationships to more distant Northern Andean so-called
"Chorreroid" styles (Fig. 119).
As they did during the
Initial Period, northern montane forest societies in
Cajamarca remain at the forefront of interaction with
Northern Andean societies.
Evidence for sparse populations
in Huamachuco continues to suggest a centrifugal attraction
to the northern edges of the Central Andean massif.
Archaeological data from the Early Horizon northCentral Andes demonstrate that the Suitacocha Phase
transpires against a backdrop of accelerating construction
of public monuments and long-distance interaction between
growing population centers.
In breadth the Early Horizon
interaction sphere transcends the earlier Northern Montane
Forest Ceramic Interaction Sphere which has splintered into
a number of smaller discrete and overlapping interaction
spheres distinguished by relatively localized decorated bowl
and bottle styles.
The local spheres correspond to Early
Horizon centers acting as nodes in redistributive systems
controlled by emergent elite (Shady and Rosas 1979; Burger
1984a, 1992; Topic and Topic 1985).
The Early Horizon interaction sphere was consolidated
by an over-arching exchange network that distributed "trade
pottery" among local centers such as Naiiafiique, Kuntur Wasi,
Cerro Blanco, Bagua, Kotosh and Chavin de Huantar.
Burger
likens this integrative web to Renfrew's "peer-polity
interaction," involving exchange of exotic "prestige items
572
with a high information content"
(Burger 1984a:46, 1993:74).
Decorated bowls were apparently perceived as particularly
well-suited for conveying gifts between population centers
(Mohr-Chavez 1981b:343).
The elaborately decorated bowl
represented by the single Paste B5 sherd is typical of
vessels likely to circulate in the Early Horizon northCentral Andes.
In the Early Horizon context of nested or concentric
Central Andean boundaries, outlying centers with complex
local styles and a relative abundance of associated exotic
wares like Nafiafiique and Bagua seem to have developed as
"gateway communities"
(as defined by Hirth 1978), channeling
interregional exchange between the lowlands and the Central
Andes (Burger 1992:101).
The data pertinent to Manachaqui
Cave's function at this time, and especially the
extraordinary resemblance of the Suitacocha assemblage to
Ecuadorian highland and lowland styles suggests that the
rockshelter also served within a similar context of
channeled interregional exchange.
Despite its Central
Andean geographic context, the Suitacocha Phase Paste A
assemblage exhibits a primary relationship to Northern
Andean and Oriente styles, rather than to neighboring
northern montane forest Cajamarca and Bagua styles.
While Suitacocha is clearly more "Chorreroid" than
Cajamarca's northern montane forests styles, all of these
seem to reflect articulation with the Chorrera interaction
573
sphere.
Especially illustrative of Suitacocha's "Chorrera
connection" is the appearance of indisputably Northern
Andean incised motifs and iridescent paint.
The Shape F
jars with decorated shoulders and the distinctive basal
joints of Suitacocha carinated bowls (Fig. 80e-h) are
likewise unequivocally "Ecuadorian."
The overall paucity of
Suitacocha Phase Paste Group B pottery is perhaps less
supportive of the interaction interpretation, although it is
likely that such important vessels were protected from
breakage during transport.
Evidence for interregional interaction between the
Central and Northern Andes in the form of intrusive pottery
is exceedingly rare in both regions.
If the Suitacocha
Phase Paste B4 sherds do originate in Ecuador's eastern
montane or premontane forests,
then they might constitute
the first documented incidence of Northern Andean pottery
recovered from Early Horizon stratigraphic contexts in the
Central Andes.
Direct contact between societies in these
two regions may likewise be indicated by the presence of
intrusive Early Horizon sherds at Chaullabamba recently
reported by Hocquenghem et al.
(1993: Fig. 3).
Based upon
perceived similarities between the Bagua and Cerro Narrio
styles, Shady (1987b:480) contends that the Marafion Valley
is a locus of Early Horizon inter-Andean interaction.
The
Suitacocha evidence lends support to her observation.
Shady and Rosas (1979), and Burger (1988, 1992) point
574
out that socio-economic status differentiation becomes most
evident in this context of heightened north-Central Andean
interaction.
The rich Early Horizon tombs excavated at
Cerro Blanco, Poro-Poro and Shillacoto may represent
interment of a socio-economically privileged class.
Ultimately, it is upon the north-Central Andean Early
Horizon interaction Sphere that the Chavin horizon is
superimposed.
At Manachaqui Cave, maize first appears
within this increasingly politicized field of interaction
that emphasizes public gathering and ritual at monumental
centers.
It may be that Suitacocha Shape F jars with tall,
spout-like polished rims are associated with maize and
chicha.
Mohr-Chavez (1981b:345) observes that Marcavalle
Phase C-D jars "behave as bowls" in prestige spheres, and
perhaps within south-Central Andean gift exchange.
More
evidence is necessary to test these functional associations.
There are indications that Initial Period east-west
interaction networks conjoining the Central Andes and
Amazonia persisted during the Early Horizon.
Lathrap has
discussed specific similarities between Kotosh-Kotosh and
Shakimu styles
(1971:87-88), and Mohr-Chavez (1981b) cites
attributes shared between Marcavalle B-D, eastern-slope
Cochabamba and lowland Amazonian styles.
Other highland
Central Andean assemblages like Urabarriu at Chavin de
Huantar and the Huaricoto style at Huaricoto lack evidence
of Amazonian affiliation (Burger 1992:154-155; 1985b:529).
575
Again, interchange with Amazonian Central Ucayali societies
is apparently restricted to the upper Huallaga and Urubamba
drainages.
Material patterning indicative of these
relationships and with the Suitacocha-Chorrera linkage
suggests that Early Horizon interregional interaction was
more narrowly focused, or point-specific than Initial Period
interaction.
Marcavalle A-D short-necked ollas and jars with zoned
punctation on vessel shoulders provide suggestive evidence
of a continuing Northern Andean connection in the southCentral Andes.
Interregional exchange networks that
distributed carved stone mortars like those found at San
Isidro, Cotocollao, Huayurco, Bagua, Pacopampa, Kotosh and
Marcavalle (Rojas 1969; Lathrap 1970:108; Peterson 1984;
Zeidler 1988) probably trended north and south along the
eastern face of the Andean cordillera, and trade in strombus
and spondylus shells also followed similar over-land routes
(Hocquenghem 1993).
The same kind of decorated short-necked
jars occurring in the south-coastal Rio Grande de Nazca
Valley Tajo style
(Silverman 1994: Figs. 7, 8) suggest
similar affiliations although Silverman proposes Chavin
horizon dates for these vessels.
Of course later Paracas
assemblages from this area are well-known for exhibiting
Amazonian characteristics (Tello 1942:631-632).
The Chavin Horizon Hiatus
During the mid-Early Horizon, north-Central Andean
576
societies were united within a pan-regional sphere of
interaction corresponding to the Chavin horizon.
After 500
b.c., Manachaqui Cave seems to have fallen into relative
disuse for two or three centuries.
It might be hypothesized
that societies in the northeastern montane forests were
excluded from the Chavin interaction sphere, but additional
evidence from the Chachapoyas area is needed to be sure.
What is clear is that Manachaqui Cave no longer serviced
systematic, or even frequent interregional interaction.
Simultaneous with the spread of the Chavin cult, Burger
(1984a; 1992:216) observes that a dearth of evidence for
interregional interaction with the Northern Andes and
Amazonia indicates the solidification of cultural boundaries
surrounding the Central Andean or Peruvian "co-tradition."
As interacting Central Andean societies turned inward,
the gateway community at Nafiafiique was virtually abandoned,
Bagua re-directed interaction toward other eastern lowland
areas and the ancient tradition of highland-lowland economic
interaction evident in the early Kotosh assemblages
disappeared (Burger 1992:219).
Likewise, Manachaqui Cave's
role as a node in gateway interaction terminated until
approximately 200 b.c.
The beginning of the Early
Intermediate Period marks the formation of new interaction
spheres functioning in increasingly politicized contexts.
Radiocarbon evidence from Gran Pajaten attests to the
nucleation of populations at hilltop settlements within the
577
montane forest of the Abiseo drainage, probably for the
express purpose of facilitating renewed interregional
interaction (Church 1994:292).
Colpar Phase Evidence
for Interaction
During the Early Intermediate Period Colpar Phase,
Manachaqui Cave and Gran Pajaten yield evidence for
occupation by local populations participating in the same
interaction spheres.
To judge by Paste A's limited
repertoire of vessel shapes and surface decorations, and
evidence provided by Pastes B6 , B8 , B10 and C1 ,
the study
area's clearest connections are to montane forest societies
east and north of Manachaqui Cave (Fig. 120).
Preceding
discussions of the Manachaqui and Suitacocha Phase data have
demonstrated that this northwest-focused interaction sphere
is ancient.
The austerity of the Colpar Paste A assemblage,
the lack of broadly distributed montane forest styles, and
especially the reduced repertoire of surface decorations,
indicates that montane forest populations around the study
area remained largely outside of the realigned and more
rigidly-bounded post-Chavin Central Andean interaction
spheres, interacting primarily among themselves.
Evidence from Manachaqui indicates that the rockshelter
was still utilized expediently as a wayside station by
infrequent travelers.
Like their Suitacocha assemblage
predecessors, the small Colpar Paste A necked-jars and
578
restricted bowls were well-suited for transport by
individuals shouldering cloth bags.
Camelids are clearly in
evidence by this time, yet there are no clear indications
that they were utilized as cargo animals, at least in any
large-scale fashion.
The lack of evidence for carnelid
transport systems coincides with the paucity of ceramic
evidence for study area interaction with neighboring Central
Andean societies that had already adopted camelid transport
technologies (Miller and Burger 1995).
By the beginning of the Early Intermediate Period,
populations in the far north-Central Andean highlands
producing the Layzon style had developed systematic
interaction with coastal Salinar societies (Mujica 1984},
probably utilizing llama caravans to transport commodities
(Sechtin 1986:33) including bulky pottery vessels.
The
Colpar Paste A assemblage's formal resemblances to the
Cajarnarca Layzon style are more likely residual than
indicative of habitual or systematic interaction.
In
neighboring Huamachuco, land-use studies by McGreevy and
Shaughnessy (1983) determined that evidence for intensive
herding activities in the surrounding punas is scant until
the Late Intermediate Period (after a.d. 900).
The absence
of corrals and the low frequency of carnelid bones in Early
Intermediate Period faunal assemblages not only suggest that
large herds were not kept (Ibid.:241), but that carnelid
transport was unimportant to Huarnachuco's Early Intermediate
579
Period regional economy.
During the Colpar Phase, foot traffic maintained
contact between the Pataz-Abiseo area and populations with
access to warm-water estuarine environments of the interAndean littoral.
From mixed deposits,
fish remains that
Kent assigns to the Gobidae family of sleepers probably
represent long-distance interaction with groups situated
between the Piura and Guayas River basins, as do Pastes B7
~1d
B9 •
Previously overlooked similarities between Pre-
Abiseo Phase sherds from Gran Pajaten and finds in the
inter-Andean area include an unusual modeled rim (Church
1994: Fig. 11m) identical to a Sechura B example (Lanning
1963: Fig. 231), and a body sherd with circles rendered by
the negative-resist technique (Church 1988:41f).
The circulation of beveled-rim bowls apparently
transcended Central Andean boundaries and followed expansive
networks established prior to the Chavin horizon.
Evidence
from Huacaloma and Pandanche intimates that these bowls of
Ecuadorian origin began to first circulate at northern
montane forest centers around the time of the Chavin
horizon.
By the post-Chavin Colpar Phase,
the network had
expanded to include the upper Zafia and Piura valleys, the
upper Utcubamba, the upper Montecristo (Abiseo), Huamachuco
and the upper Santa valley Cabana-Corongo area.
The bowl
shape persisted in the Northern Andes during the coeval
Early Guangala Phase.
In effect,
the beveled-rim bowl's
580
distribution replicates preceding Early Horizon northCentral Andean interaction spheres.
A functional
interpretation is difficult without more information on the
specific contexts of these finds, but the bowls were
probably carried by travelers engaged in continued gift
exchange, perhaps with political nuances.
Empedrada Phase Evidence
for Interaction
The mid-Early Intermediate Period Empedrada Phase
transpired against a background of hardened Central Andean
political boundaries, inter-group conflicts, restricted
travel and the systematic exchange of goods by llama caravan
transport (Burger 1992:229).
The comparative analyses
presented in Chapter 9 demonstrated that mid-Early
Intermediate Period Pataz-Abiseo area pottery styles shared
a remarkably close relationship with Recuay styles from the
upper Santa tributary Cabana River (Fig. 121).
Overall
resemblances between the Empedrada Paste A and Pashash's
Recuay assemblages are no less striking than the well-known
correspondences between pottery assemblages at highland
Layz6n and coastal Cerro Arena.
Pataz-~~iseo
Additional survey in the
and Cabana areas is needed in order to
determine just how geographically specific this relationship
is.
More likely than verticality to account for the
Pashash-Empedrada relationship, and the dense late
581
prehistoric Montecristo Valley settlement distributions
described in Chapter 3, is the access that the Abiseo
drainage offers to the lowland Huallabamba fluvial "axis"
which brought Amazonian riverine trade to the doorstep of
the north-Central Andes.
The seminal event that
precipitates the birth of this exchange linkage and the
beginning of the Empedrada Phase appears to be the
introduction of systematic llama caravan transport
technology in the study area.
Evidence for its debut at
Manachaqui Cave consists of elevated quantities of camelid
remains, and unprecedented shifts in the Paste Group A
vessel shape inventory.
The strong correspondences between
Pashash and Empedrada pottery, coupled with the lack of
evidence for herding at Manachaqui Cave and in the
surrounding Tropical Alpine Zone, invites speculation that
most of the llamas and the herders frequenting Manachaqui
Cave during the Empedrada Phase were native to the Callej6n
de Huaylas.
The notion that "foreign" Central Andean societies
widened their interaction networks by directly introducing
llama-based transport technology into adjacent areas has
important local and regional implications.
Further
exploration of this hypothesis at the local level requires
more rigorous comparative analyses of Pashash, Empedrada and
Gran Pajaten (pre-Abiseo Phase) ceramics and continued study
of prehistoric study area land-use.
In this case, the
582
Ernpedrada Phase Paste C assemblage may represent the
physical presence of upper Santa Valley highlanders at the
rockshelter.
Although the Empedrada Phase Paste Group A
assemblage clearly exhibits attributes that are locally
derived, it does contain more than a single ware.
It may be
significant that Gran Pajaten's small pre-Abiseo Phase
sample does not exhibit some important Empedrada Paste A
attributes like applique decoration and Shape H vessels.
Nor does it share other Ernpedrada Paste A affinities to the
Pashash styles.
New east-west interaction linkages utilizing cargo
animals required the construction or refurbishment of paved
roads in the Pataz-Abiseo area.
It remains unclear how far
into the montane forest llama caravans penetrated, but
traces of paved roads observed in the forested Montecristo
Valley apparently reflect linkage between Manachaqui and
Gran Pajaten.
Also, recall from Chapter 3 that Savoy (1970)
located paved roads far below in the Central Huallaga
subtropical forests surrounding the abandoned mission of
Jesus de Pajaten.
The lowland Huayabamba Complex at the
eastern end of the Pataz-Abiseo study area cross-dates to
the Early Intermediate Period, and its Central Andean aspect
clearly reflects the linkage of lower Abiseo Valley
societies into Central Andean interaction spheres.
Nascent highland-lowland exchange systems incorporating
Chachapoyas intermediaries cross-cut the ancient and
583
persistent interregional interaction sphere trending northsouth along the eastern face of the Central Andes.
This
"longitudinal" interaction sphere is evident in the
Empedrada Paste A assemblage's similarities to the nearby
pre-Abiseo phase assemblage directly east, the Tarma valley
Malarnbo and Camonal complexes to the south and the Cancharin
(at Cuelap) and Cosanga-Panzaleo styles to the north.
The
Huacrapuquio sherds and Quispisisa obsidian almost certainly
reached Manachaqui Cave via eastern montane forest routes,
as did the Paste B11 Panzaleo, or Panzaleo-like, pottery
which originated from twice as far away.
The large jar with
orifice diameter of 30 em (Fig. lOOf) was probably
transported to Manachaqui Cave by llama caravans on roads
long ago lost beneath montane forest regrowth.
Boundary Interaction and the Central Andean
Tropical Montane Forests
The previous section provided an interpretation of
evidence for interregional interaction from Manachaqui Cave
during nearly two thousand years of prehistoric occupation.
The following paragraphs will take a broader view of
evidence for interregional interaction in an effort to infer
specific modes of interaction responsible for stylistic
patterning in and around the Central Andean montane forests.
The failure to find support for postulated society-level
population movements at Manachaqui Cave is predictable if we
consider the assumptions inherent within unitary origins
584
frameworks that material culture and styles react passively
to the passage of time as do languages and gene pools.
Archaeologists narrowly regarding the stylistic
distributions as reflections of "cultural heritage" await
the eventual discovery of evolutionary "stepping stones"
(Meggers et al. 1965:174),
and grandmothers"
"collateral relatives," "mothers
(Lathrap 1970:107, 1974:142), unknown
donors (Braun 1982) and other missing links to corroborate
their interpretive models.
While in certain contexts style
may behave passively as it is learned and re-learned from
generation to generation, scholars investigating how style
functions in broad contexts of inter-group interaction now
recognize style's proactive functions in "informationexchange"
(e.g. Wobst 1977; Hodder 1982; Weissner 1990;
Hegmon 1992) .
Material patterning previously cited as evidence for
migration can be explained alternatively as conscious
manipulation of styles and style attributes for the
furtherance of a variety of social, religious, economic
and/or political ends.
This thesis assumes the position of
Lightfoot and Martinez (1995) that the "ambiguity" in
material patterning that attends processes of "boundary
arbitration" is itself substance for productive
investigation: "This ambiguity in material culture along
frontiers
[i.e. boundaries] provides an ideal opportunity
for archaeologists to study the process of creolization -
585
how people modify, create and syncretize material objects in
culture contact situations"
(Ibid.:482).
Thus, to account
for Central Andean material patterning ostensibly supportive
of migration hypotheses, the investigative emphasis shifts
from study of the mobilization of people to study of the
mobilization of style.
Archaeological evidence recounted in
the following paragraphs demonstrates that societal
boundaries at the margins of the Central Andes were loci of
stylistic innovations at crucial moments in prehistory.
The term "boundary interaction," might be utilized to
refer to inter-group interactions across mutually recognized
social or socio-political boundaries that involve the overt
manipulation of material culture, and especially stylistic
innovation.
The term "innovation" is employed here
following Hantman and Flog's (1982:240) definition of the
process as a systematic recombination of style attributes or
attribute clusters.
Of course all human interactions
involve boundaries of some form or other, and most can be
accompanied by some intentional manipulation of signs and
symbols.
At the edges of the Central Andes, however, the
theoretical development of a boundary interaction concept
can provide a useful interpretive tool for archaeological
investigations because of: 1) the ancient and persistent
coincidence of cultural boundaries with the major highlandlowland ecotones, 2) the unusually sharp juxtapositions of
disparate cultural traditions, especially between Andean and
586
Amazonian areas, and 3) the particularly abundant
archaeological evidence for style innovation in and around
the Central Andean boundaries.
Discussions of Andean and Amazonian cultural
boundaries, the Andean-Amazonian ecotone and the contrast
between Andean and Amazonian cultural traditions have been
presented by numerous scholars including Lyon (1981),
Raymond (1985), and LeMoine and Raymond (1987).
The
presence of culturally "intermediate," yet unique,
autochthonous eastern montane forest populations rendered
the pre-Hispanic juxtaposition of Andean and Amazonian
cultural traditions far less severe than it appears today.
Archaeological studies in the eastern montane and premontane
forests have uncovered numerous instances of particularly
obvious manipulation, or "hybridizing" of styles (e.g. Allen
1968:350, 352; Raymond 1976:209; Hastings 1985:589; Church
1994:290).
Stylistic innovation characteristic of boundary
interaction at the Central Andean margins can be classed
into three closely related types identified during the
course of this analysis.
The first and most basic kind of
style innovation has been repeatedly referred to in
literature as style "hybridization."
I prefer the terms
"amalgam" and "amalgamation" to "hybrid" and "hybridization"
to avoid an unintended biological analogy.
Simple style
amalgamation results in a syncretism of two or more styles
587
from identifiable sources.
The second kind of style
innovation might be termed "style displacement" because it
is a particularly sudden and severe instance of style
amalgamation.
The third style innovation is "style
amplification" because it involves style amalgamation in
which selected foreign design attributes are deliberately
amplified or exaggerated.
These kinds of style innovation
presumably accompany many of Renfrew's (1975}
exchange," and perhaps Salomon's (1985}
"modes of
"institutions of
Andean complementarity," but simple correlations between
stylistic similarities and exchange relationships cannot be
uncritically assumed (Hantman and Plog 1982}.
One potentially important mode of boundary interaction
not discussed by Lightfoot and Martinez (1995}, nor fitting
comfortably within any of Renfrew's "modes of exchange," is
the gateway community or gateway system (Hirth 1978}.
Hirth's description (1978:37-38} emphasizes the importance
of local geography, of major environmental and cultural
boundaries, and of gateway interaction's pivotal role in the
emergence of increasingly complex social and political
formations:
These communities flourish at the passage points into and
out of distinct natural or cultural regions and serve as
'gateways• which link their regions to external trade
routes ... They generally are located along natural corridors
of communication and at the critical passages between areas
of high mineral, agricultural, or craft productivity; dense
population; high demand or supply for scarce resources; and,
at the interface of different technologies or levels of
sociopolitical complexity ... Gateway communities operate as
commercial middlemen involving 'wholesale' activities.
588
Hirth also notes that gateway communities are often
autonomous.
Of particular importance to the interpretations offered
within this thesis is Hirth's (1978:43) observation that
mechanisms to express "cultural affiliation" between gateway
communities and the societies situated along external supply
lines are important to the network's functioning.
Apparently, gateway communities and systems begin to appear
at the Central Andean margins as long-distance,
interregional exchange assumes increased importance to
developing socio-political institutions during the Early
Horizon.
In theory, gateway communities should be a locus
of style innovation, as information-exchange through style
amalgamation is apparently a common mechanism for expressing
cultural affiliation.
Archaeologists examining problems related to style as
information-exchange explicitly or implicitly concur that
establishing the contexts of use for artifacts is an
imperative step in formulating viable functional
interpretations of style.
In Chapter 10, it was determined
that Manachaqui Cave is best understood as a wayside
station, perhaps utilized on occasion for habitation by
local populations.
As such, the rockshelter serviced
travelers engaged in interregional exchange during at least
part of each phase, and associated pottery assemblages may
therefore be viewed as active media for information-
589
exchange.
The early Paste A assemblages are largely
portable and thus, appropriate vehicles for conveying
information germane to inter-group exchange protocol.
This
is perhaps even more true of Paste B and C assemblages
which, by virtue of their intrusive presence, must have
conformed to acceptable norms of protocol.
In the following paragraphs, I hope to demonstrate that
some pivotal events in the north-Central Andean prehistoric
sequence previously attributed to population intrusions can
be more accurately interpreted as local developments
accompanied by style innovation in contexts of boundary
interaction.
These events include early population
nucleation, initial acquisition of pottery and the early
emergence of socio-economic stratification.
They support
Lightfoot and Martinez's (1995} assertion that cultural
boundaries are particularly dynamic social environments
where the manipulation of material culture attends the
negotiation and re-negotiation of identities within and
between groups.
Because the spread of the Chavin cult had
such a profound affect on interaction and boundaries, I will
divide the following discussion into pre-Chavin horizon and
post-Chavin horizon considerations of the evidence.
Pre-Chavin Horizon Boundary
Interaction
Migration hypotheses spotlighting Amazonian sources and
a Marafion valley entryway are not served by the late Initial
590
Period arrival of pottery at Manachaqui Cave, although
opening dates for the Manachaqui Phase remain somewhat
obscure.
Interaction with pottery-using Northern Andean
populations most likely precipitated the "dependent
invention" (as described by Clarke and Gosser 1995) of
ceramic technology on the northern slopes of the Central
Andes.
The linkage of the Manachaqui Phase Paste A pottery
to this hypothesized dependent invention obviates the search
for the style's "collateral relatives" and other common
ancestors.
The process presupposes rapid stylistic
elaboration to facilitate entry into interaction
relationships in which pottery was already functioning
symbolically as a common medium.
The acquisition of pottery
technology among montane forest societies probably spread
laterally along, rather than vertically up and down the
Central Andean slopes.
Although much montane forest
exchange must have included small-scale, kin-based
reciprocal exchange of "complementary" subsistence goods
from varied ecological contexts, the verticality framework
cannot account for the transverse dissemination of ceramic
technology and common design attributes around the northern
rim of the Central Andes.
To account for the slope-specific configuration of the
Northern Montane Forest Ceramic Interaction Sphere, models
invoking limited subsistence-related transhumance and longdistance exchange would seem most appropriate.
Hantman and
591
Plog (1982) implicate low population density as the
governing factor responsible for similar homogeneous
stylistic distributions in the U.S. southwest.
However, an
a..Tl.cient montane forest tradition of long-distance "lateral"
exchange of primarily non-subsistence commodities along the
Andean slopes constitutes a viable historical explanation
for the broad and stylistically homogenous distribution
presently observed among montane forest ceramic assemblages.
Despite the difficult terrain traversed, modes of
interaction approximated linear networks of "down-the-line"
or "middleman" exchange like those ethnographically
documented in Amazonia (Lathrap 1973a:172-173; Thomas 1981).
Interacting groups either possessed access to highly
localized resources or specialized in the production of
perishable commodities.
Osborn's (1989) ethnographic
description of the mobile, eastern slope-dwelling Tunebo of
Colombia suggests that the configuration of early montane
forest interaction spheres responded to an absence of fixed
lateral cultural boundaries along the slopes, relative to
rigid and exclusive altitudinal boundaries drawn along
vertical axes.
Evidence from early western montane forest
sites of Nanchoc (Dillehay et al. 1989) and Montegrande
(Tellenbach 1981) may contradict hypotheses of high
mobility.
For most of the Central Andean montane forests
however, a model combining limited transhumance and longdistance trade as basic features seems most appropriate in
592
the absence of additional data.
In previous discussions I suggested that the
lightweight, but well-fired Manachaqui Phase Paste A
restricted vessel shapes were ideally suited for transport.
These same characteristics describe other northern montane
forest assemblages which are likewise well-adapted for use
during activities requiring a degree of mobility.
In other
words, serviceability for transport may have been an
important factor conditioning formal and technological
aspects of the similar and widely-distributed northern
montane forest styles.
The everted rim stands out as a
particularly useful feature for securing a vessel's contents
with a fastened lid.
be viewed as
bo~~dary
The northern montane forest styles can
amalgams combining practical
considerations for mobility with symbolic expressions of
multiple exchange alliances with neighboring highland,
montane forest and lowland societies.
At the eastern edge of the Central Andes, different
processes would seem to account for the appearance of the
Initial Period Wairajirca pottery style at Kotosh.
Lathrap's characterization of the Wairajirca Phase style as
an intrusive welding of Central Andean and Amazonian ceramic
traditions
(Lathrap 1970:106} does not recognize a local,
eastern slope stylistic component.
Of the
~mall
number of
attributes shared between the Wairajirca and Manachaqui
styles, the short-necked jar shape with reinforced rim is
593
most significant because, between Manachaqui Cave and Cuzco,
it occurs only on the eastern slopes.
Rather than a Type A3
site unit intrusion (Lathrap ed. 1956:15-18; Lathrap
1971:94), Wairajirca pottery can be viewed as a local
montane forest assemblage demonstrating stylistic
amalgamation in a context of boundary interaction.
Clarke
and Blake's (1994) model describing pottery's dependent
invention within a Mesoamerican Formative Period context of
factional competition may provide a useful analogy to
explain the acquisition of pottery at Kotosh and the
Wairajirca amalgamation of Amazonian style attributes.
The Wairajirca style assemblages, especially those
illustrated by Kano (1972, 1979) from Shillacoto, convey
lowland Amazonian symbology even more explicitly than their
presumed Tutishcainyo prototypes.
The Wairajirca case
demonstrates an exaggeration of foreign style attributes or
"style amplification" that Lightfoot and Martinez (1995:485)
would contend is not uncommon under conditions of
competition between social segments in proximity to cultural
boundaries.
As in Chiapas, we might hypothesize that
dependent invention occurred in the context of competitive
attempts by one or more social factions to enhance their
position by adopting a foreign technology and symbology for
ostentatious use.
In Chiapas, and probably at Kotosh, the
manipulation of technology and style by emergent leaders
announces exclusive access to foreign alliances, special
594
esoteric knowledge and supernatural power in a manner
ethnohistorically and ethnographically documented in Central
America, Amazonia and cross-culturally by Helms (1991,
1993) .
While there is little direct evidence for social
organization among Wairajirca Phase upper Huallaga Valley
populations, Burger and Salazar-Burger (1986:77) have
suggested that the more elaborate temples at Kotosh and
Shillacoto were constructed with corporate labor directed by
"recognized leaders" wielding coercive powers, and that some
Kotosh temples appear to have been operating simultaneously.
A trend of increasing temple size during the Preceramic
Period Mito Phase culminates in the appearance of both
Wairajirca pottery and the inordinately large temple at
Shillacoto.
Unusually elaborate, but mostly looted,
Shillacoto Initial Period tombs (Izumi et al. 1972) were
apparently prepared to inter these emergent elites.
La Galgada provides another example of a ceremonial
center with elaborate temples growing steadily through the
late Preceramic and Initial Period transition (Grieder and
Bueno 1985; Burger and Salazar-Burger 1986).
Pottery again
arrives in a context of architectural elaboration and
comparatively rich interments.
While Grieder (1988:190)
finds no Amazonian aspects in the Initial Period
assemblages, a relationship to northern montane forest
styles was established in Chapter 6 of this thesis.
Again,
595
we might hypothesize that ceramic technology and design
attributes were being consciously manipulated to advance the
position of competing social groups.
Undoubtedly, important
differences between the Chiapas, Kotosh and La Galgada
processes would be revealed by more penetrating analyses
beyond the purview of this thesis.
Late Manachaqui Phase remains associated with the
Initial Period "flash horizon" around 800 b.c. may be the
earliest direct evidence for a long tradition of Central
Andean pilgrimage.
By now the rockshelter had begun
servicing travelers, traders, religious specialists and
other pilgrims, entering and leaving the Central Andes.
By
the Early Horizon, broad fields of style-sharing in montane
forest interaction spheres had fragmented into competitive
nodes within an expanding interaction network.
Rather than
reflecting migration, the Suitacocha Phase Paste A
assemblage exhibits style displacement consistent with
boundary conditions (Lightfoot and Martinez 1995:478) where
positioning at a cultural interface augments a society's
options for rapidly changing its social, economic and
political alliances.
The same dynamics may have been
responsible for similar style shifts at Bagua that Shady
(1987b:464) attributes to population invasions.
The Suitacocha Phase amalgamation of styles originating
in the Northern Andean and Oriente lowlands constitutes
unprecedented evidence of a close articulation between
596
Central Andean Early Horizon and Northern Andean Late
Formative (i.e. Chorrera) interaction spheres.
Some
archaeologists working in the north-Central Andes
(Hocquenghem et al. 1993:253-254) have recently pointed out
correspondences between the two regional chronologies, and
postulated exchange relations facilitated by border
intermediaries.
At this time pottery styles in the two
regions do share numerous formal and decorative features.
Some featureshave temporal priority in one or the other
region, while others may have appeared simultaneously during
a synergistic co-development of the interaction spheres.
Perhaps most illustrative of the amalgamation of
Chorrera attributes evident in the Suitacocha style is the
application of unmistakably "Ecuadorian" motifs and
iridescent paint to the preeminent Paste A bowl shape (Shape
C Rim 7 in Figs. 71t, u; 72a).
The only Central Andean
bowls comparable to the Suitacocha Phase carinated variety
are Montegrande's Shape A7 bowls, and those too are likely
derived from interaction with the north.
Additional Central
Andean amalgamation of Chorrera attributes can be discerned
by comparing Montegrande's unusual Shape A4 bowls (Ulbert
1994: Tafel 10 top, 15 top) with Penon del Rio's Forms 9 and
14 (Nieves Zedefio n.d.: Figs. 17d, 19c,d, 24a) from the
Guayas Basin.
Kaulicke (1981: Abb. 8:5-14 Shape 5) has
reported the same shapes from Pandanche, although they have
not yet been reported from Huacaloma.
Perhaps the most
597
"Chorreroid" Central Andean assemblages are found on the
Andean slopes.
At first glance, an absence of chonological agreement
between the Suitacocha Phase, the Chorrera horizon and
ostensibly earlier Cajamarca styles appears to contradict
the preceding arguments for stylistic relationships.
However, the Montegrande assemblages are probably later than
the single 1800 b.c. thermoluminescence date (Ulbert
1994:150).
Likewise, the Pandanche sequence begins no
earlier than Kaulicke's 1490 ± 340 and 1390 + 340 b.c.
radiocarbon dates.
Suitacocha Paste A's relatively drastic
amalgamation of Chorrera and Upano style attributes can be
interpreted not only as a symbolic expression of exchange
alliance with Northern Andean or Oriente groups, but as an
instance of "style amplification" perhaps intended to offset
the competitive edge held by Cajamarca groups situated
closer to northern "suppliers."
Again, verticality cannot account for regional
demography or the absence of stylistic correspondence
between the Paste A and adjacent lowland and highland coeval
assemblages.
Suitacocha Phase Paste A stylistic evidence
suggests that study area societies were fully engaged in the
mediation of north-south longitudinal exchange.
Exchange
mechanisms may not have differed much from prior "down-theline" or "middleman" modes, but regional stylistic
distributions were probably more powerfully influenced by
598
the exchange of ritual paraphernalia and status markers by
"home-base reciprocity exchange (Renfrew 1975:41).
The
flash horizon and Suitacocha Phase Paste B assemblages
likely constitute direct evidence of long-distance
expeditions (Lurnbreras 1993:364) that characterized Early
Horizon interregional exchange networks.
During the Suitacocha Phase, Manachaqui Cave apparently
served as a conduit channeling interregional travel between
northern lowland societies and Central Andean ceremonial
centers connected within a web of intensifying and
increasingly politicized ritual exchange of gifts.
Scholars
(Lathrap 1971; Lathrap et al. 1975; Raymond 1988:295; Burger
1992:102-103) have repeatedly observed that Early Horizon
vessel shapes utilized for Central Andean ritual like the
stirrup-spout bottle originated in the northern and eastern
lowlands.
Burger notes additionally that cylinder and stamp
seals, napkin ring ear spools and modeled anthropomorphic
and zoomorphic bottles found in north-Central Andean
mortuary contexts "may have been inspired by Ecuadorian
cultures."
Lack of fieldwork around suspected boundaries
and inadequate chronological control in both regions may be
masking the impact of the articulation between Early Horizon
and Chorrera interaction spheres in both areas.
Ultimately, the north-Central Andean network of "peer
polity interaction," many of its material accoutrements and
a tradition of long-distance travel and pilgrimage were
599
subsumed within the Chavin horizon interaction sphere.
The
tropical lowland's "upward thrust" identified by Tello and
Lathrap in Chavin de Huantar's iconographic expression has
been explained by Burger (1992:155-156) as an intentional
manipulation of foreign, esoteric images and symbols.
That
the overt manipulation of Amazonian imagery is limited to
this single site during the mid-Early Horizon punctuates the
focal position assumed by Chavin de Huantar within regional
interaction networks.
At this time Central Andean
boundaries were not "closed" so much as neglected, as
interest in foreign commodities was eclipsed by interest in
the regional cult.
Post-Chavin Horizon Boundary
Interaction
Manachaqui Cave seems to have infrequently served local
populations for some seven to nine centuries during which
time the earliest Central Andean expansionist states
emerged, and boundaries became increasingly rigid and
politicized (J. Topic and T. Topic 1983a, 1985; Burger
1992).
During the Colpar Phase, at least some of the Paste
Group B and C vessels were likely carried to Manachaqui Cave
by travelers moving relatively freely around the northern
and eastern peripheries of the bounded Central Andean
interaction spheres.
The Colpar Phase assemblages generally
attest to a lack of intense boundary interaction and style
innovation in the Pataz-Abiseo area.
600
By a.d. 400, Manachaqui Cave had entered into the
service of an interaction alliance between highland Central
Andean, local intermediary montane forest and lowland
Amazonian societies involving llama caravan transport.
Southern Chachapoyas societies assumed the role of montane
forest intermediaries in linear interregional interaction.
For the burgeoning Recuay political economy, this longdistance exchange served a need not so much for maize and
coca which could be grown closer to home, but for exotic
goods and information produced and traded by lowland
Amazonian specialists.
Raymond's (1988) list of lowland
exotica likely traded to Central Andean highlanders during
prehistoric times includes tapir's feet,
feathers, pets, and
hallucinogenic and curative herbs.
Amazonian lowlanders facilitating this east-west
exchange were border intermediaries (cf. Reeve 1994) on the
Huallabamba providing a crucial link to Huallaga river
trade.
The research presented in Chapter 3 suggests that
salt and valued minerals were also exchanged.
Highland
stone and metal tools, in great demand by lowlanders during
the 16th and 17th centuries (DeBoer 1981), were likely
traded deep into Amazonia during prehistory.
The Recuay
relationship with the Pataz-Abiseo area was probably severed
by Huari incursion into the upper Santa Valley around a.d.
700.
Within this scenario of east-west oriented
interaction, the role of groups in the neighboring
601
Huamachuco area remains somewhat enigmatic.
While the amalgamated Ernpedrada Paste A style can be
thought of as the local expression of boundary shifts and
incorporation within an expanding Central Andean interaction
sphere, the assemblage also provides evidence of a pivotal
shift in Andean transport technology.
Changes in formal
attributes signify new socio-economic alliances but, as in
the case of Manachaqui Paste A,
the ceramic technology is
conditioned by the constraints of a new transport
technology.
Larger globular vessels made resistant to
breakage by thickening the walls may have been judged most
appropriate for llama transport.
As the larger Ernpedrada Phase Paste A assemblage jars
are no longer portable for long-distance foot transit, they
likely ceased to function as primary media for informationexchange.
This austere assemblage was given relatively
little decorative attention.
Paste B and C assemblages, on
the other hand, continue to convey information related to
exchange protocol.
In fact, it seems likely that the
primary function of protocol is responsible for the high
degree of stylistic homogeneity among kaolin-ware pottery
traditions across the Early Intermediate Period northCentral Andean highlands (T. Topic 1987:9).
The so-called
Recuay "positive-painted" kaolin wares in particular appear
to be functionally specialized for exchange protocol as
their spatial distributions extend well beyond the upper
602
Santa Valley into coastal and montane forest areas.
These
vessels bear relatively innocuous geometric decorative
motifs to the exclusion of militaristic and mythological
imagery common in the upper Santa Valley Recuay homeland.
While Paste A style displacement at the onset of the
Empedrada Phase inaugurates southern Chachapoyas' entry into
Central Andean interaction spheres, a similar transition
seems to have occurred earlier in northern Chachapoyas where
Central Andean neighbors in Cajamarca had adopted llama
transport technologies during the Layz6n Phase.
Not only
are resemblances between Huayabamba and Empedrada shapes
clearly evident, but stylistic correspondences are
startlingly precise between the Huayabamba Complex and Early
Intermediate Period Cajamarca non-kaolin vessel shapes,
especially given the small Huayabamba sample.
Huayabamba
Form A is comparable to Cajamarca Light-colored Form 1 jars
(Onuki and Terada 1982: Pl. 92:1-6) and Cajamarca Coarse Red
Form 2 jars (Ibid.: Pl. 94:2-10).
Ravines' Forms
c
and D
have close highland counterparts in Cajamarca Light-colored
Forms 3 and 4 jars (Ibid.: Pls. 92:14-17 and 93:1-13).
The
unusual modeled profiles of the Cajamarca "upturned" and
"irregularly thickened" rims are identical to the Huayabamba
vessel rims.
Surface decorations based upon applique and
incision are also common to the two assemblages.
Applique and incised decorations are likewise typical
of Cancharin Phase styles documented in the upper Utcubamba
603
Valley of northern Chachapoyas (Ruiz 1972).
The Huayabamba
Complex style amalgamation can be interpreted as a
reflection of systematic, linear interaction between
Cajamarca and the Huallabarnba waterway by way of Chachapoyas
intermediaries that controlled the mountain passes between
Cajamarquilla (now Bolivar) and Leimebarnba.
Reichlen and
Reichlen (1950:241-242) who worked in both areas first drew
attention to the stylistic correspondences between Cajamarca
and northern Chachapoyas ceramic assemblages.
This east-
west exchange relationship probably began by the beginning
of the Early Cajamarca Period around a.d. 200 (Matsumoto
1993:188), and lasted with varying degrees of intensity
through the Inca-dominated Late Horizon.
Such east-west
exchange systems ultimately left fine kaolin-ware bowls
distributed widely across northern and southern Chachapoyas.
Evidence for the Cajamarca-northern Chachapoyas
relationship is particularly contrary to the expectations of
verticality frameworks because it demonstrates specific
highland interest in Amazonian resources, despite the close
proximity of tropical forest resources west and north of the
Cajamarca Basin.
In fact,
the archaeological data show that
Early Intermediate Period highland polities already
controlled lowland ecological zones on the western slopes
(J. Topic and T. Topic 1985; Proulx 1982).
In this
scenario, the eastern premontane and subtropical forest
style amalgams like Huayabarnba, Camonal and Simariba
604
conceivably represent societies with strategic access to
lowland riverine exchange networks as border intermediaries.
Archaeological data from the Early Intermediate Period
Huallaga Basin attest to similar processes of technoeconomic
change at the edge of the montane forests farcher south.
The Kotosh-Higueras Phase faunal assemblage (Wing 1972:
Table 5) exhibits a sudden rise in the relative abundance of
camelids from 43 percent during the prior Sajara-patac Phase
to 72 percent.
Despite Wing's potentially conflicting
interpretation of camelid utilization, the coincidence of
this shift with the dramatic changes in ceramic technology
and settlement patterns described by Lathrap and Isbell
suggest a close parallel to events in the Pataz-Abiseo area.
The Higueras "Coarse Brown" ware vessels are morphologically
analogous to Empedrada Phase jars and bowls, while the
sudden shift to hilltop settlement locations above the upper
Huallaga likely corresponds to the increased importance of
securing improved, secure access to high-altitude pasturage
for cargo animals.
From the Huanuco Basin, the head of
Huallaga River canoe navigation near Tinge Maria can be
accessed (see Chapter 3).
In the far north-Central Andes, Sechura E (Lanning
1963:209-210) and Tamarindo C (Kaulicke 1991:417)
assemblages may likewise represent the same technoeconomic
transition to llama caravan transport seen in the Empedrada
Phase Paste A assemblage.
However, Lanning and Hocquenghem
605
et al. {1993:455) implicate expansion of the adjacent
Mochica empire.
In Cajamarca, the Layz6n Phase shift to
llama caravan transport apparently predates the widespread
upward settlement relocation that occurred during the
Initial Cajamarca Phase {Sechtin 1986).
Browman {1989:263)
asserts that "the shift from a meat-orientation to a
cargo/wool orientation does not occur until the Middle
Cajamarca Phase c. A.D. 500."
He cites M. Shimada's report
on Cajamarca faunal remains, but Shimada's findings are
inconclusive {M. Shimada 1985:289).
Poor chronometric control and other problems frustrate
attempts to determine whether comparable shifts in ceramic
technologies, settlement patterns and animal utilization
documented elsewhere are approximately synchronous and
indicative of similar local and regional processes.
The
evidence from Huamachuco suggests that such changes were
complex and uneven across the north-Central Andes.
Huamachuco's prehistoric occupations had been situated on
promontories since the Early Horizon {J. Topic and T. Topic
1986: Fig. 2, Cuadra 9; Zaki 1983) but, as previously noted,
archaeological data from the area indicate that camelid
pastoralism was unimportant until much later {McGreevey
1989) .
Eastern montane forest modes of exchange during the
Early Intermediate Period can be effectively characterized
utilizing Hirth's gateway model.
The interpretation of
606
Andean yunga settlements as uniquely permeable boundary
sites functioning in the context of interregional
interaction is not unprecedented.
At the lower Meche Valley
Early Intermediate Period site of Cruz Blanca, J. Topic and
T. Topic (1983a} documented highland kaolin pottery
amounting to 20 to 50 percent of assemblage totals, cooccurring with pottery from the upper western slopes near
Otuzco. · Local pottery shows attributes diagnostic of the
coastal Gallinazo polity, but the site should not
necessarily be interpreted as a Gallinazo-controlled node
for that reason.
Apparently Cruz Blanca served as a western
slope Andean gateway community, and provides an analogy for
the interpretation of eastern montane forest sites and
settlement systems.
Comprising an estimated 16 percent of the total
assemblage, Manachaqui Cave's Empedrada Phase large sample
of Paste C pottery attests that the rockshelter served to
channel interregional interaction as part of a gateway
system.
Deep within the adjacent montane forest,
the Early
Intermediate Period settlement at Gran Pajaten can be
characterized as a gateway community functionally analogous
to Cruz Blanca.
Foreign kaolin ware pottery from the
adjacent highlands comprises seven percent of the pre-Abiseo
Phase assemblage from Building No. 1's fill deposits (Church
1994}.
The gateway model may serve as well to approach an
understanding of late prehistoric northern and southern
607
Chachapoyas settlement patterns, but only continued research
can evaluate the effectiveness and weaknesses of such
models.
In the central subdivision of the eastern montane
forest,
the Tarma Valley Tranca Phase settlement system
conforms to some expectations of the gateway model.
Apparently, reciprocal exchange and redistribution of local
produce took place at the paramount settlement of Tranca
(cf. Renfrew's 1975:48 central place mode of exchange).
However, Tranca also probably functioned as an autonomous
gateway community coordinating interregional exchange
between Andean, local montane forest and Amazonian
societies.
The greater diversity of wares relative to most
other coeval sites supports this interpretation.
The Tranca
settlement system, especially at Tranca and Pomohuain (Site
H15), shows clustered associations between Huacrapuquio,
Malambo and Camonal assemblages despite the otherwise
discrete regional distributions of the three wares.
Some
sites like Site CH41 in the lower montane forest
additionally show mixing of Andean and Amazonian
architectural techniques (Hastings 1985:639).
As exotic elements like the Huacrapuquio pottery do not
cluster particularly strongly at Tranca, it may be that the
system hierarchy reflects mostly local redistribution while
mediation of interregional exchange was carried out at
several nodes within a gateway system.
On the other hand,
608
small surface and subsurface pottery samples from the Tarma
Valley sites may not adequately reflect real distributions.
While present evidence for the gateway analogy is not
particularly strong, neither do the limited data support
other interpretations any more convincingly.
Ostensibly
"weak" patterning may be a function of interaction's
comparatively low levels of integration and intensity in the
Tarma Valley and adjacent regions.
Most likely such
interpretations will remain debatable until more evidence is
available.
Future investigations in the eastern montane
forest should also target ethnohistorically described trade
nodes at strategic lowland locations to explore other models
of interaction that may or may not resemble the early
Historic Period Urubamba River trade fairs below Cuzco
described by Gade (1972), Camino (1977) and Lyon (1981)
Conclusion
Through the course of this thesis, we have seen how
scholars have presented a variety of perspectives regarding
prehistoric settlement in the Central Andean montane
forests.
Their viewpoints are usually consistent with the
either the "unitary origins" or the "verticality" ruling
theoretical frameworks.
Few, if any of these perspectives
are truly free of all of the assumptions listed by Raymond
and included in my Chapter 1.
Most scholars presume that
insurmountable difficulties posed by this purportedly
inhospitable ecological zone inhibited independent cultural
609
development, and permitted only subsidized permanent
settlement.
Consequently, they have viewed population
movement and colonization as the salient montane forest
settlement process.
In this study I have attempted to evaluate the diverse
migration hypotheses bearing on the problem of montane
forest settlement and cultural development.
Detailed
examination of these hypotheses in Chapter 2 has revealed
that they were formulated post-hoc to account for anomalous
settlement distributions (Adams et al. 1978), and to further
larger theoretical frameworks.
Rather than ignoring these
migration hypotheses, or categorically rejecting them on
theoretical grounds, I have weighed each one separately,
first with the new data from Manachaqui Cave, and second
with extant data from the Central Andes and adjacent
regions.
Virtually all of the hypotheses can be rejected on
purely empirical grounds utilizing the new data from
Manachaqui Cave.
That these postulated migrations were
judged improbable does not mean that prehistoric migrations
did not occur in the Central Andes or elsewhere.
It does
suggest however, that advocates of migration explanations
still must demonstrate that some other interpretation does
not better suit the data (Rouse 1958).
Chapter 3's examination of documentary and
archaeological evidence offered a glimpse of late
prehistoric northeastern montane forest societies
610
participating in interregional interaction connecting the
highland Andes and Amazonia.
This evidence suggests Inca
incorporation of a pre-existing exchange system linking
northeastern montane forest and Amazonian Central Huallaga
societies within the imperial interaction network.
Subsequent chapters presented a long, and possibly unbroken
prehistoric sequence of human occupation in the Pataz-Abiseo
study area.
Manachaqui Cave's sequence stands as a record
of autochthonous northeastern montane forest cultural
development, and all of the evidence for prehistoric
population movements and colonization can be accounted for
within an alternative framework of interregional
interaction.
Terminal Preceramic and Initial Period material and
contextual evidence indicates that Manachaqui Cave served as
a wayside station, and perhaps alternately as a temporary
habitation.
Initial Period north-Central Andean population
nucleation on the upper slopes of the cordillera runs
counter to evolutionary expectations of population
clustering within ecological zones most favorable for early
Andean mid-elevation and high altitude crop cultivation.
Yet rather than accepting Onuki's (1985) ecological model,
or replacing it with another ecologically-based explanation,
it may be more appropriate to examine the archaeological
evidence for other factors potentially governing regional
demography.
611
Clear evidence for interregional interaction in the
northern montane forests appears with north-Central Andean
ceramic technology displaying style attributes that have
temporal precedents in distant lowland regions to the north.
However, the Alca source obsidian from northern Arequipa
recovered from Manachaqui Cave's Preceramic Period deposits
offers a rare, imperishable bit of evidence suggesting that
such long-distance exchange was an ancient feature of
prehistoric montane forest economies.
Populations had
already begun clustering around the northern rim of the
Central Andes prior to the onset of the Initial Period.
With new patterns of interregional interaction introduced
during the late Initial Period flash horizon,
the
rockshelter became functionally specialized as a wayside
station where travelers rested, ate and warmed themselves.
While interregional interaction began to fuel
burgeoning Initial Period and Early Horizon political
economies, Manachaqui Cave became a node in a gateway system
channeling long-distance interchange with the Northern AI1des
into the heart of the Central Andes.
Early Horizon
pedestrian traffic probably included religious specialists,
pilgrims and other travelers involved in ritualized gift
exchange between population centers, and perhaps in less
systematic trading and questing for knowledge and experience
(Helms 1991, 1993), much as Amazonian shamans and curanderos
undertake today (Oberem 1980; Langdon 1981; Bastien 1987).
612
Archaeological data from Kotosh, La Galgada and Chavin
de Huantar strongly suggest a correlation between the
amalgamation of local Andean styles with "foreign"
attributes, symbols and imagery, and the early emergence of
economically privileged, socio-political leaders in the
Central Andes.
Hirth's gateway model emphasizes the
catalytic role of long-distance interregional exchange in
stimulating increased social stratification, and the
evidence for emerging sociopolitical complexity at Central
Andean sites strategically situated to participate in
montane forest interaction spheres (e.g. Kotosh, La Galgada,
Poro-Poro and Cerro Blanco} supports his position.
Although Manachaqui Cave is situated at the edge of the
eastern montane forest, almost all of the archaeological
patterning prior to the mid-Early Intermediate Period can be
understood in terms of north-south, rather than east-west
interaction.
At present, the strength and importance of the
connection between the Chorrera and Early Horizon
interaction spheres may be under-appreciated.
In fact,
future research may demonstrate that interregional
interaction with northern regions had a far more substantial
role in the emergence of Central Andean civilization than
did interchange with the Amazonian lowlands to the east.
The ethnohistorically documented modes of east-west exchange
between Andean and Amazonian societies may have been
infrequent prior to the Early Intermediate Period, or highly
613
localized like the Kotosh-Tutishcainyo linkage.
Between a.d. 200 and 400, interregional interaction
involving study area societies underwent a techno-economic
revolution with the introduction of llama-caravan transport
technology.
Interaction now connected centralized highland
Central Andean polities and lowland Amazonian riverine
exchange networks via montane forest intermediaries.
Manachaqui Cave again seems to have functioned in the
context of a gateway system of exchange.
If it is
determined that portions of the Empedrada Phase Paste A
assemblage are intrusive from the Callej6n de Huaylas, then
the archaeological evidence may reflect Recuay access to
"wholesale"
(Hirth 1978:38) commodities (i.e. bypassing
additional intermediaries) at one or more autonomous gateway
communities within the montane or premontane forests.
During both the Suitacocha and Ernpedrada Phases,
long-
distance exchange would in theory have been tightly directed
between sources and destinations.
Continued regional
survey, subsurface sampling and comparative ceramic analyses
in the Pataz-Abiseo area, and especially in the Montane Rain
Forest Zone, are all crucial to confirm or reject such a
hypothesis.
Mcst frustrating is t:he paucity of direct evidence for
the commodities
excha~ged,
although virtually all goods from
the middle and lower Amazon basin were perishable.
Material
evidence for these interactions abounds in the form of style
614
manipulation.
During the Early Intermediate Period,
highland amalgamation of lowland design attributes is no
longer seen at Manachaqui Cave as Andean-Amazonian
boundaries moved down-slope.
At this time the most common
pattern of style amalgamation seems to be the grafting of
lowland style attributes like surface decorations onto
montane forest vessel shapes.
This pattern reflects the
upslope movement of information rather than subsistence
produce.
Thus, commodities with the greatest impact on
Central Andean's developmental trajectory may have been
ideas, symbols and experiences from distant regions.
Pan-regional changes in interaction during the Chavin
and Huari horizons resulted in Manachaqui Cave's relative
disuse.
Chavin horizon interaction spheres emphasized
intra-regional over interregional interaction.
Middle
Horizon events north of the Huari-controlled upper Santa
drainage are still poorly understood.
Lack of evidence for
the rockshelter's Middle Horizon use is therefore not
unexpected.
There is no reason to believe that the montane
forests were depopulated at these times, although
populations previously clustered at principle exchange nodes
would be expected to disperse.
The Late Horizon prosperity indicated by Gran Pajaten's
monumental architecture and elaborate lithic sculpture can
be understood in terms of the lucrative material advantages
acquired by "middlemen" (Lathrap 1973a:l72-173; Oberem
615
1974:350).
Montane forest intermediaries at gateway
communities are situated at the interface of two distinct
regional economies.
Hirth (1978:37) describes such
interfaces as "economic shear lines where cost factors
change."
Gran Pajaten's extraordinary architectural
achievements, the associated ornate funerary complex of Los
Pinchudos and other evidence suggestive of a ranked social
organization can be accounted for in terms of the PatazAbiseo area's privileged geographic position and gateway
interaction's common association with centralized sociopolitical authority.
This Andean-Amazonian exchange network
was apparently so important to the maintenance of Andean
social and political structures that it was left intact,
albeit overseen, by the Inca conquerors.
Historical documents reviewed in Chapter 3 offer scant,
but unequivocal evidence for autonomous late prehistoric
montane forest societies and Andean-Amazonian exchange
activities in the Pataz-Abiseo area.
Thus, Sucos, Puymal
and the Late Horizon monumental architectural complexes
encountered in the Pataz-Abiseo Montane Rain Forest Zone
represent intermediary components of a gateway system, while
the elaborate late prehistoric northern Chachapoyas
settlements arrayed along the forest edges probably
represent Hirth's "multiple gateway case"
(1978: Fig. 2c).
With the arrival of Europeans, unfamiliar strains of Old
World diseases coursed through interregional exchange
616
network vectors, and montane forest gateway communities like
Gran Pajaten would have been disproportionately buffeted.
Exchange networks disintegrated as the indigenous Central
Andean regional economy shrank, and an ancient tradition of
montane forest trade and trade mediation dwindled.
Ultimately, the eastern montane forests were abandoned, not
for ecological reasons as archaeologists usually suppose,
but for historical reasons.
Montane forest families not
wiped out by pestilence were hauled out under Viceroy
Toledo's mid-16th century mandate.
Those families escaping
Toledo's torch either dispersed to attempt a living on their
own, or departed the montane forest in search of community
elsewhere.
The agricultural potential of tropical montane forest
ecological zones warrants future investigation.
However,
the chief limiting factor in all but the high Tropical
Montane Rain Forest Life Zone (see Chapter 3) was probably
the capacity of societies to mobilize communal labor for the
necessary forest clearing and soil conservation.
Archaeological evidence from the Central Andes demonstrates
that native societies have possessed that capacity since the
Preceramic Period.
Montane forest agriculture may have had
advantages only dimly appreciated by archaeologists.
Parsons and Hastings (1988:214) observe that "a greater
variety of crops can be grown than in the highlands and
usually in a significantly shorter period of time."
617
Kauffmann's (1987:6} conclusion
tha~
intensive (non-
shifting) agriculture cannot be sustained in the montane
forest due to soil erosion and rapid invasion of secondary
regrowth is based upon his observation of modern, rather
than pre-Hispanic, methods of cultivation, and may therefore
be unwarranted.
Reasons why the eastern montane forests are only
lightly populated today are numerous and complex, but the
current lack of infrastructure joining them to the rest of
Peru contrasts sharply with the abundant evidence for
prehistoric economic integration.
In response to gloomy
commentaries regarding the inability of prehistoric montane
forest societies to "evolve" independently, it might be
appropriate to consider Clarke and Blake's observation
quoted on the first page of this chapter.
In short,
prehistoric communities in the Pataz-Abiseo area were both
instigators and outgrowths of regionally evolving change.
Within this thesis, I have attempted to show that the
Pataz-Abiseo area is more profitably investigated as a
cultural interface, than as a frontier.
The term "frontier"
has been avoided because of its core-periphery connotations
(e.g. Lerner 1984:67; Green and Perlman 1985:4).
At this
juncture we might conclude that the prehistoric montane
forests constitute one or more archaeological culture areas
as Lumbreras and Morales have suggested.
However, it might
be argued on the basis of academic conceptions of "culture
618
areas"
(e.g. Kroeber 1939) that the montane forests lacked a
center or "culture core."
Also, our archaeological
knowledge of the Central Andean montane forests is still so
sketchy that utilization of the culture area concept seems
premature.
On the other hand, referring to the montane
forests as one or more sub-regions of the Central Andes
analogous to the "north coast" or "south highlands"
(cf.
Kauffmann 1989:6) is fully justifiable.
Decisions to refer to the Central Andean montane
forests as a frontier, culture area or interaction sphere
have been based upon scholars' theoretical perspectives, and
terms will probably continue to reflect disparate research
agendas.
I have attempted to interpret evidence for what
appear to be shifting complementary and nested ceramic
interaction spheres that have alternately coincided with and
cross-cut montane forest ecological boundaries.
The montane
forest interaction spheres circumscribing the Central Andes
from north to south were loci of constant lateral cultural
interaction contributing cohesion to the Peruvian cotradition throughout prehistory.
An approach highlighting
the montane forests' connectedness, rather than their
marginality, provides an alternative perspective on dynamic
prehistoric societies situated at montane forest cultural
crossroads, and a sense of interregional interaction's
catalytic role in the development of Central Andean
civilization.
REFERENCES CITED
Acosta, Jose de
1958 El P. Jose de Acosta, Prov. al P. Everardo
Mercuriano [1577]. In Monurnenta Peruana. Edited by
A. de Egafia, Vol. 2, pp. 210-286. Monurnenta
Historica Societatis IESU. Rome.
Adams, William Y.
1979 On the Argument from Ceramics to History: a
Challenge Based on Evidence fro Medieval Nubia.
Current Anthropology 20 (4) :727-744.
Adams, William Y., Dennis P. Van Gerven and RichardS. Levy
1978 The Retreat from Migrationism. Annual Review of
Anthropology 7:483-532.
Adovasio, J.M. and Robert F. Maslowski
1980 Cordage, Basketry and Textiles. In Guitarrero Cave:
Early Man in the Andes. Edited by T. Lynch, pp.253292. Academic Press. New York.
Aldenderfer, Mark S. and Charles Stanish
1993 Domestic Architecture, Household Archaeology, and
the Past in the South-Central Andes. In Domestic
Architecture. Ethnicity. and Complementarity in the
South-Central Andes. Edited by M. Aldenderfer,
pp. 1-12. University of Iowa Press. Iowa City.
Allen, William L.
1968 A Ceramic Seqyence from the Alto Pachitea. Peru:
Some Implications for the Development of Tropical
Forest Culture in South America.
Ph.D.
dissertation, University of Illinois, Urbana.
University Microfilms International. Ann Arbor.
Alva, Walter
1986 FrUhe Keramik aus dem Jequetepeque-Tal, Nordperu.
Materialen zur Allgemeinen und Vergleichenden
Archaologie, Band 32. Verlag c. H. Beck. Mtinchen.
1988a Excavaciones en el Santuario del Tiempo Formative
Udima-Poro Poro en la Sierra del Norte del Peru.
Beitrage zur A1lgemeinen und Vergleichenden
Archaologie 8:301-352. Mainz.
619
620
1988b Investigaciones en el Complejo Formative con
Arquitectura Monumental de Purulen, Costa Norte del
Perli. Beitrage zur Allgemeinen und Vergleichenden
Archaologie 8:283-299. Mainz.
Arnich, Jose
1854 Compendia Hist6rico (1768). Libreria de Rosa y
Bouret. Paris.
Ammerman, A. J. and L. L. Cavalli-Sforza
1973 A Population Model for the Diffusion of Early
Farming in Europe. In The Explanation of Culture
Change. Edited by C. Renfrew, pp. 343-357.
University of Pittsburgh Press. Pittsburgh.
Anthony, David
1986 The "Kurgan Culture," Indo-European Origins, and the
Domestication of the Horse: A Reconsideration.
Current Anthropology 27:291-313.
1990
Migration in Archaeology: the Baby and the
Bathwater. American Anthropologist 92 (4) :895-914.
APECO (Asociaci6n Peruana para la Conservaci6n de la
Naturaleza)
1988a Results of 1987 Fieldwork, Faunal Inventory,
Rio Abiseo National Park. Peru. Report submitted to
the David and Lucile Packard Foundation. Lima, Peru
and Boulder, Colorado.
1988b Results of 1988 Fieldwork, Faunal Inventory, Rio
Abiseo National Park. Peru. Report submitted to the
David and Lucile Packard Foundation. Lima, Peru and
Boulder, Colorado.
1991
Results of 1989 Fieldwork, Faunal Inventory, Rio
Abiseo National Park. Peru. Report submitted to the
David and Lucile Packard Foundation. Lima, Peru and
Boulder, Colorado.
Arriaga, Pablo Joseph de
1968 The Extirpation of Idolatry in Peru [1621]
Translated and edited by L. Clark Keating.
University of Kentucky Press. Lexington.
ARSI (Archivurn Romanurn Societatis Iesu, Novi Regni et
Quitensis)
1636~37 Carta Annua de la Provincia del Peru; Missiones de
Infieles.
621
Athens, Steven
1986 The Site of Pumpuentsa and the Pastaza Phase in
Southeastern Lowland Ecuador. Nawpa Pacha 24:111124. Institute of Andean Studies. Berkeley.
Bandelier, Adolf
1907 The Indians and Aboriginal Ruins Near Chachapoyas in
Northern Peru. Historical Records and Studies. New
York.
Bankmann, Ulf
1979 Meche und Recuay. Baessler-Archiv. Neue Folge Band
27:253-271. Berlin.
Bard, E. M. Arnold, R. G. Fairbanks and B. Hamelin
1993 230 Th-, 234 U and 14 C Ages Obtained by Mass Spectrometry
on Corals. Radiocarbon 35:191-199.
Bastien, Joseph
1987 Healers of the Andes: Kallawaya Herbalists and their
Medicinal Plants. University of Utah Press. Salt
Lake City.
Bautista de Escobar, Juan
1899 Memorial de Juan Bautista de Escobar, en Nombre de
las Ciudades de Chachapoyas y Moyobamba sabre que No
Haga Don Martin de la Riba por Dichas Ciudades la
Entrada que Tiene Capitulada. In Autos, Fechos y
Actuados por el General D. Martin de la Riva Herrera
[1653]. Revista de Archives y Bibliotecas Nacionales
Afio II, Vol. III, entregas 3 y 4, pp.81-83. Emprenta
de El Tiempo.
Lima.
II
II
Benevides C., Mario
1971 Analisis de la Ceramica Huarpa. Revista del Museo
Nacional 37:63-88. Lima.
Bennett, Wendell C.
1944 The North Highlands of Peru: Excavations in the
Callej6n de Huaylas and at Chavin de Huantar.
Anthropological Papers of the American Museum of
Natural History Vol. 39, Part 1. New York.
1946
Excavations in the Cuenca Region, Ecuador. Yale
University Publications in Anthropology No. 35. New
Haven.
1948
The Central Andean Co-tradition. In b
Reappraisal of Peruvian Archaeology. Memoirs
of the Society for American Archaeology No.4,
pp. 1-7. Salt Lake City.
622
Bennett, Wendell C. and Junius B. Bird
1949 Andean Culture History. American Museum of Natural
History. Handbook Series No. 15. New York.
Binford, Lewis R.
1983 The Archaeology of Place. Journal of Anthropological
Archaeology 1:5-31.
Binford, Lewis R. and Jeremy Sabloff
1982 Paradigms, Systematics and Archaeology. Journal of
Anthropological Research 38:137-153.
Bingham, Hiram
1930 Machu Picchu: A Citidel of the Incas.
Books. New York.
Hacker Art
Bird, Junius B.
1963 Pre-ceramic Art from Huaca Prieta, Chicarna Valley.
Nawoa Pacha 1:29-34. Institute of Andean Studies.
Berkeley.
Bird, Robert McK., David L. Browrnan and Marshall E. Durbin
1983-84 Quechua and Maize: Mirrors of Central Andean
Culture History. Journal of the Steward
Anthropological Society 15 (1-2) :187-240. Urbana.
Birkeland, Peter W., Donald T. Rodbell, Daniel C. Miller and
Susan K. Short
1989a Investigaciones Geologicas en el Parque Nacional Rio
Abiseo, San Martin. Boletin de Lima 64:11:55-64.
Lima.
Birkeland, Peter W., Donald T. Rodbell and Susan K. Short
1989b Radiocarbon Dates on Deglaciation, Cordillera
Central, Northern Peruvian Andes. Quaternary
Research 32:111-113. Seattle.
Bischof, Henning
1982 La Fase Engoroy: Periodos, Cronologia y Relaciones.
In Primer Simposio de Correlaciones Antropol6gicas
Andino-Mesoamericano. Edited by J. Marcos and P.
Norton, pp.135-176. Escuela Superior Politecnica del
Literal, Guayaquil.
Bonavia, Duccio
1964 Investigaciones en la Ceja de Selva de Ayacucho.
Arqueol6gicas No. 6. Institute de Investigaciones
Antropologicas, Lima.
1968a Nucleos de Poblaci6n en la Ceja de Selva de Ayacucho
(Peru). Aetas y Memorias de XXXVII Congreso de
Americanistas Vol. 1, pp.75-83. Mar de Plata.
623
1968b Las Ruinas del Abiseo. Universidad Peruana de
Ciencias y Tecnologia, Lima.
1967-68 Investigaciones Arqueol6gicas en el Mantaro Media.
Revista del Museo Nacional 35:211-294. Lima.
1972a La Ceja de Selva: Colonizadores y Avanzadas. In
Pueblos y Culturas de la Sierra Central del PerU,
pp. 91-99. Cerro de Pasco Corporation. Lima.
1972b Reconocimiento Arqueol6gico en el Area del Mantaro.
Argyeol6gicas 14, pp. 13-38. Institute de
Investigaciones Antropologicas, Lima.
1978
Ecological Factors Affecting the Urban
Transformation in the Last Centuries of the PreColumbian Era. In Advances in Andean Archaeology,
ed. D. L. Browman, pp. 393-411. Mouton Publishers.
The Hague.
1981
Tello y la Arqueologia de la Ceja de Selva.
Hist6rica 5 (2) :149-158. Lima.
1990
Les Ruines du Abiseo. Inca-Peru: 3000 Ans
d'Histoire, pp. 248-261. Musees Royaux d'Art et
d'Histoire. Imschoot, uitgevers. Bruxelles.
1991
Peru Hombre e Historia: de los Origenes al Sigle XV.
Ediciones Edubanco.
Bonavia, Duccio and Rogger Ravines
1967 Las Fronteras Ecol6gicas de la Civilizaci6n Andina.
Amaru 2:61-69. Lima.
1968
Villas del Horizonte Tardio en la Ceja de Selva del
Peru: Algunas Consideraciones. Aetas y Memorias del
XXXVII Congreso Internacional de Americanistas,
Vol. 1, pp.153-160. Mar de Plata.
Bonnier, Elisabeth
1983 Piruru: Nuevas Evidencias de Ocupaci6n Temprana en
Tantamayo, Peru. Gaceta Argyeol6gica Andina 2 (8) :810. Lima.
1988
Aquitectura Preceramica en la Cordillera de los
Andes: Piruru Frente a la Diversidad de los Datos.
Antropol6gica 6 (6) :336-361. Lima.
624
Bonnier, Elisabeth, Julio Zegarra and Juan Carlos Tello
1985 Un Ejernplo de Crono-estratigrafia en un Sitio con
Superposicion Arquitectonica: Piruru-Unidad I/II.
Bulletin de L'Institut Francais D'Etudes Andines
14 (3-4) :80-101. Lima.
Bray, Tamara L.
1995 The Panzaleo Puzzle: Non-local Pottery in Northern
Highland Ecuador. Journal of Field Archaeology
22:137-156.
Braun, Robert
1982 The Formative as Seen from the Southern Ecuadorian
Highlands. In Primer Simposio de Correlaciones
Antropologicas Andino-Mesoamericano, Edited by J.
Marcos and P. Norton, pp. 41-100. Escuela Superior
Politecnica del Litoral, Guayaquil.
Brennan, Curtis T.
1978 Investigations at Cerro Arena. Peru: Incipient
Urbanism on the Peruvian North Coast, Vols. 1 and 2.
Ph.D. dissertation. University of Arizona.
University Microfilms International. Ann Arbor.
Brockington, Donald L., David M. Pereira Herrera, Ramon
Sanzetenea Rocha and Maria de los Angeles Mufioz C.
1995 Estudios Arqueologicos del Periodo Formative en el
Sur-este de Cochabamba 1988-1989. Cuadernos de
Investigacion. Serie Argueologia No.8. Institute
Antropologico y Museo de la Universidad Mayor de San
Simon. Cochabamba, Bolivia.
Brockington, Donald L. and Ramon Sanzetenea Rocha
1989 Decorated Formative Period Pottery from Cochabarnba,
Bolivia: Ryden's Observation. Ethnos 54 (1-2) :63-68.
Folkens Museum Etnografiska. Stockholm.
Browman, David
1970 Early Peruvian Peasants: The Culture History of a
Highland Valley. Ph.D. Dissertation. Department of
Anthropology, Harvard University. Cambridge, MA.
1975
Trade Patterns in the Central Highlands of Peru in
the First Millennium B.C. World Archaeology
6 (3) :322-329.
1989
Origins and Development of Andean Pastoralism: An
Overview of the Past 6000 Years. In The Walking
Larder: Patterns of Domestication. Pastoralism and
Predation. Edited by J. Clutton-Brock, pp. 256-268.
Unwin Hyman. London.
625
Bruhns, Karen 0.
1989 Intercambio entre la Costa y la Sierra en el
Forrnativo Tardio: Nuevas Evidencias del Azuay.
Proceedings of the 46th International Congress of
Americanists. Edited by J. F. Bouchard and M.
Guinea, pp. 57-74. BAR International Series 503.
Brush, Stephen B.
1974 El Lugar del Hombre en el Ecosistema Andino. Revista
del Museo Nacional 40:277-299. Lima.
1977
Mountain Field and Family: the Economy and Ecology
of an Andean Valley. University of Pennsylvania
Press. Philadelphia.
Burger, Richard L.
1981 The Radiocarbon Evidence for the Temporal Priority
of Chavin de Huantar. American Antiquity 46:592-602.
1984a Archaeological Areas and Prehistoric Frontiers: the
Case of Formative Peru and Ecuador. In Social and
Economic Organization in the Prehispanic Andes.
Edited by D. L. Browman, R. L. Burger and M. A.
Rivera, pp.33-71. BAR International Series No. 194.
Manchester.
1984b The Prehistoric Occuoation of Chavin de Huantar,
Peru. University of California Press. Berkeley.
1985a Concluding Remarks: Early Peruvian Civilization and
its Relation to the Chavin Horizon.
In Early
Ceremonial Architecture in the Andes. Edited by C.
B. Donnan, pp. 269-289. Durnbarton Oaks Research
Library and Collection. Washington, D. C.
1985b Prehistoric Stylistic Change and Cultural
Development at Huaricoto, Peru. National
Geographic Research 1 (4) :505-534.
1988
Unity and Heterogeneity within the Chavin
Horizon. In Peruvian Prehistory, edited by R.
Keatinge, pp. 99-144, Cambridge University Press.
Cambridge.
1992
Chavin and the Development of Andean Civilization.
Thames and Hudson. London.
1993
The Chavin Horizon: Stylistic Chimera or
Socioeconomic Metamorphosis. In Latin American
Horizons. Edited by D. Rice, pp. 41-82. Dumbarton
Oaks Research Library and Collection. Washington,
D.C.
626
Burger, Richard L. and Lucy Salazar-Burger
1980 Ritual and Religion at Huaricoto. Archaeology
33:26-32.
1986
Early Organizational Diversity in the Peruvian
Highlands: Huaricoto and Kotosh. In Andean
Archaeology: Papers in Memory of Clifford Evans.
Monograph XXVII, Institute of Archaeology,
University of California, Los Angeles. Edited by R.
Matos, S. A. Turpin and H. H. Eling Jr., pp.65-82.
Los Angeles.
Burger, Richard L., Michael Glascock and Hector Heff
1995 The Sources of the Obsidian Artifacts from
Manachaqui Cave: A Preliminary Consideration.
Unpublished Manuscript in possession of the authors.
Bushnell, George H. S.
1946 An Archaeological Collection from Macas, on the
Eastern Slopes of the Ecuadorian Andes. Man
46 (2) :2-6. London.
1951 The Archaeology of the Santa Elena Peninsula in
Southwest Ecuador. Cambridge University Press.
1966
Some Archaeological Discoveries from the Frontier
Region of Peru and Ecuador, Near Jaen. Aetas y
Memorias del XXXVI Congreso de Americanistas, 1964
Vol. 1, pp. 501-507. Sevilla.
Bush, M. B. and P. A. Colinvaux
1988 A 7000-year Pollen Record from the Amazon Lowlands,
Ecuador. Vegetation 76:141-154.
Caldwell, Joseph R.
1964 Interaction Spheres in Prehistory. In Hopewellian
Studies. Scientific Papers Vol. XII, edited by J. R.
Caldwell and R. L. Hall.
Springfield.
Cameron, Catherine M.
1995 Migration and the Movement of Southwestern Peoples.
Journal of Anthropological Archaeology 14:104-124.
Campana M., Roosevelt
1992 Litoescultura en Nunamarca. Yunga No. 3 (1988), pp.
13-20. Universidad Nacional de Trujillo. Trujillo,
Peru.
Camino, Alejandro
1977 Trueque, Correrias e intercambios entre los Quechuas
Andinos y los Piro y Machiguenga de la Montana
Peruana. Amazonia Peruana 1 (2) :123-140. Lima.
627
Cardich, Augusto
1973 Excavaciones en la Caverna de Huargo, Peru. Revista
del Museo Nacional 39:11-29. Lima.
1985
The Fluctuating Upper Limits of Cultivation in the
Central Andes and their Impact on Peruvian
Prehistory. Advances in World Archaeology 4:293-333.
Academic Press. New York.
de las Casas, Enrique
1930a Provincia de Huallaga. Boletin de la Sociedad
Geografica de Lima Torno XLVII, Trimestre cuarto, pp.
162-210. Lima.
1930b Monografia de la Provincia de Huallaga. Boletin de
la Sociedad Geografica de Lima Torno XLVII, Trimestre
cuarto, pp. 368-390. Lima.
1932
Provincia del Huallaga: Apuntes Monograficos
(continuaci6n) . Boletin de la Sociedad Geografica
de Lima Tomo XLIX, Trimestre tercero, pp. 167-215.
Lima.
1933
Provincia del Huallaga: Apuntes Monograficos
(continuaci6n) . Boletin de la Sociedad Geografica
de Lima Torno L, Trimestre primero, pp. 55-71.
Lima.
1934
La Mejor Ruta a Iquitos por el Rio Huallaga. Boletin
de la Sociedad Geografica de Lima Tomo LI, Trimestre
primero, pp. 51-69. Lima.
1935a Monografia de la Provincia de Huallaga
(Continuaci6n) . Boletin de la Sociedad Geografica
de Lima Tomo LII, Trimestre tercero, pp. 234-256.
Lima.
1935b Monografia de la Provincia de Huallaga
(Continuaci6n) . Boletin de la Sociedad Geografica
de Lima Tomo LII, Trimestre cuarto, pp. 335-354.
Lima.
Cedr6n Goicochea, Elke
1989 Cronologia e Identificaci6n de Funci6n en Tres
Edificios Prehispanicos del Sitio La Playa.
Departamento de San Martin. Bachelor's thesis,
Universidad Nacional de Trujillo. Trujillo, Peru.
Chapman, John and Pavel M. Dolukhanov
1992 The Baby and the Bathwater: Pulling the Plug on
Migrations. American Anthropologist 94 (1) :169-174.
628
Church, Warren B.
1988 Test Excavations and Ceramic Artifacts from Building
No.1 at Gran Pajaten. Department of San Martin,
Peru. Unpublished Master's thesis, Department of
Anthroplogy, University of Colorado, Boulder.
1994
Early Occupations at Gran Pajaten, Peru. Andean Past
4:281-318. Cornell Latin American Studies Program.
Ithaca.
Cieza de Leon, Pedro de
1976 The Incas (1553). Translated by H. de Onis and
edited by V. von Hagen. University of Oklahoma
Press. Norman.
Clark G. A.
1994 Migration as an Explanatory Concept in Paleolithic
Archaeology. Journal of Archaeological Method and
Theory 1 (4) :305-343.
Clark G. A. and J. M. Lindly
1991 On Paradigmatic Biases and Paleolithic Research
Traditions. Current Anthropology 32 (5) :577-587.
Clarke, John E. and Michael Blake
1994 The Power of Prestige: Competitive Generosity and
the Emergence of Rank Societies in Lowland
Mesoamerica. In Factional Competition and Political
Development in the New World. Edited by E. M.
Brumfiel and J. W. Fox, pp. 17-30. Cambridge
University Press. Cambridge.
Clarke, John E. and Dennis Gosser
1995 Reinventing Mesoamerica's First Pottery. In The
Emergence of Pottery: Technology and Innovation in
Ancient Societies. Edited by W. K. Barnett and J. W.
Hoopes, pp.209-222. Smithsonian Institution Press.
Washington, D. C.
Cobo, Padre Bernabe
1890 Historia del Nuevo Mundo [1653], Torno I. Edited by
D. M. Jimenez de la Espada. Sociedad de Bibliofilos
Andaluces. Sevilla.
Coe, Sophie D.
1994 America's First Cuisines. University of Texas Press.
Austin.
Collier, Donald and John V. Murra
1943 Survey and Excavation in Southern Ecuador.
Anthropological Series, Vol. 35, Field Museum of
Natural History. Chicago.
629
Cook, Noble David
1981 Demographic Collapse. Indian Peru 1520-1620.
Cambridge University Press. Cambridge.
1992 Impact of Disease in the Sixteenth-century Andean
World. In Disease and Demography in the Americas.
Edited by J. W. Verano and D. H. Ubelaker, pp. 20713. Smithsonian Institution Press. Washington, D.C.
Cooper, John M.
1946 The Araucanians. In Handbook of South American
Indians, Vol. 2: Andean Civilizations. Edited by J.
H. Steward, pp.687-761. Bureau of American Ethnology
Bulletin 143. Smithsonian Institution. Washington,
D.C.
Cordell, Linda S.
1984 Prehistory of the Southwest. Academic Press. San
Diego.
1995
Tracing Migration Pathways from the Receiving End.
Journal of Anthropological Archaeology 14:104-124.
Correal Urrego, Gonzalo
1979 Investigaciones Arqueol6gicas en Abrigos Rocosos de
Nemoc6n y Sueva. Fundaci6n de Investigaciones
Arqueol6gicas Nacionales, Banco de la RepUblica.
Bogota.
Curtin, Philip D. and Gordon R. Willey
1951 A Survey of New Archaeological Sites in Central
Pataz, Peru. Journal of the Washington Academy of
Sciences 41 (2) :49-63.
Cussler, Clive
1994 Inca Gold. Simon and Schuster. New York.
DeBoer, Warren R.
1972-74 Bin6 Style Ceramics from Iparia. Nawpa Pacha 1012:91-107. Institute of Andean Studies. Berkeley.
1981
The Machete and the Cross: Conibo Trade in the Late
Seventeenth Century. In Networks of the Past,
Proceedings of the XII Chacmool Conference, edited
by P. Francis, F. Kense and P. Duke, pp. 31-48.
University of Calgary Archaeological Association.
Calgary.
1984
Archaeological Reconnaissance in the Central
Huallaga, Department of San Martin, Northeastern
Peru. Unpublished report on 1983 field season in
possession of the author.
630
n.d.
Ceramic Assemblage Variability in
Ecuador and Peru. Paper presented
Archaeology of Formative Ecuador.
Raymond and R. L. Burger, October
Durnbarton Oaks. Washington, D.C.
the Formative of
to the session
Organized by J. s.
7-8, 1995, at
DeBoer, Warren R. and J. Scott Raymond
1987 Roots Revisited: The Origins of the Shipibo Art
Style. Journal of Latin American Lore 13:1:115-132.
Denevan, William M.
1992 The Aboriginal Population of Amazonia. In The Native
Population of the Americas in 1492, edited by W.
Denevan, 2nd edition, pp. 205-234. University of
Wisconsin Press. Madison.
Deza Rivasplata, Jaime
1975-76 La Playa: Un Complejo Arqueol6gico de la Cuenca
del Abiseo. Boletin del Seminario de Arqueologia,
No.17-18, pp.43-50. Institute Riva Aguero,
Pontificia Universidad Cat6lica del Perli. Lima.
Diez de San Miguel, Garci
1964 Visita Hecha a la Provincia de Chucuito por Garci
Diez de San Miguel en el Afio 1567. Casa de la
Cultura Peruana. Lima.
Dillehay, Tom D. and Patricia J. Netherly
1983 Exploring the Upper Zafia Valley of Peru. Archaeology
37 (4) :23-30.
Dillehay, Tom D., Patricia J. Netherly and Jack Rossen
1989 Middle Preceramic Public and Residential Sites on
the Forested Slope of the Western Andes, Northern
Peru. American Antiquity 54 (4) :733-759.
Dobyns, Henry F.
1992 Native American Trade Centers as Contagious Disease
Foci. In Disease and Demography in the Americas.
Edited by J. W. Verano and D. H. Ubelaker, pp. 215222. Smithsonian Institution Press. Washington, D.C.
1993
Disease Transfer at Contact. Annual Review of
Anthropology 22:273-291.
Drewett, Peter
1993 Review of Migrations in Prehistory, by Irving Rouse.
Antiquity 67 (255) :456-457.
631
Earle, Timothy, Terence D'Altroy and C. LeBlanc
1978 Arqueologia Regional de los Periodos Prehispanicos
Tardios en el Mantaro. In III Congreso Peruano El
Hombre y la Cultura Andina: Aetas y Trabajos. Edited
by R. Matos, Torno II, pp. 641-672. Editora Lasontay.
Lima.
Earle, Timothy, Terence D'Altroy, Christine Hastorf,
Catherine Scott, Cathy Costin, Glenn Russell and Elsie
Sandefur
1987 Archaeological Field Research in the Upper Mantaro,
Peru, 1982-1983: Investigations of Inka Expansion
and Exchange. Institute of Archaeology. Monograph 28
University of California, Los Angeles. Los Angeles.
Eddy, J.
1977 Climate and the Changing Sun. Climatic Change 1:173190.
Espinoza Soriano, Waldemar
1967 Los Sefiorios Etnicos de Chachapoyas y la Alianza
Hispano-Chacha. Revista Hist6rica 30:224-332. Lima.
Erickson, Jonathon E. and Clement W. Meighan
1984 Boundaries, Alliance and Exchange in California. In
Exploring the Limits: Frontiers and Boundaries in
Prehistok¥. Edited by S. P. De Atley and F. J.
Findlow, pp. 143-152. BAR Series 223.
Estrada, Emilio
1958 Las Culturas Pre-Clasicas. Formativas o Arcaicas del
Ecuador. Publicaci6n del Museo Victor Emilio Estrada
No. 5. Guayaquil.
Estrada, Emilio, Betty J. Meggers and Clifford Evans
1964 The Jambeli Culture of South Coastal Ecuador.
Proceedings of the U.S. National Museum Vol. 115,
pp. 483-558. Washington, D.C.
Evans, Clifford and Betty J. Meggers
1968 Archaeological Investigations on the Rio Napo,
Eastern Ecuador. Smithsonian Institution Press.
Washington, D.C.
1982
Tecnicas Decorativas Diagn6sticas y Variantes
Regionales Chorrera, un Analisis Preliminar. In
Primer Simposio de Correlaciones Antropol6gicas
Andino-Mesoamericano, Edited by J. Marcos and P.
Norton, pp. 121-134. Escuela Superior Politecnica
del Literal, Guayaquil.
632
Farrand, William R.
1985 Rockshelter and Cave Sediments. In Archaeological
Sediments in Context. Edited by J. K. Stein and W.
R. Farrand. Peopling of the Americas Edited Volume
Series : Vol. 1. Center for the Study of Early Man.
Institute for Quaternary Studies, University of
Maine at Orono. Orono.
Fejes, Paul
1944 Archaeological Explorations in the Cordillera
Vileabamba. Southeastern Peru. Viking Fund
Publication.
Flores-Ochoa, Jorge A.
1979 Pastoralists of the Andes: The Alpaca Herders of
Paratia. Institute for the Study of Human Issues.
Philadelphia.
Fogel, Reidy
1993 Settlements in Time: A Study of Social and Political
Development during the Gallinazo Occupation of the
North Coast of Peru. Unpublished Ph.D. dissertation.
Department of Anthropology, Yale University. New
Haven.
Ford, James
1969 A Comparison of the Formative Cultures in the
Americas: Diffusion or the Psychic Unity of Man.
Smithsonian Institution Press. Washington, D.C.
Fung, Rosa
1972 Las Aldas: Su Ubicaci6n dentro del Proceso Hist6rico
del Peru Antigua Dedalo 5:9-10. Sao Paolo.
1975
Excavaciones in Pacopampa, Cajamarca. Revista del
Museo Nacional. Torno XLI, pp. 129-207. Lima.
1982
El Temprano Surgimiento en el Peru de los Sistemas
Socio-politicos Complejos: Planteamiento de una
Hip6tesis de Desarrollo Original. In Primer Simposio
de Correlaciones Antropol6gicas AndinoMesoamericano, Edited by J. Marcos and P. Norton,
pp. 457-495. Escuela Superior Politecnica del
Literal, Guayaquil.
Gade, Daniel W.
1967 The Guinea Pig in Andean Folk Culture. The
Geographical Review 57 (2) :213-224.
633
1972
Comercio y Colonizaci6n en la Zona de Contacto Entre
la Sierra y las Tierras Bajas del Valle del
Urubamba, Perli. Aetas y Memorias del XXXIX Congreso
Internacional de Americanistas Vol. 4, pp.207-221.
1975
Plants. Man and the Land in the Vilcanota Valley of
Peru. Dr. W. Junk B. V., Publishers. The Hague.
1979
Inca and Colonial Settlement, Coca Cultivation and
Endemic Disease in the Tropical Forest. Journal of
Historical Geography 5 (3) :263-279.
Gambini E., Wilfreda
1984 Santa y Nepefta: Dos Valles. Dos Culturas. Imprenta
M. Castillo R. Lima.
Garcilaso de la Vega, Inca
1966 Royal Commentaries of the Incas and General History
of Peru. Part One (1609). Translated by Harold V.
Livermore. University of Texas Press. Austin.
Gentry, Alwyn
1992 Diversity and Floristic Composition of Andean
Forests of Peru and Adjacent Countries: Implications
for their Conservation. In Memorias del Museo de
Historia Natural. Edited by K. R. Young and N.
Valencia, No. 21, pp. 11-30. Universidad Nacional
Mayor de San Marcos. Lima.
Goldstein, Paul
1993 House, Community, and 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.
Green, Stanton W. and Stephen M. Perlman
1985 Frontiers, Boundaries, and Open Social Sysytems. In
The Archaeology of Frontiers and Boundaries. Edited
by S. W. Green and S. M. Perlman, pp.3-13. Academic
Press. New York.
Greenberg, Joseph H.
1987 Language in the Americas. Stanford University Press.
Stanford.
Grieder, Terence
1978 The Art and Archaeology of Pashash. University of
Texas Press, Austin.
634
1988
Ceramics. In La Galgada. Peru: A Preceramic Culture
in Transition. By T. Grieder, A. Bueno M., C. E.
Smith, Jr. and R. M. Malina, pp. 185-191. University
of Texas Press. Austin.
Grieder, Terrance and Alberto Bueno M.
1985 Ceremonial Architecture at La Galgada. In Early
Ceremonial Architecture in the Andes. Edited by C.
B. Donnan, pp. 93-109. Dumbarton Oaks Research
Library and Collection. Washington, D. C.
Grieder, Terrence, Alberto Bueno M., C. Earle Smith, Jr. and
Robert M. Malina
1988 La Galgada. Peru: A Preceramic Culture in
Transition. University of Texas Press. Austin.
Grossman, Joel W.
1972 Early Ceramic Cultures of Andahuaylas. Apurimac.
Peru. Unpublished Ph.D. dissertation, University of
California, Berkeley.
1985
Demographic Change and Economic Transformation in
the South Central Highlands of Pre-Huari Peru. Nawga
Pacha 21:45-126. Institute of Andean Studies.
Berkeley.
Guaman Poma de Alaya, Felipe
1936 Nueva Cor6nica y Buen Gobierno <Codex Peruvien
Illustrel . Travaux et Memoires de l'Institut
D'Ethnologie. Vol. XXIII. Universite de Paris.
Paris.
Guffroy, Jean
1989 Un Centro Ceremonial Formative en el Alto Piura.
Bulletin de l'Institut Francais d'Etudes Andines
18 (2) :161-208. Lima.
1992
Las Tradiciones Culturales Formativas en el Alto
Piura. In Estudios de Argueologia Peruana. Edited by
D. Bonavia, pp. 99-121. Fomciencias. Lima.
Guffroy, Jean, N. Almeida, P. Lecoq, C. Caillavet, F.
Duverneuil, L. Emperaire and B. Arnaud
1987 Loja Prehispanique: Recherches Archeologigues dans
les Andes Meridionales de L'Equateur. Institut
FranGais D'Etudes Andines. Paris.
Guillet, David
1985 Hacia una Historia Socio-Econ6mica de los Basques en
los Andes Centrales del Peru. Boletin de Lima
38:7:79-84. Lima.
635
Hampe M., Teodoro
1978 Relaci6n de los Encomenderos y Repartimientos del
PerU en 1561. Historia y Cultura No. 12, pp. 76-116.
Lima.
Hantman, Jeffrey L. and Stephen Plog
1982 The Relationship of Stylistic Similarity to Patterns
of Material Exchange. In Contexts for Prehistoric
Exchange. Edited by J. Ericson and T. Earle, pp.
237-263. Academic Press. New York.
Harlan, Jack
1975 Crops and Man. American Society of Agronomy.
Madison.
von Hagen, Victor W.
1955 Highway of the Sun. Little, Brown and Co. Boston.
Harner, Michael J.
1972 The Jivaro: People of the Sacred Waterfall.
Doubleday/Natural History Press. Garden City, NY.
Hartshorn, Gary
1983 Introduction to Chapter 7, Plants. In Costa Rica
Natural History. Edited by D. Janzen, pp. 118-157.
University of Chicago Press. Chicago.
Hastings, Charles M.
1985 The Eastern Frontier: Settlement and Subsistence in
the Andean Margins of Central Peru, Vols. 1 and 2.
Ph.D. dissertation, University of Michigan.
University Microfilms International. Ann Arbor.
1987
Implications of Andean Verticality in the Evolution
of Political Complexity: a View from the Margins. In
The Origins and Development of the Andean State,
edited by J. Hass, S. Pozorski and T. Pozorski,
pp. 145-157. Cambridge University Press. Cambridge.
Hegmon, Michelle
1992 Archaeological Research on Style. Annual Review of
Anthropology 21:517-536.
Helms, Mary W.
1991 Esoteric Knowledge, Geographic Distance, and the
Elaboration of Leadership Status: Dynamics of
Resource Control. In Profiles in Cultural Evolution
Papers from a Conference in Honor of Elman R.
Service. Edited by A T. Rambo and K. Gillogly, pp.
333-349. Anthropological Papers, Museum of
Anthropology, University of Michigan No. 85. Ann
Arbor.
636
1993
Craft and the Kingly Ideal: Art. Trade and Power.
University of Texas Press. Austin.
Henrickson, Elizabeth F. and Mary M. A. McDonald
1983 Ceramic Form and Function: An Ethnographic Search
and an Archeological Application. American
Anthropologist 82:630-643.
Herndon, William L.
1952 EXPloration of the Valley of the Amazon. Edited by
H. Basso. McGraw-Hill Book Co. New York.
Hilbert, Peter Paul
1968 Archaologische Untersuchungen am Mittleren Amazonas.
Dietrich Reimer/Andrews and Steiner. Berlin.
Hill, Betsy
1972-74 A New Chronology of the Valdivia Ceramic Complex
from the Coastal Zone of Guayas Province, Ecuador.
NawPa Pacha 10-12:1-39. Institute for Andean
Studies. Berkeley.
Hirth, Kenneth G.
1978 Interregional Trade and the Formation of Prehistoric
Gateway Communities. American Antiquity 43 (1) :3546.
Hocquenghem, Anne Marie
1993 Rutas de Entrada del Mullu en el Extreme Norte del
Perli. Bulletin de l'Institut Francais d'Etudes
Andines 22 (3) :701-719. Lima.
Hocquenghem, Anne Marie, Jaime Idrovo, Peter Kaulicke and
Dominique Gomis
1993 Bases del Intercambio entre las Sociedades
Norperuanas y Surecuatorianas: Una Zona de
Transici6n entre 1500 A.C. y 600 D.C. Bulletin de
l'Institut Francais d'Etudes Andines 22 (2) :443-466.
Lima.
Hodder, Ian
1982 Symbols in Action. Cambridge University Press.
Cambridge.
Holdridge, L. R.
1967 Life Zone Ecology. Revised edition. Tropical Science
Center. San Jose, Costa Rica.
637
Hoopes, John
1994 Ford Revisited: A Critical Review of the Chronology
and Relationships of the Earliest Ceramic Complexes
in the New World, 6000-1500 B.C. Journal of World
Prehistory 8 (1) :1-49.
Horkheimer, Hans
1959 Algunas Consideraciones Acerca de la Arqueologia en
el Valle del Utcubamba. Aetas del II Congeso
Nacional de Historia del PerU: Epoca Pre-hispanica
Vol. 1, pp. 71-90. Lima.
Hyslop, John
1984 The Inca Road System. Academic Press. New York.
Ibarra Grasso, Dick Edgar and Roy Querejazu Lewis
1986 30.000 Afios de Prehistoria en Bolivia. Editorial Los
Amigos del Libra. La Paz.
INC (Institute Nacional de Cultura, PerU)
1983 Inventario de Monumentos Argyeologicos del Peru:
Zona Norte (Primera Aproximacion). Compiled by R.
Ravines and R. Matos. Departamento de Investigacion
del Patrimonio Monumental. Lima.
Institute Geografico Nacional
1984 Mapa Fisico Politico del PerU (1:1'000,000).
Ministerio de Guerra. Lima
Institute de Geologia y Mineria
1975 Mapa Geologico del Perti. Institute de Geologia y
Mineria, Ministerio de Energia y Minas. Lima.
Isbell, William H.
1968 New Discoveries in the Montana of Southeastern Peru.
Archeology 21 (2) :108-114.
1970 Review of Las Ruinas del Abiseo, by Duccio Bonavia.
American Antigyity 35 (3) :237.
1971 Un Pueblo Rural Ayacuchano durante el Imperio Huari.
Aetas y Memorias del 39 Congreso Internacional de
Americanistas, Vol. 3, pp.89-105. Lima.
1974
Ecologia de la Expansion de los Quechua-hablantes.
Revista del Museo Nacional 40:139-155.
1977
The Rural Foundation for Urbanism. Illinois Studies
in Anthropology No. 10. University of Illinois
Press. Urbana.
638
1983-84 Andean Linguistics and Culture History: An
Examination of Competing Interpretations. Journal of
the Steward Anthropological Society 15 (1-2) :241258.
Isbell, William H. and Katherine Schreiber
1978 Was Huari a State? American Antiquity 43 (3) :372389.
Izaguirre, Bernardino
1923 Historia de las Misiones Franciscanas y Narraci6n de
la Geografia en el Oriente del Peril. Vol. 2.
Talleres Tipograficos de la Penitenciaria. Lima.
Izumi, Seiichi
1971 Development of the Formative Culture in the Ceja de
Montana: A View Point Based on the Materials from
the Kotosh Site. In Dumbarton Oaks Conference on
Chavin edited by Elizabeth Benson, pp.49-72.
Washington D.C.
Izumi, Seiichi and Toshihiko Sono
1963 Andes 2: Excavations at Kotosh.
Kadokawa Publishing. Tokyo.
Peru. 1960.
Izumi, Seiichi and Kazuo Terada
1966 Andes 3: Garbanzal and Pechiche. Tumbes Valley,
Peru. 1960. Kadokawa Publishing. Tokyo.
1972
Andes 4: Excavations at Kotosh. Peru. 1963 and 1966.
University of Tokyo Press. Tokyo.
Izumi, Seiichi, Pedro Cuculiza and Chiaki Kana
1972 Excavations at Shillacoto. Huanuco. Peru. University
Museum, University of Tokyo, Bulletin No. 3. Tokyo.
Jakobsen, Jan, J. Balslev Jorgensen, L. Kempfner Jorgensen
and Inge Schjellerup
1986-87 "Cazadores de Cabezas" en Sitios Pre-Inca de
Chachapoyas. Revista del Museo Nacional 48:139-186.
Lima.
Jij6n y Caamafto, Jacinto
1927 Puruha: Contribuci6n al Conocimiento de los
Aborigenes de la Provincia de Chimborazo de la
Republica de Ecuador. Tipografia y Encuadernaci6n
Salesianas. Quito.
1943
El Ecuador Interandino y Occidental antes de la
Conguista Castellana, Vol. 3. Editorial Ecuatoriana.
Quito.
639
Jouanen, Jose
1941 Historia del Campania de Jesus en 1a Antigua
Provincia de Quito 1570-1774, Vol. 1. Editorial
Ecuatoriana. Quito.
Julien, Catherine J.
1985 The Historical Geography of Chachapoyas.
Ethnohistory of Chachapoyas, Part 1. Unpublished
manuscript in possession of the author.
Julien, Daniel
1993 Late Pre-Inkaic Ethnic Group in Highland Peru: An
Archaeological-Ethnohistorical Model of the
Political Geography of the Cajamarca Region. Latin
American Antiquity 4 (3) :246-273.
Kana, Chiaki
1971 Excavaciones en Shillacoto, Huanuco. Revista del
Museo Nacional 37:52-62. Lima.
1972
Pre-Chavin Cultures in the Central Highlands of
Peru: New Evidence from Shillacoto, Huanuco. In
The Cult of the Feline. Edited by E. P. Benson,
pp. 139-152. Dumbarton Oaks Research Library and
Collection. Washington, D.C.
1979
The Origins of the Chavin Culture. Dumbarton Oaks
Studies in Pre-Columbian Art and Archaeology No. 22.
Washington, D.C.
Kauffmann Doig, Federico
1976 El Perli Argueol6gico. Kompaktos. Lima.
1980
"Los Pinchudos": Exploraci6n de Ruinas Intactas en
la Selva. Boletin de Lima 7:26-31. Lima.
1983
Manual de Argueologia Peruana. Octava edicion,
Peisa. Lima.
1986
Sarcofagos Pre-Incas en los Andes Amaz6nicos. Kuntur
1:4-9. Lima
1987
Andes Amaz6nicos: Investigaciones Argueol6gicas
1980-1986 (Expediciones Antisuyol . Publication of
the Banco Continental, Lima.
1991
Sobrepoblaci6n en los Andes: Una Explicaci6n del
Origen y Proceso de la Cultura Andina. l'imaginaire,
1 (3) :45-48. Publicaci6n Trimestral de la Alianza
Francesa. Lima.
640
1992a Historia del Perli AntiqUo: Una Nueva Perspectiva.
Vol. 1. Editorial Monterrico. S.A.
1992b Historia del Perli AntiqUo: Una Nueva Perspectiva.
Vol. 2. Editorial Monterrico. S.A.
Kauffmann Doig, Federico and Giancarlo Ligabue
1990 Les Andes Amazoniennes. In Inca-Peru: 3000 Ans
d'Histoire, pp. 262-337. Musees Royaux d'Art et
d'Histoire. Imschoot, uitgevers. Bruxelles.
Kauffmann Doig, Federico, Miriam Salazar, Daniel Morales,
Iain Mackay and Oscar Sacay
1989 Andes Amaz6nicas: Sitios Intervenidas por la
Expedici6n Antisuyo/86. Arqyeol6gicas No. 20.
Publicaci6n del Institute de Investigaciones
Antropologicas. Museo Nacional de Antropologia y
Arqueologia. Lima.
Kaufman, Terrence
1990 Language History in South America: What We Know and
How to Know More. In Amazonian Linguistics. Edited
by D. Payne, pp. 13-74. University of Texas Press.
Austin
Kaulicke, Peter
1975 Pandanche: Un Caso Formative en los Andes de
Cajamarca Seminario de Historia Rural Andina.
Universidad Nacional Mayor de San Marcos. Lima.
1981
Keramic der Fruhen Initial-periode aus Pandanche,
Dept. de Cajamarca, Peru. Beitrage zur Allgemeinen
und Vergleichende Archaeologie. 3:363-389.
Munchen.
1991
El Periodo Intermedio Temprano en el Alto Piura:
Avances del Proyecto Arqueol6gico "Alto Piura (1987990). Bulletin de l'Institut Francais d'Etudes
Andines 20 (2) :381-422. Lima.
Kendall, Ann
1976 Preliminary Report on Ceramic Data and the Pre-Inca
Architectural Remains of the (Lower) Urubamba
Valley, Cuzco. Baessler Archiv Neue Folge, Band
XXIV, pp. 41-160. Berlin.
Kent, Jonathan D.
1993 A New Research Facility in Andean Zooarchaeology and
Recent Investigations in Ecuador and Peru. Paper
presented at the 21st Annual Midwest Conference on
Andean and Amazonian Archaeology and Ethnohistory.
St. Louis.
641
1994
The Faunal Remains of Manachaqui Cave, North
Highlands, Peru. Unpublished manuscript in
Possession of the author.
Kalata, Alan
1983 The South Andes. In Ancient South Americans. Edited
by J. D. Jennings, pp. 241-285. W. H. Freeman and
Company. San Francisco.
1993
The Tiwanaku: Portrait of an Andean Civilization.
Blackwell, Cambridge, Massachusetts and Oxford.
Kroeber, Alfred L.
1939 Cultural and Natural Areas of Native North America.
University of California Publications in American
Archaeology and Ethnology Vol. 38. University of
California Press. Berkeley.
Krzanowski, Andrzej
1986 The Cultural Chronology of Northern Andes of Peru
(The Huamachuco-Quiruvilca-Otuzco Region) . Acta
Archaeologica Carpathica Torno XXV, pp. 231-263.
Landar, Herbert
1977 South and Central American Indian Languages.
In
Native Languages of the Americas. Edited by T. A.
Sebeok, pp. 401-527. Plenum Press. New York.
Langdon, E. Jean
1981 Cultural Bases for Trading of Visions and Spiritual
Knowledge in the Colombian and Ecuadorian Montana.
In Networks of the Past, Proceedings of the XII
Chacmool Conference, edited by P. Francis, F. Kense
and P. Duke, pp. 101-116. University of Calgary
Archaeological Association. Calgary.
Langlois, Louis
1939 Utcubamba: Investigaciones Arqueologicas en Este
Valle del Departmento de Amazonas. Revista del Museo
Nacional 8 (2) :224-249. Lima.
1940a Utcubamba: Investigaciones Arqueol6gicas en Este
Valle del Departmento de Amazonas. Revista del Museo
Nacional 9 (1) :34-71. Lima.
194Gb Utcubamba: Investigaciones Arqueol6gicas en Este
Valle del Departmento de Amazonas. Revista del Museo
Nacional, 9 (2) :191-228. Lima.
642
Lanning, Edward P.
1963 A Ceramic Sequence for the Piura and Chira Coast,
North Peru. University of California Publications in
American Archaeology and Ethnology, Vol. 48, No. 2,
pp. 135-284. Berkeley.
1967
Peru Before the Incas. Englewood Cliffs, NJ.
Larco H. , Rafael
1941 Los Cupisniques. Casa Editora "La Cr6nica y
Variedades" S .A., Ltda. Lima.
1962
La Cultura Santa. Valverde, S.A. Lima.
Lathrap, Donald W.
1958 The Cultural Sequence at Yarinacocha, Eastern Peru.
American Antiquity 23 (4) :379-388.
1962
Yarinacocha: Stratigraphic Excavations in the
Peruvian Montana. Unpublished Ph.D. dissertation.
Department of Anthropology. Harvard University.
Cambridge.
1963
Possible affiliations of the Machalilla complex
of coastal Ecuador. American Antiquity, 29:239-241.
1965
Origins of Central Andean Civilization: New
Evidence. Science Vol. 148, pp. 796-798.
1970
The Upper Amazon. Thames and Hudson. London.
1971
The Tropical Forest and the Cultural Context of
Chavin. In Dumbarton Oaks Conference on Chavin.
Edited by E. P. Benson, pp. 73-100. Dumbarton Oaks
Research Library and Collection. Washington, D.C.
1973a The Antiquity and Importance of Long-distance Trade
Relationships in the Moist Tropics of Pre-Columbian
South America. World Archaeology 5 (2) :170-186.
1973b Gifts of the Cayman: Some Thoughts on the
Subsistence Basis of Chavin. Variation in
Anthropology: Essays in Honor of John C. McGregor.
Edited by D. W. Lathrap and J. Douglas, pp. 91-103.
Illinois Archaeological Survey. Urbana.
1974
The Moist Tropics, the Arid Lands and the Appearance
of Great Art Styles in the New World. In Art and
Environment in the Native America. Edited by I. R.
Traylor and M. E. King. Texas Technological
University Press. Lubbock.
643
1977
Our Father the Cayman, Our Mother the Gourd: Spinden
Revisited, or a Unitary Model for the Emergence of
Agriculture in the New World. In Origins of
Agriculture. Edited by C. Reed, pp. 713-752. Mouton
Publishers. The Hague.
1982
Complex Iconographic Features by Olmec and Chavin
and Some Speculations on their Possible
Significance. In Primer Simposio de Correlaciones
Antropol6gicas Andino-Mesoamericano, Edited by J.
Marcos and P. Norton, pp. 301-327. Escuela Superior
Politecnica del Literal, Guayaquil.
1985
Jaws: The Control of Power in the Early Nuclear
American Ceremonial Center. In Early Ceremonial
Architecture in the Andes. Edited by C. B. Donnan,
pp. 93-109. Dumbarton Oaks Research Library and
Collection. Washington, D. C.
Lathrap, Donald (editor)
1956 An Archaeological Classification of Culture Contact
Situations. Memoirs of the Society for American
Archaeology No. 11. Salt Lake City.
Lathrap, Donald W., Donald Collier and Helen Chandra
1975 Ancient Ecuador: Culture, Clay and Creativity 3000300 B.C. Field Museum of Natural History. Chicago.
Lathrap, Donald W., Angelika Gebhart-Sayer and Ann Mester
1985 The Roots of the Shipibo Art Style: Three Waves on
Imiriacocha or There Were "Incas" Before the Incas.
Journal of Latin American Lore 11 (1) :31-119.
Lathrap, Donald W., Angelika Gebhart-Sayer, Ann Mester and
Thomas P. Myers
1987 Further Discussion of the Roots of the Shipibo Art
Style: A Rejoinder to DeBoer and Raymond. Journal of
Latin American Lore 13 (2) :225-271.
Lathrap, Donald W, Jorge G. Marcos, James A. Zeidler
1977 Real Alto: An Ancient Ceremonial Center. Archaeology
3 0 ( 1) : 2-13 .
Lathrap, Donald W. and Lawrence Rays
1963 The Archaeology of the Cave of the Owls in the Upper
Montana of Peru. American Antiquity 29 (1) :27-38.
Lavallee, Daniele
1977 Telarmachay: Campamento de Pastores en la Puna de
Junin del Periodo Formative. Revista del Museo Torno
43, pp. 61-96. Lima.
644
Lavallee, Daniele, Michele Julien and Jane Wheeler
1982 Telarmachay: Niveles Preceramicos de Ocupaci6n.
Revista del Museo Nacional Torno 46, pp. 55-127.
Lima.
Lavallee, Daniele, Michele Julien, Jane Wheeler and Claudine
Karlin
1985 Telarmachay: Chasseurs et Pasteurs Prehistorigues
des Andes Torno 1. Institut Fran9ais d'Etudes
Andines. Editions Recherche sur las Civilisations
No. 20. Paris.
Lecoq, Patrice
1987 La Periode des developpements Regionaux. In Loja
Prehispanigue: Recherches Archeologigues dans les
Andes Meridionales de L'Eguateur by J. Guffroy, N.
Almeida, P. Lecoq, C. Caillavet, F. Duverneuil, L.
Emperaire and B. Arnaud, pp. 217-258. Institut
Fran9ais D'Etudes Andines. Paris.
Ledergerber Crespo, Paulina
1995 Factores Geograficos en la Localizaci6n de Sitios
Arqueol6gicos: El Caso de Morena-Santiago, Ecuador:
Un Informe Preliminar. In Cultura y Medio Ambiente
en el Area Andina Septentrional. Co-ordinated by M.
G. Guinea, J. Bouchard and J. Marcos, pp. 343-375.
Colecci6n Biblioteca Abya-Yala 21. Ediciones Abyaala. Quito.
LeMoine, Genevieve and J. Scott Raymond
1987 Leishmaniasis and Inca Settlement in the Peruvian
Jungle. Journal of Historical Geography 13 (2) :113129.
Lennon Thomas J., Warren B. Church and Miguel Cornejo G.
1989 Investigaciones Arqueol6gicas en el Parque Nacional
Rio Abiseo, San Martin. Boletin de Lima 62:11:43-56.
Leo, Mariella
1992 Problematica del Parque Nacional Rio Abiseo.
Amazonia Peruana 11 (21) :109-144. Lima.
Leon, Blanca, Kenneth R. Young and Lois Brako
1992 Analisis de la Composici6n Floristica del Bosque
Montano Oriental del Peru. Memorias del Museo de
Historia Natural 21:141-154. Universidad Nacional
Mayor de San Marcos. Lima.
Lerche, Peter
1995 Los Chachapoya y los Simbolos de su Historia.
Servicios Editoriales Cesar Gayoso. Lima.
645
Lerner, Shereen
1984 Defining Prehistoric Frontiers: A Methodological
Approach. In Exploring the Limits: Frontiers and
Boundaries in Prehistory. Edited by S. P. De Atley
and F. J. Findlow, pp. 67-80. BAR International
Series 223. Oxford.
Lightfoot, Kent G. and Antoinette Martinez
1995 Frontiers and Boundaries in Archaeological
Perspective. AnnUal Review of Anthropology
24:471-492.
Lippi, Ronald
1983 La Ponga and the Machalilla Phase of Coastal
Ecuador. Doctoral dissertation. Department of
Anthropology, University of Wisconsin-Madison.
University Microfilms International. Ann Arbor.
Lira, Jorge A.
1970 Breve Diccionario Kechwa-Espafiol. Edici6n Popular.
Cusco.
Lopez de Alvarado, Domingo
1907 Carta a S. M. del Cabildo de Moyobamba sobre las
Conquistas de la Riva Herrera [1656] . In Juicio de
Limites entre el Peru and Bolivia: Prueba Peruana.
Presented by v. M. Maurtua. Vol. 3, pp.361-365.
Imprenta de Henrich y Comp. Barcelona.
Loredo, Rafael
1958 Los Repartimientos: Bocetos para la Nueva Historia
del PerU. Lima.
Lumbreras, Luis G.
1974 The Peoples and Cultures of Ancient Peru.
Smithsonian Institution Press. Washington.
1981 Arqueologia de la America Andina. Editorial Milla
Batres. Lima.
1993
Chavin de Huantar: Excavaciones en la Galeria de las
Ofrendas. Materialen zur Allgemeinen und
Vergleichenden Archaologie, Band 51. Verlag Philipp
von Zabern. Mainz am Rhein.
Lynch, Thomas F.
1981 Zonal Complementarity in the Andes: A History of the
Concept. In Networks of the Past, Proceedings of the
XII Chacmool Conference, edited by P. Francis, F.
Kense and P. Duke, pp. 31-48. University of Calgary
Archaeological Association. Calgary.
646
Lynch, Thomas F. (editor)
1980 Guitarrero Cave: Early Man in the Andes. Academic
Press. New York.
1992
Summary of Investigations in Bolivar submitted by
Warren Church. Current Research: Andean South
America. American Antiquity
57 (1) :156.
Lynch, Thomas F. and Susan Pollock
1981 La Arqueologia de la Cueva Negra de Chobshi.
Miscelanea Antropol6gica Ecuatoriana No. 1, pp.92119. Guayaquil.
Lyon, Patricia J.
1981 An Imaginary Frontier: Prehistoric Highland Lowland
Interchange in the Southern Peruvian Andes. In
Networks of the Past, Proceedings of the XII
Chacmool Conference, edited by P. Francis, F. Kense
and P. Duke, pp. 3-18. University of Calgary
Archaeological Association. Calgary.
MacNeish, RichardS., Thomas C. Patterson and David L.
Browman
1975 The Central Peruvian Interaction Sphere. Papers of
the Robert S. Peabody Foundation for Archaeology
Vol. 7. Phillips Academy. Andover.
MacNeish, RichardS., Angel Garcia Cook, Luis G. Lumbreras,
Robert K. Vierra and Antoinette Nelken-Turner
1981 Prehistory of the Ayacucho Basin. Peru, Vol. II:
Excavations and Chronology. University of Michigan
Press. Ann Arbor.
Maguifia c., Teofilo
1988 Chavin: La Epopeya Jamas Contada. Editorial
"Impulse." Lima.
Mariategui, Jose Carlos
1971 Seven Interpretive Essays on Peruvian Reality
Translated by M. Urquidi. University of Texas Press.
Austin.
Mason, J. Alden
1950 The Languages of South American Indians. In Handbook
of South American Indians, Vol. 6: Physical
Anthropology, Linguistics and Cultural Geography of
South American Indians. Edited by J. H. Steward,
pp. 157-318. Bureau of American Ethnology Bulletin
143. Smithsonian Institution. Washington, D.C.
647
Masucci, Maria
1992 Ceramic Change in the Guanga1a Phase. Southwest
Ecuador: A Typology and Chronology. Doctoral
dissertation. Department of Anthropology, Southern
Methodist University. University Microfilms
International. Ann Arbor.
Masuda, Shozo, Izumi Shimada and Craig Morris (editors}
1985 Andean Ecology and Civilization. University of Tokyo
Press. Tokyo.
de la Mata, Fr. Pedro
1923 Arte de la Lengua Cholona. Inca Vol. I, pp. 690-700.
Lima.
Matos, Ramiro and John W. Rick
1981 Los Recursos Naturales y el Poblamiento Preceramico
de la Puna de Junin. Revista del Museo Nacional Torno
44, pp. 23-68. Lima.
Matsumoto, Ryozo
1993 Dos Modos de Proceso Socio-Cultural: El Horizonte
Temprano y el Periodo Intermedio Temprano en el
Valle de Cajamarca. In El Mundo Ceremonial Andino.
Edited by L. Millones andY. Onuki, pp. 169-202.
Senri Ethnological Studies No. 37. National Museum
of Ethnology. Osaka
Mayer, Enrique
1985 Production Zones. In Andean Ecology and
Civilization. Edited by S. Masuda, I. Shimada and C.
Morris, pp. 45-84. University of Tokyo Press. Tokyo.
McCown, Theodore D.
1944 Pre-incaic Huamachuco: Survey and Excavations in the
Region of Huamachuco and Cajabamba. University of
California Publications in American Archaeology and
Ethnology Vol. 39, No. 4, pp. 223-400. University of
California Press. Berkeley.
McGreevy, Tom
1989 Prehispanic Pastoralism in Northern Peru. In The
Walking Larder: Patterns of Domestication,
Pastoralism and Predation. Edited by J. CluttonBrock, pp. 231-239. Unwin Hyman. London.
McGreevy, Tom and Roxanne Shaughnessy
1983 High Altitude Land Use in the Huamachuco Area. In
Investigations of the Andean Past. Edited by D.
Sandweiss, pp. 226-242. Latin American Studies
Program. Cornell University. Ithaca.
648
Means, Philip A.
1931 Ancient Civilizations of the Andes. Charles
Scribner's Sons. New York.
Meggers, Betty J.
1954 Environmental Limitation on the Development of
Culture. American Anthropologist 56:801-824.
1966
Ecuador. Frederick A. Praeger,
Inc. New York.
1971 Amazonia: Man and Culture in a Counterfeit Paradise.
Aldine-Atherton. Chicago.
1975
Application of the Biological Model of
Diversification to Cultural Distributions in
Tropical Lowland South America. Biotropica 7
( 3) : 141-61.
1977
Vegetational Fluctuation and Prehistoric Cultural
Adaptation in Amazonia: Some Tentative Correlations.
World Archaeology 8 (3) :287-303.
1979
Climatic Oscillation as a Factor in the Prehistory
of Amazonia. American Antiquity 44 (2) :252-266.
Meggers, Betty and Clifford Evans
1961 An Experimental Formulation of Horizon Styles in the
Tropical Forest Area of South America. In Essays in
Pre-Columbian Art and Archaeology. By S. K. Lothrop
and others, pp. 372-388. Harvard University Press.
Cambridge.
1983
Lowland South America and the Antilles. In Ancient
South Americans. Edited by J. D. Jennings, pp. 287335. W. H. Freeman. San Francisco.
Meggers, Betty J., Clifford Evans and Emilio Estrada
1965 Early Formative Period of coastal Ecuador: The
Valdivia and Machalilla phases. Smithsonian
Contributions to Anthropology, Vol. 1. Washington,
D.C.
Mejia X., M. Toribio
1956 Appendice. Historia de la Expedici6n Arqueologica al
Marafion de 1937. In Arqueologia del Valle de Casrna.
By J. C. Tello, pp. 321-337. Publicaci6n
Antropol6gica del Archive "Julio C. Tello" de la
Universidad Nacional Mayor de San Marcos, Vol. I.
Lima.
649
Metraux, Alfred
1963 Jesuit Missions in South America. In Handbook of
South American Indians. Vol. 5: The Comparative
Ethnology of South American Indians [1949], edited
by J. H. Steward, pp. 645-655. Bureau of American
Ethnology Bulletin 143. Cooper Square Publishers.
New York.
Miasta G., Jaime
1979 El Alto Amazonas: Arqueologia de Jaen y San Ignacio,
Perli. Seminario de Historia Rural Andina,
Universidad Nacional Mayor de San Marcos. Lima
Middendorf, Ernst
1895 Peru. Beobachtungen und Studien uber das Land und
seine Bewohner wahrend eines 25 jahrigen Aufenhalts, III, Das Hochland von Peru. Robert Oppenheim.
Berlin.
Miller, George R.
1979 An Introduction to the Ethnoarchaeology of Andean
Camelids. Ph.D. dissertation. Department of
Anthropology, University of California at Berkeley.
University Microfilms International. Ann Arbor.
Miller, George and Richard L. Burger
1995 Our Father the Cayman, Our Dinner the Llama: Animal
Utilization at Chavin de Huantar. American Antiquity
60 (3) :421-458.
Miranda, Cristobal de
1906 Relaci6n de los corregimientos y otros oficios que
se proveen en los reynos e provincias del Piru, en
el distrito e gobernaci6n del vissorrey dellos
[1583]. Juicio de Limites entre el Peru y Bolivia.
Prueba Peruana. Torno I, pp. 168-280. Edited by
Victor M. Maurtua. Imprenta de Henrich y Comp.
Barcelona.
1925
Relaci6n hecha por el Virrey D. Martin Enriquez de
los oficios que se provee en la gobernacion de los
reinos y provincias de Peru, 1583. Gobernantes del
Peru. Cartas y Papeles. Siglo XVI. Edited by Roberto
Levillier, Torno IX, pp. 114-230. Imprenta de Juan
Pueyo. Madrid.
650
Mogrovejo, Toribio Alfonso de
1921 Libra de Visitas 1593. Diario de la Segunda Visita
Pastoral que Hizo de su Arquidiocesis el Ilustrisimo
Senor Don Toribio Alfonso de Mogrovejo, Arzobispo de
los Reyes (1593). Revista del Archive Nacional del
PerU Torno II, Entrega I, pp. 37-77. San Martiny Ca.
Lima.
Mohr-Chavez, Karen L.
1980 The Archaeology of Marcavalle, An Early Horizon Site
in the Valley of Cuzco, Peru. Part I. BaesslerArchiv, Neue Folge, Band XXVIII, pp. 203-329.
Berlin.
1981a The Archaeology of Marcavalle, An Early Horizon Site
in the Valley of Cuzco, Peru. Part II. BaesslerArchiv, Neue Folge, Band XXVIII, pp. 107-205.
Berlin.
1981b The Archaeology of Marcavalle, An Early Horizon Site
in the Valley of Cuzco, Peru. Part III. BaesslerArchiv, Neue Folge, Band XXVIII, pp. 241-386.
Berlin.
Morales, Edmundo
1995 The Guinea Pig: Healing. Food and Ritual in the
Andes. University of Arizona Press. Tucson.
Morales Ch., Daniel
1978 Excavaciones en
In III Congreso
Andina: Aetas y
I, pp. 325-344.
las Salinas de San Blas (Junin) .
Peruano El Hombre y la Cultura
Trabajos. Edited by R. Matos, Torno
Editora Lasontay. Lima.
1992
Chambira: Alfareros Tempranos de la Amazonia
Peruana. In Estudios de Argueologia Peruana. Edited
by D. Bonavia, pp. 149-176. Fomciencias. Lima.
1993
Compendia Hist6rico del Perli Torno I: Historia
Arqueol6gica del PerU. Editorial Milla Batres. Lima.
Morris, Craig
1978 The Archaeological Study of Andean Exchange Systems.
In Social Archaeology: Beyond Subsistence and
Dating. Edited by C. Redman et al., pp. 315-327.
Academic Press. New York.
Moseley, Michael E.
1983 Central Andean Civilization. In Ancient South
Americans. Edited by J. D. Jennings, pp. 179-240.
W.H. Freeman and Co. San Francisco.
651
1992 The Incas and Their Ancestors: The Archaeology of
Peru. Thames and Hudson. New York.
Mujica, Elias
1984 Cerro Arena-Layzon: Relaciones Costa-Sierra en el
Norte del PerU. Gaceta Arqueologica Andina 10:12-15.
Lima.
1985 Altiplano-Coast Relationships in the South-Central
Andes: From Indirect to Direct Complementarity. In
In Andean Ecology and Civilization. Edited by S.
Masuda, I. Shimada and C. Morris, pp.l03-140.
University of Tokyo Press. Tokyo.
Murra, John V.
1964 Una Apreciacion Etnologica de la Visita. In Visita
Hacha a la Provincia de Chucuito por Garci Diez de
San Miguel en el Afio 1567. Casa de la Cultura del
Peru. Lima
1967 La Visita de los Chupachu como Fuente Etnol6gica. In
Visita de la Provincia de Leon de Huanuco en 1562
Torno I. Edited by J. Murra. Universidad Nacional
Hermilio Valdizan. Huanuco.
1972
El "Control Vertical" de un Maximo de Pisos
Ecologicos en la Economia de las Sociedades Andinas.
In Ifiigo Ortiz de Zuniga, Visita de la Provincia de
Leon de Huanuco en 1562, 2:429-476. Universidad
Nacional Hermilio Valdizan. Huanuco.
1975 Formaciones Economicas y Politicas del Mundo
Andino. Institute de Estudios Peruanos. Lima.
1980 The Economic Organization of the Inka State.
Research in Economic Anthropology Supplement 1.
JAI Press. Greenwich, CT.
Muscutt, Keith, Vincent R. Lee and Douglas Sharon
1994 Vira Vira: A Chachapoyas Site. Sixpac Manco
Publications. Wilson, WY.
Myers, Thomas P.
1976 Formative Period Occupations in the Highlands of
Northern Ecuador. American Antiquity 41 (3) :353-360.
198la Aboriginal Trade Networks in Amazonia. In
Networks of the Past, Proceedings of the XII
Chacmool Conference, edited by P. Francis, F. Kense
and P. Duke, pp. 19-30. University of Calgary
Archaeological Association. Calgary.
652
1981b Ceramics from the Hacienda Tarapoto, Department of
San Martin, Peru. Nawpa Pacha 19:155-166. Institute
of Andean Studies. Berkeley.
1988
El Efecto de las Pestes Sobre las Poblaciones de la
Amazonia Alta. Amazonia Peruana 8 (15) :61-82. Lima.
Narvaez Vargas, Luis Alfredo
1988 Kuelap: Una Ciudad Fortificada en los Andes Nororientales de Amazonas, Peru. Argyitectura y
Argueologia, compiled and edited by Victor Rangel
Flores, pp.115-142. Consejo Nacional de Ciencia y
Tecnologia. Chiclayo.
NASA (National Aeronautics and Space Administration, USA)
1976 Landsat image ERTS E-2518-14285-7, June 23, 1976.
Nieves, z., Maria
n.d. Analisis de Ceramica Chorrera del Sitio Penon del
Rio. Tesis titular. Escuela de Arqueologia,
Escuela Superior Politecnica del Literal. Guayaquil.
Niles, Susan A.
1987 Callachaca:Style and Status in an Inca Community.
University of Iowa Press. Iowa City.
Norton, Presley, Richard Lunnis and Nigel Nayling
1983 Excavaciones en Salango, Provincia de Manabi,
Ecuador. Miscelanea Antropol6gica Ecuatoriana
Vol. 3, pp. 9-72. Banco Central del Ecuador.
Guayaquil.
Nufiez del Prado, Oscar
1958 El Hombre y la Familia: su Matrimonio y su
Organizaci6n Politico-social en Q'ero. Revista
Universitaria No. 114, primer semestre. Cuzco.
Oberem, Udo
1974 Trade and Trade Goods in the Ecuadorian Montana.
In Native South Americans: Ethnology of the Least
Known Continent. Edited by P. Lyon, pp. 346-358.
Little, Brown and Co. Boston.
1980
Los Ouijos: Historia de la Transculturaci6n de un
Grupe Indigena en el Oriente Ecuatoriano. Institute
Otavaleno de Antropologia. Otavalo, Ecuador.
Olivier, Jose
1992 Donald W. Lathrap: Approaches and Contributions in
New World Archaeology. Journal of the Steward
Anthropological Society 20 (1-2) :283-344.
Urbana.
653
ONERN (Oficina Nacional de Evavulaci6n de Recursos
Naturales)
1976 Mapa Ecol6qico del PerU. ONERN. Lima.
Onuki, Yoshio
1985 The Yunga Zone in the Prehistory of the Central
Andes: Vertical and Horizontal Dimensions in Andean
Ecological and Cultural Processes. In Andean Ecology
and Civilization. Edited by S. Masuda, I. Shimada
and C. Morris, pp. 339-356. University of Tokyo
Press. Tokyo.
1993
Las Actividades Ceremoniales Tempranas en la Cuenca
del Alto Huallaga y Algunos Problemas Generales. In
El Mundo Ceremonial Andino. Edited by L. Millones
andY. Onuki, pp. 69-96. Senri Ethnological Studies
No. 37. National Museum of Ethnology. Osaka.
Osborn, Ann
1989 Multiculturalism in the Eastern Andes. In
Archaeological Approaches to Cultural Identity.
Edited by S. Shennan, pp. 141-156. Unwin Hyman.
London.
Otte, Marcel and Lawrence H. Keeley
1990 The Impact of Regionalism on Paleolithic Studies.
Current Anthropology 31 (5) :577-582.
Pardo, Luis A.
1957 Clasificaci6n de la Ceramica Cuzquefia del Antigua
Imperio de los Incas, Parts 1 and 2. Historia y
Arqueoloqia del Cuzco, (II:541-561, 565-596). Cuzco.
Parker, Gary J.
1963 La Clasificaci6n Genetica de los Dialectos Quechuas.
Revista del Museo Nacional Torno 32, pp. 242-252.
Lima.
Parsons, Jeffrey and Charles M. Hastings
1988 The Late Intermediate Period. In Peruvian
Prehistory. Edited by R. W. Keatinge, pp. 190-229.
Cambridge University Press. Cambridge.
Patterson, Thomas C.
1966 Pattern and Process in the Early Intermediate Period
Pottery of the Central Coast of Peru. University of
California Publications in Anthropology Vol. 3.
Berkeley.
654
1971 Chavin: An Interpretation of its Spread and
Influence. In Dumbarton Oaks Conference on Chavin.
Edited by Elizabeth Benson, pp.29-48. Washington
D.C.
1988 Merchant Capital and the Formation of the Inca
State. Dialectical Anthropology 12:217-227.
1991
Early Colonial Encounters and Identities in the
Caribbean: A Review of Some Recent Works and their
Implications. Dialectical Anthropology 16:1-13.
Paulsen, Allison C.
1970 A Chronology of Guangala and Libertad Ceramics of
the Santa Elena Peninsula in South Coastal Ecuador.
Ph.D. dissertation. Faculty of Political Science,
Columbia Universtiy. University Microfilms. Ann
Arbor.
1977
Patterns of Maritime Trade Between South Coastal
Ecuador and Western Mesoamerica, 1500 B.C.- A.D.600.
In The Sea in the Pre-Columbian World. Edited by E.
P. Benson, pp.141-160. Dumbarton Oaks Research
Library and Collection. Washington.
Pearsall, Deborah
1980 Pachamachay Ethnobotanical Report: Plant Utilization
at a Hunting Base Camp. In Prehistoric Hunters of
the High Andes. Edited by J. W. Rick, pp. 191-232.
Academic Press.
1989
Paleoethnobotany. A Handbook of Procedures. Academic
Press. San Diego.
1992
The Origins of Plant Cultivation in South America.
In The Origins of Agriculture: An International
Perspective. Edited by C. W. Cowan and P. J. Watson,
pp. 173-205. Smithsonian Institution Press.
Washington D.C.
Pearson, G. W. and M. Stuiver
1993 High-precision Bidecadal Calibration of the
Radiocarbon Time Scale 500-2500 B.C. Radiocarbon 35:
25-33.
Pefia Mesa, E.
1935 Pajaten. In "Monografia de la Provincia de Huallaga
(Continuaci6n) . " By Enrique de las Casas. Boletin de
la Sociedad Geografica de Lima Torno LII, Trimestre
cuarto, pp. 343-344. Lima.
655
Perez c., Ismael
1988 Monumentos Arqueologicos de Santiago de Chuco, La
Libertad. Boletin de Lima No. 60, pp.33-44. Lima.
Perez v., Jose Luis
1969 Petroglifos en Pataz. Boletin del Seminario de
Argueologia Afio I, No. 1, pp. 97-100. Instituto Riva
Agliero, Pontificia Universidad Catolica del Perli.
Lima.
Peterson, Emil
1984 Morteros Ceremoniales: the Early Development and
Distribution of a Decorated Stone Bowl Tradition in
Northwest South America. In Social and Economic
Organization in the Prehistoric Andes. Edited by D.
L. Browman, Burger, R. L. and M. A. Rivera, pp. 2131. BAR International Series No. 194. Manchester.
Pimentel G., Victor
1967 "Pajaten". Fenix, Revista de la Biblioteca Nacional
17:34-38. Lima.
Plowman, T.
1984 The Origin, Evolution and Diffusion of Coca,
Erythroxylum spp., in South and Central America. In
Pre-Columbian Plant Migrations. Edited by D. Stone,
pp. 125-163. Papers of the Peabody Museum of
Archaeology and Ethnology No. 76. Cambridge.
Porras G., Pedro I.
1974 Historia y Argueologia de la Ciudad Espanola Baeza
de los Ouijos. Siglo XVI. Centro de Publicaciones de
la Pontificia Universidad Catolica del Ecuador.
Quito.
1975a Fase Cosanga. Estudios Cientificos sobre el Oriente
Ecuatoriano, Torno II. Centro de Publicaciones de la
Pontificia Universidad Catolica del Ecuador. Quito.
1975b Fase Pastaza: El Formativo en el Oriente
Ecuatoriano. Separata de la Revista de la
Universidad Catolica Afio III, No. 10, pp. 75-134.
Quito.
1982 Argueologia de Quito. Fase Cotocollao. Centro de
Investigaciones Arqueologicas, Universidad Catolica
del Ecuador. Quito.
1987
Investigaciones Arqueologicas a las Faldas del
Sangay: Tradicion Upano. Quito.
656
1989
Investigations at the Sangay Mound Complex. National
Geographic Research 5 (3) :374-381.
Pozorski, Thomas
1983 The Caballo Muerto Complex and its Place in the
Andean Chronological Sequence. Annals of Carnegie
Museum Vol. 52, Article 1, pp. 1-40.
Proulx, Donald
1982 Territoriality in the Early Intermediate Period: the
Case of Moche and Recuay. Nawpa Pacha No. 20, pp.
83-96. Institute of Andean Studies. Berkeley.
1985 An Analysis of the Early Cultural Sequence in the
Nepena Valley. Research Report Number 25. Department
of Anthropology. University of Massachusetts,
Amherst. Amherst.
Pulgar Vidal, Javier
1987 Geografia del Peru. Novena edici6n. PEISA. Lima.
Raimondi, Antonio
1876 El PerU. Torno II. Imprenta del Estado. Lima.
1900a Itinerario de los Viajes de Raimondi en el Peru: De
Cajamarca a ... Bambamarca (1860). Boletin de la
Sociedad Geoqrafica de Lima Torno X, Nos. 1, 2 y 3,
pp. 1-40. Lima.
1900b Itinerario de los Viajes de Raimondi en el Peru: De
Bambamarca a ... Huallaga y Regreso por Pizana (1860).
Boletin de la Sociedad Geoqrafica de Lima Torno X,
Nos. 4, 5 y 6, pp. 123-171. Lima.
1942
Notas de Viaj e para su Obra El PerU.
Imprenta Torres Aguirre. Lima.
II
II
Torno I.
Ramp6n S., Lino M.
1959 Sitio Argueol6qico F.P. Cuadernos de Investigaciones
Cientificas No. 1. Misiones Cat6licas de la
Amazonia. Quito.
Ravines, Rogger
1967-68 Review of Las Ruinas del Abiseo, by Duccio
Bonavia. Revista del Museo Nacional. Torno 35, pp.
335-336. Lima.
1971
Grupos de Tradici6n Cazadora en las Tierras Altas de
Huancavelica, Peru. Revista del Museo Nacional Torno
37, pp. 17-27. Lima
657
1972
Los Caciques de Pausamarca: Alga Mas sabre las
Etnias de Chachapoyas. Historia y Cultura No. 6, pp.
217-248.
1978
Antiguos Sitios de Ocupacion en el Rio Huayabamba,
Perli. Historia. Problema y Promesa. Homenaje a Jorge
Basadre Vol. 1, pp. 521-532. Pontificia Universidad
Cat6lica del Perli. Lima.
1981a Los Reines y Sefiorios Locales de los Andes
Centrales: 800-1470 DC. In Historia del Peru. Torno
II Perli Antiquo, pp. 93-184. Editorial Juan Mejia
Baca. Lima.
1981b Yacimientos Arqueologicos de la Region Nororiental
del Peru. Amazonia Peruana 4 (7) :139-175.
Lima.
1982
Panorama de la Argyeologia Andina. Institute de
Estudios Andinos. Lima.
1986-87 Colan: Evidencias Arqueologicas. Revista del Museo
Nacional Torno 48, pp. 55-118. Lima.
Raymond, J. Scott
1972 The Cultural Remains from the Granja de Sivia, Peru:
An Archaeological Study of Tropical Forest Culture
in the Montana. Ph.D. dissertation, Department of
Anthropology, University of Illinois-Urbana.
University Microfilms International. Ann Arbor.
1976
Late Prehistoric and Historic Settlements in the
Upper Montana of Peru. In Canadian Archaeology
Abroad, edited by P. L. Shinnie, J. H. Robertson and
F. Kense, pp. 205-213. University of Calgary
Archaeological Association. Calgary.
1985
Quechuas y Chunchos: Ethnic Boundaries in Eastern
Peru. In Status. Structure and Stratification:
Current Archaeological Reconstructions, Proceedings
of the Sixteenth Annual Chacmool Conference, edited
by M. Thompson, M. T. Garcia and F. J. Kense, pp.
39-50. The Archaeological Association of the
University of Calgary. Calgary.
1988
A View from the Tropical Forest. In Peruvian
Prehistory. Edited by R. W. Keatinge, pp. 279-300.
Cambridge University Press.
658
1992
Highland Colonization of the Peruvian Montana in
Relation to the Political Economy of the Huari
Empire. Journal of the Steward Anthropological
Society 20 (1-2) :17-36. Urbana.
Raymond, J. Scott, Warren R. DeBoer and Peter G. Roe
1975 Cumancaya: A Peruvian Ceramic Tradition. Occasional
Paper No. 2, Department of Archaeology, University
of Calgary. Calgary.
Raymond, J. Scott, Augusto Oyuela-Caycedo and Patrick
Carmichael
1994 Una Comparaci6n de las Tecnologias de la Ceramica
Temprana de Ecuador y Colombia. In Tecnologia y
Organizaci6n de la Producci6n Ceramica Prehispanica
en los Andes. Edited by I. Shimada, pp. 33-52.
Pontificia Universidad Cat6lica del Perti, Fonda
Editorial. Lima.
Reeve, Mary-Elizabeth
1994 Regional Interaction in the Western Amazon: the
Early Colonial Encounter and the Jesuit Years.
Ethnohistory 41 (1) :106-138.
Reichlen, Henry and Paule
1950 Recherches Archeologiques dans les Andes du Haut
Utcubamba, Journal de la Societe des Arnericanistes,
39:219-246. Paris.
Renard-Casevitz, F. M., Th. Saignes and A. C. Taylor
1988 Al Este de los Andes: Relaciones entre las
Sociedades Amaz6nicas y Andinas entre los Siglos
XV y XVII. 2 Volumes. Institute Frances de Estudios
Andinos. Ediciones ABYA-YALA. Quito.
Renfrew, Colin
1975 Trade as Action at a Distance: Questions of
Integration and Communication. In Ancient
Civilization and Trade. Edited by J. A. Sabloff and
C. C. Lamberg-Karlovski, pp. 3-59. University of New
Mexico Press.
1987
Archaeology and Language: The Puzzle of IndoEuropean Origins. Jonathan Cape. London.
Rice, DonS., Charles Stanish and Phillip R. Scarr (editors)
1989 Ecology. Settlement and History in the Os~ore
Drainage. Peru. Parts 1 and 2. BAR International
Series 545. London.
659
Rice, Prudence M.
1987 Pottery Analysis: A Sourcebook. University of
Chicago Press. Chicago.
Richardson, III, James B.
1973 The Preceramic Sequence and the Pleistocene and
Post-Pleistocene Climate of Northwest Peru. In
Variation in Anthropology: Essays in Honor of John
C. McGregor. Edited by D. W. Lathrap and J. Douglas,
pp. 199-211. Illinois Archaeological Survey. Urbana.
Rick, John W.
1980 Prehistoric Hunters of the High Andes. Academic
Press. New York.
1988
The Character and Context of Highland Precerarnic
Society. In Peruvian Prehistory. Edited by R. W.
Keatinge, pp. 3-40. Cambridge University Press.
Rios, Manuel A., C. F. Ponce, P. Vasquez and A. Tovar
1982 Planificaci6n Para el Establecimiento de Unidades de
Conservaci6n en el Bosque Nublado del Noreste del
Perli. Informe Final del Proyecto WWF-IUCN/1792.
Departamento Manejo Forestal, Universidad Nacional
Agraria La Molina. Lima.
de la Riva Herrera, Martin
1907 Testimonio de lo que Ha Obrado el Gobernador Don
Martin de la Riva Herrera en la Reducci6n de
Differentes Provincias de Indios [1655]. In Juicio
de Limites del Perli: Prueba Peruana. Presented by V.
M. Maurtua. Vol. 3, pp. 259-346. Imprenta de Henrich
y Comp. Barcelona.
Rivera, Mario A.
1975 Una Hip6tesis sobre Movimientos Poblacionales
Altiplanicos y Transaltiplanicos a las Costas del
Norte de Chile. Chungara No. 5, pp. 7-32.
Universidad del Norte, Depto. de Antropologia.
Arica.
1991
The Prehistory of Northern Chile: A Synthesis.
Journal of World Prehistory 5:1:1-47.
Rivera, Mario A. and Francisco Rothhammer
1986 Evaluaci6n Biol6gica y Cultural de Poblaciones
Chinchorro: Nuevas Elementos para la Hip6tesis de
Contactos Transaltiplanicos, Cuenca Amazonas-Costa
Pacifico. Chungara No. 16-17, pp. 295-306.
Universidad de Tarapaca. Arica.
660
Rivera S. , Raul
1957 Libro Primero de Cabildos de San Juan de la Frontera
de Chachapoyas. Penix. Revista de la Biblioteca
Nacional 12:280-336. Lima.
Rivet, Paul
1968 Les Elements Constitutifs des Civilisations du NordOuest et de L'Ouest Sud-americain. International
Congress of Americanists, pp. 1-20. Gothenburg 1924.
Kraus Reprint. Nendeln/Liechtenstein.
Robinson, Roger W.
1994 Recent Excavations at Hacha in the Acari Valley,
Peru. Andean Past 4:9-37. Latin American Studies
Program. Cornell University. Ithaca.
Rodbell, Donald
1991 Late Quaternary Glaciation and Climatic Change in
the Northern Peruvian Andes. Ph.D. dissertation.
Department of Geology, University of Colorado,
Boulder.
Rodbell, Donald T. and Barbara C. S. Hansen
n.d. A Late Pleistocene-Holocene Palynologic and
Sedimentologic Record from the Eastern Andes of
Northern Peru. Unpublished manuscript in possession
of the authors.
Rojas Ponce, Pedro
1966 Un Informe sobre las Ruinas del Pajaten. Cuadernos
Americanos 148 (5) :119-127. Mexico City.
1967
The Ruins of Pajaten, Archaeology 20 (1) :9-17.
1969
La "Huaca" Huayurco, Jaen Cajamarca. Boletin del
Seminario de Arqueologia 4:48-56. Institute Riva
Agliero, Pontificia Universidad Cat6lica del Peru.
Lima.
Roosevelt, Anna C.
1980 Parmana: Prehistoric Maize and Manioc Subsistence
along the Amazon and Orinoco. Academic Press. New
York.
1995
Early Pottery in the Amazon: Twenty Years of
Scholarly Obsurity. In The Emergence of Pottery:
Technology and Innovation in Ancient Societies.
Edited by W. K. Barnett and J. W. Hoopes, pp. 115132. Smithsonian Institution Press. Washington, D.C.
661
Roosevelt, Anna C., R. A. Housley, M. Imazio da Silviera, S.
Maranca and R. Johnson
1991 Eighth Millennium Pottery from a Prehistoric Shell
Midden in the Brazilian Amazon. Science 254:16211624.
Rosas, Hermilio
1976 Investigaciones Arqueol6gicas en la Cuenca del
Chotano. Aetas del XLI Congreso de Americanistas,
1974, Vol. 3, pp. 564-578. Mexico.
Rosas, Hermilio and Ruth Shady
1970 Pacopampa: Un Complejo Temprano del Periodo
Formative Peruano. Argueologia y Sociedad. No. 3,
pp. 3-16. Museo de Arqueologia y Etnologia de la
Universidad Nacional Mayor de San Marcos. Lima.
1974
Sobre el Periodo Formative en la Sierra del Extreme
Norte del Perli. Argueol6gicas 15:6-35. Institute
Nacional de Cultura. Lima
Rostoker, Arthur G.
1988 An Archaeological Collection from Eastern Ecuador.
Unpublished Master's thesis. Department of
Anthropology, San Francisco State University.
Rostworowski de Diez Canseco, Maria
1977 Coastal Fishermen, Merchants, and Artisans in PreHispanic Peru. In The Sea in the Pre-Columbian
World. Edited by E. P. Benson, pp. 167-183.
Dumbarton Oaks Research Library and Collection.
Washington.
Rowe, John H.
1946 Inca Culture at the Time of Spanish Conquest. In
Handbook of South American Indians. Bureau of
American Ethnology Bulletin, 143, Vol. 2, Part 2,
edited by J. H. Steward, pp.182-330. Washington,
D.C.
Rowe, John H. and Dorothy Menzel
1967 Introduction. Peruvian Archaeology. Selected
Readings, edited by J. Rowe and D. Menzel, pp.v-x.
Peek Publications. Palo Alto.
Rozenberg, Catherine
1982 Le Materiel Archeologique de Piruru II: La
Collection Louis Girault. Bulletin de l'Institut
Francais d'Etudes Andines 11 (3-4) :115-141. Lima.
662
Rozenberg, Catherine and Maurice Picon
1990 Circulation, Echange et Production de Poteries dans
les Andes Centrales au Deuxieme Millenaire Av. J.C.
Bulletin de l'Institut Francais d'Etudes Andines
19 (1) :1-14. Lima.
Rouse, Irving
1958 The Inference of Migrations from Anthropological
Evidence. In Migrations in New World Culture History
Edited by R. H. Thompson, pp. 63-68. University of
Arizona Press. Tucson.
1986 Migrations in Prehistory: Inferring Population
Movement from Cultural Remains. Yale University
Press. New Haven.
Rouse, Irving and Jose Cruxent
1963 Venezuelan Archaeology Yale University Press. New
Haven.
Ruiz Estrada, Arturo
1969 Alfareria del Estilo Huari en Cuelap. Boletin del
Seminario de Arqueologia, No. 4, pp. 60-65.
Pontificia Universidad Cat6lica del Peru. Lima.
1972
La Alfareria de Cuelap: Tradici6n y Cambio,
Bachelor's thesis, Universidad Nacional Mayor de San
Marcos, Lima.
Salazar, Ernesto
1983 El Hombre Temprano en el Ecuador. In Nueva Historia
del Ecuador. Vol. 1: Epoca Aborigen. Edited by E.
Ayala Mora, pp. 73-128. Corporaci6n Editora Nacional
Grijalbo.
Salomon, Frank
1985 The Dynamic Potential of the Complementarity
Concept. In Andean Ecology and Civilization. Edited
by S. Masuda, I. Shimada and C. Morris, pp. 511-532.
University of Tokyo Press. Tokyo.
1986 Native Lords of Quito in the Age of the Incas: The
Political Economy of North Andean Chiefdoms.
Cambridge University Press. Cambridge.
Sahlins, Marshall D.
1961 The Segmentary Lineage: An Organization of Preditory
Expansion. American Anthropologist 53:322-345.
Sauer, Carl
1952 Agricultural Origins and Dispersals. American
Geographical Society. New York.
663
Savoy, Gene
1965 El Gran Pajaten Expedition: a Lost Pre-Inca
Civilization in the Eastern Andes. Andean Air Mail
and Peruvian Times Vol.25, No. 1294. Lima.
1970
Antisuyo: The Search for the Lost Cities of the
Amazon, Simon and Schuster, New York.
Schaedel, Richard P.
1948 Stone Sculpture in the Callejon de Huaylas. In b
Reappraisal of Peruvian Archaeology. Edited by W. C.
Bennett, pp. 66-79. Memoirs of the Society for
American Archaeology, No. 4. Menasha.
Schiffer, Michael B.
1983 Toward the Identification of Formation Processes.
American Antiquity 48 (4) :675-706.
Schortman, Edward M. and Patricia A. Urban
1987 Modeling Interregional Interaction in Prehistory. In
Advances in Archaeological Method and Theory Vol.
11, pp. 37-95. Academic Press. New York.
1992
Current Trends in Interaction Research. In
Resources, Power. and Interregional Interaction.
Edited by E. M. Schortman and P. Urban, pp. 235-255.
Plenum Press. New York.
Schjellerup, Inge
1979-80 Documents on Paper and in Stone: A Preliminary
Report on the Inca Ruins in Cochabamba, Province of
Chachapoyas, Peru. Folk 21-22:299-311.
1984
Cochabamba: An Incaic Administrative Centre in the
Rebellious Province of Chachapoyas. Current Research
in the Andes. Edited by A. Kendall. BAR
International Series No. 210, pp. 161-187. Oxford.
1985
Observations on Ridged Fields and Terracing Systems
in the Northern Highlands of Peru. Tools and
Tillage Vol. V, No. 2, pp. 100-121. Copenhagen.
1990
Recherches Archeologiques et Historiques au
Chachapoyas - Perou. In Inca-Peru: 3000 Ans
d'Histoire, pp. 236-247. Musees Royaux d'Art et
d'Histoire. Imschoot, uitgevers. Bruxelles.
1992 Patrones de Asentamiento en las Faldas Orientales de
los Andes de la Region de Chachapoyas. Estudios de
Argueologia Peruana. Edited by D. Bonavia, pp. 355373. Fomciencias. Lima.
664
Science News
1985 A 'Lost and Found' City in Peru Gets New
Perspective. Science News 127 (8) :117. Washington,
D.C.
Sechlin C., Jorge
1986 Patrones de Asentamiento en e1 Proceso Cultural
Prehispanico del Valle de Cajamarca. Serie:
Materiales para la Arqueo1ogia de Cajamarca-3.
Editorial Sudamericana. Trujillo, Perli
Seltzer, G. and C. Hastorf
1990 Climatic Change and its Affect on Prehispanic
Agriculture in the Central Andes. Journal of Field
Archaeology. 17:397-414.
Shady, Ruth
1976 Investigaciones Arqueologicas en la Cuenca del
Utcubamba. Aetas del XLI Congreso de Americanistas,
1974. Vol. 3, pp.579-589.
1983 Una Aproximaci6n al Mundo de las Creencias Andinas:
la Cultura Pacopampa. Patrones de Enterramientos
durante el Formative en la Sierra Norte del Peru.
Museo Nacional de Antropologia y Arqueologia Boletin
No. 8, pp. 17-24. Lima.
1987a La Arqueologia del Norte Peruano. In II Seminario de
Investigaciones Sociales en la Region Norte, pp.7191. Concytec. Lima.
1987b Tradici6n y Cambia en las Sociedades Formativas de
Bagua, Amazonas, Perli. Revista Andina 5 (2) :457-488.
1993
Del Arcaico al Formative en los Andes Centrales.
Revista Andina 11 (1) :103-132.
Shady, Ruth and Hermilio Rosas
1979 El Complejo Bagua y el Sistema de Establecimientos
durante el Formative en la Sierra Norte del Peru.
Nawpa Pacha No. 17, pp. 109-142. Institute of Andean
Studies. Berkeley.
1982 Algunas Consideraciones sobre las Relaciones de
Bagua con otros Centres Formativos del Area Andina.
In Primer Simposio de Correlaciones Antropol6gicas
Andino-Mesoamericano, Edited by J. Marcos and P.
Norton, pp. 383-398. Escuela Superior Politecnica
del Literal, Guayaquil.
665
Shimada, Izumi
1987 Horizontal and vertical Dimensions of Prehistoric
States in North Peru. In The Origin and Development
of the Andean State. Edited by J. Haas, S. Pozorski
and T. Pozorski, pp. 130-144. Cambridge University
Press. Cambridge.
Shimada, Izumi and Adriana Maguifia
1994 Nueva Vision sobre la Cultura Gallinazo y su
Relaci6n con la Cultura Moche. In Moche: Propuestas
y Perspectivas. Edited by S. Uceda and E Mujica.
Travaux de l'institut FranGais d'Etudes Andines No.
79, pp. 31-58. Lima.
Shimada, Izumi, Carlos Elera and Victor Chang J.
1990 Excavaciones en Hornos de Ceramica de la Epoca
Formativa en Batan Grande, Costa Norte del Perli.
Gaceta Argyeol6gica Andina 5 (20) :19-44. Lima.
Shimada, Izumi, Carlos Elera y Melody Shimada
1982 Excavaciones Efectuadas en el Centro de la Huaca
Lucia-Ch6lope, del Horizonte Temprano, Batan Grande,
Costa Norte del Peru: 1979-1981. Argyeol6gicas
19:109-208. Lima.
Shimada, Melody
1982 Zooarchaeology of Huacaloma: Behavioral and Cultural
Implications. In Excavations at Huacaloma in the
Cajarnarca Valley. Peru. 1979: Report of the Japanese
Scientific Expedition to Nuclear America. Edited by
K. Terada andY. Onuki, pp. 303-336. University of
Tokyo Press. Tokyo.
1985
Continuities and Changes in Patterns of Faunal
Resource Utilization: Formative through Cajarnarca
Periods. In The Formative Period in the Cajarnarca
Basin. Peru: Excavations at Huacaloma and Layzon,
1982. Edited by K. Terada andY. Onuki, pp. 289-310.
University of Tokyo Press. Tokyo.
Silva S. Jorge
1988 La Alfareria de Pacharnachay, Junin. Boletin de Lima
57:21-30. Lima.
Silverman, Helaine
1994 Paracas in Nazca: New Data on the Early Horizon
Occupation of the Rio Grande Drainage, Peru. Latin
American Antigyity 5 (4) :359-382.
Snow, Dean
1995 Migration in Prehistory: The Northern Iroquoian
Case. American Antiquity 60 (1) :59-79.
666
Solecki, Ralph S.
1979 Contemporary Kurdish Winter-time Inhabitants of
Shanidar Cave, Iraq. World Archaeology 10 (3) :318330.
Sonin, Robert
1977 Appendix: The Processes Involved in the Production
of Iridescent-Painted Pottery found Archaeologically
on the South Coast of Ecuador. In The Sea in the
Pre-Columbian World. Edited by E. P. Benson, pp.
141-160. Dumbarton Oaks Research Library and
Collection. Washington, D.C.
Spinden, Herbert J.
1917 The origin and Distribution of Agriculture in
America. Proceedings of the XIX International
Congress of Americanists, pp. 269-276. Washington,
D.C.
Stanish, Charles
1992 Ancient Andean Political Economy. University of
Texas Press. Austin.
Stark, Louisa R.
1985 Ecuadorian Highland Quechua: History and Current
Status. In South American Indian Languages. Edited
by H. E. M. Klein and L. R. Stark, pp. 443-481.
University of Texas Press. Austin.
Steward, Julian H.
1948 Tribes of the Montana: An Introduction. In
Handbook of South American Indians, Vol. 3: The
Tropical Forest Tribes. Edited by J. H. Steward,
pp.507-533. Bureau of American Ethnology Bulletin
143. Smithsonian Institution. Washington, D.C.
1949
South American Cultures: An Interpretive Essay. In
Handbook of South American Indians. Vol. 5: The
Comparative Ethnology of South American Indians,
edited by J. H. Steward, pp. 669-772. Bureau of
American Ethnology Bulletin 143. Washington, D.C.
Steward, Julian H. and Louis C. Faron
1959 Native Peoples of South America. McGraw Hill. New
York.
Steward, Julian H. and Alfred Metraux
1948 Tribes of the Peruvian and Ecuadorian Montana. In
Handbook of South American Indians, Vol. 3: The
Tropical Forest Tribes. Edited by J. H. Steward,
pp.535-656. Bureau of American Ethnology Bulletin
143. Smithsonian Institution. Washington, D.C.
667
Stewart, T. D. and Marshall T. Newman
1950 Anthropometry of South American Indian Skeletal
Remains. In Handbook of South American Indians, Vol.
6: Physical Anthropology, Linguistics and Cultural
Geography of South American Indians. Edited by J. H.
Steward, pp.19-42. Bureau of American Ethnology
Bulletin 143. Smithsonian Institution. Washington,
D.C.
Stothert, Karen E.
1993 Un Sitio de Guanga1a Temprano en el Suroeste del
Ecuador. Smithsonian Institution and Banco Central
del Ecuador. Washington, D.C. and Guayaquil.
Straus, Lawrence Guy
1979 Caves: A Paleoanthropological Perspective. World
Archaeology 10 (3):331-339.
1990 Underground Archaeology: Perspectives on Caves and
Rockshe1ters. In Archaeological Method and Theory
Vol. 2. Edited by M. B. Schi~ier, pp. 255-304.
University of Arizona Press.
Strong, William Duncan and Clifford Evans, Jr.
1952 Cultural Stratigraphy in the Viru Valley, Northern
Peru: The Formative and Florescent Epochs. Columbia
University Press, New York.
Stuiver, M. and G. W. Pearson
1993 High-precision Bidecadal Calibration of the
Radiocarbon Time Scale, A.D. 1950-500 B.C. and 2500000 B.C. Radiocarbon 35:1-23.
Sued-Badillo, Jalil
1992 Facing Up to Caribbean History. American Antiquity
57 (4) :599-607
Tarnawiecki, M. C.
1927 Monografia Minera de la Provincia de Patas. Boletin
de la Sociedad Geograf;ca de Lima. Torno XLIII,
Trimestre segundo de 1926, pp. 165-182. Lima.
Tellenbach, Michael
1981 Volbericht tiber die erste Kampagne der Ausgrabung
bei Montegrande im Jequetepeque-Tal, Nordperu.
Beitrage zur Allgemeinen und Vergleichende
Archaeologie. 3:415-435. Munchen.
Tello, Julio C.
1921 Introducci6n a la Historia Antigua del Peru.
Editorial Euforion. Lima.
668
1923 Wira Kocha. Inca Vol. I, No. 1, pp. 93-320; No. 3,
pp. 583-606. Organa del Museo de Arqueologia de la
Universidad Mayor de San Marcos. Lima.
1929 Antiquo PerU: Primera Epoca. Editado para la
Comisi6n Organizadora del Segundo Congreso
Sudamericano de Turismo. Lima.
1938 Objeto y Prop6sito de la Expedici6n Arqueol6gica al
Marafion. Boletin de la Sociedad Geografica de Lima.
Torno LV, Trimestre segundo y tercero, pp. 123-131.
Lima.
1942 Origen y Desarrollo de las Civilizaciones
Prehist6ricas Andinas. Aetas y Trabajos Cientificos
del XVII Congreso Internacional de Americanistas,
Lima 1939. Torno I, pp. 589-720. Lima.
1943 Discovery of the Chavin Culture in Peru. American
Antigyity 9:1:135-160.
1960 Chavin: Cultura Matriz de la Civilizaci6n Andina.
Publicaci6n Antropol6gica del Archive "Julio C.
Tello" de la Universidad Nacional Mayor de San
Marcos, Volumen 2. Lima.
1967
Paginas Escoaidas. Edited by T. Mejia X. Universidad
Nacional Mayor de San Marcos. Lima.
Terada, Kazuo
1979 Excavations at La Pampa in the North Highlands of
Peru. 1975, Report No. 1 of the Japanese Scientific
Expedition to Nuclear America. University of Tokyo
Press, Tokyo.
Terada, Kazuo and Yoshio Onuki (Editors)
1982 Excavations at Huacaloma in the Cajamarca Valley,
Peru. 1979, Report No. 2 of the Japanese Scientific
Expedition to Nuclear America. University of Tokyo
Press, Tokyo.
1985 The Formative Period in the Cajamarca Basin. Peru:
Excavations at Huacaloma and Layzon. 1982. Report 3
of the Japanese Scientific Expedition to Nuclear
America, University of Tokyo Press. Tokyo.
1988
Las Excavaciones en Cerro Blanco y Huacaloma.
Cajamarca. Peru. 1985. Andes Chosashitsu. Tokyo.
Tessmann, Gililter
1930 Die Indianer Nordost-Perus. Friederichsen, de
Gruyter & Co. M.B.H. Hamburg.
669
Thatcher, John P.
1972 Continuity and Change in the Ceramics of Huamachuco,
North Highlands. Peru. Ph.D. dissertation,
University of Pennsylvania. University Microfilms
International. Ann Arbor.
1972-74 Early Intermediate Period and Middle Horizon 1B
Ceramic Assemblages of Huamachuco, North Highlands,
Peru. Nawpa Pacha Nos. 10-12, pp. 109-127, Institute
of Andean Studies. Berkeley.
1979
Early Ceramic Assemblages from Huamachuco, North
Highlands, Peru. Nawpa Pacha, No. 17, pp. 91-106.
Institute of Andean Studies. Berkeley.
Thomas, David J.
1981 Cultural Homogeneity, Resource Distribution and
Exchange Patterns in the Guiana Highlands of South
America. In Networks of the Past, Proceedings of the
XII Chacmool Conference, edited by P. Francis, F.
Kense and P. Duke, pp. 85-100. University of Calgary
Archaeological Association. Calgary.
Thompson, Donald E.
1973 Archaeological Investigations in the Eastern Andes
of Northern Peru, XL Congresso Internazionale Degli
Americanisti 1:363-369. Roma-Genova.
1976
Prehistory of the Uchucmarca Valley in the North
Highlands of Peru. Aetas del XLI Congreso
Internacional de Americanistas Vol. 2, pp. 99-106.
Mexico City.
1980
The Precolumbian and Colonial Heritage of Rapayan.
Archaeology 32 (2} :45-51.
1984 Ancient Highland Connections with Selva and Coast:
Evidence from Uchucmarca, Peru. Current Research in
the Andes. Edited by A. Kendall, pp. 153-160. BAR
International Series, 194:153-160. Oxford.
Thompson L., E. Mosley-Thompsen, M. Davis, P. Lin, K.
Henderson, J. Cole-Dai, J. Bolzan and K. Liu
1995 Late Glacial Stage and Holocene Tropical Ice Core
Records from Huascaran, Peru. Science 269:46-50.
Thompson, Raymond (Editor}
1958 Migrations in New World Culture History. University
of Arizona Press. Tucson.
670
Topic, John R. and Theresa Lange Topic
1982 Huamachuco Archaeological Project Preliminary Report
on the First Season. July-August 1981. Unpublished
report on file at the Department of Anthropology,
Trent University. Peterborough.
1983a Coast-Highland Relations in Northern Peru: Some
Observations on Routes, Networks and Scales of
Interaction. Civilization in the Ancient Americas,
Essays in Honor of Gordon R. Willey, edited by R. M.
Leventhal and A. L. Kalata, pp. 237-260. University
of New Mexico Press and Peabody Museum.
1983b Huamachuco Archaeological Project Preliminary Report
on the Second Season. June-August 1982. Unpublished
report on file at the Department of Anthropology,
Trent University. Peterborough.
1985
Coast-Highland Relations in Northern Peru: the
Structure and Strategy of Interaction. Status,
Structure and Stratification. Current Archaeological
Reconstructions, Proceedings of the Sixteenth
Annual Chacmool Conference, edited by M. Thompson,
M. T. Garcia and F. J. Kense, pp. 55-65. The
Archaeological Association of the University of
Calgary. Calgary.
Topic, Theresa Lange
1987 The Kaolin Ceramic Tradition in Northern Peru. Paper
presented at the Fourth Annual Northeast Conference
on Andean Archaeology and Ethnohistory, SUNY-Albany,
November 2-3, 1985. Revised November, 1987.
Topic, Theresa Lange and John R. Topic
1987 Huamachuco Archaeological Project: Preliminary
Report on the 1986 Field Season. Trent University
Occasional Papers in Anthropology No.4.
Peterborough.
Torero, Alfredo
1964 Los Dialectos Quechuas. Annales Cientificos de la
Universidad Agraria Vol. II, No. 4, pp. 447-478.
Lima.
1974
El Ouechua y la Historia Social Andina. Universidad
Ricardo Palma. Lima.
Torres c., German
1903 La Via Fluvial del Huayabamba. Boletin del la
Sociedad Geografica de Lima. Torno XIII, Trirnestre
tercero, pp. 295-309. Lima.
671
Tosi, Joseph A. Jr.
1960 Zonas de Vida Natural en el Perli. Institute
Interamericano de Ciencias Agricolas, Boletin
Tecnico No. 5, Lima.
Trigger, Bruce
1989 A History of Archaeological Thought. Cambridge
University Press. Cambridge.
Troll, Carl
1958 Las Culturas Superiores Andinas y el Medio
Geografico [1931]. Monografias y Ensayos Geograficos
No. 1. Facultad de Letras, Universidad Nacional
Mayor de San Marcos. Lima.
Turner, Christy G. and Laurel Lofgren
1966 Household Size of Prehistoric Western Pueblo
Indians. Southwestern Journal of Anthropology
22:117-132.
UC and FPCN (University of Colorado and the Fundaci6n
Peruana para la Conservaci6n de la Naturaleza)
1991 Plan Operative 1990-1992. Parque Nacional Rio
Abiseo. Volumen I. Parques Nacionales del Peru.
Lima.
Ulbert, Cornelius
1994 Die Keramik der forrnativzeitlichen Siedlung
Montegrande, Jequetepequetal, Nord-Peru. Materialen
zur Allgemeinen und Vergleichenden Archaologie, Band
52. Verlag Philipp von Zabern. Mainz am Rhein.
Uhle, Max
1922a Influencias Mayas en el Alto Ecuador. Boletin de la
Academia Nacional de Historia, Vol. IV, Nos. 10-11,
pp. 202-240. Quito.
1922b Origenes Centroamericanos. Boletin de la Academia
Nacional de Historia, Vol. IV, No. 9, pp. 1-6.
Quito.
Unanue, Hipolito and Manuel Sobreviela
1963 Historia de las Missiones de Cajarnarguilla v
Reducci6n de la de Manoa. Biblioteca Tenanitla:
Libros Espafioles e Hispanoarnericanos 6. Ediciones
Jose Porrua Turanzas. Madrid.
Valcarcel, Luis E.
1964 Etnohistoria del Peru AntiqUo. Universidad Nacional
Mayor de San Marcos. Departamento de Publicaciones.
Lima.
672
Valencia, Niels
1992 Los Bosques Nublados Secos de la Vertiente
Occidental de los Andes del PerU. In Memorias del
Museo de Historia Natural. Edited by K. R. Young and
N. Valencia, No. 21, pp. 155-170. Universidad
Nacional Mayor de San Marcos. Lima.
Van Buren, Mary
1996 Rethinking the Vertical Archipelago: Ethnicity,
Exchange, and History in the South Central Andes.
American Anthropologist 98 (2) :338-351.
Vargas Ugarte, Ruben
1941 Los Jesuitas del Peru (1568-1767}. A. Castefieda.
Lima.
1959
Historia de la Iglesia en el Peru. Torno II (15701640). Burgos. Spain.
Vasquez de Espinoza, Antonio
1969 Description of the Indies (c.1620}. Translated by C.
U. Clark. Smithsonian Miscellaneous Collections Vol.
102. Smithsonian Institution Press. Washington, D.C.
Vega Ocampo, Abel
1978 Complejo Arqueol6gico de Uchucmarca: Descripci6n del
Elemento Ceramica de Pirca-Pirca. Investigaci6n
Argueol6gica, No. 2, pp. 8-19. Universidad Nacional
de Trujillo. Trujillo, Peru.
1979
Irnportancia Argueol6gica de Bolivar. Unpublished
"Trabajo de Habilitaci6n." Departamento Academico de
Ciencias Sociales. Universidad Nacional de Trujillo.
Trujillo, Peru.
Villabla 0., Marcelo
1988 Cotocollao: Una Aldea Formativa del Valle de Quito.
Miscelanea Antropol6gica Ecuatoriana. Serie
Monografica No. 2. Museos del Banco Central del
Peru. Quito.
Weberbauer, Augusto
1920 La Salida de Patas al Huallaga Estudiada en la Ruta
de Pajaten. Boletin de la Sociedad Geografica de
Lima Torno XXXVI, Trimestre prirnero, pp. 5-13. Lima.
1945
El Mundo Vegetal de los Andes Peruanos. New edition.
Ministerio de Agricultura. Lima.
673
Wegner, Steven A.
1981 Identifying Recuaydom: An Expanded Ceramic Inventory
for Identifying Recuay Settlement Sites. Revised
version of a paper given at the 46th Annual Meeting
of the Society for American Archaeology, San Diego.
Weissner, Polly
1990 Is there a Unity to Style? In The Uses of Stvle in
Archaeology. Edited by M. Conkey and C. Hastorf, pp.
105-112. Cambridge University Press. Cambridge.
Wheeler, Jane
1985 De la Chasse a l'Elevage. In Telarmachay: Chasseurs
et Pasteurs Prehistorigues des Andes. Torno 1. By D.
Lavallee, M. Julien, J. Wheeler and C. Karlin.
Institut Fran~ais d'Etudes Andines. Editions
Recherche sur las Civilisations No. 20, pp. 61-80.
Paris.
Wiener, Charles
1884 Amazone et Cordilleres. Le Tour du Mende. 48
(2) :385-416. Paris
Willey, Gordon R.
1945 Horizon Styles and Pottery Traditions in Peruvian
Archaeology. American Antiquity 11:49-56.
1971 An Introduction to American Archaeology, Vol. 2
South America. Prentice Hall Inc. Englewood Cliffs,
New Jersey.
Willey, Gordon R. and Jeremy A. Sabloff
1974 A History of American Archaeology. W. H. Freeman and
Co. San Francisco.
Wilson, David J.
1988 Prehispanic Settlement Patterns in the Lower Santa
Valley. Peru. Smithsonian Institution Press.
Washington, D.C.
Wilson and Reyes
1964 Geologia del Cuadrangulo de Pataz. Carta
Geologica Nacional Serie A, Boletin No. 9. Comision
de la Carta Geologica Nacional. Lima.
Wing, Elizabeth
1972 Utilization of Animal Resources in the Peruvian
Andes. Appendix IV in Andes 4: Excavations at
Kotosh. Peru. 1963 and 1966. By S. Izumi and
K. Terada. University of Tokyo Press. Tokyo.
674
1980
Faunal Remains. In Guitarrero Cave: Early Man in the
Andes. Edited by T. F. Lynch, pp. 149-172. Academic
Press. New York.
Wise, Karen Ruth
1976 Apuntes sabre la Influencia Inca entre los Arnuesha.
Factor que Oscurece la Clasificaci6n de su Idioma.
Revista del Museo Naciona1 Torno 42, pp. 355-366.
Lima.
1985
Indigenous Languages of Lowland Peru: History and
Current Status. In South American Indian Languages:
Retrospect and Prospect. Edited by H. E. Manelis
Klein and L. R. Stark, pp. 194-223. University of
Texas Press. Austin.
Wobst, Martin
1977 Stylistic Behavior and Information Exchange. In
For the Director: Research Essays in Honor of James
B. Griffin. Edited by C. E. Cleland, pp. 317-342.
Museum of Anthropology Anthropological Paper 61. Ann
Arbor.
Wright Jr., H. E., G. 0. Seltzer and B. C. S. Hansen
1989 Glacial and Climatic History of the Central Peruvian
Andes. National Geographic Research 5 (4) :439-445.
Young, Kenneth R.
1990 Biogeography and Ecology of a Timberline Forest in
North-central Peru. Unpublished Doctoral thesis,
University of Colorado, Boulder.
1991
Floristic Diversity on the Eastern Slopes of the
Peruvian Andes. Candollea 46:125-143.
1992a Biogeografia y Conservaci6n de los Basques Montanos
Tropicales. In Memoria del Decima Congreso Nacional
de Biologia. Edited by E. Castillo de Maruenda, pp.
57-62. Consejo Nacional, Colegio de Bi6logos del
Peru.
1992b Biogeography of the Montane Forest Zone of the
Eastern Slopes of Peru. In Memorias del Museo de
Historia Natural. Edited by K. R. Young and N.
Valencia, No. 21, pp. 119-140. Universidad Nacional
Mayor de San Marcos. Lima.
1993
National Park Protection in Relation to the
Ecological Zonation of a Neighboring Human
Community: An Example from Northern Peru. Mountain
Research and Development 13 (3) :267-280.
675
n.d.
Una Perspectiva Dinamica de la Vegetaci6n del Parque
Nacional Rio Abiseo.
In Memorias del Simposio
Biodiversidad. Historia Cultural y Futuro del Pargue
Nacional Rio Abiseo.
In Press. Asociaci6n Peruana
para la Conservaci6n de la Naturaleza.
Lima.
Young, Kenneth R. and Blanca Leon
1988 Vegetaci6n de la Zona Alta del Parque Nacional Rio
Abiseo, San Martin, Peru. Revista Forestal del Perli,
16:1. Lima.
1990
Catalogo de las Plantas de la Zona Alta del Parque
Nacional Rio Abiseo, Peril. Publicaciones del Museo
de Historia Natural (B) 34:1-37. Universidad
Nacional Mayor de San Marcos. Lima.
1993
Distribution and Conservation of Peru's Montane
Forests: Interactions between the Biota and Human
Society. In Tropical Montane Cloud Forests. Compiled
and edited by L. Hamilton, J. Juvik and F. Scatena.
East-West Center. Honolulu.
Young, K. R., W. B. Church, M. Leo and P. F. Moore
1994 Threats to Rio Abiseo National Park, Northern Peru.
Ambio 23 (4-5) :312-314. Stockholm.
Young, Kenneth R. and Neils Valencia
1992 Introducci6n: Los Bosques Montanos del Peril. In
Memorias del Museo de Historia Natural. Edited by K.
R. Young and N. Valencia, No. 21, pp. 5-10.
Universidad Nacional Mayor de San Marcos. Lima.
Zaki, Andrzej
1983 Cultura Pelon. Una Desconocida Cultura en la Sierra
Norte. Boletin de Lima 5 (29) :13-19. Lima.
Zevallos Q., Jorge
1982 Omnastica Prehispanica de Chachapoyas. Investigaci6n
Argueol6gica, No. 4, pp. 3-18. Universidad Nacional
de Trujillo. Trujillo, Peru.
1987
Introducci6n al Estudio Etnohist6rico de
Chachapoyas. Unpublished manuscript in possession of
the author.
Zeidler, James A.
1988 Feline Imagery, Stone Mortars, and Formative Period
Interaction Spheres in the Northern Andean Area.
Journal of Latin American Lore 14 (2) :243-283.
APPENDIX A
ILLUSTRATIONS AND ARTIFACT PROVENIENCE
676
677
SOUTH
AMERICA
PACIFIC
OCEAN
Scale
500km
Fig. 1. Map of the Central Andes, western Amazonia and the
Northern Andes (after Institute Geografico Nacional 1984).
678
! ECUADOR
COLOMBIA
SOUTH
r'/
AMERICA
I
10 o
PACIFIC
OCEAN
Legend
(jjJ Western Slope
•
Northern
0 Eastern Slope
Scale
500km
Fig. 2. Map of the Central Andean Montane Forests in Peru
(after Young· and Leon 1993: Fig. 1). The highland montane
forest is not shown.
679
SOUTH
AMERICA
PACIFIC
OCEAN
Scale
500km
Fig. 3. Map of some sites mentioned in the text.
680
Legend
•
0
Towns
Archaeological
~
Rivers
Scale
50km
f
Fig. 4. Map of the northeastern Peruvian Andes. Shaded area
is above 3,000 m {after Institute Geografico Nacional 1984
and NASA 1976).
00'
Legend
- ......... Study Area Boundaries
ffiTI Dry Forest Zone
0 Moist Montane Zone
D Tropical Alpine Zone
~ Montane Rain Forest Zone
Ill Premontane Forest Zone
Town
o Arcpaeolog!cal !;lite
Study Area Macro-ecological
Scale
25km
Fig. 5. Map of the Pataz-Abiseo study area and macroecological zones described in the text (after Young 1993:
_Fig. 3 and Young et al. 1994: Fig. 1).
0'1
00
f-'
Manachaqul Cave
(3625ml
4500 m
I
Cran Pojalen
12850ml
I
Tropical Alpine Zone
I
3500 m
,
Montane Rain Forest
Upper Montane
Forest
2500m
Lower Montane
Forest
Maraflon-H uallaga
Divide
1500m
4000m
I
3000m
Montane Rain
Forest Zone
2000m
Premontane Forest
lOOOm
500m
Okm
50km
IOOkm
Fig. 6. Profile of the Maranon-Huallaga divide showing
distribution of macro-ecological zones.
0"1
00
1:\.)
~
N
I
-
•
.,.___..__
Approx. :J km
1\taral'lon·Huallag;t
river dl\,dc
1\toncane Rain
forest Zone
lakes and rivers
I
~
Fig. 7. Map of the Manachaqui and Montecristo River valleys
and the pre-Hispanic roads.
"'w
00
684
D
\
,
\
\
\
I
I
\
\
\.
I
\
SECTOR
B
2m
\
.........
- Shelter interior
Fig. 8. Plan map of the Manachaqui Cave site complex (Site
M-1).
685
.--.29
_-·-·-·30
26
·-·-.
5
32
31
6
........
.. -....
27.
28
........
.......
18
20
21
4
17
3
22
./
/
\
24
,... 23
N
I
1 m
J
\
\
\
\
\
\
\
'\
''
- · - · - Area "cl.eaned"
Edge of roof
-··· ··· ....
Drip line
Fig. 9. Plan map with detail of Manachaqui Cave (Site M-lA)
interior. The principal interior space is shaded.
686
2m
· ____, __
',9..., ~--·-.
.. ,......
..' \ ' It
'\,.
Contour interval
=0.50 rn
\ .....' t,
'·\ ~
~...
~
l
'\, . . _:.:=
_....---.. ,
Shelter interior
Cl
Test pit No. 1
Fig. 10. Map of Manachaqui Cave excavation units 1988-1990.
687
Legend for Stratigraphy Illustrations
Charcoal or carbon-rich
lenses
Lenses of yellow-brown
fill
Gravel
ROCK
~-
Ul
U2
U3
UNEXCAVATED
~~-
~~!!'~'&'-
~~
Fig. 11. west profile of Units 1, 2 and 3.
m
(X)
(X)
us
U4
~a
<:::::>
<J
to.~
' - - - - - R-3
~ (]:!---~-----
U35
U6
~
....
0 ao o ~
----~~~
~
---~~-c::>
""'
2A/B c.
0
(/
c::::?\:7
c:;J
Doao
<10
o
b
()
Q
cJ
Oa::::?o"'c::::::J
6
~
°
1::::>
Fig. 12. West profile of Units 4, 5, 6 and 35. Features R-2 and
R-3 are shown.
0"1
(X)
lO
U38
U39
U7
us
c.
D ()
0
()
ROCK
Fig. 13. West profile of Units 38, 39, 7 and 8.
0'\
~
0
(
ROCK
U3
Ul
U2
(]
0
2A
2A/B
Fig. 14. East profile of Units 3, 2 and 1.
~
-0\
~
I-'
us
U6
U35
oJ)
U4
c:;.
0'\:>
~
Q
c:J
=o
-c::::>o
()
c()<)o_g-_1,_
0...
Q....
""
[J
...
--
0
r:::::.o
~
~~
C:P~
co
0[]
c:J
tV
~
2A/B o
<»
2A
~ODcv
a
2s6
Q
(l
CJ
Fig. 15. East profile of Units 35, 6, 5 and 4. Feature R-4 is
shown.
3A/B
0"1
1.0
1'0
U8
U7
U39
PEDISTAL
Fig. 16. East profile of Units 7, 8, 39 and 38.
"'w
\0
694
U37
U28
U32
2A/B
F'lg.
17. East profile of Units 37, 32 and 28.
Ull
Ul
Ul3
ROCK
Fig. 18. South profile of Units 11, 1, 13 and 12. Feature R-5 is
shown.
0"1
\.0
Ul
696
\
U12
\
U13
-;;;~~~~~c=~--~--~-
-
\
A
~~....--~15~;;;;;;;::;::::===~~zg--=2:A~~[:
w
X
y
z
AA
Fig. 19. North profile of units 12 and 13. Feature R-5 is
shown and sector A floors are indicated in right margin.
Ul9
U20
U21
~
Ul8
~:;!_
Qc::J
CJ
Q
c:::J
r::::,.(;> ~
~
<::r
eo
cO
c:::>
2A
c:=o
()
0
"V
-·o
lA/B
______
~
-<::>
~
t;>
~
p
0
()~
0
2A/B
0
0
'\:::>
-=-
0
Q
0
0
()
3A
~
~~~-
£2
Fig. 20. South profile of Units 21, 20, 19 and 18. Feature R-3
and Floor Z's Hearths 1 and 2 are shown.
0'\
lO
..,_]
U29
~l
I
(
c::>
~, '
'
I
FALLEN ROCK ·, c:::::2
Pre-1986 Surfia
, _____(ji' ______. :.,~~Cl
CJ
I
~
----
4
--~
--=-==--
<:::::..
0
C7
,;:>
~
0
!R t;>
0 qC),()
\)Dg'<l~
0
d
Cl
(;P
Qb'
=
U32
ce
:::::::::::::g2;:=:::.
•
R
8:;
U31
U6
U30
c
2A/B
C7
Fig. 21. North profile of Units 29, 30, 35, 31 and 32.
0"1
\.0
00
699
U5
U6
LEVEL
16
18
20
22
c:::::::::::::. 24
26
0
0
0
0
2A/B
0
CJ
Cl
~
D
CJ
28
32
34
36
38
40
Fig. 22. Schematic diagram illustrating relationships
between 1988 excavation levels and strata as seen in Unit 5
and Unit 6 east profiles.
700
U32
\
\
2A/B
17
19
30
~..---::...--~
~-=~~~~~~~32
\
\
\
Fig. 23. schematic diagram illustrating relationships
between 1990 excavation levels and strata as seen in Unit 31
and unit 32 north profiles.
701
od
3
I
·-·-·-·-·-·-,·-·
i
I
i
I
~~
i
14
2
--·-·-·~-·-
--- ----1
12
I
I
0.50m
-------
WALLS OF
DRAIN
~DIRECTION OF
FLOW
R-7
10
Fig. 24. Plan view of Feature R-5, Sector A.
702
Ul6
Ul7
......._
' ..;
,
' fl.. /
I
,
I
'\
I
'
ROCK
'' ,,'
,
I
ROCK
Ul4
Ul5
Ul6
Ul7
A
<JeD
D(}
C)
f3:J
()0
ROCK
ROCK
Ul4
Ul5 B
Fig. 25. Sector A floors in Units 14-17. a: Floor FF and
associated rock-filled hearth; b: Floor EE and associated
rock-filled hearth.
703
Ul6
Ul7
ROCK
ROCK
Ul4
Ul5 A
Ul7
Ul6
ROCK
ROCK
Ul4
Ul5 B
Fig. 26. Sector A floors in Units 14-17. a: Floor CC and
associated rock-filled hearth; b: Floor BB and associated
hearth.
704
A
UlO
U9
Fig. 27. Feature R-7 in rear of Sector A, Units 9 and 10. a:
north face of stone wall in Unit 10. b: east profile of
Feature R-7. Cave wall protrusion is shaded.
705
SECTORB
,,-..:;.-:.":.. .:::::-.::\
;I'
~......•
k
t!''
•••
N
SECTOR A
.,
•
...,_~ • • ~::-:
+ .. r:··
.......~
......
•. "••
~
······
)
lrn
Fig. 28. Plan view of Manachaqui Cave interior showing rock
wall features R-6, R-7 and R-8.
706
U16
U17
f
'
,
,,
,,- ......
''
\
,I
\.._ ,,-'HEARTH
ROCK
U14
t:J16
\)
U15 A
c=:>O U17
I
I
I
t?
I
I
I
\
,
HEARTH 1''----------'
0(}
0
0
/
,'
t
\
HEARTH2-
tJ
ROCK
U14
U15 B
Fig. 29. Sector A floors in Units 14-17. a: Floor AA and
associated hearth; b: Floor Z and associated Hearths 1 and
2.
707
U16
Ul7
,..,..,..,---- ....... ,,
/
'q
HEARTH\
{
I
I
\
\
U14
\
QI'-.J
I
I
I
r
I\.,
' ..... _.,,
'I
'./
U15
Fig. 30. Sector A Floor P and associated hearth in Units 1417.
708
UNIT 17 A
------..
0
' ___
r -"
I
I
I
I
..................
_,, ,.
/
I
1
I
I
I
UNIT 16
8
Fig. 31. Sector A floors containing hearths with embedded
stones. a: FloorS in Unit 17; b: Floor M? in Unit 16.
709
U28
U19 8
Fig. 32. Plan views of Sector Brock wall features. a:
Feature R-6 in Units 23 and 28 (Rockshelter exterior is
shaded). b: Feature R-8 in Units 18 and 19.
710
U31 A
A---.J:)_.......
- ...- .- .-. .-
;:r--F=>"<--)- A '
0
8
Fig. 33. Sector B Feature R-4 in Unit 31, Level 16. a: Plan
view of hearth. b: Profile of hearth.
711
Provenience of Illustrated Artifacts:
Unit#-Leve1# (rim diameters)
Fig.34a:
b:
c:
d:
e:
f:
g:
h:
i:
j:
k:
1:
m:
n:
o:
p:
q:
r:
s:
t:
Fig.35a:
b:
c:
d:
e:
f:
g:
h:
i:
j:
k:
1:
m:
n:
o:
p:
Fig.38a:
b:
c:
d:
e:
24-20
24-15
2-21
5-32
31-14
25-8
2-20
2-21
6-33
6-33
1-20
4-29
30-14
28-19
5-30
37-9
4-25
27-16
35-17
19-12
f:
g:
h:
i:
j:
K:
1:
m:
n:
o:
p:
34-10 (10)
31-20 (12)
1-24 (11)
28-16 (12)
34-11 (11)
27-17 (11)
30-20 (10)
2-21 (9)
1-24 (11)
32-19 (18)
32-18 (18)
35-16 (17)
39-7 (14)
28-17 (14)
29-12 (13)
36-14 (12)
26-18 (14)
29-14 (19)
36-12 (10)
5-33 (26)
34-11 (26)
f: 28-18 (24)
g: ?
(23)
h: 39-7 (14)
Fig.36a:
b:
c:
d:
e:
Fig.37a:
b:
c:
d:
e:
f:
g:
h:
i:
30-16
37-8
36-12
24-19
26-19
21-6
28-13
28-16
28-14
(14)
(12)
(14)
(11)
(13)
(14)
(25)
(24)
(16)
4-29 (24)
6-29 (30)
6-34 (25)
5-33 (23)
6-30 (22)
22-18 (14)
38-10 (13)
22-17 (12)
34-11 (12)
25-8 (14)
30-16 (13)
20-10 (14)
22-19 (14)
22-17 (14)
34-12 (13)
32-17 (12)
Fig.39a:
b:
c:
d:
e:
21-13 (14)
20-12 (13)
20-11 (13)
19-15 (12)
37-11 (11)
f: 24-15 (11)
g: 2-23 (13)
h: 13-9 (16)
i: 31-13 (16)
j: 30-18 (16)
k: 22-14 (14)
1: 31-16 (14)
m: 28-15 (13)
n: 31-14 (12)
o: 31-14 (12)
p: 22-16 (11)
q: 31-13 (16)
r: 22-16 (15)
s: 5-24 (14)
t: 3-21 (14)
u: 31-14 (14)
v: 30-15 (13)
w: 31-15 (12)
x: 16-04 (10)
y: 6-30 (14)
z: 1-22 (12)
aa: 13-9 (?)
Fig.40a:
b:
c:
d:
e:
f:
g:
h:
i:
j:
k:
1:
18-10 (12)
31-15 (12)
28-21 (19)
30-19 (16)
4-28 (14)
31-15 (14)
30-19 (13)
31-20 (12)
28-22 (12)
14-34 (17)
5-31 (14)
5-32 (13)
m:
n:
o:
p:
5-32
13-13
1-21
5-28
(12)
(11)
(11)
(10)
Fig.41a: 5-26 (12)
b: same
c: 5-32 (18)
d: 4-30 (16)
e: 30-17 (15)
f: 25-6 (14)
g: 5-32 (13)
h: 5-30 (13)
i: 5-33 (11)
j: 30-15 (16)
k: 30-16 (15)
1: 30-20 (14)
m: 12-9 (13)
n: 28-23 (11)
o: 31-14 (11)
p: 30-19 (11)
q: 5-31 (18)
r: 5-31 (18)
s: 31-17 (15)
t : 5-32 (14)
u: 16-45 (11)
v: 25-6 (ll)
w: 30-19 (10)
Fig.42a: 32-16 (19)
b: 2-18 (15)
c: 31-16 (15)
d: 28-18 (14)
e: 31-16 (14)
f: 31-12 (12)
g: 30-18 (10)
h: 22-17 (12)
i: 4-21 (20)
j: 6-30 (16)
k: 4-26 (13)
1: 4-27 (12)
m: 5-28 (11)
n: 31-16 (14)
o: 22-17 (11)
p: 30-16 (10)
Fig.43a: 29-17 (15)
b: 26-17 (16)
c: 29-17 (15)
Fig.44a:
b:
c:
d:
29-11 (12)
28-19 (13)
32-17 (17)
36-14 (17)
Fig.45a: 17-43 (11)
b: 28-20 (15)
712
c:
d:
e:
f:
g:
h:
i:
j:
k:
1:
m:
n:
o:
30-17 (14)
30-14 (12)
31-14 (11)
31-18 (11)
30-12 (?)
30-16 (13)
31-14 (13)
22-16 (13)
5-31 (13)
30-16 (12)
1-18 (14)
24-18 (12)
6-30 (12)
Fig.46a: 28-15 (18)
b: 31-17 (18)
c: 22-19 (14)
Fig.47a:
b:
c:
d:
e:
f:
30-15 (13)
28-19 (13)
1-23 (14)
28-21 (15)
5-29 (19)
28-14 (24)
Fig.48a:
b:
c:
d:
e:
31-14
30-17
37-12
28-17
30-15
k: 35-18 (13)
1: 30-20 (13)
m: 6-35 (12)
n: 20-13 (12)
o: 18-18 (13)
Fig.52a:
b:
c:
d:
e:
f:
g:
32-15 (11)
20-12 (11)
30-16 (12)
31-17 (12)
30-16 (14)
4-27 (28)
2-23 (12)
Fig.53a:
b:
c:
d:
e.
f:
g:
h:
i:
2-23 (?)
6-27 (26)
13-12 (23)
32-14 (18)
6-27 (15)
1-21 (13)
1-20 (13)
11-7 (11)
24-21 (18)
37-9 (18)
28-16 (14)
20-10 (13)
1-20 (15)
j:
k:
(16)
(14)
(14)
(16)
(15)
Fig.49a: 17-44 (16)
b: 5-31 (16)
c: 28-17 (17)
Fig.50a:
b:
c:
d:
e:
f:
g:
h:
i:
j:
k:
1:
m:
n:
o:
p:
q:
21-16 (18)
11-6 (17)
31-17 (17)
13-11 (15)
30-18 (14)
26-19 (12)
30-19 (12)
32-20 (11)
13-E2 (18)
28-23 (13)
19-13 (13)
19-14 (?)
25-9 (18)
29-15 (17)
25-9 (16)
1-20 (13)
31-17 (12)
4-28 (18)
26-19 (16)
12-10 (16)
32-17 (14)
18-18 (13)
5-33 (12)
36-15 (12)
5-30 (11)
22-15 (12)
j : 4-26 (18)
Fig.51a:
b:
c:
d:
e:
f:
g:
h:
i:
1:
m:
1-20 (15)
5-28 (14)
5-27 (13)
36-12 (12)
6-35 (19)
31-20 (18)
19-16 (14)
24-15 (18)
27-18 (17)
j: 36-9 (17)
k: 5-31 (10)
1: 34-12 (9)
Fig.54a:
b:
c:
d:
e:
f:
g:
h:
i:
Fig.55a:
b:
c:
d:
e:
f:
g:
h:
i:
26-21 (16)
16-45 (14)
3-26 (13)
28-14 (13)
31-17 (12)
28-22 (14)
22-14 (18)
32-14 (14)
14-42 (16)
j: 31-17 (14)
k: 38-10 (16)
Fig.56a: 25-L (18)
b: 16-44 (22)
c: 30-17 (18)
b: 26-15
C: 6-33
d: 6-34
e: 1-E
f: 2-23
g: 22-17
h: 6-33
i: 30-14
j: 4-30
k: 20-10
1: 31-12
m: 36-14
Fig.59a:
b:
c:
d:
29-18
19-12
15-26
5-30 and
27-20
e: 5-30
f: 22-15
g: 6-32
Fig.60a:
b:
c:
d:
e:
f:
g:
h:
i:
14-41
34-7
36-15
1-19
29-19
14-33
29-11
36-14
34-11.
j: 15-30
k: 24-13
Fig.61a:
b:
c:
d:
e:
f:
g:
22-17
37-9
35-15
22-16
1-23
23-17
4-29
Fig.62a: 29-15 and
35-16
b: 22-11
c: 27-16
d: 26-21
e: 30-18
f: 5-28
Fig.63a:
b:
c:
d:
22-16
24-19
37-11
30-16
Fig.57a: 27-16 and
32-17
b: 27-15
c: 24-14
Fig.64a: 21-12
b: 26-9
C: 6-28
d: 36-11
e: 31-17
f: 30-15
Fig.58a: 28-19
Fig.65a: 27-17 and
713
24-17 (13)
b: 36-12 (14)
c: 38-9,
36-13 and
6-30 (13)
Fig.66a: 28-23 (?)
b: 27-14,
22-16 and
22-14
c: 6-30 (9)
Fig.67a: 6-32,
34-10 and
30-16 (13)
b: 23-17
c: 6-24
f:
g:
h:
i:
j:
k:
1:
m:
n:
o:
p:
q:
r:
s:
t:
U:
v:
W:
Fig.68a: 1-19 (11)
b: 6-25 (10)
C: 5-23 (10)
d: 12-4 (13)
e: 30-10 (12)
f: 11-8 (10)
g: 36-8 (12)
h: 22-9 (12)
i: 25-5 (17)
j: 12-7 (15)
k: 31-11 (13)
1: 30-11 (12)
m: 32-8 (11)
n: 31-13 (10)
o: 16-2 (8)
p: 12-10 (7)
q: 28-14 (14)
r: 31-13 (10)
s: 4-22 (14)
Fig.69a: 6-25 and
31-13 (12)
b: 4-22 (13)
c: 21-8 (10)
d: 31-9 (14)
e: 31-9 (12)
f: 31-7 (12)
g: 4-15 (11)
h: 30-9 (10)
i: 24-6 (?)
j: 6-23 (8)
k: 28-6 (14)
1: 4-20 (15)
m: 4-21 (14)
n: 25-5 (13)
o: 4-23 (12)
p: 28-5 (11)
q: 28-6 (10)
r: 16-10 (7)
Fig.70a:
b:
c:
d:
e:
30-11
5-24
5-23
6-25
6-23
(14)
(13)
(12)
(11)
(10)
5-23 (10)
6-25 (10)
6-26 (8)
3-21 (11)
14-21 (12)
4-19 (11)
36-11 (12)
6-27 (16)
30-11 {14)
4-23 (13)
4-22 (13)
31-11 (12)
5-24 (10)
6-24 (10)
6-25 (14)
6-24 (12)
4-23 (11)
4-20 (10)
c:
d:
e:
f:
g:
h:
i:
j:
k:
1:
m:
n:
0:
p:
q:
r:
s:
t:
u:
Fig.71a:
b:
c:
d:
e:
f:
g:
h:
i:
j:
k:
1:
m:
n:
0:
p:
q:
r:
s:
t:
u:
31-11 (?)
30-9 (18)
30-10 (15)
6-26 (12)
30-10 (12)
22-8 (10)
6-26 (15)
31-12 (14)
31-13 (12)
31-12 (11)
31-12 (10)
13-9 (10)
24-11 (16)
28-7 (14)
4-17 (13)
4-17 (13)
31-11 (12)
28-11 (12)
4-23 (11)
22-13 (16)
4-22 (9)
Fig.72a: 22-11 (14)
b: 20-9 (14)
C: 28-14 (20)
d: 21-9 (15)
Fig.73a:
b:
c:
d:
e:
f:
g:
h:
i:
17-11 (16)
28-10 (16)
27-10 (14)
34-7 (16)
28-10 (18)
28-9 (10)
3-20 (?)
14-33 (13)
30-11 (12)
j: 22-11 (27)
k: 20-4 (12)
1: 11-8
Fig.74: 21-8
Fig.75a: 30-12 (14)
b: 4-20 (13)
v:
Fig.76a:
b:
c:
d:
e:
f:
g:
h:
i:
j:
k:
1:
m:
n:
Fig.77a:
b:
c:
d:
e:
f:
13-4 (12)
4-21 (12)
30-11 (12)
25-1 (11)
5-24 (7)
1-18 (11)
30-10 (16)
30-10 (?)
30-10 {12)
5-23 {10)
28-14 { 8)
26-8 (10)
6-26 (?)
30-10 (17)
24-12 (14)
6-24 (12)
15-32 (12)
1-16 (11)
31-12 (11)
26-12 (15)
13-8 (14)
1-18 (13)
36-7 (12)
31-11 (11)
36-8 (11)
24-8 (10)
36-7 ( 9)
6-25 (10)
22-13 (10)
28-12 (15)
6-24
24-11 (?)
22-11 (?)
4-20 (13}
24-9
22-15
24-12
36-12
28-12
11-10
(18)
(10)
(12)
(18)
(14)
(14)
5-23 (14)
30-10 (14)
13-9 (10)
4-22 (9)
31-14 ( 10)
35-9 (12)
29-9 (15)
30-11 (10)
29-11 (15)
j: 6-26 (13)
k: 27-11 (21)
Fig.78a:
b:
c:
d:
e:
f:
g:
h:
i:
Fig.79a:
b:
c:
d:
e:
f:
g:
h:
16-3
6-29
31-13
28-10
3-15
28-11
17-10
31-9
(15)
(15)
(12)
(14)
(13}
(12)
(12)
(12)
714
i: 20-7 (11)
j: 4-17 (9)
k: 30-11 and
31-12 (18)
1: 5-24 (18)
m: 30-10 (12)
n: 35-8 (?)
0: 28-12 (15)
p: 14-42 (14)
q: 4-21 (12)
r: 3-21 (11)
Fig.80a: 6-26 ( 2 0)
b: 31-9 (12)
c: 13-1 (16)
d: 5-26
e: 25-7
f: 32-10
g: 2-19
h: 22-11 and
28-12
Fig.81a: 37-7
b: 32-9
C: 39-6
d: 31-10
e: 29-9
f: 27-10
g: 6-23
h: 28-9
i: 27-12
Fig.82a:
b:
c:
d:
e:
f:
g:
h:
i:
j:
38-6
32-9
5-24
1-19
4-22 and
6-24
30-10
24-12
22-10
22-14
21-7
Fig.83a: 4-23
b: 6-25
C: 30-11
d: 35-9
e: 4-19
f: 4-24
g: 31-7
h: 28-11
i: 22-11
j: 20-8
k: 35-10
1: 30-13
Fig.84a:
b:
c:
d:
e:
f:
4-20
22-9
6-27
5-24
22-15
36-10
g: 22-12
h: 22-13
i: 28-12
j: 34-6
k: 4-20
1: 19-9
Fig.85a:
b:
c:
d:
e:
f:
g:
4-21
3-22
4-23
17-11
20-9
11-5
22-11
h: 24-12
30-12
11-6
5-24
28-12
6-27
27-11
6-17
6-25
4-22
j: 19-11
k: 27-12
1: 2-24
Fig.86a:
b:
c:
d:
e:
f:
g:
h:
i:
Fig.87a:
b:
c:
d:
e:
f.
g:
h:
i:
j:
30-11
26-9
24-13
26-10
6-25,
27-11,
35-10 and
3-17
4-22
28-13
22-11
4-22
21-8
Fig.88a:
b:
c:
d:
e:
22-12
25-8
17-11
23-13
21-7 and
6-25
f: 6-25
g: 25-5
h: 1-22
Fig.89a: 30-10
b: 19-13
C: 20-10
d: 36-9
e: 30-10
Fig.90a: 35-10
b: 30-10
C: 30-12
d: 30-10
e:
f:
g:
h:
26-10
6-26
30-12
11-8
Fig.91a:
b:
c:
d:
e:
f:
g:
h:
i:
21-7
5-20
6-22
12-10
32-6
32-8
6-23
4-19
6-21
32-3
3-20
31-9
28-9
13-8
3-15
26-5
19-10
2-20
32-8
20-4
13-4
5-20
j:
k:
1:
m:
n:
o:
p:
q:
r:
s:
t:
u:
V':
Fig.92a:
b:
c:
d:
e:
f:
g:
h:
i:
j:
k:
1:
m:
n:
o:
p:
31-9
30-9
30-9
1-19
28-9
28-9
17-12
29-6
28-10
29-9
20-7
13-9
14-17
18-7
37-5
6-22
Fig.93a: 36-8
b: 36-8
c: 2-20
d: 28-7
e: 5-22
f: 2-21
g: 1-17,
and 2-19
Fig.94a:
b:
c:
d:
e:
f:
g:
1-14
12-7
2-19
2-19
13-9
13-10
3-22
(?)
(16)
(12)
(10)
(10)
(12)
(10)
(?)
(16)
(?)
(19)
(14)
(13)
(12)
(?)
(14)
(16)
(13)
(12)
(?)
(18?)
(12)
(16)
(14)
(14)
(17)
(16)
(15)
(15)
(14)
(14)
(16)
(17)
(17)
(11)
(?)
(19)
(15)
(?)
(20)
2-18
(20)
(?)
(18)
(13)
(13)
(13)
(16)
(17)
Fig.95a: 29-2 (16)
715
b: 15-32 (15)
c: 13-5 (13)
d: 24-3 (13)
e: 3-19 (14)
f: 3-16 (10)
g: 30-7 (20)
h: 27-4 (18)
i: 6-22 (16)
j: 28-9 (14)
k: 4-20 (14)
1: 13-8 (16)
m: 2-18 (13)
n: 23-8 (14)
o: 3-13 and
3-14 (14)
p: 2-22 (12)
q: 24-3 (10)
r: 25-1 (9)
s: 14-30 ( 6)
t:2-16 (?)
u: 1-14 (17)
V: 3-15 (14)
w: 5-20 (12)
Fig.96a:
b:
c:
d:
e:
f:
g:
h:
i:
j:
k:
1:
m:
n:
o:
p:
q:
r:
S:
14-21 (18)
32-6 (16)
5-20 (16)
16-45 (15)
13-12 (23)
1-19 (17)
31-8 (13)
25-6 (11)
23-3 (20)
17-10 (12)
4-18 ( 8)
28-8 (14)
28-7 (14)
32-6 (10)
17-5 (3 0)
6-22 (20)
12-5 (13)
6-21 (?)
19-4 (?)
Fig.97a: 5-18
b: 30-3
c: 17-6
d: 35-6
e: 32-13
f: 35-5
g: 30-6
h: 30-4
i: 23-3
j: 31-3
k: 24-6
1: 18-5
m: 32-7
(22)
(20)
(18)
(15)
(14)
(13)
(10)
(10)
(?)
(11)
(?)
(?)
(10)
Fig.98a:
b:
c:
d:
e:
(19)
(19)
(17)
(16)
(15)
16-39
1-20
16-4
27-8
3-20
f: 37-5 and
24-6 (14)
g: 15-3 (14)
h: 31-6 (11)
i: 19-4 and
23-7 (24)
j: 30-7 (16)
k: 12-4 (13)
1: 2-21 (14)
m: 28-7 (13)
n: 5-21 (11)
o: 6-20 (15)
p: 21-4 (20)
q: 3-14 (14)
r: 3-21 (11)
s: 15-10 (19)
t: 34-4 (17?)
u: 11-5 (12)
v: 13-4 (?)
Fig.99a: 34-4 and
18-6 (16)
b: 31-8 (15)
c: 28-9 (14)
d: 28-7 (13)
e: 1-13 (25)
f: 5-19 (11)
g: 1-14 (10)
Fig.100a: 27-6
b: 38-3
c: 5-22
d: 1-14
e: 14-18
f: 29-6 (30)
g: 28-6 (15)
Fig.101a: 2-16 (14)
b: 21-6
c: 15-20
d: 12-7 (16)
e: 1-18 (?)
f: 13-5
Fig.102a: 11-8 (20)
b: 29-7 (13)
c: 2-18 (14)
d: 4-17
e: 1-23
f: 12-13 (18)
Fig.103a: 18-4 (?)
b: 11-8 (18)
c: 17-7 (14)
d: 6-20 (?)
e: 6-20 ( ?)
f: 5-21 (14)
g: 20-3 (14)
h: 17-7 (14)
i: 22-5 (20)
j: 26-4 (?)
k: 18-L (?)
1: 5-19 (14)
m:
n:
o:
p:
q:
r:
S:
t:
u:
v:
w:
X:
5-19 (14)
15-6 (10)
35-7 (8)
17-7 (14)
21-6 (20)
15-11 (18)
16-1 (16)
n/p (14)
1-19 (14)
3-17 (14)
1-19 (12)
2-20 (15)
Fig.104a: 2-21 (14)
b: 2-18 (16)
c: 16-10 (14)
d: 3-22 (12)
e: 3-19 (18)
Fig.105a: 21-7
b: 2-18
C: 12-5
d: 5-20
e: 5-20
(10)
(14)
(22)
(14)
(14)
Fig.106a: 16-M3
b: 2-20
C: 4-18
d: 21-5
e: 1-20
f: 2-16 (6)
g: 20-3
h: 6-21 (24)
i: 20-3 (14)
j: 6-21 (14)
k: 39-3 (10)
1: 18-6 ( 7)
m: 17-6 (14)
n: 11-7 (12)
Fig.107a: 6-20 (18)
b: 2-21 (10)
C: 15-29 (16)
d: 3-18 (18)
e: 15-15
Fig.108a: 11-2 (14)
b: 1-18
c: 20-3 (16)
d: 17-7 (14)
e: 2-16 (12)
f: 20-3 (14)
g: 15-12 (18)
Fig.109a: 12-13
b: 15-8
c: 2-18
d: 4-17 and
4-18 (12)
e: 31-9 (16)
f: 21-6 (12)
g: 26-7 (14)
h: 30-7 (14)
716
Fig.l.l.Oa: l.l.-l.
b: 12-7
c: 34-7 (14)
d: 10-2
717
Legend for Ceramic Illustrations*
Red paint or slip
D
Orange paint or slip
Brown paint or slip
* except for Figs. 103x-106e and 109d, e; see captions.
718
e
a
c
b
i
f
\\g
h
d
~)
1
j
)
q
m
f T
r
s
Fig. 34. Manachaqui Phase Paste A body profiles. a-h:
globular, semi-carinated and carinated profiles. i-p: same,
with applique medial ribs. q-t: rare "stepped" shoulder
profiles with shoulder ribs.
719
, ,"
''
/,
/
'
/
''
/
'\
/
/
'\
I
\
\
I
\
I
\
I
I
I
\
\
I
I
I
I
I
I
I
I
I
I
I
\
\
\
I
I
I
\
'
\
I
I
\
/
\
\
I
'
' ' ' ......
/
/
' ' .... ,
.......
/
/
/
;'
- ....
;'
... ___
.....
__
----- ---
--
..... ...-"'
/
i
,,
0
m
p
n
Fig. 35. Manachaqui Phase Paste A, Shape A rim profiles. ah: Rim 1. i: hypothetical Rim 1 jar. j-p: Rim 2.
720
/~I
····
.
f'·: ·. . .
~·.
.
.. .. ..
.
.
'
a
)
b
-, -,
d
g
e
c
f
,
,
'II
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
\
\
I
\
I
\
I
\
\
''
/
''
I
I
/
' ' ''",
'~
... '-,
............
____ ....
---
_.. .
,,/
,, .... ""
/
-==--==5 em
h
Fig. 36. Manachaqui Phase Paste A, Shape A rim profiles. a,
b, g: Rim 2 decorated rims. c-f: Rim 2 vessels with large
orifices. h: hypothetical Rim 2 jar.
721
\
)
'
)
a
b
'
c
~--, ,
._J
•
I
~
f
I
e
f
5 em
g
i
Fig. 37. Manachaqui Phase Paste A, Shape A rim profiles. ac: Rim 3. d-f: Rim 4. g-i: Rim 5.
722
e
b
f
g
h
i
.
•
\
I
I
I
\
I
\
•
/ J
'\ \
/
''
/
/
''
/
/
' ' .....
/
/
' ..... .... ....
.... ........ ....
_
-- --- -------
------- ---
----
..,""
~~,...
/
-I
k
5 em
Fig. 38. Manachaqui Phase Paste A, Shape A rim profiles. ae: Rim 6. f-i: Rim 7. j: hypothetical Rim 7 jar. k-m: Rim 8.
n-p: Rim 9.
723
a
g
h
p
X
u
t
v
q
5 em
y
z
w
aa
Fig. 39. Manachaqui Phase Paste A and Brim profiles. a-e:
Shape A Rim 10. f: Shape A Rim 11. g: Shape A Rim 12. h:
Shape A Rim 13. i-p: Shape BRim la. q-r: Shape BRim lb. yaa: Shape B Rim lc.
724
a
I
\ b
I
I
\
(
\
(
\
I
'
I
I
I
(
I
\
I
I
\
\
I
I
\
''
I
I
I
'' '
/
/
''
/
' ..... .....
.................................
-- --- -- ---
_..... .....
/
/
/
//
/'
i
1
5 em
Fig. 40. Manachaqui Phase Paste A, Shape B rim profiles. a:
Rim la with handle. b: hypothetical Rim lb jar. c-p: Rim 2a.
725
b (top view)
5 em
Fig. 41. Manachaqui Phase Paste A, Shape B rim profiles. a,
b: Rim 2a profile and top view. c-i: Rim 2b. j-p: Rim 3a. qw: Rim 3b.
726
'~''''r
a
c
d
e
f
g
h
r
i
m
j
~
p
0
n
5 em
Fig. 42. Manachaqui Phase Paste A, Shape Brim profiles. ah: Rim 4. i-p: Rim 5.
727
a
b
/
,
/
/
I
I
~
5 em
----
--- --
Fig. 43. Manachaqui Phase Paste A, Shape Brim profiles. a,
b: Rim 5 with applique decoration. c: hypothetical Rim 5
jar.
728
'
a
c
d
5 em
Fig. 44. Manachaqui Phase Paste A, Shape B rim profiles. a:
Rim 5 with rare notched decoration. b-d: Rim 5 with incision
and applique decoration.
729
l (. ~
a
b
~~ lf '
, ,
h
i
g
r
j
1
k
m
n
o
Fig. 45. Manachaqui Phase Paste A, Shape B rim profiles. ag: Rim 6. h-1: Rim 7. m-o: Rim 8.
730
)
' \.
\.
'
/
\.
/
a
/
''
/
''
/
/
./
' .... ... ...
.... ......... ....
__
....
....
_
---- ---
------ --
_-
.,.......--"'
,./
/
./
/
,./
......
5 em
Fig. 46. Manachaqui Phase Paste A, Shape Brim profiles. a:
Rim 8 partially reconstructed jar. b, c: Rim 9.
731
1"
a
I
b
c
{
d
Ej
I
.,;:.,-.·'/;· ..
···.
e
(
VI
.., ...:-.
·
.
f
5 em
Fig. 47. Manachaqui Phase Paste A, Shape BRim 9 profiles.
732
,
\0
f
a
b
I
I
I
c
/
\
'
/
'' '
/
''
' ' ,,
/
/
/
/
/
,,-.......
.............
__
_
/
.,/ /
__, ~~
~----
5 em
Fig. 48. Manachaqui Phase Paste A, Shape Brim profiles. a,
b: Rim 9. c: Rim 9 partially reconstructed vessel. d, e:
Rim 10.
733
""',
" ' ...
' ...
./a
,., "'"'
' ...... ......
/'
,.
,.-""
..........
--- --- --- ------ --------------., ....
,'"'
b
5 em
Fig. 49. Manachaqui Phase Paste A, Shape Brim profiles. a:
Rim 10 partially reconstructed jar. b, c: Rim 10 profiles
with top views.
-,- -,c-, -,~\X\h
a
d
734
g
b
i
! ([I I
ILJI
..:
.. ·.;
~
'
. '
'
l
..
j
k
n
q
I
I
I
I
I
I
r
I
(i
I
\
I
\
f
I
/
\
''
I
' '- ' .... _
'
•-==--==-·
. . ----5 em
/
------
/
/
.... ·-""'
Fig. 50. Manachaqui Phase Paste A, Shape C rim profiles. ak: Rim la. 1-p: Rim lb. q: hypothetical Rim la bowl.
735
h
c
i
-, -\k_\_,
.
J
IQ
1
I
m
\
n
\ 0
/
/
\
I
'\
I
'\
\
I
\
I
\
I
\
I
I
I
I
q
p
\
I
\
I
\
/
''
I
/
' ' ',
/
/
'
..........
............
---- -----
----~--r-----
---- ---
...-...,..,..,,
......
//
5 em
Fig. 51. Manachaqui Phase Paste A, Shape C rim profiles. ai: Rim 2a. j-n: Rim 2b. o: hypothetical Rim 2a bowl.
736
1.
(
(
c
/
5 em
'
d
-,
f
Fig. 52. Manachaqui Phase Paste A, Shape C rim profiles. a,
b: Rim 3. c-e: Rim 4. f: Rim 5. g: Rim 6.
737
c
k
1
i
j
I
\ m
f
I
I
I
I
I
I
(
\
I
''
I
)
/
''
/
__ __ _
''
'-.. ..... _
.......
-----
------- ---
--
/
/
/
.//
5 em
Fig. 53. Manachaqui Phase Paste A, ShapeD rim profiles. ah: Rim 1. i-1: Rim 2. m: hypothetical Rim 3 bowl.
738
I
c
b
d
e
f
g
a
h
i
k
1
5 em
Fig. 54. Manachaqui Phase Paste A, Shape D rim profiles. ad: Rim 3. e-g: Rim 4. h-j: Rim 5. k, 1: Rim 6.
739
c
b
a
-,
e
~I
1
h
g
j
5 em
Fig. 55. Manachaqui Phase Paste A, Shape E rim profiles. a,
b: Rim 1. c, d: Rim 2. e-h: Rim 3. i, j: Rim 4. k: Rim 5.
740
a
5 em
Fig. 56. Manachaqui Phase Paste A, Shape E rim profiles. a:
Rim 5. b: Rim 6. c: Rim 7.
741
I
I
•
I
I
C.
c
5 em
Fig. 57. Manachaqui Phase Paste A miscellaneous shapes. a:
spoon fragment. b, c: fragments of unidentified artifacts.
742
a
c
b
'\?~
e
f
g
/\-\
~
).
~
h
~
.
.
\
I>
k
j
-)
<:;;;-\
.
!
m
1
5 em
Fig. 58. Manachaqui Phase Paste A applique rib decorations.
a-d: unembellished medial ribs. e-g: unembellished shoulder
ribs. h-m: Notched A medial ribs.
743
~
d
·,
f
g
5 em
Fig. 59. Manachaqui Phase Paste A applique rib decorations.
a, b: Notched A medial ribs. c: Notched B shoulder rib. d:
Notched B medial rib. e, g: Incised A medial ribs. f:
Incised A shoulder rib.
744
c
~~/
~ Ye
d
h
\
i
j
5 em
Fig. 60. Manachaqui Phase Paste A applique rib and band
decorations. a: Incised A medial rib. b-e: Incised B medial
and shoulder flanges. f-k: notched bands.
745
~
c
e
5 em
Fig. 61. Manachaqui Phase Paste A applique band decorations.
a, b: notched bands with Incised A ribs. c, d: bands with
circular punctations. e, f: bands with ovoid incisions. g:
applique serpent head.
746
a
c
I
~=~ ~})
d
e
J
5 em
r
f
Fig. 62. Manachaqui Phase Paste A assorted applique
decorations. a: Incised A rib/band. b: lug with notched
medial rib and notched band. c: incised button on
unembellished rib. d: incised button on Incised A medial
rib. e, f: snake adornos on unembellished ribs.
747
a
.·
c
-13
d
5 em
Fig. 63. Manachaqui Phase Paste A adornos. a: head of
amphibean or fish. b: bird wing or fish fin. c: zoomorphic
heads.
748
a
d)-'
b
.
.
-\
c
d
~
f
5 em
e
Fig. 64. Manachaqui Phase Paste A incised decoration. a-c:
Incised Vessel 1 with zoned punctation. d, e: Incised Vessel
2 with zoned punctation. f: incised sherd.
749
b
c
5 em
Fig. 65. Manachaqui Phase Paste B1 vessels. a: Vessel 1, b:
Vessel 2, c: Vessel 3.
750
-,.
,.....----"
\
I
I
II
I
'I
I
I
I
b
\.
\.
5 em
Fig. 66. Manachaqui Phase Paste Group B. a: Paste B1 , Vessel
4. b: Paste B1 Vessel 5. c: Paste B2 vessel.
751
I
c
5 em
Fig. 67. Manachaqui and Suitacocha Phase artifacts. a:
Manachaqui Phase Paste B3 vessel. b: Manachaqui Phase
incised slate disk. c: Suitacocha Phase sherd with net
impressions on interior surface.
752
-,a-,~,c
-,
-,
h
i
g
f
1
j
k
f
-,
-( -(
m
1
n
r
p
0
q
r
~:c
-· . . :.--::·--:·.>
-~- ~ ~·.: ~~·~-,
r._.'.::.._. .:· ·..;. .:./·:_:;·.-..-.. ~~~ . :_~·;~.~ \
''
,'
s
5 em
Fig. 68. Suitacocha Phase Paste A, Shape A and B rim
profiles. a-c: Shape A Rim 14. d-f: Shape A rim 15. g, h:
Shape A Rim 16. i: Shape A Rim 17. j-s: Shape BRim 11a.
753
Jr:::JI
~b
) tid
t
-l -, -, -, -, 1.
··.· ...... .
.
;·. ~-
... :· .·.
I
c
l
d
f
e
h
g
i
J
?~
~1
-\
-t
p
n
r
0
q
m
5 em
Fig. 69. Suitacocha Phase Paste A, Shape B rim profiles. ac: Rim lla. d-k: Rim llb. 1-r: Rim llc.
754
-,
-, -t
-~
a
b
c
-,
d
~--::-~~.
g
:~
---: .,_ ']
. :' ·-·- .-_
·._:!
-, -,
h
j
k
i
1
0
-, -,
q
v
s
w
5 em
Fig. 70. Suitacocha Phase Paste A, Shape B rim profiles. ah: Rim lld. i-1: Rim lle. rn-s: Rim llf. t-w: Rim llg.
755
-(
l r
-~
a
b
d
e
c
-,
f
k
-,m -t -,
1
-~
0
-,q -r
-~
n
t
u
5 em
Fig. 71. Suitacocha Phase Paste A, Shapes B and C rim
profiles. a-f: Shape BRim 12. g-1: Shape BRim 13. m-u:
Shape C Rim 7.
756
a
b
c
d
5 em
Fig. 72. Suitacocha Phase Paste A, Shape C rim profiles. a:
Rim 7. b-d: Rim 8.
757
-,
a
b
c
d
-, -, -, -, -,
i
f
g
j
h
e
(
81
1
5 em
Fig. 73. Suitacocha Phase Paste A, Shape C and E rim
profiles. a-c: Shape C Rim 9. d: Shape C Rim 10. e-h: Shape
E Rim 8. i: Shape E Rim 9. j: Shape E Rim 10. k, 1: Shape E
Rim 11.
758
Fig. 74. Suitacocha Phase Paste A partially reconstructed
Shape F Rim 1 vessel.
759
-, -,
a
-, -, -, -,
b
d
f
e
g
c
h
-,
7
j
m
i
-, -, -, 1
n
r
s
t
u
q
v
5 em
Fig. 75. Suitacocha Phase Paste A, Shape F rim profiles. an: Rim la. o-v: Rim lb.
760
-1
d
c
f
e
g
h
J ltiT]
\ tj;j;j
I
l.
~
r
I
j
J e~?J
;;:;.~----~-
;.·~~~·.::·.:-.
lk
.· r
: .. ---
-:~.;;._~
1
----------
5 em
•••
• ••
.
1
m
r
.n
Fig. 76. Suitacocha Phase Paste A, Shape F rim profiles. ak: Rim 2. 1-n: Rim 3.
761
a
r
b
r
! WI
c
r
\~]1
)
.....·.
..
I
d
r
f
5 em
Fig. 77. Suitacocha Phase Paste A, Shape F Rim 3 profiles.
762
a
b
e
f
r r
1
i
h
J
5 em
Fig. 78. Suitacocha Phase Paste A, Shape F rim profiles. a:
Rim 3. b-d: Rim 4a. e, f: Rim 4b. g, h: Rim 5. i, j: Rim 6.
k: Rim 7.
763
-,
c
b
-, -,
a
e
d
j
k
-,
m
n
1
5 em
-,
-, 11
r
0
p
q
Fig. 79. Suitacocha Phase Paste A, Shape F and X rim
profiles. a-c: Shape F Rim 8. d-j: Shape X Rim 1. k-n: Shape
X Rim 2. o-r: Shape X Rim 3.
764
a
b
c
5 em
Fig. 80. Suitacocha Phase Paste A rim profiles and basal
sherds. a: Shape X Rim 4. b: Shape X Rim 5. c: Shape X Rim
6. d: jar or bowl base. e-h: basal angles from carinated
bowls.
765
- L)..
.
~
c
b
a
~.
I
f
e
h
g
5 em
Fig. 81. Suitacocha Phase Paste A miscellaneous shapes. a-c:
fragments of mammiform vessel legs. d-h: handle fragments?
i: unidentified fragment.
766
b
a
c
d
I
Q~
I
f
g
e
I
i
h
5 em
Fig. 82.
punctate
notched.
j: round
Suitacocha Phase Paste A sherds with notched and
rib or band decoration. a-d: notched. e: parallel
f, g: high relief notched. h, i: ovoid punctation.
punctation.
j
767
I
I
@L
1
~
d
I
I
.
.
.
__
:.
I
c
~
t!l~
I
fifijW ~e
f
I
I
@I g\
.
.
. . ~--.~-:-
b
a
e( '
·.
g
I
h
i
'
1
5 em
Fig. 83. Suitacocha Phase Paste A sherds with applique and
incised decoration. a: round punctation. b, c: unembellished
applique. d-f: flanges. g-i: incised lines. j: divergent
arrays of parallel lines. k, 1: cross-hatching.
768
I
-
.
<fJ···~. J
l
I
o ...
I
.
r::;;;:::;;;;;
c
b
~~.) ~1··1
I
I
I
a
~
.
.
.
·.·:
..
•.· ''.0:::<~:.-••
d
I
'd!.
e
·
.
.
.
·<:. <) ....
I
f
I
~,
~
I
g
i
5 em
Fig. 84. Suitacocha Phase Paste A decorated sherds with
stamped circles. a: in row. b, h: with paint. c-f, i-1:
with incision. j, k: with punctation.
769
a,
I
,..,.
Q'
.
-
~
I
.
I
a
c
I
~l
~
I
f
e
d
h
5 em
Fig. 85. Suitacocha Phase Paste A decorated sherds. a-f:
stamped circles and applique with round punctations. g, h:
punctation and incision.
770
a
b
~
~e
c
d
I
I
Af-2l €91
~I
I
At
~
.
I
g
i
h
f
I
Q\
j
I
I
k
5 em
Fig. 86. Suitacocha Phase Paste A decorated sherds with
punctation and incision.
1
771
~1.
c
\
.
,
.
'
.
a\,
I
I
.
·. ··.
..
'
I
f
I
e
.
.
g
.
'
h
I
~~~
~
I
5 em
i
j
Fig. 87. Suitacocha Phase Paste A sherds with punctation. ah: with punctation and incision. i: punctation and applique
with round punctation. j: punctation, incision, notched
applique and red paint.
772
a
b
I
1)DJ
c
d
I
g
•I
I
h
5 em
Fig. 88. Suitacocha Phase Paste A assorted decorated sherds.
a, b: punctation, incision and notched applique. c, d:
combing. e: brushing. f: rouletting. g: incised boss. h:
fine-line scratched.
773
a
c
e
5 em
Fig. 89. Suitacocha Phase Paste A assorted sherds. a, b: mat
and/or fabric impressed. c: zoomorphic adorno depicting
parrot head. d: zoomorphic adorno (opposurn head?). e:
zoomorphic adorno (probably a bat's head and face).
774
a
b
c
e
\~
·I
. ..
.
.
..
~
I
h
5 em
Fig. 90. Suitacocha Phase Paste A adorno and Group B sherds.
a: anthropomorphic head adorno. b: Paste B4 • c, d: Paste B4
jar rims. e-g: Paste B4 sherds with punctation and incision.
h: Paste B5 sherd from bowl.
776
7 7 7
a
1
\
c
b
~ 1~
~-
!19.·.·.···..
·::i:·. . .
....·:.·.
7~
h
I
j
7
1
1
m
P·
n
5 em
Fig. 92. Colpar Phase Paste A, Shape E rim profiles and
decorated sherds. a: Rim 3, b, c: Rim 9. d-j: Rim 12. k, 1:
Rim 13. m: Rim 14. n-p: notched applique.
777
e
5 em
Fig. 93. Colpar Phase Paste Group B. a, b: Paste B6 rim and
basal angle profiles. c-e: Paste B, rim profiles. f: Paste B,
rim with iridescent red paint. g: Paste B8 rim with whiteon-red decoration.
778
UTUI
d
e
f
(
g
5 em
Fig. 94. Colpar Phase Paste Groups B and C. a, b: Paste B8
rim profiles. c-e: Paste B9 rims with red paint. f: Paste B10
rim with orange-red paint. g: Paste C1 rim with red paint.
779
-,
7
a
-J
b
7.
f
d
c
j
1
n
s
q
0
5 em
J
p
-I
1
t
v
w
u
Fig. 95. Empedrada Phase Paste A Shape E rim profiles. a-c:
Rim 3. d-k: Rim 9. 1-s: Rim 15. t-w: Rim 16.
780
7
e
1/
f
g
])
m
}1
p
r
q
n
1
s
0
5 em
Fig. 96: Empedrada Phase Paste A Shape E rim profiles. a-d:
Rim 17. e-h: Rim 18. i-k: Rim 19. 1-n: Rim 20. o: Rim 21. p,
q: Rim 22. r: Rim 23. s: Rim 24.
781
,,,,
e
f
g
h
j
• •• • •
•
•
7
k
r
1
5 em
Fig. 97. Empedrada Phase Paste A Shape GRim profiles. a-m:
Rim 1.
782
~,
l/1
k
1
m
n
i
I!' 17
s
t
q
v
Fig. 98. Empedrada Phase Paste A Shape G rim profiles. a-h:
Rim 2. i-k: Rim 3. 1-n: Rim 4. o: Rim 5. p, q: Rim 6. r: Rim
7. s, t: Rim 8. u: Rim 9. v: Rim 11.
783
a
f
I
g
5 em
Fig. 99. Empedrada Phase Paste A Shape H rim profiles. a-d:
Rim 1. e: Rim 2. f: Rim 3. g: Rim 4.
784
e~
•
a
I
'
c
b
d
l
g
e
5 em
Fig. 100. Empedrada Phase artifacts. a, b: Paste A spoon
fragments. c: unidentified artifact. d-e: notched applique.
f: Paste B11 jar rim profile. g: Paste B12 rim profile.
785
c
b
5 em
Fig. 101: Empedrada Phase Paste Group B. a-c: Paste B13 bowl
with red paint. d-f: Paste Bu sherds from bowls with red and
brown paint.
786
c
d
e
5 em
Fig. 102. Empedrada Phase Paste Group B. a, b: Paste B15 bowl
with dark red paint, pale orange slip on exterior. c: Paste
B15 : bowl with iridescent red paint. d, e: Paste B17 sherds
with orange and white paint on brown slip. f: Paste B18 bowl
with orange slip and negative-resist decoration.
787
-, ~I
f
e
i
g
n
7
w
s
X
5 em
Fig. 103. Empedrada Phase Paste Group C2 • a-p: unpainted. qw: traces of red paint. x: Paste C2 a with red paint.
788
a
J
b
I
c
}
\lJ
d
\~
J
-----
1
e
5 em
Fig. 104. Empedrada Phase Paste Group C2 • a-e: Paste C2 a with
red paint.
789
I
\
i
a
I
[))
b
d
5 em
Fig. 105. Empedrada Phase Paste Group C2 • a-e: Paste C2 a with
red paint.
790
Lli[J_
b
c
0
~
-
-,
j
~-<
2
)
1 -,
g
f
e
d
~
k
1
m
n
5 em
Fig. 106. Empedrada Phase Paste Group Cl. a-e: Paste Cla with
red paint. f-n: Paste Cla with orange slip.
791
J
b
(
~I
)
c
\~I
J
d
5 em
e
Fig. 107. Empedrada Phase Paste C2 • a-e: Paste C2b with black
paint.
792
1.
--b
l
<
---
( \Ji
I
( Fti7
I
d
J.
J
f
g
5 em
Fig. 108. Empedrada Phase Paste C2 • a-d: Paste C2b with black
paint. e-g: Paste C2 c with red and black paint.
793
I~~
a
I
b
r~1
I
VI
I
c
)f
I
\
(
g
IS j
h
5 em
Fig. 109. Ernpedrada Phase Paste C2 • a-c: Paste C2 C with red
and black paint. d-f: Paste C2 d with tan paint. g, h: Paste
C2 e with black paint on orange slip.
794
a
c
5 em
d
Fig. 110. Empedrada Phase Pastes C2 • a: Paste C2 f with brown
slip (bottom), negative-resist smudging (top} and white
painted circles on natural tan paste (middle) . b: Paste C2g
with red paint on light brown paste. c: Paste C2h unpainted
jar rim. d: notched ground slate disk from Unit 10 Level 2.
795
PRE-LAVASEN PHASE (N=102)
Q.
(/)
z
50
45
40
35
30
25
20
15
10
5
0
SMALL
SMALL·
MED
MEDIUM
MEDLARGE
LARGE
Animal Size
Fig. 111. Pre-Lavasen Phase animal size distribution
(NISP=Number of Individual Specimens).
LAVAS EN PHASE (N=21}
6
5
4
Q.
~
z
3
2
SMALL
SMALL·
MED
MEDIUM
MEDLARGE
LARGE
Animal Size
Fig. 112. Lavasen Phase animal size distribution
(NISP=Number of Individual Specimens).
796
MANACHAQUIPHASE(N=~)
16
c.
en
z
10
8
SMALL
SMALLMED
MEDIUM
MED-
LARGE
LARGE
Animal Size
Fig. 113. Manachaqui Phase animal size distribution
(NISP=Number of Individual Specimens).
SUITACOCHA PHASE (N=29)
c.
en
z
18
16
14
12
10
8
6
4
2
0
SMALL
SMALLMED
MEDIUM
MED-
LARGE
LARGE
Animal Size
Fig. 114. Suitacocha Phase animal size distribution
(NISP=Number of Individual Specimens).
797
COLPAR PHASE (N::36)
25
20
a.
en
z
15
10
5
0
SMALL
SMALLMED
MEDIUM
MEDLARGE
LARGE
Animal Size
Fig. 115. Colpar Phase animal size distribution (NISP=Number
of Individual Specimens).
EMPEDRADA PHASE (N=211)
100
90
eo
a.
en
z
70
60
50
40
30
20
10
0
SMALL
SMALLMED
MEDIUM
MEDLARGE
LARGE
Animal Size
Fig. 116. Empedrada Phase animal size distribution
(NISP=Number of Individual Specimens) .
798
•
•
MOS
CJPfTVLODgLQSMAJORDD
MAlO 2)0MOSMI[I1 E
'
'e#.S 'YI'raj""'-f~tq,k&.&.i)14.fdoStu~;JoUJ1 ..
- - - ,~~syJcy·ytfpMc.lJfL!y•tuf~or.-o-3
...::=---
Fig. 117. Guaman Poma's rendering of a camelid with a cloth
or net alforja containing jars for transport (from Guaman
Poma 1936:524).
799
SOUTH
AMERICA
Scale
500km
Fig. 118. Map showing relative degrees of stylistic
similarity to the Manachaqui Phase Paste A style, and the
probable sources of Paste Group B sherds. Darkest shading
indicates closest similarity to Manachaqui Paste A.
800
SOUTH
AMERICA
PACIFIC
OCEAN
Scale
500 krn
7
Fig. 119. Map showing relative degrees of stylistic
similarity to the Suitacocha Phase Paste A style, and the
probable sources of Paste Group B sherds. Darkest shading
indicates closest similarity to Suitacocha Paste A.
801
SOUTH
AMERICA
PACIFIC
OCEAN
Scale
500km
Fig. 120. Map showing relative degrees of stylistic
similarity to the Colpar Phase Paste A style, and the
probable sources of Paste Group B and C sherds. Darkest
shading indicates closest similarity to Colpar Paste A.
802
SOUTH
AMERICA
PACIFIC
OCEAN
Scale
500 km
Fig. 121. Map showing relative degrees of stylistic
similarity to the Empedrada Phase Paste A style, and the
probable sources of Paste Group Band C sherds. Darkest
shading indicates closest similarity to Ernpedrada Paste A.
APPENDIX B
PLATES
803
804
Plate I. View of the Manachaqui Cave site complex from the
north side of Manachaqui Valley facing south.
805
Plate II. Manachaqui Cave and 1988 trench looking east.
Plate III. Manachaqui Cave 1990 work in progress.
806
Plate IV. Manachaqui Cave Unit 6 north profile 1988.
807
Plate V. Manachaqui Cave Unit 13 north profile 1990.
Plate VI. Manachaqui Cave Units 14-17 Floor CC, 1990.
808
7
Plate
9
10
II
12
13
Manachaqui Phase
Plate VIII. Manachaqui Phase Paste B2 (see Fig. 66c).
809
Plate IX. Ground slate points. m: Manachaqui Phase. K, 1:
Suitacocha/Colpar Phase. f, g: Colpar Phase. a-e, h-j:
Empedrada Phase.
810
Plate X. Suitacocha Phase $hape F jar (see Fig. 74).
811
Plate XII. Ernpedrada Phase Paste C2 a kaolin sherds .
...
,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,,,, 111f111111111f11
13
14
15
16
17
18
Plate XIII. Empedrada Phase projectile points.
APPENDIX C
TABLES
812
813
UNIT
1.5
Stratum
Color
pH
Organic
%
Gravel
Silt
Sand
%
%
%
Clay
%
2A
10YR2/1
6.60
5.80
18.5
68.25
0.95
12.30
2B
7.5YR3/2
6.95
3.72
23.3
62.42
0.91
9.65
3A
10YR3/2
6.30
4.41
22.7
61.66
0.89
10.34
3B
10YR3/3
6.80
3.60
28.5
57.34
1.03
9.53
Fill
Brown
8.20
0.0
29.6
62.70
1.50
7.20
Table 1.
UNIT
Soil sediments from Unit 15, Sector A.
36
Stratum
Color
pH
Organic
Gravel
%
%
Silt
%
Sand
%
Clay
%
2A
10YR2/1
5.6
1.77
18.3
69.47
1.40
9.06
2A
10YR2/1
5.6
1. 79
18.9
67.69
1.42
10.2
2B
10YR2/1
6.4
1. 78
19.2
68.22
1.45
9.35
2B
10YR2/1
6.4
1. 80
23.1
65.55
1.56
7.99
3A
10YR3/2
6.8
1.28
39.0
50.17
1. 75
7.80
3B
10YR3/3
6.85
1.30
39.4
50.24
1. 76
7.30
3B
10YR3/3
6.85
1.30
41.5
48.96
1. 74
6.50
3B
7.5YR3/2
5.80
1.32
42.7
47.60
1. 78
6.60
Bedrock
Brown
7.80
0.0
84.2
9.20
1.10
5.50
Table 2. Soil sediments from Unit 36, Sector B.
814
Floor/Unit
Class
Diam.
Depth
Phase
B
I 14
s
38 em
5 em
Empedrada
J
I 15
B
12 em
1.5 em
Empedrada
L I 14
B
23 em
6 em
Empedrada
M? I 16
s
28 em
4.5 em
Empedrada
I 17
s
19 em
5 em
Empedrada
15
B
35 em
5 em
Empedrada
p
I 14
s
45 em
13 em
Col par
z
z
I 16
R?
48 em
11 em
Manaehaqui
I 17
.S?
30 em
5 em
Manaehaqui
AA
I 17
B
26 em
2 em
Manaehaqui
BB
I 15
B
38 em
2 em
Lavas en
cc I 17
R
50 em
5 em
Lavas en
DO I 15
R
26 em
3 em
Lavas en
EE I 16
R
50 em
13 em
Lavas en
FF I 15
R
so em
7 em
Lavas en
s
T I
Table 3. Hearth classification, measurements and
chronology. Hearth classes are: R = rock-filled;
s = single embedded stone; and B = Simple basin (no
stones).
815
Lab I
I
Fie1d I (Floor)
I Context I
Sector A:
C-14 h.p.
I
calih.*
I
Phase
Shelter Interior
I-16,980
I-16,975
I-16,973
M1A-14-ll (H)
M1A-14-15 (L)
M1A-15-25 (T)
Hearth
Hearth
Hearth
1500 + 80
1380 + 80
1460 !: 80
AD 619
AD 671
AD 646
Empedrada
Empedrada
Empedrada
I-16,978
I-16,979
M1A-14-21 (P)
M1A-15-30 (W)
Hearth
Floor
1840 + 80
2110 + 80
AD 239
50 BC
Col par
Col par
Beta-75,230
I-17,756
M1A-15-35 (X)
M1A-15-38 (Y)
Floor
Floor
2560 :!:. 100
2450 :!:. 90
765 BC
409 BC
Suitacocha
Suitacocha
I-17,428
I-16,976
M1A-16-41 (Z)
M1A-17-43 (AA)
Hearth
Hearth
2800 + 90
2850 + 90
902 BC
927 BC
Manachaqui
Manachaqui
I-17,318
I-17,487
I-16,974
M1A-15-54 (BB)
M1A-16-63 (££)
M1A-15-66 (FF)
Hearth
Hearth
Hearth
3670 + 100
3830 + 100
3520 + 100
1973 BC
2197 BC
1758 BC
Lavas en
Lavas en
Lavas en
Sector B:
-
Shelter Exterior
I-17,321
I-17,319
M1A-22-ll
M1A-6-25
Lens
Lens
2630 :!:. 100
2740 :!:. 90
J95 BC
827 BC
suitacocha
Suitacocha
I-17,320
M1A-31-16
Hearth
2810 :!:. 100
906 BC
Manachaqui
I-17,322
I-17 ,429
M1A-31-25/26
M1A-31-30
Strat 3A
Strat 3B
4120 + 130
4280 :!:. 110
2586 BC
2881 BC
Unnamed
Unnamed
CAMS-13,151
M1A-31-35
strat 3C
10270 :!:. 60
10152 BC
Unnamed
• Pearson and Stuiver 1993; Stuiver and Pearson 1993; Bard et al. 1993.
The laboratory measurements were adjusted for the southern hemisphere prior
to calibration.
Table 4. Radiocarbon dates from Manachaqui Cave.
816
Rim Form
I Sherds I
1
2
3
4
5
6
7
8
9
10
11
12
13
48
38
14
12
12
10
8
8
7
6
5
5
2
Table 5.
Rim Form
la
lb
lc
2a
2b
3a
3b
4
5
6
Range
6-16
10-26
9-15
9-14
13-25
22-29
12-14
13-14
12-14
11-14
11
13
12,16
I
em
em
em
em
em
em
em
em
em
em
em
em
em
Mode
I
R
12 em
25
Bi-modal 29
13
11
N/A
9
Bi-modal? 5
N/A
8
N/A
8
N/A
3
N/A
3
N/A
6
N/A
1
N/A
2
N/A
2
I Sherds I
Range
l
Mode
I l
H
8
8-16
9-18
10-14
9-19
10-18
9-17
10-18
10-16
10-20
8-17
11-13
12-18
9
10
189
109
Decorated Groups
54
10-24 em
15 em
16 em
43
10-22 em
Table 6.
I
Average
11.64 em
15.97 em
12.55 em
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
S.D.
2.45 em
4.20 em
1.44 em
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
Manachaqui Phase Shape A Rims
250
245
8
454
41
224
67
51
37
25
19
9
7
I
em
em
em
em
em
em
em
em
em
em
em
em
12 em
88
12 em
65
N/A
3
12 em 144
20
13 em
12 em
57
12,13 em 25
13,15 em 18
15,16 em 22
11 em
15
N/A
9
N/A
8
Average
12.98
12.98
N/A
13.08
13.70
13.25
13.72
13.89
14.45
11.20
N/A
N/A
J
em
em
em
em
em
em
em
em
em
14.80 em
15.42 em
S.D.
1.65 em
1.91 em
N/A
1.87 em
1. 79 em
1.91 em
2.74 em
1.45 em
2.33 em
2.45 em
N/A
N/A
2.70 em
2.18 em
Manachaqui Phase Shape B Rims.
* Calculations were not made for sample sizes of less than
10 sherds.
817
Rim Form
I Shards I
1a
1b
2a
2b
3
4
5
6
57
11
18
9
8
7
2
1
Table 7.
Rim Form
10-18
7-18
11-19
12-18
11
12-14
12
28
JN
Mode
12 em
N/A
12 em
N/A
N/A
N/A
N/A
N/A
em
em
em
em
em
em
em
em
I Sherds I
32
14
13
13
13
2
Table 8.
Range
11-26
12-18
12-17
14-20
15-18
9.5
I
em
em
em
em
em
em
Mode
l
40
8
13
6
4
5
1
1
Manachaqui Phase Shape
1
2
3
4
5
6
Rim Form
I
Range
c
Average
13.95 em
N/A
14.46 em
N/A
N/A
N/A
N/A
N/A
S.D.
2.18 em
N/A
2.73 em
N/A
N/A
N/A
N/A
N/A
Rims.
I I
R
Average
18
6
7
9
5
2
17.5 em
N/A
N/A
N/A
N/A
N/A
15 em
N/A
N/A
N/A
N/A
N/A
l
I
S.D.
3.99 em
N/A
N/A
N/A
N/A
N/A
Manachaqui Phase Shape D Rims.
I Shards _I
1
2
3
4
5
6
7
6
6
5
3
3
2
1
Range
10-18
11-14
13-18
14-16
16-18
22
18
em
em
em
em
em
em
em
Table 9. Manachaqui Phase
Shape E Rims.
Technique
Total
Interior Surface Treatment
Finished
Unfinished
Position
Medial
Rib ends
Shoulder
Unembellished
122
20 (16\)
102 (84\)
115 (94\)
7 (6\)
10 (8\)
Notched A
292
231 (79\)
61 (21\)
292 (100\)
0 (0\)
14 (5\)
Notched B
38
36 (95\)
2 (5\)
14 (37\)
24 (63\)
Incised A
502
Incised B
75
24 (5\)
17 (23\)
6 ( 16\)
478 (95\)
485 (97\)
17 (3\)
18 (4\)
58 (77\)
55 (73\)
20 (27\)
N/A
961 (93\)
68 (7\)
48 (
Table 10. Totals pertaining to rib and flange embellishment
techniques. Severely eroded sherds were not included for
tabulations (N=l,029).
818
Rim Form
j
14
15
16
17
I
14
8
7
1
Table 11.
Rim Form
Sherds
10-14
10-14
11-12
17
252
110
86
52
35
28
22
16
11
llf
llg
12
13
Table 12.
B
6
5
4
1
Average
I
N/A
N/A
N/A
N/A
S.D.
N/A
N/A
N/A
N/A
I
Mode
I
12 em
12 em
10 em
12 em
12 em
13 em
N/A
N/A
N/A
em
em
em
em
em
em
em
em
em
B
I
64
44
41
26
11
15
7
9
7
Average
11.58
11.57
11.44
11.54
11.18
12.0
N/A
N/A
N/A
em
em
em
em
em
em
I
S.D.
1. 73 em
1. 68 em
2.02 em
1. 95 em
1. 80 em
1. 86 em
N/A
N/A
N/A
Suitacocha Phase Shape B rims.
I Sherds I
7
8
9
10
37
9
7
1
Range
9-16
14-20
14-17
16
em
em
em
em
I
Mode
I I
Average
13/14 em
N/A
N/A
N/A
14
6
5
1
13.07 em
N/A
N/A
N/A
H
suitacocha Phase Shape
Rim Form
I Sherds I
8
9
10
11
7
6
1
1
Table 14.
J l
N/A
N/A
N/A
N/A
em
em
em
em
Range
7-15
8-15
7-15
8-16
8-15
9-16
9-12
10-18
10-15
Rim Form
Table 13.
Mode
Suitacocha Phase Shape A rims.
I Sherds I
lla
llb
lle
lld
lle
I
Range
Range
11-18 em
8-12 em
27 em
12 em
I
Mode
N/A
N/A
N/A
N/A
I
S.D.
1. 91 em
N/A
N/A
N/A
c rims.
I I
H
Average
4
2
1
1
N/A
N/A
N/A
N/A
suitacocha Phase Shape E rims.
I
S.D.
N/A
N/A
N/A
N/A
819
Rim Form
I Sherds I
109
29
79
26
7
3
6
6
4
3
1a
1b
2
3
4a
4b
5
6
7
8
Table 15.
Rim Form
7-16
10-17
5-15
10-18
7-14
10-12
10-15
13-15
21
12-15
11
71
27
14
10
9 •
6
4
4
3
2
2
Table 16.
Mode
11 em
11
11
14
N/A
N/A
N/A
N/A
N/A
N/A
em
em
em
em
em
em
em
em
em
em
I I
Average
31
15
19
17
5
2
2
4
1
3
11.52 em
12.33 em
10.84 em
13.94
N/A
N/A
N/A
N/A
N/A
N/A
R
I
S.D.
2.09
1. 92
2.54
2.55
N/A
N/A
N/A
N/A
N/A
N/A
em
em
em
em
Range
I
Mode
12-17 em
N/A
9-20 em 13/14 em
6-20 em
14 em
N/A
12-20 em
N/A
15-18 em
N/A
11-23 em
N/A
8-13 em
N/A
10-14 em
30 em
N/A
13-20 em
N/A
N/A
N/A
N/A
N/A
I I
Average
5
29
19
7
4
5
4
4
1
2
0
0
N/A
14.06 em
13.11 em
't:T/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
R
I
S.D.
N/A
3.02 em
3.58 em
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
Empedrada Phase Shape E rims.
Rim Form
1Sherds I
1
2
3
4
5
6
7
8
9
10
56
43
21
12
8
4
3
2
1
1
1
11
I
Suitacocha Phase Shape F rims.
I Sherds I
3
9
15
16
17
18
19
20
21
22
23
24
Range
Range
9-24 em
11-23 em
12-21 em
11-20 em
15-20 em
12 em
11 em
19 em
7 em
12 em
N/A
I
Mode
10 em
17 em
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
I I
Average
21
25
7
8
3
1
1
1
1
1
0
15.48 em
15.96 em
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
R
Table 17. Empedrada Phase Shape G rims.
I
S.D.
4.79 em
2.88 em
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
820
Phase
Lavas ell
Mallachaqui
suitacocha
C:olpar
C:ol/E=p
Floors
PF - BB
AA- z
Y, Z:
'tf
v, u
l4
4
4
2
2
14
Sample Weight
565.01
139.91
98.98
50.76
13.25
527.2
Wooc! Weight
430.32
54.58
20.985
19.745
3.14
291.84
\ of Sample
Weight
85.2
39
21.2
38.9
23.7
55.4
Maize Kel:'llals
37
46
1
236
Maize Cob Frags
42
15
4
555
Phaseolus
2
No. of Samples
E=pec!rac!a
or-
5
cf. Sapot. Fruit
76
8
6
Fruit Rille!&
57
20
5
1
'l!uber/Root Frags
40
5
1
2
17
1
1
LupillUS
1
6
cf. Ribes
Rubus
Festuca
9
1
1
1
3555
C:hellopoc!/amerallt
3
Polygollum/Rumex
5
2
12
32
3
4
11
6
123
1
1
2
2
Table 18. Botanical remains by phase (see Appendix F).
3
G
821
Class
Order
Family
Genus
Species
Mammalia ••••••••
• ••••.••.•. . 19
Rodentia . ...................... .
Caviidae •••••••••••
Cavia •..•...•..•.
Artiodactyla ••••••••••••
cf. Artiodactyla ••••••••
Aves.
RISP
• ........•.•. 2
. ............. 1
. ............. l
. ........•••. 2
• ••..•...•.•. 1
.•••.......... 1
. •••.......... 1
Strigiformes •••••
Table 19. Fauna of the Lavasen Phase.
Class
Order
Family
Genus
Species
RISP
15
Marmnalia • .................•.
Rodentia ••••••••••••••
S igmodon tinae ••
. .....•. 1
. •........•... 1
Table 20. Fauna in LavasenjManachaqui Strata
822
Class
Order
Family
Genus
Species
NISP
............................
51
. .............•............... 1
Mammalia ••••••
cf. Mammalia ••
Rodentia •••••••••••••
Caviidae ..•••.•
Cavia .•
Agoutidae ...•.•..........
Agouti ... ....... .
cf. taczano.
..5
• •2
• •2
. ........... 1
. ....... 1
. ...... . 1
Artiodactyla •..••••••••••••.••••
Cervidae . ............... .
• .•...•• 3
• •••••••• 2
Odocoileus •••••••
. ......... 1
virginianus •••••••••
. ...•.. . 1
cf. Odocoileus •••••••••••••••
. ....... 1
cf. virginianus ••••••• . . . . . . . . . . 1
Xenarthra •••••.•••••••.•
Dasypodidae ••••••
Dasypus .•••••.••.
•••••••••• 5
• •••.••••• 5
.......... 1
Table 21. Fauna of the Manachaqui Phase
Class
Order
Family
Genus
Species
Mammalia ••••
NISP
. ••.•••. 6
Rodentia.
• •2
Artiodactyla . ................ .
Cerv-idae . ............. .
.
.
.
.
Odocoileus • •..•
virginian us ••
.1
.1
.1
.1
. .1
Osteichthyes .•••.
Table 22. Fauna in ManachaquijSuitacocha Strata.
Class
Order
Family
Genus
Species
Mammalia •.••
NISP
• •••••. 28
. •••.•...••• 8
. .1
Rodentia ••••..•••
cf. Rodentia ...•.
Caviidae .•
Cavia.
cf. porcellus • ••
Sigmodontinae.
. ......... 3
• •3
.• 1
. .1
Osteichthyes.
.2
Aves ...•••
.1
Reptilia . ................................................... 1
Table 23. Fauna of the Suitacocha Phase.
823
Class
Order
Family
Genus
Species
HISP
.......................
Mammalia ••••
6
Artiodactyla . ......................................... 2
Cervidae . ..................................... . 1
cf. Odocoileus ••••.••
.1
c f . Camelidae . ....................•............ 1
Table 24. Fauna in SuitacochajColpar Strata.
Cl.ass
Order
Family
Genus
Species
Mammal.ia ••••
HISP
•••.•....... 33
Rodentia ••
cf. Rodentia •••••••••
Caviidae •••••
• .11
... 2
. •• 6
Cavia ••••••••
cf. Cavia.
3
. ..•.......... 2
Artiodactyla ••.•.••••••••
. .....•. 3
Cervidae . . . . . . . . . .
cf. Mazama ••••
cf. Camelidae •••••••••
. ........ 2
...... ~ . 2
. ....... 1
Xenarthra.
Dasypodidae •••••••••••
cf. Dasypus ••
cf. Dasypodidae ••••••
. .......... 1
. ........ 1
. ..... 1
Osteichthyes ••••
........ 2
. ..... 1
Table 25. Fauna of the Colpar Phase.
Class
Order
Famil.y
Genus
Species
Mammalia ••••
HISP
• • 28
Rodentia ••••••••.••••
Caviidae •••••••
. .11
Cavia.
porcellus •••
cf. porcellus.
Aves •• •••••••
Osteichthyes •••••••••••••••
cf. Eleotridae ••
Table 26. Fauna in ColparjEmpedrada Strata.
. •. 6
. •. 6
.•• 1
.1
.1
.1
. .. 1
824
Class
Order
Family
Species
Genus
Mammalia •••••••
cf. Mammalia •••
RISP
. •••••.•.•..•••.••• • 19 9
. •••.••....••••.•••• 3
Rodentia.
cf. Rodentia •••••
Caviidae ••
cf. Caviidae ••••
Cavia ••.
cf. Cavia.
porcellus.
cf. porcellus ••
Sciuridae ••
Sigmodontinae •••
• .•••••••••••• 4 3
• •••.•••••••••• 2
• •• 16
• .••.•.•.•••••. 6
• •• 12
. ••. 2
.3
.6
.1
.1
Ar-tiodactyla. . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . 22
cf. Ar-tiodactyla. .
. ......................... 2
Cervidae. . .
.................
. ... 7
cf. Cervidae. . . .
. ........................... 2
Odocoileus. . . . . . . . .
. .............. 4
virginianus • •
. •• 4
cf. Pudu •.
..•• 1
Camelidae •••••••••
• ••• 7
cf. Camelidae •••••
• ••• 2
cf. Lama.
.1
Xenarthra •••••••••••••
Dasypodidae ••••
•••• 6
. .•• 6
Carnivora ••••••••
cf. Carnivora ••••
Canidae..
. ...... .
Dusicyon •••
culpaeus andinus.
.3
.3
.1
.1
.1
. .• 1
. ..... 1
Primates . .......... .
Hominidae ••.••
Homo ••
sapiens ••••
Aves.
Strigiforrnes .••••
Strigidae.
cf. Reptilia ••••••••
cf. Amphibia .•
cf. Osteichthyes •••
Table 27. Fauna of the Empedrada Phase.
... 1
. .1
• •• 11
. .. 1
. ... 1
•••• 2
. ... 1
.1
825
I
i!HA.SE~.SIZE
I
.SHALL
I .SHALL-MED I MEDIJ.lM I HED-LABGE I LAB!:ZE I :t~AL I
EMPEDRADA
98
43
25
12
33
211
COLPAR
21
5
4
2
4
36
SUITACOCHA
17
3
2
3
4
29
MANACHAQUI
11
16
10
5
6
48
LAVASEH
4
2
4
6
5
21
PRE-LAVASEN
7
10
17
48
20
102
Table 28. Animal size by phase. Sums are NISP.
APPENDIX D
MACROCHRONOLOGY
826
827
Unit 1
Unit 2
Level Shards Carin Incised Kaolin Lithics
Level Shards Carin Incised Kaolin Lithics
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
10
12
30
84
172
136
141
128
183
201
180
150
83
193
80
14
6
Total 1803
2
2
9
1
6
1
3
2
3
1
5
2
6
9
20
13
1
26
1
2
10
28
31
37
36
20
28
22
4
8
3
9
10
11
2
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
8
54
66
73
46
43
63
85
142
117
154
58
170
54
38
11
3
1
1
1
2
7
1
4
2
2
2
1
3
1
2
3
4
4
1
2
1
10
41
41
45
21
18
7
8
10
14
4
7
5
7
8
286
Total 1185
20
3
22
246
Unit 3
Unit 4
Level Shards Carin Incised Kaolin Lithics
Level Shards Carin Incised Kaolin Lithics
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
3
25
140
111
30
35
16
75
3
99
98
76
83
82
47
12
4
Total 1004
3
3
5
2
3
1
1
1
27
45
12
2
4
8
8
5
7
4
5
6
19
141
1
1
1
2
2
2
6
4
1
1
1
1
2
2
1
4
18
9
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
so
161
79
112
88
155
96
123
4
10
2
1
1
1
3
4
5
1
lOS
100
80
72
36
80
65
152
205
74
18
8
4
3
2
2
4
4
2
Total 1860
23
1
53
62
17
18
11
17
4
2
5
2
1
1
3
5
6
13
24
12
23
13
18
279
Note: "Carin" refers to sherds from medial and shoulder
angles, and sherds bearing applique ribs. It excludes the
16 Suitacocha Phase sherds from basal angles. "Incised"
refers only to line-incised, but excludes the 31 Manachaqui
Phase sherds with line-incision.
828
Unit 5
Unit 6
Level Shards Carin Incised Kaolin Lithics
Level Shards Carin Incised Kaolin Lithics
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
67
140
105
118
160
137
118
255
172
138
85
eo
120
95
149
172
120
135
110
15
1
Total 2492
1
1
3
1
5
11
13
11
6
4
1
55
1
1
1
5
4
4
2
2
2
17
7
6
1
2
51
69
41
25
23
16
14
32
4
2
1
3
6
2
19
37
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
54
87
101
11
75
66
1
lOB
95
147
190
157
134
150
290
104
98
87
130
150
142
162
88
58
7
4
1
1
2
2
3
8
12
3
4
10
23
14
5
10
5
3
2
2
l
l
ll
l
l
l
5
4
2
2
289
Total 2543
37
58
68
38
23
8
4
48
26
56
439
Unit 7
Unit 8
Level Shards Carin Incised Kaolin Lithics
Level Shards Carin Incised Kaolin Lithics
1
2
3
4
5
6
7
8
9
10
12
13
22
7
20
90
142
97
126
85
130
60
36
15
5
Total
842
ll
5
l
2
10
3
8
32
26
2
3
l
l
5
4
3
1
2
12
14
5
2
3
4
5
6
7
8
9
10
11
12
5
6
19
28
31
82
8
105
45
34
11
30
Total
404
l
3
2
4
5
5
l
l
2
1
1
l
1
l
4
5
27
102
Unit 11
Unit 12
Level Shards Carin Incised Kaolin Lithics
Level Shards Carin Incised Kaolin Lithics
l
2
3
4
5
6
7
8
9
10
11
12
13
70
65
93
120
102
118
115
188
104
130
18
1
3
Total 1148
l
l
1
3
2
2
3
l
l
2
2
3
1
2
4
1
l
l
l
2
13
9
14
3
2
12
8
7
13
17
18
12
15
4
1
6
120
12
13
15
21
27
62
38
25
95
23
118
56
14
18
3
Total
515
l
2
3
4
5
6
7
8
9
10
ll
l
l
l
1
l
6
2
2
1
3
1
1
3
3
8
12
17
6
l
2
1
l
12
4
9
so
829
Unit 13
Unit 15
Level Sherds Carin Incised Kaolin Lithics
Level Shards Carin Incised Kaolin Lithics
14
10
30
17
77
27
39
12
43
105
53
43
48
5
14
Total
523
1
2
3
4
5
6
7
8
9
10
11
12
13
5
1
2/A
3/B
4/C
5
6/0
7
8/E
9/F
10/G
11
10
26
2
10
2
1
1
1
4
1
1
6
4
2
1
1
11/Ei
15
78
2
1
Unit 14
Level Sherds Carin Incised Kaolin Lithics
1
2/A
3/B
4/C
5
6/0
7
8/E
9/F
10/G
11/Ei
12/I
13/J
14/K
15/L
16/LL
17/M
18/N
19/N
20/0
21/P
22/Q
23/R
24/S
25/T
26/U
27
28/V
29
30/W
31
32
33
34
35/X
36
37
38/Y
39
40
41/Z
42
43/AA
44
45
46
47
48
49
50
Total
2
10
20
18
20
23
3
5
4
7
2
1
5
3
5
4
2
3
2
7
5
7
2
2
4
3
4
31
4
52
5
27
32
53
24
1
10
11
4
8
16
3
15
I
1
1
1
6
3
3
1
1
1
Total
1
1
1
2
11
2
8
16
4
2
1
1
1
486
12/I
13/J
14/K
15/L
16/LL
17/M
18/N
19/N
20/0
21/P
22/Q
23/R
24/S
25/T
26/U
27
28/V
29
30/W
31
32
33
34
35/X
36
37
38/Y
39
40
41/Z
42
43/AA
44
45
46
47
48
49
50
51
52
53
54/BB
55
14
5
1
32
4
12
2
11
27
10
3
14
2
8
5
4
1
1
4
3
1
4
4
2
2
4
5
3
2
4
3
4
5
3
12
4
29
1
12
20
23
39
7
6
15
2
18
8
1
1
1
1
1
1
1
1
2
1
2
1
1
1
1
1
1
10
11
2
7
1
1
2
2
1
2
2
1
2
370
6
1
11
30
830
Unit 16
Unit 17
Level Shards Carin Incised Kaolin Lithics
Level Shards Carin Incised Kaolin Lithics
1
2/A
3/B
4/C
5
6/D
7
6/E
9/F
10/G
4*
5*
6*
7•
B*
9*
10*
11*
36/Y
39
40
41/Z
42
43/AA
44
45
46
47
46
49
50
53
46
93
25
64
6
12
5
6
11
262
63
67
133
lOB
96
46
46
7
4
73
39
33
1
35
33
45
11
9
7
3
Total 1444
1
1
19
14
25
6
1
2
3
1
9
3
17
6
6
5
7
1
2
1
2
1
7
1
3
1
1
2/A
3*
4*
5•
6•
7*
B*
9*
10*
11*
12*
36/Y
39
40
41/Z
42
43/AA
44
45
46
47
46
49
50
51
52
53
4
66
233
142
40
53
63
104
49
34
94
31
3
26
25
11
1
19
25
37
10
6
2
2
1
1
1
1
3
5
1
1
1
2
4
1
1
4
3
4
1
1
2
1
1
1
1
1
1
1
2
5
2
Total 1104
4
18
41
15
3
10
10
14
101
2
16
3
13
119
Note: Totals do not include artifacts contained in soil samples.
* Denotes arbitrary levels
Unit 16
Unit 19
Level Shards Carin Incised Kaolin Lithics
Level Shards Carin Incised Kaolin Lithics
4
110
96
325
136
276
256
279
139
132
102
47
161
71
93
96
26
33
16
4
3
1
1
1
3
1
3
2
2
7
1
5
2
Total 2410
29
1
2
3
4
5
6
7
6
9
10
11
12
13
14
15
16
17
16
19
20
21
22
2
1
1
2
4
1
1
3
6
6
6
4
2
23
31
45
17
46
26
19
6
4
6
5
3
3
11
2
5
1
10
29
1
2
3
4
5
6
7
6
9
10
11
12
13
14
15
16
17
16
19
20
21
22
23
5
59
190
506
121
165
247
326
164
150
236
166
295
205
119
194
53
12
2
3
1
3
1
5
25
13
5
6
2
1
1
2
3
6
1
2
1
1
19
5
6
1
62
119
16
22
25
32
4
6
17
1
10
13
12
11
7
1
2
4
4
3
8
1
256
Total 3245
62
16
35
401
831
Unit 20
Unit 21
Level Shards Carin Incised Kaolin Lithics
Level Shards Carin Incised Kaolin Lithics
1
2
3
4
5
6
7
8
9
10
11
12
13
14
78
211
188
151
239
268
320
255
170
235
283
185
88
19
Total 2690
1
1
2
2
3
2
5
6
4
1
4
6
8
1
4
11
28
6
3
2
1
1
2
1
27
31
28
33
19
21
21
20
9
7
1
6
1
20
56
199
1
2
3
4
5
6
7
8
9
10
ll
12
13
14
15
16
17
18
50
194
132
175
134
170
248
314
167
82
93
77
87
51
51
25
1
3
1
2
5
6
1
1
2
5
15
35
7
1
10
2
3
3
13
3
1
3
2
5
1
2
4
3
Total 2053
20
13
56
9
25
19
10
25
3
4
4
67
35
183
Unit 22
Unit 23
Level Shards Carin Incised Kaolin Lithics
Level Shards Carin Incised Kaolin Lithics
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
51
85
84
78
70
101
105
166
86
182
496
224
248
166
236
257
280
161
119
71
Total 3266
1
1
2
7
9
16
12
13
3
6
5
10
5
6
2
1
1
13
6
2
1
37
51
26
19
24
24
16
14
12
8
1
2
3
4
5
6
7
8
9
10
11
ll
1
2
12
13
14
15
16
17
18
19
20
1
3
5
1
1
6
38
77
40
ll
24
292
6
3
42
92
56
86
107
78
62
82
43
70
145
125
43
40
65
70
59
72
1
2
4
4
2
Total 1346
16
1
1
2
l
3
2
1
4
6
22
66
71
40
34
32
27
22
13
17
34
9
1
1
7
1
399
832
Unit 24
Unit 25
Level shards Carin Incised Kaolin Lithics
Level Shards Carin Incised Kaolin Lithics
l
2
3
4
5
6
7
8
9
10
ll
12
13
14
15
16
17
18
19
20
21
22
23
24
8
39
128
94
152
200
360
321
200
171
160
265
100
124
172
166
193
123
252
135
47
36
17
1
l
2
3
4
5
7
11
7
20
7
2
3
2
4
7
5
3
18
5
13
8
4
ll
6
12
l
2
3
4
5
6
7
8
9
10
Total
l
1
82
59
11
28
125
159
136
100
46
7
2
2
l
l
l
l
3
l
l
3
3
3
2
l
4
2
5
12
3
8
12
17
3
ll
9
9
7l
l
l
l
754
4
3
1
1
3
4
6
12
ll
Total 3474
56
71
154
87
85
84
125
2
5
5
5
73
46
.
25
710
Unit 26
Unit 27
Level Shards Carin Incised Kaolin Lithics
Level Shards Carin Incised Kaolin Lithics
1
2
3
4
5
6
7
8
9
10
ll
12
13
14
15
16
17
18
19
20
21
100
104
124
140
182
176
146
98
172
135
25
43
94
133
128
151
220
145
73
24
9
Total 2422
3
3
4
2
2
4
3
4
2
4
2
3
2
7
3
3
44
46
35
17
19
17
9
4
10
6
7
8
9
10
ll
1
ll
10
1
ll
l
3
4
l
l
63
l
2
3
4
5
l
23
16
12
13
14
15
16
17
18
19
20
21
22
55
138
181
93
109
145
162
164
95
192
218
179
153
103
130
156
121
200
172
l
l
l
2
11
8
15
10
6
3
2
5
44
53
69
31
29
33
14
15
14
4
6
3
l
4
2
4
10
18
12
9
66
6
57
10
2
Total 2899
67
21
193
24
51
327
833
Unit 2B
Unit 29
Level shards Carin Incised Kaolin Lithics
Level Shards Carin Incised Kaolin Lithics
l
2
3
4
5
6
7
B
9
lO
ll
12
13
14
15
16
17
lB
19
20
21
22
23
24
25
eo
ll4
150
104
109
173
155
lOB
66
66
10
21l
172
91
103
67
llB
51
162
1l
61
29
36
13
l
Total 2503
l
7
19
13
6
l
l
l
l
6
6
3
lO
5
3
l
2
36
4
14
2
2
43
40
76
47
35
37
54
40
21
lO
2
l
6
2
l
l
ll
33
23
46
l
2
3
4
5
6
7
B
9
lO
ll
12
13
14
15
16
17
lB
19
20
17
66
240
69
127
159
62
169
122
270
301
192
324
242
225
166
lOB
21
19
4
Total 2949
l
5
l
5
6
15
B
ll
6
B
l
l
6
2
3
1
2
2
3
3
l
3
l
l
2
1
l
66
11
75
76
24
65
62
42
93
44
27
lB
l
2
9
21
576
460
Unit 30
Unit 31
Level Shards Carin Incised Kaolin Lithics
Level Shards Carin Incised Kaolin Lithics
l
2
3
4
5
6
1
B
9
lO
ll
12
13
14
15
16
17
lB
19
20
21
22
40
74
209
166
69
ll3
137
206
273
497
492
276
206
322
342
322
252
297
236
40
lO
6
Total 46ll
l
l
2
16
9
2
l
l
5
16
lO
ll
6
12
B
2
72
6
19
ll
4
40
31
lB
46
100
49
22
34
50
44
64
65
20
17
15
l
2
ll
6
B
15
5
14
19
1
2
3
4
5
6
7
B
9
lO
ll
12
13
14
15
16
17
lB
19
20
21
22
23
65
66
42
69
l2B
90
120
162
161
104
276
206
227
212
64
383
221
l2l
106
54
39
12
B
l
l
3
l
7
B
15
6
7
l
l
12
4
5
5
2
6
7
2
32
16
24
34
ll2
27
46.
47
41
17
6
3
4
2
2
l
l
4
2
4
16
ll5
2
l
667
Total 3022
50
19
29
557
834
Unit 32
Unit 34
Level Sherds Carin Incised Kaolin Lithics
Level Sherds Carin Incised Kaolin Lithics
l
2
3
4
5
6
7
8
9
10
ll
12
13
14
15
16
17
18
19
20
21
22
23
46
61
26
58
79
164
l
3
9
2
2
us
129
267
365
264
234
226
237
218
201
254
136
96
64
l
l
l
l
5
16
14
3
12
4
3
2
6
3
2
4
2
l
l
2
6
7
2
10
5
68
13
ll
4
4
Total 3259
63
15
27
19
42
49
66
48
47
34
21
4
4
5
6
17
19
l
2
3
4
5
6
7
8
9
10
ll
12
13
14
15
193
7
97
247
96
98
520
182
135
176
145
51
18
8
5
Total 1977
1
7
3
4
5
2
14
5
1
7
3
l
3
6
6
15
l
34
86
14
130
79
25
20
54
8
11
439
500
Unit 35
Unit 36
Level Sherds Carin Incised Kaolin Lithics
Level Sherds Carin Incised Kaolin Lithics
l
2
3
4
5
6
7
8
9
10
ll
12
13
14
15
16
17
18
19
20
52
89
91
6
24
138
161
165
206
298
250
178
34
91
107
136
92
27
20
10
12
l
l
5
4
10
6
6
4
15
12
34
4
10
36
55
38
13
l
2
2
l
l
4
5
7
8
l
6
10
2
2
17
l
2
3
4
5
6
7
8
9
10
ll
12
13
14
15
16
17
18
19
8
28
115
35
78
240
195
205
447
218
236
303
308
366
210
75
ll
46
17
12
260
4
8
8
l
3
4
5
12
12
8
10
34
19
73
7l
44
29
4
6
2
l
l
4
l
Total 3082
Total 2175
2
l
49
13
22
281
835
Unit 37
Unit 38
Level Shards Carin Incised Kaolin Lithics
Level Shards Carin Incised Kaolin Lithics
l
2
3
4
5
6
7
8
9
10
11
12
13
14
15
133
26
120
237
164
220
201
123
182
157
374
188
110
53
6
Total 2295
2
57
34
67
lOS
l?
l
35
29
2
l
4
7
13
21
2
6
l
1
l
5
56
3
3
333
l
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
40
33
3
11
57
191
135
328
386
144
60
Laval Shards Carin Incised Kaolin Lithics
1
2
3
4
5
6
7
8
9
10
81
16
201
75
270
180
216
128
35
1
Total 1203
1
ll
1
l
2
7
10
2
1
21
1
14
40
7
118
69
25
2
8
1
3
17
290
5
1
3
11
5
1
9
1
16
12
9
6
2
Total 1443
Unit 39
1
10
5
4
1
2
24
10
12
5
1
7
2
3
15
31
35
21
11
7
167
APPENDIX E
VESSEL SHAPES AND RIM FORMS BY PHASE
836
837
Shs:b!~
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
Bim
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
~risnt
I2ism :U:nit
6
6
9
9
10
10
10
11
11
11
12
12
12
12
12
12
12
13
13
14
14
14
15
15
16
2
2
2
2
2
2
2
10
10
11
2
7
7
7
7
7
7
8
19
20
25
26
27
29
29
29
33
33
34
34
35
37
37
2
28
2
24
1
30
30
1
27
34
6
7
24
28
31
32
34
17
30
4
7
11
6
22
7
2
21
26
28
36
36
38
38
38
30
36
20
MANACBAQUJ: PHASE R:IMS
I.!~v~l
24
7
8
9
9
9
9
7
7
13
7
16
17
15
15
16
19
22
11
11
11
11
12
18
12
21
20
19
16
20
24
17
11
28
9
20
16
20
9
10
48
18
30
9
8
32
20
9
19
13
19
21
8
11
9
10
12
17
12
10
Shsg~
Bim
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
2
2
2
2
2
2
12
12
12
13
13
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
13
14
14
14
14
15
15
16
16
16
17
18
18
18
18
19
23
24
26
26
~~igDt ~ism
13
9
12
12
12
12
13
13
13
13
13
14
15
9
9
9
10
10
11
11
13
14
Unit
29
31
36
1
5
29
29
26
28
34
39
13
39
8
18
24
35
4
25
32
32
29
0
28
5
34
21
24
26
15
30
30
31
37
5
26
30
31
31
36
30
29
34
36
25
26
34
29
35
24
36
26
21
18
19
35
36
36
36
36
l!~v~l
16
20
14
12
33
12
16
18
17
:..1
7
11
7
10
18
17
16
28
0
18
19
14
0
18
33
11
4
14
13
39
15
16
16
8
27
17
15
14
14
12
14
13
11
14
5
12
11
16
20
19
13
19
6
7
6
10
11
12
13
13
838
MANACBAQUI: PHASE RDIS
.Sh'l:r2~
Bim :ls!.I:iii!.Dt 1:2ism! Ynit
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
B
B
B
B
5
5
5
5
5
6
6
6
6
6
6
6
6
6
6
7
7
7
7
7
7
7
7
8
8
8
8
8
8
8
8
9
9
9
9
9
9
9
10
10
10
10
10
10
12
13
14
11
12
13
13
14
14
11
11
11
11
11
11
13
15
16
24
25
19
22
22
22
23
24
25
27
29
12
12
12
12
13
13
14
14
13
14
14
12
12
12
12
12
13
13
12
16
13
13
1
1
1
1
a
a
a
a
19
24
28
28
28
20
4
4
6
27
5
4
6
20
6
22
25
34
34
38
38
22
25
5
20
26
26
27
30
20
22
1
18
20
38
32
34
22
37
19
20
20
21
30
11
19
24
24
24
8
26
35
2
2
31
13
2
2
8
13
L~:!l~l
13
19
14
16
13
10
27
28
30
15
33
29
34
12
29
17
8
11
12
10
10
18
8
33
12
16
18
17
16.,
10
19
21
11
11
10
17
12
17
11
15
11
12
13
18
5
7
9
13
15
8
12
11
23
23
16
9
20
23
8
10
Sh'l:b!~
Rim
B
B
B
B
B
3
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
~'l.:iii!.nt
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
1:2i'lm
~nit L~~l
15
15
15
17
18
18
18
18
18
18
19
19
19
19
19
19
19
19
20
20
20
21
21
21
21
22
22
22
22
22
22
22
22
22
22
22
22
22
22
22
22
22
22
22
23
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
28
30
43
41
4
7
7
10
14
16
11
12
12
12
13
13
13
13
5
9
11
6
6
7
7
12
13
13
13
14
14
14
15
15
15
15
15
16
16
16
16
17
17
18
18
14
14
14
14
15
15
15
15
16
16
17
17
17
17
17
839
MANACBAQOJ: PHASE RIMS
Sha:Q~
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
Bim Vg:t::is.nt l:2is.m unit
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
24
24
26
26
26
26
26
26
26
26
26
27
27
27
27
27
27
27
27
27
28
28
29
29
29
29
29
29
29
29
30
30
32
32
32
32
32
32
32
32
32
32
32
32
32
33
33
34
34
34
34
34
35
35
35
35
35
35
35
35
J:.~~l
17
19
5
6
6
10
12
14
14
14
16
14
15
15
15
16
16
17
17
17
15
19
11
11
12
12
13
13
14
15
14
14
9
12
12
12
13
13
14
15
15
15
15
16
17
19
20
6
8
8
8
9
6
12
12
13
14
15
16
16
Sha:Q~
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
Bim
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Ys.~is.nt ~is.m
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
8
9
9
10
10
10
11
11
11
11
11
12
12
12
12
12
12
12
12
12
12
12
unit
35
35
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
37
37
37
37
37
37
37
37
37
37
37
38
38
38
38
38
38
39
39
5
12
30
1
5
30
1
1
6
17
28
5
5
5
6
14
18
20
28
28
28
28
I.!~v~l
17
18
9
11
11
11
12
12
12
12
12
12
12
12
12
12
12
13
14
8
9
9
9
10
10
10
11
11
11
12
8
8
8
8
8
9
6
7
26
9
15
22
30
15
21
21
30
8
17
24
29
33
27
13
10
8
13
15
15
16
840
MANACBAQUI PHASE RIMS
Shsu:2~
Bim Ysu::is:mt
l:li~m
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
13
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
15
15
15
15
15
15
15
15
15
llnit
28
30
30
30
30
30
30
30
30
31
31
31
31
31
31
1
4
4
4
5
5
5
27
28
28
28
28
30
31
31
31
3
5
5
6
6
13
17
17
17
28
28
28
30
30
30
30
31
31
31
31
5
5
6
12
16
17
17
30
30
l!~~~l
16
13
14
15
15
15
15
15
18
14
14
14
15
16
17
20
19
20
26
27
27
28
20
16
17
17
17
15
13
15
17
1
27
33
27
27
8
6
39
45
13
14
14
12
13
14
15
16
16
16
16
26
33
27
11
40
7
42
14
18
Sh~:12~
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
a
a
a
a
B
B
a
a
Rim
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
v~~i~nt Di~m
a
a
a
a
a
a
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
16
16
16
16
16
16
Qnit
6
14
28
30
31
36
2
2
2
2
3
3
3
3
3
3
3
3
3
6
6
6
7
11
11
11
11
15
18
18
18
19
19
19
19
20
20
20
20
20
20
20
20
21
21
21
21
22
22
22
22
22
22
22
22
22
22
23
23
23
~gv~l
29
26
17
18
13
13
20
20
22
22
20
20
20
21
23
23
23
24
24
19
26
33
7
5
7
9
10
35
6
8
14
4
11
14
16
5
7
9
9
10
10
11
11
10
:o
11
12
13
14
14
14
14
15
,15
...
-0
l7
17
17
17
17
841
MANACBAQUZ PHASE RIMS
Shg:b!~
Bim
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
~.:isnt
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
;QiS!J!l :Unit
23
23
23
23
23
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
26
26
26
26
26
26
26
27
27
27
27
27
27
27
27
27
27
27
27
27
29
29
29
29
29
29
29
29
29
29
30
31
L~Y:~l
17
18
19
20
20
12
12
13
13
13
13
14
14
14
16
16
16
16
16
16
16
17
17
18
18
19
19
19
5
13
14
14
15
16
16
13
14
15
15
15
16
16
16
16
16
18
19
19
9
10
10
12
12
13
13
13
13
14
14
12
SlJg:Q~
Bim
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
~~ignt
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
Qism :Unit
31
32
32
32
32
32
32
32
32
32
32
32
34
34
34
34
34
34
34
34
34
34
34
34
34
34
35
35
35
35
35
35
35
35
35
36
36
36
36
36
36
36
36
36
36
36
37
37
37
37
37
38
38
38
38
38
38
38
38
38
~~Y:~l
16
10
12
13
13
15
15
15
15
15
16
16
8
8
9
9
9
9
9
9
9
10
10
10
10
11
10
11
11
11
11
11
12
12
14
10
11
11
11
12
12
12
12
13
13
13
7
9
10
10
11
9
9
9
9
9
9
9
9
9
842
Sh5li2~
Rim
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
:sle.::iS~.Dt
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
Q;i.SU!l Unit Ls:vs:l
9
9
10
10
10
10
11
11
11
11
11
11
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
13
13
13
13
13
13
13
13
13
13
13
13
13
14
14
14
14
14
14
14
14
14
14
14
14
15
38
38
39
39
39
39
12
30
4
5
30
31
1
5
5
6
28
31
4
5
6
6
6
6
6
12
15
28
28
28
30
30
31
31
5
5
5
5
5
6
6
17
28
28
30
30
31
5
5
5
6
14
17
28
30
30
31
31
31
16
9
10
5
6
6
7
9
15
24
26
15
13
22
28
30
25
15
16
26
27
29
30
30
31
31
4
54
16
17
18
13
15
14
15
25
26
28
30
30
28
30
43
14
17
15
15
14
24
29
31
30
40
45
15
15
18
10
14
17
42
MANACBAQU:I PHASE RDIS
:2he!2~
Eim
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
:sla~iSlnt
b
b
b
b
b
b
b
b
b
b
b
c
c
c
c
c
c
c
c
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
Qiam Qnit
15
15
15
15
15
16
16
16
17
18
18
10
12
14
17
22
22
31
31
6
28
31
17
1
6
13
27
30
32
32
5
1
6
2
2
2
2
2
3
3
4
5
6
6
7
7
7
7
7
7
7
8
8
8
8
8
8
8
8
8
8
8
8
11
11
11
11
16
18
18
18
18
18
18
:bl~v~l
12
15
18
13
14
32
13
13
44
22
30
9
18
13
13
14
30
22
30
19
19
20
22
23
20
23
22
33
25
30
7
7
8
9
9
9
10
8
8
8
8
9
9
9
10
10
10
10
:.2
2
5
7
9
49
7
7
8
8
9
9
843
MAHACHAQUX PHASE RDIS
Shial2~
Bim
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
:ila..::i~an:t
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
Dia.m Qnit
18
18
18
18
18
18
18
18
18
18
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
20
20
20
20
20
20
20
20
20
20
20
20
20
20
21
21
21
21
21
22
22
22
22
22
22
22
22
22
22
22
22
l:i~!l~l
10
10
12
13
13
14
15
15
16
16
4
8
9
9
10
11
11
11
13
13
13
13
13
14
14
14
15
16
16
7
7
10
10
11
11
11
12
12
12
12
12
13
13
6
8
9
10
12
15
15
15
16
16
17
17
18
18
18
18
18
Sha.t2~
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
E
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
Bim
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
~~i~ant
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
Dia.m Qnit
22
22
23
23
23
23
23
23
23
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
25
25
25
26
26
26
26
26
26
26
26
26
26
26
26
26
26
26
26
26
26
26
26
26
26
L~:sz:~l
18
19
13
18
18
19
20
20
20
10
12
14
15
16
16
17
17
18
18
18
18
18
19
19
19
19
19
19
19
19
19
20
20
20
20
6
7
8
6
6
8
10
12
14
14
15
16
16
16
16
16
16
16
16
16
16
17
17
17
17
844
MANACHAQUI: PHASE RIMS
Sbaa~
Bim
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
2
2
2
2
2
2
2
2
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
~;z;:iant
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
I:! ism
Ynit
26
26
26
26
26
26
26
27
27
27
27
27
27
27
27
27
27
27
27
27
27
27
27
27
27
27
27
27
27
28
28
28
28
29
29
29
29
29
29
29
29
29
29
29
29
29
29
29
29
29
29
29
30
30
31
31
31
31
31
32
L~vd
17
17
17
17
18
18
19
13
13
14
14
15
15
15
15
15
16
17
17
17
18
18
18
18
18
19
19
20
22
18
19
19
22
11
12
13
13
14
14
14
15
15
15
15
16
16
16
16
17
17
17
18
17
20
3
14
16
16
18
14
Shaa~
Bim
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
Va~iant ~ism
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
Ynit
32
32
32
32
32
32
32
32
32
32
32
32
32
32
32
33
33
33
33
33
34
34
34
34
34
34
34
34
34
34
34
34
34
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
36
36
36
36
36
36
36
36
36
36
36
36
:b!~:l~l
14
14
14
14
16
16
17
17
17
18
19
19
20
20
20
16
17
18
18
19
8
8
9
9
10
10
10
10
11
11
11
11
11
3
3
11
11
12
12
14
15
15
15
16
16
16
16
17
9
9
10
11
12
13
13
13
13
13
14
14
845
MANACHAQO::t PHASE R:tMS
Shsi2~
Bim va:risnt DiSlm Ilnit
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
9
9
10
10
10
10
10
10
10
11
11
11
11
11
11
11
11
11
11
11
11
11
11
12
12
12
12
12
12
12
12
36
36
36
36
36
36
37
37
37
37
37
37
37
38
38
38
38
38
38
38
38
39
39
39
39
39
39
39
39
5
25
1
4
5
6
6
28
30
1
1
1
4
4
5
5
13
18
25
27
28
30
30
1
1
1
5
5
5
6
6
I.!~:ll~J.
14
14
14
15
16
16
10
10
11
11
11
13
14
9
9
9
9
10
10
11
11
7
7
7
7
7
8
9
9
30
5
20
25
28
25
32
22
16
18
20
21
19
27
28
29
13
9
21
19
15
18
17
19
20
26
32
33
24
28
Sllsi2~
Bim YSl:tisnt Qigm Unit
B
B
B
B
B
B
B
B
B
B
B
B
B
B
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
6
6
13
13
14
14
14
24
24
24
28
28
28
28
28
30
30
30
30
30
30
31
31
31
31
31
31
32
1
1
1
3
4
4
4
4
5
5
5
5
5
6
6
6
8
16
17
19
24
24
25
25
28
30
30
30
31
31
31
31
~~:ll~J.
30
30
10
11
33
35
40
19
22
23
17
19
19
21
22
15
15
16
16
17
19
14
15
15
16
18
20
20
17
20
23
25
22
24
29
29
28
29
29
31
32
25
32
34
10
46
11
16
10
20
1
6
19
19
19
19
15
19
20
20
846
MANACHAQUI PHASE RDIS
~hai2~
Bim 1!a:r::isnt Dism :unit
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
b
b
b
b
b
b
b
13
13
13
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
15
15
15
15
15
15
15
15
15
15
15
16
16
16
16
16
17
17
17
17
17
18
18
19
19
31
36
36
4
5
5
5
6
6
6
13
13
16
17
25
25
25
27
28
28
30
30
30
30
31
31
31
37
1
1
5
6
11
12
15
16
21
25
30
4
11
11
30
35
1
14
24
30
31
6
30
5
28
5
5
7
7
7
14
19
~!;i1j!!ii1l
21
14
15
28
31
31
31
29
30
33
13
14
39
11
6
6
7
11
19
19
16
19
19
19
15
17
18
12
19
24
31
33
10
10
34
41
15
7
17
26
8
10
19
3
23
34
23
19
18
30
15
27
21
32
34
7
7
10
34
9
~hal2!ii1
Bim
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
Vs~isnt
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
J:lism Ynit
10
11
12
12
13
13
13
13
13
13
14
14
14
14
15
15
15
16
16
18
19
19
20
24
24
26
26
29
32
35
37
37
37
38
30
5
1
4
5
5
5
5
6
21
5
24
25
28
1
30
30
4
5
5
2
2
3
3
3
3
5
5
6
6
7
7
7
7
7
7
7
~!;;1V!;i1:!,
11
11
7
16
19
6
18
14
16
18
11
11
11
10
18
33
23
24
30
30
32
32
29
15
33
22
6
21
24
17
17
30
31
33
25
26
22
22
22
23
33
34
29
33
7
7
7
7
7
8
8
7
11
8
8
8
8
8
8
8
8
8
8
8
9
9
9
9
10
847
MAHACHAQU:I PHASE R:IMS
Sba:bl~
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
8
8
B
B
8
B
B
B
8
B
B
B
B
B
B
B
B
B
B
8
8
B
8
8
B
B
8
B
B
B
B
B
B
Bim Jla:z:iant 1:2iam IIDit
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
8
10
11
14
15
16
18
18
18
18
18
19
19
19
19
19
19
19
20
20
20
20
20
20
20
20
20
20
21
21
21
22
22
22
22
22
22
22
22
23
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
26
26
26
26
tt~3l:~l
12
10
9
29
29
16
4
11
14
16
16
8
12
13
13
13
14
15
6
6
7
9
9
10
11
12
12
13
10
12
14
15
16
18
18
18
18
19
19
20
7
12
12
15
16
17
17
18
19
19
20
20
20
20
21
22
6
7
8
16
Sba:bl~
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
8
8
8
B
B
8
8
B
B
8
8
8
8
B
B
8
B
8
8
8
B
B
B
B
8
8
8
8
B
B
B
8
B
B
B
B
B
B
Bim Jla:z:iant Qiam Yoit
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
26
26
26
26
26
26
26
26
26
27
27
27
27
28
28
29
29
29
29
29
29
29
29
29
31
32
32
32
32
32
32
32
33
34
34
34
34
34
34
34
34
35
35
35
35
35
35
36
36
36
36
36
36
36
36
36
36
36
36
36
~~3ld
16
16
17
17
17
17
17
18
19
18
18
19
21
18
19
14
15
15
15
15
15
16
16
19
16
12
15
16
16
17
17
20
14
9
9
10
10
10
11
11
11
8
8
10
13
15
19
4
10
11
12
12
13
13
13
14
14
14
14
15
848
MANACHAQUI: PHASE RI:MS
Shii!.l2~
Bim :slsu::iii!.nt Diii!.m lhlit
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
9
10
10
10
11
11
11
11
11
12
12
12
12
12
12
12
12
12
12
12
12
12
12
13
13
13
13
13
13
13
13
13
13
13
14
14
14
14
14
36
36
37
37
37
37
37
37
37
37
37
37
37
38
38
38
38
39
39
39
39
4
13
28
31
5
28
28
30
31
4
5
6
6
6
13
17
22
28
30
30
30
30
31
1
4
4
5
5
5
6
12
25
28
30
5
6
28
30
30
L~v~l
15
16
10
11
11
11
11
11
11
11
12
12
13
9
10
10
10
7
7
9
9
20
9
18
11
32
21
22
19
14
29
32
31
32
32
10
43
11
19
16
16
18
18
14
20
27
29
30
32
34
32
9
6
24
15
29
32
13
14
20
Shii!.l2~
Bim
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
:slii!.~iii!.nt
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
Diii!.m
14
14
15
15
15
15
15
15
16
16
16
16
16
16
16
16
17
17
~nit L~:2:~l
36
36
4
4
5
22
30
31
1
5
6
6
15
25
30
31
5
6
2
2
2
3
3
5
6
7
7
7
7
7
8
15
18
18
19
19
19
19
19
20
20
21
22
22
22
24
24
26
26
29
29
32
34
34
35
35
35
36
37
39
14
17
27
30
28
19
16
15
19
30
32
33
29
7
15
13
30
32
20
21
21
24
25
31
30
7
9
9
9
9
12
3
14
15
9
11
13
13
14
7
11
12
12
16
17
18
20
10
15
13
16
14
8
12
12
16
16
13
8
7
849
:;!hgg~
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
Rim VSl.::iSlnt D;i.sm~ Ynit;
3
b
10
1
10
3
b
25
3
b
10
30
3
b
11
16
3
11
b
25
3
12
b
5
3
12
b
14
3
b
12
17
3
b
12
17
3
12
b
22
3
b
13
1
3
b
13
5
3
b
13
6
3
13
b
6
3
b
13
19
14
3
b
5
3
14
b
30
3
b
15
31
3
16
b
5
3
16
b
6
3
18
b
4
3
18
b
5
3
18
b
5
3
b
18
6
3
b
19
22
4
2
4
3
4
7
4
19
4
20
4
20
4
20
4
22
4
24
4
24
4
24
4
25
4
26
4
26
4
26
4
27
4
27
4
27
4
28
4
29
4
32
4
32
4
32
4
32
4
33
4
34
4
36
4
36
4
36
4
36
4
38
4
39
4
39
4
10
30
4
12
31
~~:sz:~l
17
5
19
45
6
29
33
11
40
17
23
27
29
34
16
32
14
17
30
34
15
31
31
34
15
18
27
9
12
8
9
11
17
15
15
15
7
6
14
18
15
15
16
18
11
12
16
16
17
20
12
9
12
12
14
9
6
7
18
12
:;!bSla~
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
MANACHAQOI PHASE RIMS
Bim V51ri51nt QiSlm Ynit ~~:sz:~l
13
29
5
4
13
22
11
4
13
22
11
4
13
18
28
4
13
18
28
4
14
5
32
4
14
5
32
4
14
30
19
4
14
16
31
4
15
1
21
4
15
28
6
4
15
12
21
4
17
15
28
4
15
16
31
4
16
10
13
4
16
12
21
4
22
3
5
11
19
5
5
20
5
20
12
5
17
22
5
20
23
5
4
26
5
4
26
5
26
15
5
27
17
5
27
19
5
14
30
5
34
11
5
14
36
5
37
10
5
10
30
16
5
28
11
5
5
27
12
4
5
12
20
22
5
12
29
11
5
26
13
4
5
13
28
19
5
20
13
28
5
13
12
35
5
14
16
31
5
15
18
1
5
15
17
29
5
15
31
13
5
14
15
36
5
16
26
5
5
16
32
6
5
17
16
24
5
17
16
26
5
17
23
1
5
17
17
31
5
17
17
32
5
21
20
4
5
26
2
6
6
20
6
9
20
6
11
21
6
16
22
6
17
23
6
16
26
6
850
MANACBAQOJ: PHASE RIMS
Shs:!.:l;!~
Bim
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
6
6
6
6
6
6
6
B
B
B
B
B
B
B
B
B
B
B
B
B
B
6
6
6
6
6
6
6
6
6
6
6
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
8
8
8
8
8
8
8
8
8
9
9
9
9
9
9
9
9
9
9
9
9
9
9
~su:::is:!.nt
l2is:!.m Ynit
8
8
9
9
10
10
11
11
11
11
12
12
14
15
17
11
12
12
12
13
13
13
13
13
12
12
12
12
13
14
18
18
29
30
35
30
31
11
32
28
28
17
21
31
31
20
30
30
28
34
18
20
20
22
27
27
27
35
38
39
5
19
20
30
5
21
22
30
31
29
6
7
24
26
1
1
28
31
2
2
3
3
3
4
6
6
6
7
7
7
7
7
t.~~~l
11
12
11
15
12
6
10
17
19
43
11
14
18
11
14
17
20
9
14
6
13
11
18
18
20
14
11
6
27
16
12
16
31
13
16
16
14
10
30
11
18
13
18
18
15
15
19
25
17
25
25
25
24
32
33
7
8
8
9
9
Slls:!.:l;!~
Bim
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
B
B
B
B
B
B
9
9
9
9
9
9
9
9
9
9
9
9
9
B
9
Vs:!.~is:!.nt ~is:!.m
Ynit
7
7
7
7
8
8
8
8
8
10
13
14
14
18
19
19
19
19
20
20
20
20
20
20
21
21
21
22
22
23
24
24
24
24
24
24
24
24
24
24
24
24
24
26
26
26
26
26
26
26
26
26
27
27
27
27
27
28
28
28
L~~d
9
9
12
13
9
9
10
10
12
10
12
29
33
13
13
14
15
17
8
10
11
11
11
19
12
12
13
15
18
18
14
14
14
15
15
18
18
19
19
19
20
20
21
6
8
10
13
15
15
16
17
19
18
19
20
20
21
19
20
21
851
MANACRAQtn PRASE RXMS
Sh~:&!~
Bim
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
Y~;r;:ignt I;!i~m
Ilnit
29
29
29
29
29
29
29
30
30
31
32
32
32
32
32
32
32
32
32
32
33
33
34
34
34
34
34
34
34
34
34
34
34
35
35
35
36
36
36
36
36
36
36
36
36
36
36
36
36
37
37
37
37
37
37
38
38
38
38
39
~~::ll:~l
10
11
13
15
15
16
17
12
12
20
9
9
13
15
16
17
17
18
19
19
17
18
9
9
10
10
11
11
11
11
11
12
12
14
16
16
1.1
13
13
13
13
14
14
14
14
14
14
14
15
11
11
11
12
12
15
4
9
10
10
7
SbgJ;l~
Bim
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
10
10
10
10
10
yg;r;:i~nt Qi~m
10
12
12
12
12
12
12
12
12
12
12
12
13
13
13
13
13
13
13
14
14
14
14
14
14
14
14
14
15
15
15
15
15
15
15
15
15
16
16
16
16
16
17
17
17
17
17
18
18
18
19
21
22
24
Unit
~~v~l
39
30
6
24
28
28
30
30
30
31
31
31
36
4
4
6
28
28
30
31
1
5
6
30
30
30
31
31
31
1
4
6
6
12
24
24
25
31
4
14
22
25
31
1
1
7
15
31
22
20
22
17
17
19
18
19
20
14
25
29
32
19
21
15
17
23
28
24
17
17
18
18
22
23
24
27
13
11
14
28
1
22
31
5
6
6
28
2
3
3
11
11
16
19
20
20
17
29
32
33
14
22
23
24
3
8
31
32
10
19
19
9
18
28
44
19
0
14
20
24
852
MAHACHAQtn PHASE RXMS
Sbsl2~
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
Ri.m Veriant Qimn IDlit
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
11
14
15
17
18
18
19
19
19
19
19
20
20
21
21
21
22
22
23
23
24
24
24
24
24
24
26
26
27
27
27
27
27
27
27
27
28
28
29
29
29
29
29
29
29
30
34
34
35
35
35
35
36
36
36
36
36
36
37
37
w~:2:~l
9
21
3
44
13
15
13
13
13
14
14
10
10
5
9
10
15
15
16
18
12
13
13
16
17
17
6
8
14
15
15
16
16
16
17
18
15
17
10
12
12
14
14
14
14
13
8
9
9
13
14
16
9
11
12
12
12
13
9
12
Sh212~
Rim
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
Va~i.ant
Qism Ynit
10
12
12
12
13
13
13
14
14
14
14
14
14
14
15
15
15
15
15
15
15
16
16
16
16
16
16
16
16
16
16
16
16
17
17
17
18
18
18
18
19
19
22
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
38
1
6
22
28
4
5
22
6
6
13
16
22
22
30
5
6
15
28
28
30
30
5
5
5
5
6
15
22
22
28
30
30
37
16
28
32
5
5
6
12
1
5
5
1
4
6
7
13
18
19
19
22
24
29
29
30
36
36
36
L~:ll~1
9
18
22
16
15
25
30
16
29
31
11
44
13
14
14
29
31
38
12
17
15
17
28
28
31
31
27
44
15
16
17
13
16
12
40
17
17
29
31
27
9
19
29
31
23
14
32
9
11
6
16
17
13
10
7..7
7..7
19
12
l3
14
853
MANACHAQO:I PHASE RIMS
SbsU;l~
Bim Y5!.:ti"'nt I:lism IInit
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
b
b
b
b
b
b
b
b
b
b
b
a
a
a
a
a
a
a
a
10
39
37
11
11
11
12
12
12
12
12
12
12
12
12
12
12
32
6
6
13
19
19
21
24
26
29
30
30
13
13
11
13
13
13
14
14
14
14
14
14
14
15
15
15
15
16
16
16
17
17
17
18
18
18
18
7
12
13
14
16
17
18
18
19
19
28
29
16
21
22
23
25
30
34
6
6
13
29
6
19
19
11
25
31
6
13
19
21
7
19
25
13
11
1
21
25
29
4
25
11
11
12
12
19
19
22
34
37
5
22
l!~:ll~l
8
12
9
20
32
34
12
15
16
14
19
19
11
18
19
10
13
14
23
17
46
14
14
20
6
18
10
30
30
11
17
33
15
15
6
10
17
28
7
16
11
14
5
4
10
20
13
9
15
27
9
9
13
16
17
12
11
33
15
Sbsl2~
Bim
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
5
5
6
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
Y5!.~isnt
a
a
a
a
a
a
a
a
a
a
b
b
b
b
b
b
b
b
b
I:lism Unit
12
12
13
14
14
16
16
18
19
19
12
12
13
13
14
18
11
11
11
11
12
12
12
14
14
12
28
26
36
18
19
32
12
26
4
4
13
6
32
35
6
20
30
35
35
4
7
36
39
39
20
22
32
36
30
30
18
30
31
17
30
1
2
4
8
18
20
20
21
22
23
24
26
27
27
32
32
35
11
11
13
13
14
15
15
15
16
16
1
1
25
3
6
12
6
14
l:i~:ll~l
17
15
18
17
17
10
19
29
30
33
20
14
35
13
20
18
14
26
9
14
7
10
12
18
15
14
16
16
4
16
17
46
16
21
23
27
10
8
12
12
15
13
17
17
15
17
18
l3
16
17
7
20
21
10
23
27
9
34
36
854
MANACHAQU:I PHASE R:IMS
Sbs:!.:g~
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
Bim :llii!..::is:!.Dt I:liam Ilnit
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
3
3
3
3
17
18
18
19
21
22
23
23
26
12
13
13
14
18
18
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
5
12
13
14
14
15
17
17
14
14
15
16
18
19
20
20
20
15
17
18
10
14
32
4
31
5
5
13
6
22
27
29
32
32
32
38
39
28
20
29
28
24
37
26
26
26
27
34
39
36
5
5
24
1
27
39
7
8
11
39
19
31
1
25
31
6
6
8
8
18
24
24
24
27
34
38
38
5
36
24
L~:ll:~l
11
31
14
26
18
30
30
12
27
15
19
10
15
16
17
8
8
7
10
11
16
21
9
6
10
13
15
9
8
12
27
28
17
20
19
8
9
9
6
7
16
17
21
6
20
35
27
8
8
15
15
15
19
18
12
9
9
28
9
15
Shag~
Bim :llii!..::is:!.nt I:liam Unit
D
D
D
D
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
5
5
6
6
1
1
1
1
1
1
2
2
2
2
2
2
3
3
3
3
3
4
4
4
5
5
5
6
6
7
18
18
10
10
10
14
14
18
11
11
12
14
13
14
16
18
14
16
16
16
16
18
22
18
24
24
5
34
34
34
5
32
35
22
6
23
26
31
31
28
11
3
16
26
17
31
14
25
31
38
25
27
16
30
I.i~:ll:~l
15
15
31
12
11
11
31
14
14
14
34
20
17
18
17
22
7
26
45
21
41
17
42
6
19
10
0
20
44
17
855
stJITACOCHA PHASE RIMS
Sluu;~~
Rim
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
14
14
14
14
14
14
14
14
14
14
14
14
14
14
15
15
15
15
15
15
15
15
16
16
16
16
16
16
16
17
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
:..1
11
11
11
~~ism.t
t!iam IInit
10
10
10
11
11
14
10
13
13
14
14
11
12
12
12
17
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
4
4
18
19
22
24
26
36
5
6
27
1
1
24
5
30
30
11
12
27
30
30
7
24
36
35
22
31
36
25
1
1
2
2
3
3
4
5
5
5
6
6
6
6
7
7
7
7
8
10
11
11
11
11
11
11
11
11
15
15
I.!~v~l
20
20
8
6
12
12
8
9
23
25
14
19
19
12
21
10
10
8
4
10
10
10
6
8
7
11
9
10
8
5
20
21
21
25
21
22
19
21
22
22
25
26
26
26
4
4
7
7
7
7
2
4
4
6
8
8
8
9
18
30
Sha:r2~
Bim
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
3
3
B
B
B
B
B
B
3
3
3
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
:sla~i.Sl.nt
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
t!iam IInit
18
18
18
18
18
18
18
18
18
18
18
18
18
18
18
18
18
19
19
19
19
19
19
19
19
19
19
20
20
20
20
20
20
20
20
21
21
21
21
21
21
21
21
21
21
21
21
22
22
22
22
22
22
22
22
22
22
23
23
23
L~v~l
6
6
6
7
7
7
7
8
8
8
8
9
10
10
11
11
12
8
8
8
8
9
9
10
10
11
12
6
6
7
7
8
8
10
12
6
7
7
7
8
8
8
8
9
11
11
11
8
9
10
11
11
11
12
12
13
13
14
16
17
856
SUXTACOCHA PHASE RD!S
Sh5!.!2~ Bim YS!.;(iS!.nt I:2iS~.m Qnit ~~v~l
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
E
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
23
24
24
24
24
24
24
24
24
24
24
24
24
24
24
25
26
26
26
26
26
26
26
26
26
27
27
27
27
27
27
29
29
29
29
29
29
29
29
29
29
30
31
31
32
32
32
32
32
32
32
32
33
33
34
34
34
34
34
35
20
1
9
9
9
10
10
11
11
12
12
12
12
12
13
5
5
6
7
7
8
9
9
9
10
11
11
12
12
12
14
7
10
10
10
11
11
11
11
11
13
10
13
14
8
10
11
12
12
12
13
14
14
15
7
7
7
7
7
8
Sh5!.!2~ R;i.m
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
Va;riS~.nt
D;i.gm Qnit I.!~v~l
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
7
8
9
9
9
9
9
10
10
10
10
10
10
10
10
10
10
10
11
11
11
11
35
35
35
35
35
35
35
35
35
35
36
36
36
36
36
36
36
36
36
37
37
37
37
37
38
38
38
38
38
38
38
38
38
38
38
38
39
39
12
16
1
5
5
30
31
28
28
28
30
30
30
30
31
31
31
31
4
4
5
5
9
9
9
10
10
11
11
11
11
12
8
9
9
10
10
10
10
12
12
6
7
7
7
8
6
8
8
8
8
8
8
8
8
8
9
9
6
6
10
2
15
24
25
13
11
13
13
13
11
12
13
14
11
12
13
13
22
22
22
22
857
~hga~
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
Rim yarignt Diam ];lnit
5
11
a
11
6
11
11
a
6
11
11
a
6
11
11
a
14
11
a
11
28
11
a
11
30
11
a
11
31
11
a
11
31
11
a
11
12
1
11
a
6
11
a
12
6
11
12
a
25
11
12
a
28
11
a
12
28
11
a
12
30
a
12
11
31
12
11
a
31
12
11
a
31
12
a
11
31
a
12
11
12
31
a
11
31
12
a
11
31
12
11
a
4
13
11
a
4
13
11
a
4
11
a
13
12
a
13
11
30
13
a
11
31
13
a
11
31
a
13
11
4
14
a
11
4
14
11
a
14
4
a
11
14
4
a
11
4
14
11
a
14
5
11
a
24
14
11
a
28
14
11
a
30
14
11
a
30
11
a
14
14
31
a
11
12
15
11
a
2
11
b
2
b
11
4
b
11
6
b
11
6
b
11
10
b
11
10
11
b
18
b
11
b
18
11
18
11
b
18
11
b
18
b
11
19
b
11
19
b
11
19
b
11
20
11
b
20
b
11
21
b
11
L~v~l
25
26
27
27
36
12
11
11
13
22
25
25
6
11
12
11
10
11
13
13
13
13
14
22
23
23
11
11
11
14
19
22
22
22
22
23
16
14
10
11
11
7
19
23
15
17
25
5
5
7
8
12
13
13
8
9
10
7
7
3
~hga~
B
3
3
3
B
B
3
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
StJ:t'l'ACOCBA PHASE R:tMS
Rim Vsrignt Digm J;lnit L~v~l
b
22
9
11
22
10
b
11
22
11
b
11
22
11
b
11
12
b
22
11
14
22
b
11
b
23
13
11
23
14
b
11
24
3
b
11
5
b
24
11
7
b
24
11
24
9
b
11
24
9
b
11
3
b
26
11
3
b
26
11
26
9
b
11
10
26
b
11
27
9
b
11
10
27
b
11
27
13
b
11
7
29
b
11
8
b
29
11
8
b
29
11
10
b
29
11
29
11
b
11
11
29
b
11
11
29
b
11
13
29
b
11
2
30
b
11
2
30
b
11
1
32
b
11
7
32
b
11
9
b
32
11
9
32
11
b
9
32
b
11
10
32
b
11
10
32
b
11
11
32
b
11
12
32
b
11
13
32
b
11
12
b
33
11
8
b
35
11
10
b
35
11
10
35
b
11
7
37
b
11
37
8
11
b
37
8
11
b
37
11
b
11
23
8
6
b
11
25
6
b
9
11
7
28
b
9
11
7
28
b
9
11
24
10
6
b
11
8
21
b
10
11
15
23
b
1.0
11
24
3
b
10
11
4
24
b
10
11
9
10
30
b
11
11
10
31
b
11
11
10
31
b
11
858
StnTACOCHA PHASE RDIS
Sh5ll2~
B
B
B
B
8
8
B
B
B
B
B
B
3
3
B
B
3
3
3
3
B
3
B
3
3
B
8
B
B
B
B
B
B
3
B
B
B
B
B
B
3
3
3
B
B
B
3
3
3
3
3
3
3
3
3
B
B
B
B
3
Bim
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
~;r.:i5lnt
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
;Qis;!,m Ynit
10
10
11
11
11
11
11
11
11
12
12
12
12
12
12
12
12
12
12
12
12
13
13
14
14
14
14
14
14
14
14
15
31
32
3
4
5
6
24
31
35
6
6
6
18
30
30
30
31
31
31
31
31
3
37
3
4
6
28
30
31
32
32
24
2
4
6
6
6
6
6
6
7
8
12
17
18
19
19
19
20
20
20
20
20
21
21
22
23
24
24
26
l:!~:!.!:~l
12
8
22
15
24
25
9
11
10
24
27
27
12
10
12
16
7
8
9
10
11
22
6
22
20
25
6
6
9
10
10
11
18
18
2
18
23
24
25
25
3
6
9
10
7
4
8
8
4
4
5
5
8
4
9
11
13
7
7
4
Sh2l2~
Bim
B
B
B
11
8
8
8
8
8
8
8
8
8
B
8
8
8
B
8
8
8
B
B
B
8
8
8
B
B
B
8
B
8
B
8
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
8
B
B
B
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
~~i2nt
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
d
d
Oi2m :unit
7
8
8
9
9
9
10
10
10
10
10
10
10
10
10
11
11
11
11
11
11
11
12
12
12
12
12
12
13
13
13
13
13
13
14
14
14
15
15
15
15
27
29
29
29
30
30
30
32
32
32
34
35
37
37
37
38
38
22
24
28
5
17
31
4
6
14
19
21
28
28
30
30
4
4
6
28
28
28
35
3
4
4
28
30
30
4
4
4
4
4
25
4
16
20
4
5
6
15
3
6
J:t~v~l
10
6
10
10
8
10
11
9
9
9
6
8
7
8
11
6
7
12
8
13
23
5
11
19
25
33
7
5
6
10
10
11
15
23
27
5
10
11
12
14
19
23
12
9
10
19
20
20
22
23
5
21
39
9
20
26
25
17
21
23
859
Sh~:b!!i:
Bim
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
11
11
11.
11
11
11.
11.
11.
11
11
11
11.
11
11.
11.
11.
B
B
B
B
B
B
B
B
B
B
B
B
B
B
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
:L~;r;:i~nt
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
e
e
e
e
e
e
e
e
e
e
Qigm IDlit
8
8
8
10
10
10
10
10
11
11
11
12
12
12
12
12
12
12
12
12
13
14
14
14
14
16
l:!!i:~!i:l
10
19
20
21
23
23
24
24
24
24
24
24
27
29
30
31
32
32
34
35
36
38
38
39
6
30
8
8
4
6
13
14
6
7
7
7
9
9
12
8
8
16
10
13
7
7
8
7
8
6
26
12
31
11
5
6
6
31
37
6
23
23
25
11
31
5
6
6
6
24
24
30
31
31
5
5
6
6
30
1
2
3
6
7
11
18
19
20
22
22
11
7
25
4
13
23
23
24
26
10
10
11
11
14
24
21
25
25
11
18
18
21
25
7
2
5
8
7
8
13
Sha:12~
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
SUITACOCHA PHASE RIMS
Bim Va;r;:ignt Qigm :unit l::.!!i:~!i:l
1.1
1.1.
1.1.
1.1.
11
1.1.
1.1.
1.1
1.1.
1.1.
1.1.
1.1.
1.1
1.1
11
11
1.1.
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
11
1.1
1.1
1.1
1.1
1.1.
1.1
11
11
11
11
1.1
1.1
11
1.1
1.1
1.1
11
11
11
11
1.1
11
11
11
11
11
1.1
11
11
11
11
e
e
e
e
e
e
e
e
e
e
e
e
e
e
e
e
e
e
e
e
e
e
e
e
e
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
g
g
g
g
g
g
g
8
9
10
10
23
32
32
34
36
36
36
36
36
36
37
38
38
39
31
4
30
31
11
4
12
12
12
12
12
15
3
14
31
36
36
6
4
12
18
19
20
20
21
21
21
21
27
35
36
6
31
5
9
9
10
10
11
12
12
12
13
13
13
13
13
14
16
11
31
6
28
31
4
4
28
31
37
30
6
3
5
6
19
19
22
22
13
9
10
7
8
9
11.
11.
11.
11.
6
7
8
4
12
19
9
13
19
21
21
13
11
11
29
24
9
8
7
5
5
5
6
11
11
11
10
10
24
10
24
2
11
27
10
11
22
23
12
13
9
11
27
16
25
24
6
8
8
9
860
stnTACOCBA PHASE R:tMS
Sh~:b!Sl
B
B
B
B
B
3
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
Bim
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
13
13
13
13
13
13
13
13
13
13
13
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
~.:i~nt
g
g
g
g
g
g
g
g
g
g
g
g
g
g
g
!:!ism unit
9
9
10
11
11
12
12
10
12
12
14
14
15
15
15
18
10
10
11
12
12
14
15
22
22
22
22
27
27
39
39
4
11
4
4
6
11
31
2
22
22
23
23
23
31
22
6
30
19
30
5
22
30
30
21
23
36
36
13
31
31
20
31
31
6
4
13
15
18
18
19
19
19
22
22
22
24
24
27
27
27
32
32
LS1~~l
10
10
10
15
8
10
6
6
20
6
20
23
24
5
9
18
8
8
9
9
12
11
8
26
10
7
11
25
9
10
9
9
20
9
11
9
12
12
9
13
12
26
22
9
33
4
10
9
9
9
3
8
10
9
12
9
11
12
8
10
Sh~:b!~
Bim
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
8
8
8
8
8
8
8
8
8
9
9
9
9
9
9
9
10
8
8
8
8
8
8
8
9
9
9
9
9
9
10
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
F
F
F
F
F
F
F
F
F
:llsu:i~nt Di~m
9
11
11
12
12
13
13
13
14
14
14
15
16
16
14
14
14
15
15
20
14
14
16
16
17
16
11
13
18
18
8
11
12
27
12
11
1
1
1
1
1
1
1
1
1
a
a
a
a
a
a
a
a
a
Ynit
32
33
34
35
35
4
4
24
18
28
4
4
22
18
22
28
24
22
24
21
21
27
6
20
21
6
21
28
23
24
24
27
17
28
27
34
3
24
28
18
14
24
28
18
22
22
5
30
30
22
20
2
3
3
3
3
3
5
11
11
:b!~VS1l
11
13
7
10
10
22
23
8
8
11
17
17
14
6
11
7
12
13
11
8
8
13
18
9
9
27
9
14
10
6
11
10
11
10
8
7
20
9
9
4
33
9
10
8
12
13
23
19
11
11
4
21
18
18
20
20
22
23
3
3
861
SUZTACOCBA PHASE RIMS
Sba:g~
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
Bim Va;tiant Qis.m JJ:nit
a
11
1
11
1
a
a
11
1
11
1
a
14
1
a
a
14
1
15
a
1
a
18
1
a
19
1
19
1
a
19
1
a
19
1
a
1
a
19
19
1
a
19
1
a
20
1
a
1
a
20
20
1
a
20
1
a
20
1
a
a
20
1
a
21
1
21
1
a
21
1
a
21
1
a
21
1
a
21
1
a
21
1
a
21
1
a
21
1
a
a
21
1
22
a
1
22
a
1
a
22
1
1
a
22
24
1
a
26
1
a
1
a
26
a
26
1
27
1
a
27
1
a
27
1
a
27
1
a
27
1
a
29
1
a
1
a
29
1
30
a
1
32
a
1
32
a
32
1
a
1
32
a
33
1
a
1
a
33
33
1
a
1
a
34
1
a
34
1
a
34
1
34
a
1
a
34
34
1
a
l:!~:lld
4
5
8
10
12
34
26
9
3
8
8
9
9
9
10
4
5
5
7
9
10
4
5
5
7
8
8
9
9
9
9
11
11
12
12
12
3
7
8
6
11
12
12
13
10
12
10
10
10
10
11
15
15
17
4
7
7
7
7
9
Sbal2~
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
Birn Va.:iant I2iam J;.Init
35
a
1
35
a
1
35
a
1
36
a
1
36
a
1
36
a
1
37
a
1
a
38
1
a
39
1
7
5
a
1
28
a
8
1
1
a
9
1
16
a
9
1
a
10
5
1
11
a
10
1
26
a
10
1
10
30
a
1
10
31
a
1
32
a
10
1
1
a
11
1
3
a
11
1
4
11
a
1
11
13
a
1
17
11
a
1
25
a
11
1
25
a
11
1
4
a
12
1
12
a
12
1
13
a
12
1
25
a
1.2
1
12
30
a
1
30
12
a
1
4
13
a
1
28
a
13
1
24
14
a
1
30
a
14
1
21
a
15
1
21
15
a
1
34
15
a
1
30
a
16
1
2
b
1
2
b
1
6
b
1
14
b
1
18
b
1
19
b
1
21
b
1
22
b
1
23
b
1
26
b
1
27
b
1
32
b
1
35
b
1
36
b
1
26
b
10
1
28
b
10
1
1
11
b
1
17
11
b
1
22
b
11
1
28
b
11
1
l:!~v~l
11
11
12
8
8
9
3
8
7
24
14
18
40
23
7
8
11
12
13
18
20
19
6
6
1
5
21
8
4
10
11
20
13
12
12
8
8
7
10
21
24
26
34
7
5
8
7
13
7
12
10
10
8
8
12
16
11
8
12
862
SUXTACOCBA PHASE RZMS
Sh5!.9~
Bim Vs;t:iS!.nt Qism Ynit
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
"E'
b
b
b
b
b
b
b
b
b
11
12
12
13
13
14
14
15
17
31
6
15
27
27
5
6
26
30
2
3
3
3
3
6
6
11
11
11
15
15
15
15
18
18
19
19
19
19
19
19
19
22
23
24
24
24
26
27
27
27
27
27
27
27
27
29
29
29
29
32
33
36
36
36
36
36
36
36
36
L~v~l
12
24
32
10
10
24
24
12
10
12
15
18
18
18
24
24
3
6
8
4
6
6
9
4
7
7
7
8
8
8
8
11
9
10
7
8
8
9
9
10
10
10
11
11
12
12
9
9
10
11
8
12
7
7
7
7
8
8
8
9
Sh5!.!2~
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
Bim
~.-isnt
Dism Unit
2
2
2
2
2
2
2
2
2
2
2
2
5
7
7
9
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
9
10
10
11
11
11
11
12
12
12
13
13
14
14
15
10
10
12
12
12
13
13
13
14
14
14
14
14
18
18
18
18
a
a
a
a
a
a
36
37
37
37
37
38
39
39
39
3
15
32
11
36
6
24
4
20
31
36
6
30
36
1
29
13
15
28
2
4
22
22
24
24
24
28
38
22
31
24
24
35
4
4
25
5
5
11
11
28
24
24
36
36
19
L~Y:~l
9
4
4
7
10
8
5
6
6
18
29
10
6
8
25
8
20
5
11
8
24
11
7
18
4
8
30
12
2
23
11
11
5
11
11
12
6
15
14
12
12
9
20
20
23
23
10
10
12
9
9
12
12
9
7
9
9
24
4
4
12
8
22
22
10
13
9
22
863
SUITACOCHA PHASE Rms
Shiil:bl~
Bim JlS!.;r;:iiilnt !:liiilm Ynit
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
4
4
4
4
5
5
5
5
5
5
6
6
6
6
6
6
7
7
7
7
8
8
8
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
3
3
3
3
a
14
b
b
b
10
12
10
15
13
13
13
15
21
12
15
15
9
10
10
11
11
12
12
12
12
12
12
13
14
11
12
18
18
18
20
11
12
14
30
2
31
35
3
29
34
37
30
29
24
35
6
19
19
29
19
27
37
27
31
6
16
19
21
22
22
22
22
22
22
28
28
31
36
4
4
4
4
20
17
28
31
31
31
32
3
28
32
35
3
30
5
30
31
25
17
3
4
14
l!~vd
10
2
14
9
13
9
12
6
11
9
8
11
26
10
10
11
8
11
5
11
13
29
3
4
6
7
8
8
8
10
11
4
8
8
6
17
20
21
17
7
10
11
9
11
11
8
15
10
9
8
20
10
24
11
2
1
7
21
21
42
Sbiil:bl~
Bim V5l;r;:i5lnt Q;i.gm Unit
X
X
X
X
X
X
X
X
X
3
3
4
4
4
5
5
6
7
15
15
20
20
12
16
19
28
28
6
6
6
35
31
13
13
~:t~v~l
12
12
25
26
27
10
9
1
9
864
COLPAR PHASE R:IMS
&haoe
BUl
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
B
8
B
8
8
8
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
E
.:.
E
E
.:.
.:.
E
E
E
E
E
E
12
12
12
12
12
12
13
13
14
15
3
3
3
3
9
9
9
12
12
12
12
12
VI.;J;:i.Ut
h
h
h
h
h
h
h
h
h
h
h
h
h
Dig JZnit LeV!!l
10
12
12
12
16
i
i
i
i
i
10
10
12
j
j
j
j
16
16
12
13
14
16
18
19
12
13
14
16
18
12
14
14
16
8
20
21
29
29
32
32
32
12
6
27
27
5
4
37
4
6
32
32
38
6
6
3
24
24
27
29
36
38
38
38
38
13
28
31
7
36
3
22
27
32
2
26
19
19
20
13
5
19
20
22
28
30
30
31
24
29
36
36
36
6
6
7
6
10
6
10
10
10
22
7
7
20
19
6
22
23
8
3
8
21
22
15
10
11
9
8
6
6
6
6
6
8
9
9
4
8
20
8
4
8
20
5
10
7
4
4
20
6
3
8
12
9
9
9
8
7
5
7
8
&ha:ge Rim varigt Diam
E
E
E
E
E
E
E
E
E
E
E
12
12
12
12
12
12
12
12
14
14
15
15
16
16
17
20
13
13
14
17
11
Jlnit
28
29
17
28
28
29
1
21
20
13
14
~§vel
10
6
12
9
9
9
19
3
7
9
:..7
865
EMPEDRADA PHASE R:IMS
ShS!.I2Sl Bim ~ii!.;~;:iS~.nt I2iS!.rn Ynit l:iSl:ll:Sll
3
19
3
E
3
18
1
E
3
2
21
E
2
3
21
E
3
5
21
E
5
3
38
E
16
3
12
4
E
3
5
13
13
E
3
32
15
15
E
17
3
8
16
E
3
17
2
29
E
14
9
E
1
16
9
1
E
17
9
E
1
18
9
1
E
9
18
1
E
22
9
2
E
15
9
3
E
19
9
3
E
18
9
4
E
20
9
5
E
18
9
6
E
20
9
6
E
21
9
6
E
12
9
15
E
15
9
15
E
20
9
E
15
23
9
15
E
26
15
E
9
8
18
E
9
5
19
E
9
5
E
9
19
2
9
20
E
20
3
E
9
4
9
E
20
6
9
21
E
7
9
21
E
9
1
E
22
9
22
1
E
9
23
5
E
9
5
24
E
1
9
E
26
1
9
26
E
27
3
E
9
27
3
E
9
8
29
E
9
8
E
9
29
9
4
E
34
6
9
34
E
9
8
E
35
9
6
E
36
7
9
E
36
E
9
39
5
7
17
E
9
9
7
9
E
28
9
E
9
10
2
15
9
10
16
E
3
E
9
11
20
5
9
16
E
12
2
9
12
21
5
E
Sbii!.I2Sl Bim ~ii!.:dS~.nt I2ism! :unit l:iSlV§l1
5
12
21
E
9
5
32
12
E
9
23
13
6
E
9
3
13
11
E
9
3
13
24
E
9
5
13
26
E
9
7
13
28
E
9
14
19
3
E
9
20
14
4
E
9
10
14
23
E
9
5
14
28
E
9
9
14
28
E
9
22
16
6
E
9
3
16
19
E
9
3
16
19
E
9
5
16
24
E
9
15
4
18
E
9
27
4
18
E
9
27
4
18
E
9
4
27
18
E
9
17
8
20
E
9
7
20
30
E
9
22
6
E
15
8
15
E
15
2
20
E
15
2
20
E
15
2
20
E
15
2
20
E
15
5
35
E
15
7
35
E
15
30
14
6
E
15
23
8
5
E
15
1
25
9
E
15
3
24
10
E
15
18
11
2
E
15
20
11
2
E
15
22
12
2
E
15
17
12
31
E
15
18
13
2
E
15
5
15
13
26
E
14
14
3
E
15
9
14
11
E
15
3
14
18
E
15
8
14
23
E
15
8
14
28
E
15
8
13
E
15
16
15
19
6
E
15
9
19
13
E
15
18
20
4
E
15
20
2
E
16
23
2
E
16
15
3
E
16
10
8
E
16
1
20
E
16
12
22
E
16
7
24
E
16
20
E
16
12
5
21
5
16
12
E
3
13
21
E
16
28
14
E
13
16
866
Shs:b!~
Bim
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
16
16
16
17
17
17
17
17
17
17
17
17
17
18
18
18
18
18
18
18
18
18
19
19
19
19
19
19
20
20
20
20
21
21
21
21
22
22
22
23
23
24
24
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
~:r::iant
I:lism Ynit
15
17
20
8
12
35
1
20
19
19
19
20
20
34
16
5
32
14
6
15
21
34
25
31
18
1
13
5
23
36
4
17
13
13
10
14
14
14
32
28
28
28
21
21
21
17
6
12
6
6
6
19
19
4
7
13
15
18
18
22
22
22
23
23
23
24
24
24
24
26
15
16
16
18
11
13
14
17
23
30
13
20
I.!~vd
2
14
1
5
7
16
5
5
4
45
20
6
21
22
22
4
6
6
8
4
19
12
17
3
3
18
10
5
6
7
8
8
2
2
5
5
22
5
22
21
21
4
4
18
4
4
25
4
5
3
3
3
4
4
7
3
6
6
6
3
~hii!::b!~
BMPEDRADA PHASE RIMS
Rim Vs;r;:iii!:nt Digm Ynit L~v~l
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
27
27
31
9
10
10
10
10
11
12
12
13
14
15
16
17
18
18
19
20
21
22
24
24
11
12
12
31
32
32
33
34
35
36
36
36
37
38
38
38
39
39
6
30
30
31
32
31
28
31
35
31
22
24
28
17
34
18
36
4
5
18
18
3
6
6
20
22
22
23
24
26
27
29
32
35
35
36
37
38
39
4
27
31
3
6
7
9
7
13
12
3
6
6
7
7
5
5
5
5
3
4
17
4
6
6
7
4
6
3
5
2
4
4
7
6
3
6
8
18
18
6
8
20
22
29
4
1
7
10
6
3
B
9
6
7
7
3
5
5
3
17
7
6
867
EMPEDRADA PHASE RDIS
Shiil:b!~
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
Bim :sliil:tiiilDt l:liiilm Unit
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
13
13
14
14
14
15
15
15
15
16
16
17
17
17
17
17
18
18
19
19
22
23
12
13
14
15
16
20
21
11
11
13
13
13
14
16
20
22
24
15
26
29
17
27
27
35
27
27
2
16
24
27
36
20
30
1
16
7
14
19
19
19
19
19
19
19
22
27
27
27
29
29
34
23
12
19
19
30
23
22
19
29
29
38
5
24
28
28
31
35
25
18
2
8
8
8
20
~~:ll~l
6
6
3
2
2
12
8
10
8
7
7
14
4
4
10
7
7
5
20
39
4
14
4
4
4
4
4
4
4
5
5
6
6
3
6
4
9
4
8
7
7
7
6
4
6
6
1
21
4
4
7
8
6
5
16
3
4
5
7
Shiil:b!~
Bim Va:tiiilDt I2iiilm Ilnit
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
5
5
5
6
6
6
6
7
7
7
8
8
9
10
11
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
3
4
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
15
15
20
12
11
19
7
12
10
13
14
14
15
16
25
11
10
6
11
7
3
21
21
39
18
34
3
34
15
4
11
13
1
18
22
23
24
29
29
34
34
34
34
37
37
36
28
27
28
31
28
1
5
1
I.!~v~l
20
4
4
14
4
4
3
4
4
21
4
10
20
5
4
16
6
9
10
6
6
18
4
4
4
6
4
4
7
7
7
9
8
6
13
19
14
APPENDIX F
REPORT ON BOTANICAL REMAINS
868
869
ANALYSIS OF CHARRED BOTANICAL REMAINS FROM
MANACHAQUI CAVE, PERU
by
Deborah M. Pearsall
Universicy of Missouri
10-95
INTRODUCTION
This reporc presencs che results of analysis of charred botanical
materials recovered from Manachaqui Cave by Warren Church during
excavations in 1988 and 1990.
of three chambers.
The cave is a boulder cluster comprised
It is located at 3,625 m elevation near the eastern
edge of the subalpine grasslands of the northeastern Peruvian Andes
between the Maranon and Huallaga rivers, on the crest of the divide
(Church 1993).
The study area corresponds to the districts of Pataz (La
Libertad Department) and Huicungo (San Martin Department), and is
located jusc to che west of che Rio Abiseo National Park.
The cave is
in the paramo (wee Andean grassland) vegetation formation; montane
forest is located nearby and may have been closer to the cave in the
past, although the area is thought to have been paramo during the period
the cave was used.
Manachaqui Cave was utilized at various times from the Late
Preceramic (2200
(A.D. 1532).
b.c., uncalibrated)
to the end of the Late Horizon
Table 1 summarizes the C-14-based chronology and
assignment of excavation levels to cultural phases and periods.
hiatuses
may exist in the occupation, between 500
(beginning of Early Intermediate), a.d. 200
a.d. 700 and
Three
b.c. and 200
b.c.
and a.d. 400, and between
A.D. 1470 (beginning of Late Horizon).
Given the location of the shelter on the Maranon-Huallaga divide,
the fact that a prehistoric road passes by it, and the modern function
of the cave as a shelter for travelers, Church has proposed that
Manachaqui Cave functioned as a wayside station during prehistory,
rather than exclusively as a ritual site or hunting or herding camp.
I
870
will therefore examine the charred botanical materials recovered from
the site in light of this hypothesis, to determine if data support this
function for some or all of the cultural periods represented at the
site.
other
Further, the date of the appearance of corn, and the presence of
~exotic"
(i.e., non-paramo) resources, may shed light on regional
interactions during the periods the cave was used.
I am aware of no ethnoarchaeological or experimental data that
would help in framing botanical correlates of the hypothesis or
alternative explanations of cave function (most work of this type has
focused on crop processing, e.g., Hillman (1984).
However, discussing
the ways in which botanical materials become deposited in a site such as
Manachaqui may help in setting up a framework to test the hypothesis.
The first point co emphasize is that only those botanical materials that
became charred survived in this environment.
This included plants used
as fuels for cooking or heating fires, edible plant parts lost during
cooking, inedible plant parts burned as refuse, and any plant material,
food or nonfood, accidentally burned (bedding catching fire,
example) .
for
Any food items not cooked in the shelter are very unlikely to
become part of the deposit.
This would include "trail foods• prepared
elsewhere and consumed in the cave, as well as any foods consumed raw.
This may limit the utility of the macroremain data to address the
hypothesis of cave function.
Phytolith analysis, which can identify
plants that decay without leaving charred remains
(Pearsall 1989;
Pearsall and Piperno 1993), would be an interesting addition to the
present study.
As detailed in Pearsall (1988), there are a variety of avenues
through which botanical materials become charred and deposited in sites:
(1) wood may be collected and burned as fuel or used in construction or
tool manufacture and later burned;
(2) seeds may be gathered for food
and accidentally charred during cooking, brought
in dung burned as
fuel, brought in incidentally with root or stem foods and discarded as
the waste portion, or brought in with plants for nonfood purposes and
871
discarded as the waste portion; and (3) roots may be roasted as foods or
included in sod burned as fuel.
Peru,
At Panaulauca cave in the Junin puna of
for example. the source of many of the small seeds
recovered in the deposits was camellid dung fuel.
(<
2.0 mm)
Small. medium, and
large grass seeds. Cyperaceae, Sisyrinchium, and Calandrinia seeds were
observed in dung and therefore interpreted as nonfood items at this site
(Pearsall 1988).
It is thus important to look not only at what is
likely to become charred. but at the source(s) of those materials, since
not all "edible" remains may be foods.
It is not always possible to determine whether charred plant
remains are foods, or should be considered to represent some other
function.
Such is the case with Festuca, or ichu grass.
This useful
plant may have been used for bedding or for fire starter at the cave.
However, ichu also produces large grains that may have been an
attractive local food source.
Both interpretations will be presented
below.
Finally,
it is important to realize that not all charred materials
deposited in a site will survive or be recovered during excavation.
Fragile root remains, for example, are likely to be underrepresented
relative to more robust corn cob fragments or woody fruit rinds.
Any
food with inedible parts likely to be disposed of in the fire is likely
to be better represented than foods wholly consumed.
Smaller materials,
like the seeds of many annual plants, may be underrepresented due to the
size of screen or flotation technique used in recovering remains.
Given these considerations. the environment of the site, and the
time periods in question, the following ethnobotanical indicators might
be helpful for distinguishing use of the cave as a wayside station from
use as a residential camp for hunters/gatherers/herders:
Use of camelid dung fuel
source):
(and presence of small seeds from this
more abundant at a camp; less at a wayside shelter (sheltered
pack animals would produce little dung; a
concentrated quantities).
~anaged
herd much larger,
This indicator cannot be used in the
872
Manachaqui case, since few small seeds were recovered (see below) .
local food items:
more use at a permanent or seasonal camp; less
use at a shelter.
"exotic" food items:
might be present in either case, but would
be more abundant at a wayside shelter, less so at a habitation site.
abundance of food remains:
more abundant relative to amount of
fuel burned at a habitation site; less abundant at a wayside
shelter--i.e., difference in frequency of cooking fires (more accidental
charring of foods) versus fires for warmth (less accidental food
charring) .
In the case of corn, evidence for cooking would be indicated
by presence of kernels {lost during parching, for example), rather than
just the presence of the inedible cob, which could become deposited
either after in-situ food preparation or consumption of a "traveler's
meal" of corn still on the cob.
.
diversity of botanical remains:
low diversity at a wayside
shelter, higher diversity at an occupation site (i.e., following Yellen
1977, specialized versus generalized activities; illustrated in Pearsall
1983 for Pachamachay Cave).
This measure cannot be used in the
Manachaqui case due to the small quantities of smaller seeds recovered.
Richness (number of taxa present) will be discussed briefly.
One would expect local wood species to dominate in both situations,
since it is unlikely travelers would carry fuel.
EXCAVATION AND FIELD BOTANICAL PROCESSING PROCEDURES
Manachaqui cave is actually three small "shelters• created by a
cluster of boulders.
Shelter M-lA, the largest and most intensely used,
is the source of the samples discussed here.
was excavated,
Approximately 35% of M-lA
to a maximum depth of 2 m (Church 1993).
Excavations
proceeded in 5 em levels, in 39 1 x 1 m units placed in the interior
(Sector A) and berm (Sector 8) of the shelter.
Soil for botanical
recovery was taken from Sector A's floor deposits in Unit 15 , and from
Sector A hearths in units 14, 15, 16 and 17 , and Sector B's Unit 31,
873
for a total of 83 samples.
To reduce sample weight and volume for
transport from the site, larger soil samples were wet-screened in the
field using 1/16" mesh, and sherds, lithics, and so on were removed in
the Trujillo laboratory.
were recorded.
Weight and volume of the "reducedw samples
Residues were later chemically floated using zinc
chloride to recover botanical rnacerials, which were sent to che
University of Missouri Paleoechnobotany laboratory for analysis.
LABORATORY PROCEDURES
Because budget constraints made it impossible to analyze all 83
samples, the following sampling strategy was employed (refer to Table
1):
all samples (floors and hearths)
from the preceramic Lavasen and
Initial period Manachaqui levels were sorted (14 precerarnic, 4
Manachaqui), all samples from the Early Horizon Suitacocha period were
sorted (4), one sample from each of the Early Intermediate Colpar and
Empedrada floors was sorted, with the exception that 2 from W, 2 from N,
and none from 0 were done (18 total), and two samples were selected from
among the Late Horizon Poblano floors
(F, B).
The sample total was 42.
The large quantity of charred wood in some of the individual
samples also necessitated subsampling:
13 samples were split using a
riffle sampler (Pearsall 1989), and 25% or 50% of the sample analyzed.
Sample sorting followed standard paleoethnobotanical procedures
(Pearsall 1989).
Briefly, large samples were subsampled using the
riffle sampler, and the fraction to be analyzed passed through a 2.0 mm
screen.
The > 2.0 rnm fraction was sorted in its entirety, typically
into wood, root material,
unidentifiable fragments.
fruit rind,
large seeds, corn, and
All materials were counted and weighed.
< 2.0 rnm fraction was then examined for seeds and corn fragments.
The
Since
larger samples were wet screened using 1/16" mesh prior to chemical
flotation, relatively few small seeds were recovered.
Identifications
of seeds, fruic fragments, and roots were made by comparing unknowns to
materials in che MU Paleoethnobotany comparative collection, standard
874
reference books, and herbarium specimens curated at the Missouri
Botanical Garden, St. Louis.
Some remains are still unidentified,
however, some because the materials are highly fragmented; a few,
tallied as unknown types in the raw data table because they did not
match available comparative collections or published sources.
Corn
remains were callied as kernels, cupules (single), or rachis segments
(groups of joined cupules) . The maize remains will be studied at a later
date to determine the race or races present.
Wood was not identified.
Raw data were selected and combined in various ways to address the
main hypothesis of the study, and to observe any patterning in
occurrence of food plants over time.
For floors in which multiple
samples were analyzed, I summed the data over all samples.
Any sample
not sorted in its encirety was extrapolated to 100% before being summed
with other samples from that floor.
A series of ratios designed to
address the issue of cave function were then calculated (Figs. 1-6; see
bottom of Table 2 for values): local:exotic+local foods;'corn
kernel:corn cob fragments (corn processing question); and total
food:wood (abundance of food items) .
As mentioned above, there was some question as to how Festuca, or
~
grass, should be used in the ratios:
as food or not.
As detailed
in Table 2, "total food, 2• represents the total omitting Festuca;
"total food• contains it.
Further, it was unclear whether unidentified
fruit hull fragments should be considered as "localw or "exoticw foods
(they are considered food remains).
Ratios using "local foodsw include
these remains; those using "local foods, 2" do not include them.
\
DISCUSSION OF RESULTS
Food Plants Recovered
Food plants recovered from the excavation are summarized in Table
2.
There are probably other foods represented among the unknown seeds,
but none of the unknowns are common (most are single occurrences), and
therefore not likely co change the results significantly.
875
Maize, Phasolus beans, and thin epidermal fragments of the pit of
a fruit in the Sapotaceae are considered "exotic• food resources.
cave is located above the cultivation zone for maize and beans.
The
Trees
in the Sapotaceae {genera such as Pouteria, formerly Lucyma, Manilkara;
Bumelia} would occur wild {or tended, in the case of cultivated taxa} in
the moist.
lower elevation forest to the east of the study region.
The
few fragrnencs of the Sapotaceae fruit pic type thac had the hilum
(atcachmenc areal presenc permicced idencificacion to family, but are
too fragmencary co idencify chese conclusively as lucuma (Pouteria
spp.}.
A fruit similar co lucuma is indicated, however.
Local foods, those present in the paramo grassland formation, are
Lupinus, Ribes, Rubus, Festuca, Cheno/Am, Polyqonum/Rumex, fruit rinds
{omitted in "local foods, 2•), and tuber/root fragments.
Most of these
occurred at Panaulauca and Pachamachay caves in the Junin pyng of
central Peru (refer to Pearsall 1980, 1988 for descriptions of habitat
and use of these taxa} .
Briefly,
~
grass, Festuca, produces large
seeds that are a potential food resource.
This grass is not a preferred
forage of camellids {smaller grass seeds recovered, likely Calamaqrostis
and
~.
are not considered foods in this analysis, since camellids do
favor these} and thus less likely to be in dung.
The presence of
Festuca seeds in the cave could be the result of gathering for food, or
inclusion by che burning of ichu grass as tinder or through accident.
Lupinus seeds (lupine). though not common, are potential food resources.
These seeds are not large enough to be identified as the cultivated
species.
Small, clover-sized legume seeds are not tallied as food
plants {I have observed these in camellid dung samples from Junin}.
~ and~
seeds may represent the eating of these small berries.
Both planes are small shrubs.
Seeds tallied as Cheno/Am were too
distorted to identify precisely to genus, but are either Chenopodium or
Affiaranthus.
These are annual plants that produce large quantities of
edible seeds (and also greens}, and often favor disturbed habitats.
Seeds are not of che domescicated forms.
Polygonaceae seeds {either
876
~
or Polygonum) are also included in the local foods group, since
representatives of these genera occur in high elevation grasslands, and
are considered edible by some cultures.
The categories fruic rinds and tuber/root fragments require some
additional explanation.
Although many of the plant remains recovered
from the cave were highly fragmented, it was possible to distinguish
fragments that appeared to be root or tuber tissue.
Such tissue is very
porous in texture, and lacks the internal morphology associated with
wood or other larger plant remains like bean or corn fragments.
few rare cases, an "eye" (node) was preserved.
In a
None of these had the
clear shape of potato eyes, however; a few looked like sedge (Cyperus or
Scirpus) roots.
Given the abundant wild and domesticated root and tuber
foods found in the higher elevations of the Andes, it seemed best to
include these fragmented remains in the list of local foods.
case of fruit fragments,
fragments
this is a residual category:
(thickness variedl that
were~
In the
all curved tissue
Sapotaceae were tallied here.
None of these are squash; none have diagnostic features.
I included
these in local foods because they did not look like any of the tropical
forest tree fruits in my collection; ratios calculated with "local
foods, 2" omit these remains, for comparison.
Patterning through Time
The three "exotic" taxa, corn, bean, and Sapotaceae fruit, appear
in the record at Manachaqui cave at different points in the sequence.
Sapotaceae fruit pit fragments appear earlier, in a hearth sample (#80)
and a floor sample (#64) from floor FF, the lowest level analyzed, dated
to 1500 b.c.
Thus one food from outside the resource area of the site
was present from the earliest occupation.
throughout the sequence.
The fruits continue to occur
The largest quantity (29 fragments), occurs in
sample #58, one of the samples from the CC floor,
preceramic.
in the later
Corn and bean appear together in floor Y (sample #50), the
earliest level of the Suitacocha phase, 800 b.c ..
There are no
877
Preceramic or Initial period macroremains of corn or bean at the site.
Corn also occurs in floor X, the upper Suitacocha floor; bean does not.
Corn is consistently and often abundantly present throughout the rest of
the sequence; bean occurs only in floors Y, M, J, I, and B (1-3
fragments per occurrence) .
There are few
foods.
striki~g
patterns among the occurrence of local
Festuca seeds are present in high concentrations in several
samples:
172 in #78 !hearth in CC); 55 in #59 (floor CCI; 1566 in #77
(hearth in 88); 56 in
<floor 81.
~41
!floor Tl; 37 in #39 (floor Sl; 67 in #1
Otherwise, 1-10 ichu seeds occur in most samples.
The other
seven local plants/plant categories are present in low numbers
sporadically throughout the sequence.
in numbers of taxa present:
Samples are not especially rich
one sample has 5 local taxa, 8 samples have
4 taxa; the rest have 0-3 taxa present (Table 2).
The Late Preceramic
levels appear to be less rich than the rest of the sequence.
Looking at abundance of food remains in comparison to quantity of
wood charcoal is one way to determine how commonly food was cooked at
the site.
If fires were used primarily for warmth, and little food was
cooked, one would expect fewer "cooking accidents• and disposal of
inedible food residues than if fires were used for cooking on a daily
basis.
Fig. 1 illustrates the ratio of total food remains to wood
charcoal.
as fuel.
and S.
I am assuming that wood was brought to the site only for use
Four levels stand out with high food:wood ratios:
BB, Z, T,
In the case of 88, a 1566-seed concentration of Festuca seeds
pushes the food:wood ratio up; in floor Z a very small amount of wood is
present (0.836 g), which has the same effect.
The high ratios in floors
T and S are caused in part by low wood abundance (1.23, 2.7 g), but food
remains are also more abundant in these samples than in many examined.
Floors T and S, and perhaps 8B, represent a depositional pattern
consistent with increased cooking (relative to the other levels).
Since
it is not possible to examine these data in terms of amount of charred
material in a standard volume of matrix, differences among levels in the
878
amount of burning activity are not clearly delineated.
Higher amounts
of charcoal can also be an indicator of more intensive use of a site.
Looking at abundance of foods using "Total food,2:wood" (i.e., omitting
Festuca as a food)
food ratios:
(Fig. 21. three levels stand out as having higher
Z, T, and S.
The only change by omitting Festuca is thus
the omission of 88 as a level with a high food ratio.
Examining the data in terms of the relative abundance of local and
"exotic" foods highlights the impact of the introduction of corn into
the regional economic system (Fig. 3).
Corn appears at the site for the
first time in floor Y, and this results in a dramatic drop in the ratio
of local to exotic foods.
Although Sapotaceae fruit fragments are
present in the earlier levels, the Preceramic and Initial period
assemblages are dominated by local food items.
The pattern of local use
reverses with the introduction of corn, and reappears at only two points
in the later part of the sequence:
floors T and S (early part of the
Early Intermediate) and floors F and B (the two Incaic levels analyzed).
These two periods may represent a change in use of the cave relative to
the rest of the "post-maize" sequence.
There are a number of changes in the ratio of local to exotic
foods when "Local foods, 2" (i.e., omitting Festuca and fruit rinds) is
used in the calculation (Fig. 4).
The earlier levels of the cave are
much less even (ie, a less consistent pattern of high local foods) when
Festuca and fruit rinds are removed.
Sample size falls considerably,
which may contribute to the unevenness.
of exotics--a new pattern.
Levels EE and CC show dominance
After corn comes in, exotic foods dominate,
as before, and there are no pronounced reversals of this pattern, as
appeared using the other sum.
This weakens the argument for periodic
changes in cave use (see below).
Looking for the relative abundance of maize cob (inedible) and
kernel (edible) remains (Fig. Sl may shed additional light on whether
corn use changed during the period Manachaqui Cave was in use.
Cob
fragments are generally more abundant than kernel fragments at the site
879
(i.e., in most levels the ratio is >1).
This shows that corn was not
brought to the cave exclusively as shelled grain, and that cobs, perhaps
with a few kernels still attached, were disposed of in the fire.
Floor
S stands out with a very high cob:kernel ratio; only 2 kernel fragments
were recovered in comparison to 26 cob fragments.
Levels W and F are
the exceptions to the dominance of cob material; in these levels kernels
are three times as common as cob fragments (Fig. 6).
Little corn is
present in level F, but W has abundant remains, dominated by kernels.
Since it is unlikely kernels were burned deliberately, on site cooking
or parching of kernels (and accidental burr.ing) seems indicated.
CONCLUSXONS
Considering all the charred botanical data, and remembering that
results are impacted by loss of some smaller remains from the
assemblage, the results of this study generally support the hypothesis
that Manachaqui Cave was a camp for travelers for much of its history.
Food remains are not very abundant relative to charred wood (evidence
that fires were used more for warmth than for cooking); uexotic" food
stuffs are present throughout the sequence (evidence that foods were
carried in); once corn was introduced (Early Horizon), use of local
foods drops off (evidence that local foods were relatively unimportant);
corn kernels are generally less abundant than cob remains (evidence that
parching/cooking of corn was uncommon).
Sincd some cob remains are
always present, it is clear that travelers were not just carrying
shelled corn.
Within this overall pattern, however, there are some interesting
exceptions that suggest the cave may have changed functions for brief
periods.
The floors in question are BB and Z (Manachaqui phase); W, T,
and S (early part of Ernpedrada phase), and F and B (Poblano phase).
The
Manachaqui floors show one indicator of more intensive use of the cave:
higher food:wood ratios,
in one case (Z) based on a small sample.
are multiple indicators that
~he
There
cave was more intensely used during the
880
early part of the Empedrada phase:
a high ratio of kernels to cobs
(floor W); high food co wood ratios (floors T and S); high local to
exotic food ratios !T and Sl.
These indicators suggest the cave
functioned as a habitation site during this period.
continued to be available. however.
"Exotic" foods
The two Poblano phase samples
analyzed (F and Bl, presenc a slightly different pattern:
there is
increased use of local foods, but no overall increase in foods relative
to charred wood.
Perhaps the cave had a more specialized function
during the Late Horizon, different from a traveler's stop, but not as
intensive as a habitation site.
As mentioned above, the pattern
involving local:exotic indicators weakens if Festuca and fruit remains
are removed as local foods.
This study of charred botanical remains from Manachaqui Cave
illustrates how ethnobotanical data can be used to test hypotheses of
site function.
The shifts in cave function suggested by these data can
be tested through analysis of other macroremain samples, comparisons to
artifact and faunal data, detailed study of corn remains, and future
research, perhaps including phytolith sampling.
LITERATURE CITED
Church, W.
1993
Evidence for prehistoric highland-tropical forest
interaction from Manachaqui Cave in the northeastern
highlands of Peru:
Preliminary report. Paper presented at
the Society for American Archaeology meetings, St. Louis.
Hillman, G. C.
1984 Interpretation of archaeological plant remains: The
application of ethnographic models from Turkey. In Plants
and Ancient Man: Studies in Palaeoethnobotany, edited by W.
V. Zeist and Ttl. Casparie, pp. 1-41. A.A. Balkerna'Hubbard, R.
N. L. B., Rotterdam.
Pearsall. D. M.
1980 Pachamachay ethnobotanical report:
Plant utilization at a
hunting base camp. In Prehistoric Hunters of the High Andes,
J. W. Rick, pp. 191-231. Academic Press, New York.
881
1983
Evaluating the stability of subsistence strategies by use of
paleoethnobotanical data. Journal of Ethnobiology
3(2):121-137.
1988
Interpreting the meaning of macroremain abundance: The
impact of source and context. In Current Paleoethnobotany.
Analytical Methods and Cultural Interpretations of
Archaeological Plant Remains, edited by C. A. Hastorf and V.
s. Popper, pp. 97-118. University of Chicago Press, Chicago.
1989
Paleoethnobotany.
San Diego.
A Handbook of Procedures. Academic Press,
and D. R. Piperno (editors)
1993 Current Research in Phytolith Analysis: Applications in
Archaeology and Paleoecology. MASCA, University of
Pennsylvania Museum., Philadelphia.
Yellen, J. E.
1977 Archaeological Approaches to the Present: Models for
Reconstructing the Past. Academic Press, New York.
882
Floors/Samples
Period
~
Date Cuncalibrated l
Late Horizon
Poblano
AD 1532
A
B
C-E
F
1
AD 1470
1
Hiatus
Middle Horizon
Early Inter.
G
H
I
J
K
L
M
N
0
p
T
1
1
1
1
1
u
v
w
1
1
2
Q
R
s
Empedrada
AD 700
1
1
1
1
1
1
1
2
AD 200/400
Col par
AD 200/400
200 BC
Chavin horizon Hiatus?
Early Horizon
X
y
Suitacocha
2
2
BOO BC
Initial Period Manachaqui
z
AA
1
3
cc
DD
EE
FF
3
3
2
3
3
BOO BC
1500 BC
Late Preceram. Lavas en
BB
400 BC
1500 BC
2200 BC
Table 1. Manachaqui Cave chronology and samples analyzed.
79
63
62
80
65
64
LAB NUMBER
MIA.15.66 M1A.15.66 M1A.15.66
M1A.16/17.63 MIA.15.63 MIA.15.63
PROVENIENCE
FF-65
Total FF
EE-79-H
EE-63
EE-62
FF-64
FLOOR
FF·80-H
Total EE
3
1
2
1
4
2
FS#
0.625
0.25
1.25
2.65
LITERS
0.3
0.475
222.75
17.25
TOTAL SAMPLE WGHT
60.22
214.56
11.68
18.66
%EXAMINED
25
100
100
100
25
100
100
100
SUB-SAMPLE WEIGHT
51.71
53.33
WOOD WEIGHT
26.56
0.775
24.91
131.93
19.26
4.14
6.64
87.82
Maize kernels, total count
Maize cob fra_gs, total count
Phaseolus, total count
cf. Sapotaceae fruit
1
2
6
7
2
1
13
3
2
2
14
Fruit rinds
3
11
Tuber/root frags
3
12
3
3
~lnus (Legumlnosae}
cf. Ribas (Saxifragaceae)
Rubus (Rosaceae)
Festuca (Poaceae)
1
2
12
1
3
1
Cheno/Am
Polygonum/Rumex (Polygonac.)
1
4
4
Total local foods, 1
1
2
26
31
6
7
4
0
0
Total local foods, 2
4
0
0
16
3
3
0
Total exotic foods
1
2
6
7
13
2
1
Fruit rinds
3
14
2
3
2
11
8
37
Total foods, 1
1
4
11
5
39
8
0
8
4
36
10
Total foods 2
3
27
5
Locai:Exotlc+Local, 1
0.84
0.67
Total food:wood, 1
0.28
0.44
0.73
Locai:Exotlc+local, 2
0.19
Total food:wood 2
0.27
0.31
kernel:cob
cob:kernel
Table 2. Summary of botanical data.
(X)
00
w
1LABNUMBER
IPROVENIENCE
I
IA.OOR
FS#
UTERS
TOTAL SAMPLE WGHT
%EXAMINED
SUB-SAMPLE WEIGHT
WOOD WEIGHT
Maize kernels, total count
Maize cob frags, total count
Phaseolus, total count
cf. Sapotaceae fruit
Fruit rinds
Tuber/root frags
Luoinus (Leguminosae)
cf. Ribas (Saxlfragaceae)
Rubus (Rosaceae)
Festuca (Poaceae)
Cheno/Am
58
60
78
59
77
57
61
M1A.17.57 MIA.15.57 MIA.15.57
M1A.15.54 MIA.15.54
MIA.15.60 MIA.15.60
CC-58
DD·60
Total DD
CC-78
CC-59
Total CC BB·77·H
88·57
DD·61
2
1
2
5
2
1
0.65
2
0.19
0.45
2.225
0.25
0.225
206.18
7.64
42.5
6.3
61.77
199.15
13.13
50
100
50
100
100
100
50
100
100
100.59
20
98.58
8.472
56.633
1.568
128
22.85
2.007
18.951
1.3
13.17
2
2
8
7
3
3
16
8
14
4
29
5
2
Table 2 (continued) .
1
4
3
1
4
1
6
172
1
55
14
8
2
2
1G
12
7
3
7
3
14
13
28
14
16
8
44
38
0.64
2.32
0.47
2.01
175
3
4
60
0
0
5
60
5
Pol~gonum/Rumex {Pol~gonac.)
Total local foods, 1
]'otal local foods, 2
Total exotic foods
-------·
Fruit rinds
Total foods, 1
Total foods, 2
!:Q.c::.ai:Exotic+'=.Q.c__al, 1
Total food:wood, 1
Locai:Exotlc+local, 2
Total food:wood, 2
kernel:cob
cob:kernel
37
5
4
----
179
7
5
1
6
1
29
35
30
404
2
1
416
7
37
5
453
49
0.92
3.54
0.16
0.38
1566
1573
3
1
4
1574
8
1
1
0
1
- - -1 - -
(X)
(X)
tJ:>.
LAB NUMBER
PROVENIENCE
FlOOR
FS If
UTEAS
TOTAL SAMPLE WGHT
%EXAMINED
SUB-SAMPLE WEIGHT
WOOD WEIGHT
Maize kernels, total count
Maize cob frags, total count
Phaseolus, total count
cl. Sapotaceae fruit
Bull rinds
Tuber/root frags
Lupinus (Leguminosae)
cl. Ribas (Saxllragacealll_
Rubus_ (Rosaceae)
Festuca (Poaceae)
Cheno/Am
Polygonum/Rumex (Poly_gonac.'
Total local foods, 1
Total local foods, 2
Tot~xotlc food_s
Fruit rinds
Total foods, 1
Total foods 2
Locai:Exolic+Local, 1
Total food:wood, 1
Locai:Exolic+local, 2
Total food:wood, 2
kernel:cob
cob:kernel
76
56
55
54
53
51
MIA.15.43 MIA.17.43 M1A.15.43
MIA.15.54
M1A.15.41 M1A.15.38
AA-76-H
AA-55
AA-54
88-56
Total 88
Y-51
Total AA
2
2
3
1
1
2
0.525
0.25
0.085
0.85
0.2
0.12
44.15
5.35
87.43
40
7.13
12.69
100
100
100
100
100
100
100
100
z
16.62
63.62
2
11
1
4
19
7
1
1
1
3132
2
2
15
4
2
11
17
17
-
Table 2 (continued).
1.16
3162
11
4
1s
3166
34
1.00
49.76
0.73
0.53
-
----
-
L___. _ _ _
41.64
10.95
53.75
0.836
2
8
10
5
10
4
8
10
1
1
1
10
1
11
1
2
1
13
2
8
10
21
20
2
28
7
8
10
36
25
0.78
0.67
0.47
0.47
11
0
0
10
11
10
1.00
13.16
0.00
11.96
4
1
0
1
4
___2_ _
1
1
1
0
14
4
0
1
1
14
4
--
----
-
--
00
00
U1
50
49
48
47
46
45
LAB NUMBER
M1A.15.30 M1A.15.30
M1A.15.35 M1A.15.35
PROVENIENCE
M1A.15.38
M1A.15.28
Total X
W-47
W-46
X-49
X-48
Total W
Y-50
Total Y
v
FlOOR
3
2
3
3
4
2
FS#
0.37
0.65
0.175
0.16
0.425
0.2
UTERS
58.66
5.22
22.41
10.35
40.41
TOTAL SAMPLE WGHT
8.31
50
100
100
100
100
100
100
%EXAMINED
100
100
SUB-SAMPLE WEIGHT
30.45
11 .635
13.64
0.83
6.515
7.345
16.413
19.745
WOOD WEIGHT
3.332
2.12
26
Maize kernels, total count
13
11
11
46
46
1
24
18
18
15
Maize cob !rags, total count
12
15
4
1
2
Phaseolus, total count
6
3
6
ct. Sapotaceae fruit
6
5
1
2
1
1
Fruit rinds
1
2
2
Tuber/root fraos
1
1
Lupinus Cleouminosae)
cf. Ribes (Saxlfragaceae)
1
2
Rubus (Rosaceae)
1
6
3
4
8
2
1
Festuca (Poaceae)
3
2
1
1
1
1
Cheno/Am
!:Q!Y.gonum/Rumex {Polygonac.)
4
16
2
9
Total local foods, 1
6
2
2
11
2
3
1
1
0
1
0
4
4
!_o_!a.!_l~cal food~,_2
1
_ 2_ _
Total exotic foods
29
58
0
29
29
0
67
67
0
2
5
Fruit rinds
1
1
1
35
74
2
31
33
2
76
78
Total foods, 1
7
32
_6p_
0
30
0
72
72
I.l?tal fooqs, 2
7
i--30
Locai:Exotic+Local, 1
0.22
0.12
0.14
Q~~
4.49
Total food:wood 1
5.43
3.95
3.30
0.05
0.03
0.06
Locai:Exolic+local, 2
0.17
4.08
4.84
Total food:wood, 2
3.65
3.30
1.08
0.61
3.07
0.25
kernel:cob
0.92
1.64
0.33
cob:kernel
4.00
i
Table 2 (continued) .
00
00
0'1
43
41
39
36
34
32
31
26
LA.BNUMBER
M1A.15.26 M1A.15.25 M1A.15.24 M1A.15.23 M1A.15.22 M1A.15.21 MIA.15.18 MIA.15.18
PROVENIENCE
p
N-26
N-31
u
T
R
FLOOR
0
2
3
2
3
3
2
3
3
FS#
I
0.2
0.15
0.2
0.37
0.43
0.35
0.35
0.75
LITERS
5.19
8.91
4.94
33.19
39
38.7
18.3
90
TOTAL SAMPLE WGHT
I
100
100
100
100
100
100
100
100
%EXAMINED
I
SUB-SAMPLE WEIGHT
1.23
2.7
13.52
17.06
18.46
1.02
2.25
42.6
WOOD WEIGHT
I
7
2
5
19
16
Maize kernels, total count
41
26
14
19
35
18
Maize cob frags, total count
165
Phaseolus, total count
4
1
1
cf. Sapotaceae fruit
8
2
1
3
7
Fruit rinds
3
2
Tuber/root frags
1
Luplnus (Legumlnosae)
1
cf. Ribes (Saxlfragaceae)
Rubus (Rosaceae)
3
2
56
1
5
3
37
Festuca (Poaceae)
1
1
Cheno/Am
1
1
~gonum/Rumex (Poly_gonac.)
3
60
15
0
37
6
7
Total local foods, 1
11
0
3
0
2
2
2
0
2
IotaiJgE-!!1 foods, 2
0
21
28
28
54
35
206
Total exotic foods
1
Fruit rinds
1
3
8
2
7
81
4
0
65
34
69
42
217
Total foods, 1
0
25
28
64
215
Total foods, 2
33
39
1
1-·--0.00
0.74
0.57
0.22
0.17
Locai:Exotic+Local, 1
__9.18
Total food:wood, 1
0.00
65.85
24.07
2.51
4.04
2.28
0.00
0.13
0.00
0.07
0.04
0.05
Locai:Exotlc+local, 2
0.00
20.33
Total food:wood, 2
10.37
2.44
3.75
2.11
kernel:cob
0.00
0.50
0.08
0.26
0.54
0.89
0.00
2.00
13.00
3.80
1.84
cob:kJ~mel
-- 1.13
s
I
Table 2 (continued) .
(X)
00
-1
20
18
16
14
12
9
24
11
LA.BNUMBER
MIA.15.17 MIA.15.15 MIA.15.14 MIA.15.13 MIA.15.12 MIA.15.11 MIA.15.10 MIA.15.9
PROVENIB'JCE
K
J
I
Total N
M
L
H
G
F
FLOOR
1
2
3
3
3
3
3
3
FS#
0.575
0.5
0.5
0.5
0.4
1
0.75
1.025
LITERS
76.74
73.9
170.4
35.43
24.35
132.25
74
158.7
TOTAL SAMPLE WGHT
50
50
50
50
25
25
100
100
100
%EXAMINED
35.7
39
37.86
41.4
39.52
65
SUB-SAMPLE WEIGHT
20.728
17.719
44.85
32.43
18.55
24.385
22.79
11.545
13.07
WOOD WEIGHT
19
19
34
27
41
27
4
16
6
Maize kernels 1 total count
62
2
165
76
16
33
36
40
15
Maize cob rraos, total count
3
1
1
Phaseolus, total count
1
1
1
1
cr. Sapotaceae fruit
2
2
1
3
2
7
1
Fruit rinds
7
3
1
1
1
Tuber/root rrags
1
Luplnus (Leoumlnosae)
1
1
cr. Rlbes (Saxlrraoaceae)
2
1
1
Rubus (Rosaceae)
6
1
4
1
4
4
Festuca (Poaceae)
5
Cheno/Am
1
1
1
fQ!Y.gonum/Rumex (Polygonac.'
6
8
5
12
3
9
8
1
Total local foods, 1
14
I
2
1
2
2
10
2
3
1
0
Total local roods, 2
105
82
55
8
207
21
71
67
31
Total exotic foods
2
1
2
1
3
2
Fruit rinds
7
90
60
221
117
24
79
75
32
14
I
Total foods, 1
11 7
24
87
59
75
69
31
_10_ _1
216
ToJ.?~_ods, 2
---0.10
0.13
0.10
0.08
0.10
0.11
0.03
0.43
0.06
Locai:Exolic+Local, 1
---·-1
3.39
3.24
3.29
4.93
3.61
1.29
4.34
2.77
Total food:wood, 1
1.07 I
0.09
0.04
0.02
0.03
0.01
0.00
0.20
0.01
0.09
Locai:Exolic+local, 2
I
1.29
4.20
4.82
3.33
3.08
3.03
2.69
0.77
Total food:wood, 2
3.61
0.36
0.25
0.31
0.58
0.94
0.68
1.07
3.00
0.25
kernel:cob
4.00
3.26_
0.94
4.02
2.81
cob:kernel
-0.33 _I
1_.7_~- __1_.06 ----- .. _1.48_
I
Table 2 (continued).
00
00
(X)
LAB NUMBER
PROVENIB'JCE
1
MIA.15.3
FLOOR
FS#
UTERS
TOTAL SAMPLE WGHT
%EXAMINED
SUB-SAMPLE WEIGHT
WOOD WEIGHT
Maize kernels, total count
Maize cob frags, total count
Phaseolus, total count
cf. Sapotaceae fruit
Fruit rinds
Tuber/root frags
Lupinus (leguminosae)
cf. Ribes (Saxifragaceae)
Rubus (Rosaceae)
Festuca CPoaceae)
Cheno/Am
f2_!Y.gonum/Rumex (Polygonac.l
Total local foods, 1
Total local foods, 2
Total exotic foods
Fruit rinds
Total foods, 1
Total foods, 2
Locai:Exolic+Local, 1
Total food:wood, 1
Locai:Exotic+local, 2
Total food:wood, 2
kernel:cob
cob:kernel
8
3
1.25
274.675
50
138.35
78.86
36
41
2
4
7
1
I
I
1
67
5
81
7
83
7
164
97
0.49
2.08
0.08
1.23
0.88
1.14
'
Table 2 (continued) .
(X)
(X)
\0
890
8
:f
E)
H
I
r
}I
1
L"J
N
"'0
0
0
~
-
(ij
0
I-
r-1
d
~
0
..Q
...0en
s 0
u:
.l
'0
~
"'0
0
0
H
0
0
>,
0
0
~
0
.jJ
n
'0
1\
~
M
~
X
U)
A
·.-!
z
vv
88
a:>
(](]
33
.:H
0
0
0
0
.jJ
co
l-1
~
0
E
co
l-1
Ol
0
.jJ
U)
·.-!
:r:
.-i
0
0
0
0
0
0
r-...
co
0
0
0
lO
0
0
0
0
0
0
-.:t
one~:~
.
C")
0
0
0
0
0
C\1
0
,...
0
0
0
Ol
·.-!
r:..
891
8
:f
E)
H
I
r
>I
1
1-1
0
0
~
~
~
N
"C
0
0
~
"!.
d
>.,
..0
0
H
'0
0
0
s
"C
0
.E
l.
~
n
0
I-
A
M
...en
0
0
u:
~
0
.w
N
'
'0
0
0
~
~
0
X
(/}
A
·.-i
z
1-1
vv
88
0)
00
33
±I
0
.w
CIS
~
0
E
CIS
1-1
Ol
0
.w
(/}
·.-i
::r:
N
0
0
0
0
0
0
0
0
0
0
1.0
0
10
0
C\1
0
,....
1.0
C\J
0
0
.
,....
one~:~
Ol
·.-i
rr..
Local :Exotic+ Local
1.00
0.90
0.80
0.70
0.60
0
iii 0.50
a:
0.40
0.30
0.20
0.10
0.00
lf: I:H
8 8 fB
~
N
>-
X
3: >
:::::>
t-
(J)
a: 0 a. z
;::E ...J
~
-,
-
J: (!) U.
m
Floors
Fig. 3. Histogram of ratios of local to exotic + local by
floor.
00
1.0
1:\..)
Local ,2 :exotic+ local, 2
0.80
0.70
0.60
0.50
0
~ 0.40
a:
0.30
0.20
0.10
0.00
l:f: l:f:J
8 8 ffi
~
N
>-
X
~
>
:::> r
Cl)
0:
0
a. z
~
__.
~
-, -
:I:
CJ u.
m
Floors
Fig. 4. Histogram of ratios of local,2 to exotic + local,2 by
floor.
OJ
\.0
w
cob:kernel
14.00
12.00
10.00
0
8.00
':;:::
co
a:
6.00
4.00
2.00
0.00
~
ill
8 8
m
~
N ~ X ~
>
~
r
00 ~ 0
~
Z ~ ~ ~ ~ -
I
~
~
m
Aoo~
Fig. 5. Histogram of ratios of maize cobs to kernels by
floor.
CXl
\0
~
kernel:cob
3.50
3.00
2.50
2.00
0
:;::
m
a:
1.50
1.00
0.50
0.00
l:f: HJ 8 8 83
~
N
>-
X
3: > :::> 1- cn a: 0
a. z
:::2:
...J
~
,
-
:X:
(!)
u_
m
Floors
Fig. 6. Histogram of ratios of maize kernels to cobs by
floor.
00
1.0
lTI