Professional Documents
Culture Documents
Andean Archaeology II
Art, Landscape, and Society
Edited by
Helaine Silverman
University of Illinois at Urbana-Champaign
Urbana-Champaign, Illinois
and
William H. Isbell
State University of New York at Binghamton
Binghamton, New York
Contributors
vii
viii Contributors
Stella Nair • Department of Architecture, University of California, Berkeley,
CA 94720
Jose Ochatoma Paravicino • Universidad Nacional de San Crist6bal de
Huamanga, Ayacucho, Peru
Donald A. Proulx • Department of Anthropology, University of
Massachusetts, Amherst, MA 01003
Jean-Pierre Protzen • Department of Architecture, University of California,
Berkeley, CA 94720
Helaine Silverman • Department of Anthropology, University of Illinois,
Urbana, IL 61801
Adriana von Hagen • Institute of Andean Studies, University of California,
Berkeley, CA 94709
Preface
x Preface
businesses, knowledge about Moche is unsurpassed. And now, Moche specialists
are investigating new issues-from ethnic boundaries and interactions with
neighboring peoples to the development of proto-writing (Jackson, this volume).
Clearly, Moche is but one and, certainly, an extraordinary example of an ancient
Andean society, culture, and style.
Many of the cultures associated with less popular art styles have remained
little known or even ignored. The creators of the Lima style (see Makowski, in
Volume I) produced pottery that does not spark the imagination as do Moche and
Nasca. Although associated with impressive pyramid complexes throughout
Peru's modern capital city, where there are numerous museums and universities,
Lima culture is poorly reported and only now being systematically investigated.
Highland cultures are less studied than coastal ones, in part because of difficult
logistics, but also because their graves rarely contain complete examples of fine
pottery and textiles. Least known are the archaeological records of the eastern
Andean ceja de montana and Upper Amazon tributaries.
Today, archaeologists are beginning to study little known areas and archae-
ological cultures. New investigations of highland Huari, not even recognized as
an independent style until the early 1950s, are exposing the important city of
Conchopata. Investigations of Recuay and other north highland styles are in
progress. Pucara, Tiwanaku, and Titicaca Basin archaeology is experiencing great
advances. The outlines of a new Nasca archaeology can be discerned. Little
known styles from the Acari and distant Ocofia, Majes, and Sihuas valleys are no
longer collapsed into one culture. Vast areas of the higWands remain uncharted,
especially in Bolivia. Even the Cochabamba valley is little investigated, in spite
of its importance to the Incas and probable key role in early cultural develop-
ments in the southern sphere of the Central Andes. Many nameless archaeologi-
cal societies are just beginning to appear in the literature, or remain to be
discovered and meaningfully integrated into the sweep of Andean prehistory.
In addition to the attraction of discovery in an area rich with ruins, the
history of Andean art and archaeology has also been influenced by the vast volume
of material remains from the past, as well as their extraordinary preservation, espe-
cially in south coastal Peru and neighboring Chile. Our colleagues who work in
various other parts of the world are often astonished by our photographs of stand-
ing fieldstone and adobe architecture, intact textiles, vivid polychrome sherds,
pyroengraved gourds, and complete skeletons and mummy bundles. Yet this pro-
fusion of remains is a kind of curse since there is so much to analyze.
Overwhelming quantities of materials have hampered timely completion of reports
and publications. Pucara pottery (Chavez, this volume) is finally published fifty
years after its original excavation. Pottery newly excavated at Conchopata (Isbell
and Cook, Ochatoma and Cabrera, this volume) exceeds fifteen tons, so the task
of simply determining what collections of sherds are worth further attention in pre-
liminary efforts to reconstruct forms and designs is daunting, yet progressing.
Preface xi
The demands of field and lab work in Central Andean archaeology occupy
so much time and energy that theorizing may suffer in a perverse inversion of data
and ideas. In balance (as we consider more fully in the two section introductions
for this volume), many Andeanists have devoted careers to one or another ancient
culture and they are now, in maturity, achieving fascinating (and, above all, plau
sible) insights into ancient societies. They have "paid their dues," conducting
years of painstaking empirical research that might qualify as what ethnographers
call "thick description." They have become masters of the archaeological record,
experts in the art and all of its variations, authorities on ancient technologies,
experienced in settlement patterns and community organization, and even able to
predict where tombs are to be found (this kind of knowledge has legendarily been
attributed to looters who have spent a lifetime digging). Such depth of knowledge
promotes empathic insights into an archaeological art and culture, furthering con
ditions for breakthroughs in knowledge. The associated scholarship also creates a
marvelous foundation on which young archaeologists can build.
The early 21st century will see, we believe, a mature Andean scholarship
that will offer to colleagues outside our culture area important contributions for
comparative theorizing on processes of social inequality, prestige generation and
maintenance, ethnogenesis, materiality and the differential production of culture,
household organization, craft production, burial customs, landscape evolution,
urbanism, and rituals of power, among many other aspects of sociopolitical life in
the past. In a variety of ways the papers in this volume address these issues. Their
strength, we believe, is in their foregrounding of area, data, and empirical analy
ses rather than theoretical approaches.
The field of Central Andean archaeology is creative and healthy. In addition
to a very strong presence of established scholars active in the field, each year sees
a larger group of recent Ph.D.s and advanced graduate students conducting
research and presenting results at national and international meetings. Many are
working on major archaeological cultures and the coast is still preferred over the
highlands, but enthusiasm for venturing into unknown cultures is increasing,
especially as civil unrest in Peru has been normalized. New archaeological
research in unknown territories is necessarily designed to simply determine what
is there, and when. The importance of this basic research cannot be emphasized
enough.
We enthusiastically offer this volume, the second in what we hope will
become a series, as an example of the exciting work being done in Andean
archaeology today and as a preview of what will be addressed in the future.
Contents
PART I. INTRODUCTION
Chapter 1. From Art to Material Culture. . . . . . . . . . . . . . . . . . . . . . . 3
Helaine Silverman and William H. Isbell
xiii
xiv Contents
Chapter 7. The Gateways of Tiwanaku: Symbols or Passages? 189
Jean-Pierre Protzen and Stella Nair
Chapter 8. Religious Ideology and Military Organization in the
Iconography of a D-Shaped Ceremonial Precinct at Conchopata .... 225
Jose Ochatoma Paravicino and Martha Cabrera Romero
Chapter 9. A New Perspective on Conchopata and the
Andean Middle Horizon 249
William H. Isbell and Anita G. Cook
Chapter 10. The Correlation Between Geoglyphs and Subterranean
Water Resources in the Rio Grande de Nazca Drainage 307
David W Johnson, Donald A. Proulx, and Stephen B. Mabee
Chapter 11. Rock Art, Historical Memory, and Ethnic Boundaries:
A Study from the Northern Andean Highlands 333
Tamara L. Bray
PART IV. CONCLUSION
Chapter 12. Issues of Cultural Production and Reproduction . . . . . . . 357
Helaine Silverman and William H. Isbell
Index.................................................... 365
Part I
Introduction
Chapter 1
From Art to Material Culture
HELAINE SILVERMAN AND WILLIAM H. ISBELL
At the end of the 1980s and into the 1990s, archaeologists' thinking about the mean-
ing of ancient Andean art was revolutionized as a result of spectacular discoveries
at Sipan, a Moche culture pyramid complex in the Lambayeque Valley on Peru's far
north coast. Initially ravished by looters, excavations at the pyramid were soon
taken over by archaeologist Walter Alva (1988, 1990; Alva and Donnan 1993) who
revealed a series of royal graves containing bodies of the ancient polity's kings and
counselors, resplendent in their ceremonial attire and ritual paraphernalia.
Before Sipan, archaeologists had no idea that Moche graves could be so
elaborate and contain so much wealth. Moreover, the archaeological world was
astonished that the principal men in two of Sipan's tombs were buried with the
costumes-including even such details as the spotted dog beside his feet--of the
iconographically known Warrior Priest. A royal tomb in the back of the same
pyramid complex held the remains of another principal individual spectacularly
dressed as the iconographically known Bird Priest. A few years later, at San Jose
de Moro, a Moche center in the Jequetepeque Valley some fifty kilometers south
of Sipan, archaeologists discovered a magnificent tomb containing the costumed
body of the iconographically known Priestess (Donnan and Castillo 1992, 1994).
The recent discovery of a massacre at Huaca de la Luna, in the Moche Valley,
similarly confirms the reality of the Sacrifice Ceremony. For the first time, with
these discoveries, it was clear that Moche art's formalized, representational, com-
plex imagery depicted real people participating in the iconographically depicted
scenes: living rulers and counselors had presided over real events, appropriately
costumed as the Warrior Priest, the Bird Priest, the Priestess, and other adjudica-
tors of the temple/court. It is also apparent that several royal Moche courts existed
(see Bawden 1996; Castillo and Donnan 1994). They followed the same set of
rules and recognized a more-or-less standard hierarchy of priestly and more sec-
ular offices, but all clearly of profound ritual significance.
3
4 Helaine Silverman and William H. Isbell
It must be remembered that South American civilization developed without
writing, (Margaret Jackson's important analysis of Moche proto-writing, in this
volume, notwithstanding). As a consequence, scholars of the prehispanic past are
forever disadvantaged in their attempts to understand ancient Andean peoples and
cultures. Yet it is now clear that Andean people produced messages about them-
selves and their world, at least some of which have survived into modem times
and can be "read" more or less as they were intended to be understood if archae-
ological associations and contexts are sufficient for us to properly identify the
complex signs.
The stunning breakthroughs in understanding Moche visual art that have
come with the new discoveries at Sipan, San Jose de Moro, Huaca de la Luna, and
Huaca El Brujo can only be fully appreciated because of a century of antecedent
scholarship which, with recent field and/or museum research, has greatly advanced
Moche scholarship (chronologically and among others: Middendorf 1892; Uhle
1913, 1915 who reported the first scientific excavations; Larco Hoyle 1938, 1939,
1945c, 1946b, 1948 who created the Moche relative chronology; Benson 1972;
Donnan 1973, 1976, 1978, 1982a,b, 1988, 1996, 2001; Donnan and McClelland
1999; Hocquenghem 1987; Alva 1988, 1990, 1994; Alva and Donnan 1993;
Castillo 1989; Bawden 1996; Uceda 1997; Uceda and Chapdelaine 1998; Uceda
et al. 1997, 1998; Chapdelaine 2000; Chapdelaine, Kennedy, and Uceda 2001;
Chapdelaine, Millaire, and Kennedy 2001; Chapdelaine, Pimentel and Bernier
2001). Similarly, understanding other Andean "messages from the past" has not
been immediate or simple, and breakthroughs, when they occasionally take place,
are virtually transcendental in importance. The emerging new understanding of
Huari is almost as exemplary as the Moche case. For years archaeologists have
argued about the nature of contact between the Tiwanaku capital on the south
shore of Lake Titicaca and the south-central highland city of Huari. During the
Middle Horizon they shared a religious iconography so similar that a single origin
is indisputable. Several modem scholars proposed that devout pilgrims may have
traveled from Huari to witness ceremonies conducted at Tiwanaku. This, in tum,
would explain the appearance of symbols from the Lake Titicaca religion in high-
land Ayacucho. But in the last few years archaeologists have discovered messages
painted by Conchopata's ancient artists (Isbell and Cook, Ochatoma and Cabrera,
this volume) that represent Huari-style men, brandishing weapons and shields, and
kneeling in reed boats of the kind used to cross Lake Titicaca. Conchopata potters
may have been telling of real trips made to the far-off lake. If so, Huari travelers
were not modest pilgrims, but apparently aggressive raiders declaring their mili-
tary power far beyond the boundaries of Huari political influence.
The experience of Andean civilizations was not written but, rather, inscribed
with painted and modeled iconography, symbols, graffiti, and the places created
by ancient built environments. Andean art, iconography, and architecture are
extremely rich, providing many "messages from the past" that remain to be
REFERENCES
Alva, Walter, 1988, Discovering the New World's richest unlooted tomb. National Geographic 174
(4): 510-550.
Alva, Walter, 1990, New tomb of royal splendor. National Geographic 177 (6): 2-15.
Alva, Walter, 1994, Sipdn. Colecci6n Cultura y Artes del Peru. Cerveceria Backus & Johnston, SA,
Lima.
Alva, Walter and Christopher B. Donnan, 1993, Royal Tombs of Sipdn. Fowler Museum of Culture
History, Los Angeles.
Bawden, Garth, 1996, The Moche. Blackwell, Oxford.
Bennett, Wendell c., 1936, Excavations in Bolivia. Anthropological Papers ofthe American Museum
ofNatural History 35 (4): 329-507.
Bennett, Wendell c., 1948, A revised sequence for the South Titicaca Basin. In A Reappraisal of
Peruvian Archaeology, edited by Wendell C. Bennett, pp. 90-92. Memoir 4. Society for
American Archaeology, Menasha.
Bennett, Wendell c., 1950, The Gallinazo Group, Viru Valley, Peru. Yale University Publications in
Anthropology, Number 43. New Haven.
Benson, Elizabeth P., 1972, The Mochica: A Culture of Peru. Praeger, New York.
Benson, Elizabeth P. (ed.), 1972, The Cult of the Feline. Dumbarton Oaks Research Library and
Collection, Washington, D.C.
Bonavia, Duccio, 1994, Arte e Historia del Peru Antiguo: Colecci6n Enrico Poli Bianchi. Banco del
Sur, Arequipa.
Browne, David M., Helaine Silverman, and Ruben Garcia, 1993, A cache of 48 Nasca trophy heads
from Cerro Carapo, Peru. Latin American Antiquity 4 (3): 274-294.
Burger, Richard L., 1984, The Prehistoric Occupation of Chav{n de Hudntar, Peru. University of
California Publications in Anthropology, volume 14.
Burger, Richard L., 1988, Unity and heterogeneity within the Chavfn horizon. In Peruvian Prehistory,
edited by Richard W. Keatinge, pp. 99-144. Cambridge University Press, Cambridge.
Part II
Andean Art and Society
Introduction
HELAINE SILVERMAN AND WILLIAM H. ISBELL
Andean art encompasses aesthetic and technical masterpieces of first order. Even
before the end of the 19th century the more representational art traditions of
ancient Peru had become favorites of great museums as well as private collectors.
Ancient sites adorned with exotic sculptures attracted travelers, scholars, and
adventurers. The quality of precolumbian textiles is unsurpassed in world history.
Many precolumbian ceramics are among the most lovely, whimsical, and sensual
ever created. Given these appreciations, it is perplexing that Andean art has not
stimulated the pervasive, rigorous, detailed, systematic, art-oriented study and
analysis that is characteristic of the Old World (e.g., Classical Antiquity, Late
Antiquity, Oriental, Medieval and Renaissance Europe) as well as Mesoamerica.
Anthropologists and art historians dedicated to the Central Andes have some
significant catching up to do.
Kubler (1990: 32) has attributed underdevelopment in precolumbian art-
focused studies to the co-evolution of New World archaeology and anthropology,
as compared to humanistic Renaissance learning and appreciation that was the
source of Old World archaeology. For Kubler (1991: 158), "visual evidence [i]s
primary rather than only supportive of cultural theory." He regards anthropology
(within which New World archaeology is practiced) as a limiting force on the
study of ancient Amerindian art (Kubler 1991: 177). Yet, it can and should be
counter argued that anthropological perspectives can give the study of pre-
columbian art its greatest intellectual strength, providing a human and social con-
text for the art, and enabling scholars to achieve significant understanding of the
cosmological, symbolic, ideological, and sociopolitical systems represented in it.
We believe that contributions to this volume contribute to a better anthropological
approach to ancient Andean art and will motivate more persistent attention to it.
The anthropological study of Andean art, however, has not been without its
impediments. The analysis of art requires both a depth of archaeological knowl-
edge and a meticulous attention to details that are only achieved with many years
of dedicated study, for which there is often little academic reward. Only now are
23
REFERENCES
Alva, Walter, 1988, Discovering the New World's richest unlooted tomb. National Geographic 174 (4):
510-549.
Alva, Walter, 1990, The Moche of ancient Peru: new tomb of royal splendor. National Geographic
177 (6): 2-15.
Alva, Walter, 1994, Siptin. Colecci6n Cultura y Artes del Peru. Cervecerfa Backus & Johnston, S.A.,
Lima.
Alva, Walter and Christopher B. Donnan, 1993, Royal Tombs ofSipdn. Fowler Museum of Cultural
History, UCLA.
Bennett, Wendell C. and Junius B. Bird, 1964, Andean Culture History. The Natural History Press,
Garden City.
Bourget, Steve, 1990a, Des tubercules pour la mort: analyses preliminaires des relations entre I'ordre
natural et l'ordre culturel dans I'iconographie Mochica. Bulletin de l'lnstitut Franrais d'Etudes
Andines 19 (I): 45-85.
Bourget, Steve, 1990b, Los caracoles sagrados en la iconograffa Moche. Gaceta Arqueol6gica Andina
20: 45-58.
Bourget, Steve, 1994a, Los sacerdotes a la sombra del Cerro Blanco y del arco bicefalo. Revista del
Museo de Arqueologfa, Antropologfa e Historia 5: 81-126. Universidad Nacional de Trujillo.
Chapter 2
Identification ofthe Camelid
Woman and Feline Man
Themes, Motifs, and Designs in
Pucara Style Pottery
SERGIO J. CHAVEZ
INTRODUCTION
The materials upon which this study is based primarily consist of some 10,000
pottery specimens derived from the archaeological excavations conducted by
Alfred Kidder II in 1939 at the site of Pucara (K. Chavez 1989a: 5-6). The exca-
vations included ceremonial dumps or midden deposits along the river bank
(Excavations I, II, and Ill), complex public architecture on the plain (Excavation
IV), and two temples on terraces above (Excavations V and VI). In addition to
these contexts, pottery was also found in offerings (Excavation I) and a burial
(Excavation VI); what appeared to be a domestic structure in Excavation I was
partially excavated, but clear pottery associations could not be determined
(S. Chavez 1992: 51-83).
Although most of the collection was housed at the Peabody Museum at
Harvard, Kidder left numerous specimens in the Museo Nacional de Antropologia
y Arqueologia in Lima, others went to Cambridge, England, and Vancouver,
Canada. Thanks to Kidder's notes, sketches, drawings and photographs, Karen
Mohr Chavez and I were able to identify and document most of the collection,
including materials in Cuzco corresponding to the 1955 excavations at Pucara car-
ried out by Kidder and Chavez Ballon to obtain samples for radiocarbon
35
36 Sergio J. Chavez
, 1,j~
/,\''''-'-'.11-
J•
•;- y... \ .....,
·cuzco " A,,_ ......
\,
')
~
,.. "\
+
.....T1OIlIM.1CIUCIooUI'f -'-'
OU' fWI'1'UW' -'-
.,. fWJn e.WonM. •
~~104 •
_TaCT. Mil
i..I:'::10'.':::1,,,,;Mt:,,==~'j;"r __.,;jlr\a
Figure 2.1. Map of the Lake Titicaca Basin showing the location of Pucara, and other sites
mentioned in the text.
dating (K. Chavez in press, K. and S. Chavez n.d,a). Furthermore, additional com-
parative examples were also used in the present study derived from unpublished
excavations and/or sUIface collections conducted by Julio C. Tello, Manuel Chavez
Ballon, Karen Mohr Chavez, and myself in the northern Lake Titicaca Basin.
The site of Pucara is situated in the northwestern Lake Titicaca Basin at an alti-
tude of 3900m asl (Figure 2.1) [Note 1]. The site's name was given to the Pucara
style and archaeological culture. Pucara is characterized by multiple monumental
sunken temples and other public structures in front of a gigantic natural cliff, as well
as finely carved stone sculpture, and fancy ceremonial polychrome pottery. Pucara
was partly contemporary with the late Yaya-Mama occupation in Copacabana on
the basis of five radiocarbon dates obtained from recent excavations conducted at
the temple site of Ch'isi on the Copacabana Peninsula (from inside the sunken
temple ca. 220 BC and continuing to 10 BC, uncorrected). Because of the contem-
poraneity of these sites and others in the Basin they must be included within the
Yaya-Mama Religious Tradition (K. Chavez 1997; K. Chavez and S. Chavez 1997).
The Yaya-Mama Religious Tradition was defined for the first time in 1988
by Karen Mohr Chavez (l989b; see also K. Chavez and S. Chavez 1997) and was
38 Sergio J. Chavez
Janabarriu contexts. Such increased interaction may have been stimulated by the
Yaya-Mama Religious Tradition.
o 10
.'._._' em
5 10
em
, .,I.
Figure 2.2. (a,b) Female figurines. (c) Camelid Woman Theme, pedestal-base bowl.
42 Sergio J. Chavez
Figure 2.3. (a,b) Camelid Woman Theme, from upper portions of two pedestal-base bowls.
•',.... ':
~:c.;.,
"n;;~~:
c
o 5 10
em
Figure 2.4. (a,b,c,d) Camelid Woman Heads, upper portions of pedestal·base bowls. (e) Rayed Head
Motif, lower portion of pedestal-base bowl. (f) Rayed Head Motif. (g) Rayed Head Motif, tripod or
tetrapod vessel (polished black).
areas filled with white post-fire paint that occur on the forehead, around the eyes,
earrings, or a combination of these. Furthermore, in some cases there are addi-
tional excised areas including the chin, circular collar pendants, and the tips of the
"feathers" or streamers emanating from behind the head.
This central personage has a consistent number of associated motifs,
elements, and designs which are also standardized on a red background as follows:
1) her right arm is bent at the elbow, and her hand with four fingers holds a staff
extending from forehead height to almost the bottom of the skirt; 2) she also holds
in her right hand a round item, perhaps a bag (or a ball of wool yarn?); 3) her left
arm is also bent at the elbow, and her hand holds a camelid by a rope that is in
profile facing her. The smaller camelid is portrayed above the level of her hand.
44
II
_____ --l...._-=
a
Sergio J. Chavez
o 5 10
.'_1IIII'_1Ii em
Figure 2.5. (a) Rayed Head Motif, oversized beaker. (b) Camelid Motif (alpacas with foot prints)
lower portion of pedestal-base bowl. (c,d) Camelid Motif (alpacas with foot prints) upper portions of
pedestal-base bowls.
The abundant body wool and tuft on top of the head served to identify the camelid
as an alpaca; 4) a pair of exterior lugs at the lip and at opposite sides of the
vessel, separates the central personage and the alpaca from the plant or flower-
like motifs that are portrayed from rim to waist of the vessel; 5) finally, the
pedestal portion of the vessel has additional vertical bands or panels in zigzag
flanking the personage below the waist.
Additional attributes include face markings (Figure 2.4a-d) which can be
incised lines following the lower contour of the eye and terminating towards the
cheeks (Figures 2.2c, 2.4b), a black rectangle (Figure 2.4c) or double-stepped
blocks under the eyes (Figure 2.3a), zigzag bands originating under the eyes
and extending into the lower sides of the face (Figure 2.3b), and checkered
:1
lat::MI~- ..•.
-a.:..;:..-·· .. ··········1
.'jJ J\
45
46 Sergio J. Chavez
Figure 2.7. (a) Miniature male-headed effigy jar. (b) Miniature male-headed effigy bowl.
(c,d) Regular sized male-headed effigy bowls.
(Figure 2.4a) or nested (Figure 2.4d) crosses suspended from under the eyes.
Except for one case (Figure 2.3b), the forehead is further accentuated by excised
designs including two-stepped ones divided in two (Figure 2.4b) or joined
together (Figure 2.4a), three triangles (Figures 2.2c, 2.3a), or four vertical inci-
sions suspended from a horizontal line (Figure 2.4c). The ovoidal eyes are verti-
cally divided and surrounded by an additional ovoidal red band. Sometimes the
sides of the eyes are also outlined by excised fields tapering at both extremes-
giving the appearance of almond-shaped comers (Figures 2.2c, 2.4a, b). All diag-
nostic samples include a necklace. These necklaces can have one or two rows of
beads each alternating in a combination of three colors, and in one case in two
colors and extending from ear to ear (Figure 2.3b).
em
Figure 2.8. (a) Feline Man Theme, slightly outflaring bowl. (b) Feline Man Theme,
pedestal-base bowl.
The main garment she wears is what appears to be a black dress (perhaps
sleeveless), with a cream belt, or row of rectangles at the waist (S. Chavez 1992:
figs. 152-153). A vertical band at both sides of the waist extends along the
length of her skirt like side-flaps (Figure 2.2c). Each of these side-flaps contains
two rectangular panels of double-stepped blocks, which can also be present
at her chest (S. Chavez 1992: fig. 144) or bordering a kind of veil in one case
(Figure 2Ad). Other kinds of designs or elements can also be present at her
belt/waist (S. Chavez 1992: figs.152-l53). The skirt shows a hem-line and ends
at foot height. Most of the time she wears "shoes" which are consistently cream
in color (Figure 2.2c).
The plant motif is represented in a flowering stage, with fruits/seeds stem-
ming from the petal and/or leaves which can be circular elements and/or nested
crosses (Figures 2.2c, 2.3a), or in a non-jlowering stage indicated only by petals
alone (Figure 2.3b). It should be noted that the nested cross design is also present
as face markings below the eye (Figure 2Ad). Furthermore, in the single case the
non-flowering plant is associated with the simplest camelid woman (Figure 2.3b).
She lacks the excised geometric markings on her forehead, her hair or "cape" is
not present or visible, she has a staff with rounded rather than sharp comers, and
two small pendants hanging from a necklace, which in tum is atypically attached
from ear to ear. Conversely, the blossoming/flowering plants are associated
48 Sergio J. Chavez
Figure 2.9. (a,b) Feline Man Theme (wearing a feline pelt), from two pedestal-base bowls.
with the most ornamented women holding a scepter/staff with sharp comers
(Figures 2.2c, 2.3a). Consequently, it could be argued that these associations may
indicate either differences in age, rank, and/or season of the year.
The pedestal portion of the vessel contains a series of vertical bands in zigzag,
each alternating in three colors (Figures 2.2c, 2.4e; S. Chavez 1992: fig. 153). These
bands likely fonn independent panels which flank the skirt of the woman, and in
one case there is an additional motif preserved consisting of a head with six rayed
appendages terminating in small heads (Figure 2.4e).
oII'__5'_I11III1'0
em
Figure 2.10. (a,b) Feline Man Theme, from two pedestal-base bowls.
49
(Figures 2.4f, g, 2.5a). Note the similarities between the attributes present on the
female figurine (Figure 2.2a) and the Rayed Head Motif: protruding ears, band
surrounding the face, simple horizontal mouth, connected nose and eyebrows,
simple female-related face markings. Some similarities with the camelid woman
include: possible plant motifs (Figure 2.4f) and appendages terminating in small
feline-like heads and a ring (Figure 2.4g).
On the other hand, the Owl, some Felines, and some Feline-headed Snake
motifs are not present in the main theme, but they also share many attributes
with those of the camelid woman. Some of the attributes which serve to relate the
Owl Motif (Figure 2.6) to the camelid woman include: the stepped blocks, pairs
of stepped streamers (reminiscent of the feathers worn by the camelid woman),
band surrounding the face, and the rectangular block or bands under the eyes
(see also S. Chavez 1992: figs. 198-199 for owls on oversized bulging neck jar
50 Sergio J. Chavez
and miniature jar). Conversely, the Feline Headed Snake and Feline Motifs are
the only ones sharing attributes related to both the camelid woman and feline
man, or to each of them, implying an ambiguous and/or intermediary position.
Sometimes these two motifs are portrayed together (e.g., S. Chavez 1992: fig. 297).
Some of the feline-headed snakes are just like those on the chest of the feline man
(e.g., Figure 2.8a), but are also associated with female-related stepped blocks
(S. Chavez 1992: fig. 296), which are designs exclusively associated with the
camelid woman, present on her dress (Figure 2.2c), as face markings (Figure 2.3a),
and veil (Figure 2Ad; see the description of female related geometric designs
below, and for an expanded discussion of Feline-Headed Snake Motif see
S. Chavez 1992: 342-356).
The analysis and classification of the Feline Motif show even a greater diver-
sity with a complex of associated attributes and vessel shapes including ceremo-
nial burners, regular sized and miniature jars and effigy, oversized bulging necked
jars, and trumpets (see also the identification and description of male-related
felines below). In addition to the model forfemale-related felines (Figure 2.14a),
the Feline Motif is the only one that also can be anthropomorphized. Diagnostic
attributes include a series of geometric designs which serve to relate one group of
felines to the camelid woman, as follows: interlocking stepped blocks on Register
3, are also present on the pedestal portion of burners and jar necks (Figures 2.15,
2.16b), and checkered crosses especially on the feline's body (Figure 2.16), or
lunate-shaped designs (S. Chavez 1992: figs. 375-379). The body of such felines
also show a consistent portrayal of a series of joined nested rectangles (Registers
1 and 2) on a black background/body, a cream narrow band surrounding the chest
and lower body, and a ring pendant from the neck which is usually associated
with an inverted V-shaped mouth. Furthermore, another group of female-related
felines are anthropomorphized by possessing a human arm and hand, in one case
holding a bird by the neck (Figure 2.16b).
Finally, a series of geometric designs centering mainly on checkered crosses
and interlocking or stepped blocks are derived from, and/or related to the Camelid
Woman Theme and related motifs and elements. They can be classified as female-
related (Figure 2.19b).
52 Sergio J. Chavez
(which may be plucked faces or may not be preserved). However, attributes link-
ing with the miniature vessels include the staff element at the forehead, and the
three stylized heads on one side of the face which also begin above the eye and
continue below it. Furthermore, both bowls also possess the same kind of pro-
truding chins, thick lips in relief, and hair parted in the middle of the forehead (for
more details see also S. Chavez 1992: figs. 202-203); and in at least one speci-
men (Figure 2.7c) a red ear is preserved with a variation of the horizontal
T-shaped element in the center.
New attributes include a face which is cream on the left and red on the
right, and an impressive bird on the right side that stands in profile almost
surrounding the eye (Figure 2.7d). Based on the long neck and shape of beak, this
bird appears to represent a black flamingo. The face markings on the other effigy
bowl (Figure 2.7c) include on the right side two branching bands emerging under
the eye and surrounding its outer portions, where one terminates above the eye in
a black triangular element, and the other appears to be connected to the staff
element at the forehead. The other face marking also has a band under the mouth
terminating in the small stylized head element which is similar to that at the
forehead of the miniature jar, and two additional bands in zigzag continue above
the eye.
Based on these three effigy bowls and jar, we can generate a list of attributes
associated with the portrayal of male supematurals. Significant attributes, which
stand in clear contrast or opposition to those of the camelid woman, include:
(1) presence of moustache or moustache and goatee beard; (2) hair and braids
parted in the middle of the forehead (rather than plain hair following the hairline);
(3) thick red lips in relief and/or thick incision in the middle which is painted
black (rather than a single horizontal incision); (4) unlike the ears in the camelid
woman, those on the male heads are less pronounced with ear folds, and lack
earrings; (5) protruding chin; (6) symmetrical or asymmetrical faces with more
elaborate face markings (unlike the consistently red faces and simple face markings
on the camelid woman); (7) absence of excised areas or excised face markings.
It remains for me to make the crucial connection between the three-
dimensional effigy heads and the two-dimensional personage of the feline man.
Beginning with the face markings, most can be directly linked to or duplicated in
those on the male effigies, and hence, serve to identify the main personage as
male. For example, one class of appendage (Figure 2.7c) is virtually duplicated
on a feline man (Figure 2.8a). Since the head of the feline man is tilted upwards
with a crown and in profile, the hairlbraids are absent or not visible; hence, this
attribute will remain useful only in contrasting the male-female differences in the
previous step among the effigy bowls.
The thick lips are maintained in the feline man, although the mouth is open
with teeth and a pair of fangs, and a long appendage(s) emanates from the mouth.
The use of cream color for the face is consistently present on all feline men, as
54 Sergio J. Chavez
Another significant attribute is the kind of garment and its associated orna-
ments. The personage wears what appears to be tight-fitting pants and a long-
sleeved shirt, where one arm and one leg are black and the others alternate in cream.
A bandlbelt at the waist contains horizontal zigzag lines or triangles-a pattern
repeated at the base of some crowns and axe handles. Sometimes a feline-headed
snake motifis present in the middle of the chest (Figure 2.8a). Additional elements
include a circle at the hip and shoulder joints, and a long bent appendage emanat-
ing from the middle of each bracelet and anklet. Furthermore, suspended from the
personage's neck and extending onto the shoulders, is a flaring-shaped collar. Such
collars can also be seen on Pucara style anthropomorphic stone statues (e.g.,
S. Chavez 1982: fig.l). Some feline men are portrayed wearing actual feline pelts
with nested diamonds on a cream background, a ring feline nose, and more feline-
like ears (Figure 2.9); but the collar and belt are still also part of their garments.
Specimens with large portions of the theme preserved show that additional
human legs and arms lie by themselves in empty spaces. The legs can be cut at
the thigh or femoral region or at the hip, and are bent with a semi-circle at the
knee (Figures 2.8b, 2.l0b). There are striking similarities between the feline man
and the severed legs, arms, and some of the severed heads-a situation which
may indicate that those body parts correspond to some feline men victims (see
motifs and designs derived from the feline man below).
Based on the classification and study of the different crowns, staffs, axes, and
face markings, it can be proposed that there are different kinds of feline men, and
perhaps each holds the same or similar rank. If we consider the crown to be
an important indicator of rank, then, there could be at least seven positions or ranks
of feline men. However, considering the staff, axe, and face markings, then, there
are perhaps four or five feline men. In addition, if the six known Pucara temples
were functioning at the same time, then, we could argue that each temple also
required a feline man to control and represent it.
oI
0-
c
o 5 10
em
Figure 2.11. (a) Severed Head Motif, on the interior of slightly outflaring bowl. (b) Severed Head
Motif, miniature jar. (el Severed Head Motif, trumpet. (d) Severed Head Motif, on the interior of
slightly outflaring bowl.
portrayed in profile in a repetitive fashion, and usually have red faces. Some also
show attributes present on the feline man as well. Therefore, severed heads are
classified into plain red faces (Figure 2.11 a--e); those with a face marking (like
those on the miniature effigy jar, Figure 2.7a), and a ring at the forehead
and upper nose (Figures 2.11d, 2.l2a), or associated/alternating with severed
arms (Figure 2.12b). Additional severed heads with complex face markings
(Figures 2.l2d, f) show striking similarities with those present on some of the
feline men (Figures 2.8a, 2.9), as well as one with a flamingo face marking
(Figure 2.l2g) which is most like the one on the effigy bowl (Figure 2.7d).
Furthermore, miniature effigy jars having closed eyes may also be considered
independent severed heads (Figures 2.7a, 2.l2e).
The Guanaco and Bird Motif is portrayed on the interior and exterior of a
large outflaring bowl (Figure 2.13). The identification of the wild camelid is
based on attributes which are absent on the female-related domestic camelid/
alpaca associated with the camelid woman, and includes: lack of abundant wool,
homogenous cream-colored body, a streamer emanating from under the ear, and
guanacos are portrayed in a running or jumping position (instead of the station-
ary and controlled position with a rope at the neck). The male-related attributes
are seen in between each of three pairs of guanacos, depicted counterclockwise
on the interior of the bowl, where a smaller and simpler version of the staff with
56 Sergio J. Chavez
.,:.....5....'0
I :I 6%";; t""';":'::;:ii>.;';::'<D /
Figure 2.12. (a) Severed Head Motif, neck portion of jar. (b) Severed Head and Arm Motif, neck
portion of smaJljar. (c) Dead Human Body Motif, on the interior of outflaring bowl. (d) Severed Head
Motif, on the interior base of outflaring bowl. (e) Severed Head Motif, miniature male-headed effigy
jar. (f) Severed Head Motif, probably a slightly incurved bowl. (g) Severed Head Motif, slightly
incurved deep bowl.
Figure 2.13. Guanaco and Bird Motif (guanacos on interior. and guanacos and birds on exterior).
large slightly outflaring bowl.
Figure 2.14. (a) Model for female-related feline motifs. (b) Model for male-related feline motifs.
58 Sergio J. Chavez
Figure 2.15. (a) Feline Motif (female-related), pedestal-base bowl. (b) Feline Motif
(female-related), upper portion of pedestal-base bowl.
one which can also be anthropomorphized (Figure 2.18b, note the human arm and
band emerging from the feline's front body). Some male-related felines can also
be seen in association with a series of human severed heads (Figure 2.l8a).
The attributes used to relate this group of felines to the feline man derive
from the kind of feline pelt some feline men wear (Figure 2.9, note the feline tail
and ring nose), where a series of independent nested diamonds are present on a
cream backgroundlbody (different from the connected nested rectangle on a black
body present on female-related felines). Therefore, since this kind of design
and pattern are also duplicated on a group of feline motifs, their male-related
5 10
em
~-~
,, '' '!It ~
.~~_:...'
o 10
1...'_.....I._-J'cm
Figure 2.16. (a) Feline Motif (female-related, with additional checkered cross on the base exterior),
miniature jar. (b) Feline Motif (female-related and anthropomorphized), regular sized jar.
60 Sergio J. Chavez
o It)
'-'__'--_...'. em
o 5 10
em
Figure 2.17. (a) Feline Motif (male-related), pedestal-base bowl. (b) Feline Motif (male-related),
miniature feline effigy jar.
stand in contrast to those derived from the camelid woman. Such designs are for
the most part variations of the reclining-S designs (Figure 2.l9c).
III
o 5 10
L' "--__.' e",
"
Figure 2.18. (a) Feline Motif (male-related, associated with severed heads), trumpet. (b) Feline
Motif (male-related, anthropomorphized), oversized bulging necked jar.
•o
...
Qa
GO
...~
J.
""
ItS]..., I
C1I
...
Figure 2.19. (a) Winged Man Motif (and related birds), oversized beaker. (b) Geometric designs
derived from and/or related to the Camelid Woman
Theme. (c) Geometric designs derived from and/or related to the Feline Man Theme.
o 5 10
em
... K
...~
..
..~ CIl
.n.>.
<19.
Q
~
<1
"=','
~
64 Sergio J. Chlivez
The only other theme in Pucara is the Feline Man Theme. This male person-
age also has supernatural indicators and a different pose which I call the feline
man pose. Based on the different kinds of crowns, staffs, face markings, and
attire, it is possible that there is more than one male personage represented, and/or
these may indicate differences in rank. Unlike the camelid woman, the feline man
is always depicted in profile and in pairs (either facing or chasing each other), and
very rarely the head may be in low relief. Such a unified composition, I would
propose, also represents a myth or story different from that of the Camelid
Woman Theme, where the main supernatural male personage is associated with
his activities, life-taking attributes, power, and supernatural domain.
Other realms of the natural world may be incorporated into his domain or
sphere of activities, as follows: the black-bodied flamingo (present on half of the
face of a male-headed effigy bowl), could be the Phoenicopterus chilensis which
possesses black wings, or grey plumage when immature. This bird is known to fly
into the Lake Titicaca Basin during the dry season to feed on salt-water mollusks
and crustacea (Aparicio Paredes 1957: 29-30), and serves to assign the colder
dry season in the feline man's domain. Also, the male-related Guanaco and Bird
Motif overwhelmingly indicates an association with a wild camelid and a raptor-
ial bird. The only other motif which is even more closely associated with the
Feline Man Theme is the male-related feline, whose pelt is also worn by some
feline men. The ferocious and man-devouring qualities may have also been
sought after or acquired by the feline man, especially since he consistently pos-
sesses the kind of fanged teeth present only on some male-related felines with
body color and markings resembling those of the jaguar. This association with
wild animals and qualities is overwhelmingly different and contrasts to the
domestic alpaca in the Camelid Woman Theme. Likewise, the active and life-
taking character of the feline man stands in clear contrast with the more passive
and fertility-oriented depictions of the Camelid Woman Theme.
On a different level of interpretation, I would also propose that inherent in
the Camelid Woman Theme is an economic aspect, an association with economic
activities. The presence of wool-producing alpaca, the possible wool yam she
holds, the staff which may be a distaff, as well as plants, lends support to the argu-
ment that she may have possessed an intermediary role between these natural
realms and supernatural forces regulating them. Specifically, the uncertain nature
of agricultural production and herding as they relate to, for example, cycles of
drought, flooding, frost, reduced pasturage and water, may have been attended to
by the camelid woman to regularize, manage, or seek out the goodwill of nat-
ural/supernatural forces which ultimately controlled the outcome of production.
Conversely, the Feline Man Theme reflects a political aspect, one involving
power, control, conflict, competition, and even non-local contacts, conflicts, or
alliances with other groups. In this discussion it is also crucial to incorporate the
associated motifs containing human severed heads, legs, arms, and headless
Identification of the Camelid Woman and Feline Man 65
bodies, which bear striking similarities to those of the feline man. Based on
analysis of a number of shared attributes, I have proposed that severed heads
represent individuals who at one point possessed a rank close to the feline man
himself; that is, the resulting three to four tiered hierarchy of heads were ranked
from those least like the feline man (the most frequent) to those most like him (the
most rare). One of these groups possesses different face markings and hair. And
its paste is non-local, indicating that the high ranking individuals represented may
have been non-local, although probably within Pucara's sphere of influence.
I suggest that in Pucara politics, both local and non-local members participated in
a ruling elite of high status leaders possessing different ranks, perhaps corre-
sponding to a hierarchy of cultural centers (religious, political). They may have
been in competition or conflict with elites of other centers, including Pucara itself
such that decapitation expressed differing relations within the authority structure.
The severed head displayed publicly or portrayed in a repetitive fashion
magnifies the violent act, while at the same time suggesting social control by
threat of force or visual terrorism to prevent some act or conflict. Likewise, the
imagery is intentionally standardized, suggesting centralized control and manip-
ulation-a situation in which religion backed by actual decapitation by force pro-
vides a means of, or rationale for ruling by force or threat of force, especially
when religious or supernatural threats are no longer effective in ruling by persua-
sion (this interpretation was developed and proposed to me by Karen Mohr
Chavez). In addition, the use of force or threat of force is supported by the one
hundred mandibles/skull fragments uncovered in a public ceremonial context in
Excavation IV (S. Chavez, 1992: 63-64, fig. 2). However, this association also
implies a religious function for decapitation, perhaps involving human sacrifice.
Centralized control of decapitation that is publicly ritualized may express and
sanctify authority given to a restricted group, suggesting a monopoly over force.
This is as close as we get to declaring Pucara a pristine state on the basis of
iconography.
I also argue that in addition to this material/political interpretation, there also
is a ceremonial/mythical component to the iconography, and that practices may
have served both political as well as ceremonial ends. In this respect, the feline
man is shown carrying or ostentatiously displaying the severed parts, and never
actually severing them (hence, the term "sacrificer" was purposefully omitted
here). These body parts are depicted on fancy ceremonial shapes such as trum-
pets, ceremonial burners, and miniature vessels containing lime which is chewed
with coca-a frequent ingredient in rituals, and many miniatures occur in temple
contexts. On the other hand, the one hundred mandibles/skull fragments may
have been those of ancestors or important people, whose skulls were kept as relics
and as the focus of pilgrimages. Such pilgrimages, manifested ethnographically
as annual fairs at Catholic shrines, take place during the dry season when travel
is easier and there is no conflict with rainy season agricultural or herding activities.
66 Sergio J. Chavez
It is at these times when contact and exchanges between groups from other envi-
ronmental zones occurs (e.g., K. Chavez 1992: 69), and it has been argued that
the feline man has a dry season affiliation.
The presence of severed human heads as well as their artistic representations
are too often interpreted as evidence for warfare and termed "trophy heads:' to
imply trophies taken in battle. An assessment of Pucara sites in the region that could
provide clear indicators of warfare (e.g., weapons, fortifications, defensive earth-
works, and settlements in strategic locations), however, has failed to present evi-
dence for frequent, widespread, endemic, or reoccurring warfare. In this respect, I
have made a general survey of different periods in Andean prehistory when indica-
tors of actual and widespread militarism were present or absent. In this survey it
was concluded that periods when militarism is absent or not so clearly indicated
(that is, when actual uncontrolled fighting is less common), are times when violent
motifs and scenes are depicted (such as in Pucara), suggesting its use as a means of
social control to warn by displaying terrorist imagery in the absence of, or as a sub-
stitute for widespread warfare or rule by force. Conversely, during periods of
endemic warfare, violent iconography appears to be no longer necessary.
The documentation of two primary themes/supernaturals that are never
depicted together, and appear to represent distinct domains contributes to under-
standing the antiquity of principles of complementarity, so prevalent in Andean cul-
ture. The Pucara case constitutes the earliest evidence for the Lake Titicaca
higWands. Even when ambiguous imagery occurs, it employs felines, not the two
supernatural personages themselves. Furthermore, because both themes co-occur
in public religious contexts, both were being controlled at the same temples or pub-
lic ceremonial locations by emerging elites. Hence, I also propose that control over
the production of images (in pottery and perhaps other media as well), promoted
access to supernatural power that underwrote political and economic authority.
Based on the identification and classification of paste groups conducted by
Karen Mohr Chavez and confirmed by geological/petrographic studies (K. Chavez,
Chyi, S. Chavez 1988; S. Chavez 1992: 519-521), an overwhelming majority of
decorated fancy pottery were most certainly made locally. Furthermore, Pucara
style pottery shows a higher degree of elaboration and standardization than many
of the contemporary or partially contemporary pottery styles also associated with
Yaya-Mama sites in the southern portion of the Titicaca Basin (K. Chavez
2002)-despite their sharing the same components of the Tradition (temple archi-
tecture, ritual paraphernalia, and iconography). The handmade and labor inten-
sive techniques include: fine, broad, and regular incisions outlining polychrome
fields; excision; applique; carving of impressions for inlays; pre-fire painting and
higWy polished surfaces including the interior and exterior base bottoms; post-
fire painting within incised and excised areas; and combination of high
relief/three-dimensional figures with two-dimensional ones on a single vessel.
68
Acknowledgments
Sergio J. Chavez
I wish to dedicate this article to the memory of Karen Lynne Mohr Chavez,
my source of constant inspiration. She and John H. Rowe guided me and provided
valuable comments during the different stages of this research. Special apprecia-
tion also goes to Stanislawa Stachniewicz, who greatly helped in the preparation
of this article, and to Catherine Julien for her editorial comments on the intro-
duction. The illustrations were inked by Edwin Chavez, Geoffrey Deventier,
Teobaldo Y<ibar, and myself based on my original tracings and pencil drawings.
An earlier version of this paper was presented during the 33rd Annual Meeting of
the Institute of Andean Studies in Berkeley in 1993.
Notes
I. The color key for this chapter is as follows. Black represents black pigment, white refers to cream,
and dots are for red. Discontinuous horizontal lines/ dashes refer to excision, and small v-shaped pat-
terns refer to broken or worn surfaces. Most associated shapes were eliminated, but the reader can
refer to S. Chavez 1992 for illustrations and descriptions of shape and other details not visible here.
REFERENCES
Aparicio Paredes, S. Manuel, 1957, Aves del Titicaca. Editorial Garcilaso, Cuzco.
Burger, Richard, Karen Chavez, and Sergio J.Chavez, 2000, Through the glass darkly: prehispanic
obsidian procurement and exchange in southern Peru and northern Bolivia. Journal of World
Prehistory 14 (3): 267-362.
Chavez, Karen L. Mohr, I986a, Early Tiahuanaco-related ceremonial burners from Cuzco, Peru.
Dialogo Andino 4 (1985): 137-178.
Chavez, Karen L. Mohr, I 986b, Carved stone bowls from Kidder's excavations at Pucara, Puno, Peru.
A link between Pucara and Tiahuanaco. Paper presented at the 26th Annual Meeting of the
Institute of Andean Studies, Berkeley.
Chavez, Karen L. Mohr, 1989a, Alfred Kidder II: 1911-1984. Expedition 30 (3): 4-7.
Chavez, Karen L. Mohr, 1989b, The significance of Chiripa in Lake Titicaca Basin developments.
Expedition 30 (4): 17-26.
Chavez, Karen L. Mohr, 1992, The organization of production and distribution of traditional pottery
in south highland Peru. In Ceramic Production and Distribution: An Integrated Approach, edited
by George J. Bey III and Christopher A. Pool, pp. 49-92. Westview Press, Boulder.
Chavez, Karen L. Mohr, 1997, The temple site ofCh'isi on the Copacabana Peninsula, Bolivia: A view
of local differences and regional similarities within the Yaya-Mama religious tradition. Paper pre-
sented at the 62nd Annual Meeting of the Society for American Archaeology, Nashville.
Chavez, Karen L. Mohr, 2002, Local differences and regional similarities in pottery of the
Yaya-Mama Religious Tradition. Paper presented at the 67th Annual Meeting of the Society for
American Archaeology, Denver.
Chavez, Karen L. Mohr, in press, Alfred Kidder II and the history of Andean Archaeology: A
biographical and contextual view. In press in Andean Past.
Chavez, Karen L. Mohr, Kwo-Ling Chyi, and Sergio J. Chavez, 1988, Physico-chemical analysis of
Pucara style pottery from Pucara, Peru. Videotape presentation at the 16th Annual Midwest
Conference on Andean and Amazonian Archaeology and Ethnohistory, Ann Arbor.
Chapter 3
Differentiating Paracas
Necropolis and Early Nasca
Textiles
HELAINE SILVERMAN
INTRODUCTION
Bean (1998: 115) has argued that "[nlext to language, cloth and clothing animate
the most elaborated systems of representation in social and cultural life. In iden-
tity constructon and manipulation ... dress dominates the interface between peo-
ple and their social worlds." Vivid examples of Bean's contention abound across
the world. Indeed, one of the foundational texts in the anthropological study of
identity construction is Wobst's (1977) study of stylistic behavior and information
exchange manifested in the headdresses worn by members of different ethnic
groups in what today is the former Yugoslavia. In addition to the construction of
ethnic identity, dress is a dramatic expression of class, ideology, role, occupation,
and gender.
In this chapter I am concerned with one dimension of expressive identity
manifested in costume, that of ethnic or cultural identity. Specifically, I consider
the identity of the adult men who wore in life the extraordinarily beautiful and
technologically superb garments excavated by Julio C. Tello (1959; Tello and
Mejia Xesspe 1979) in cemeteries along Paracas Bay on the south coast of Peru.
This problem has confounded archaeologists for decades (Dwyer 1971: 220-223;
Kroeber 1944: 32-34; Peters 1991; Rowe 1995: 35-38; Silverman 1991; Strong
1957: 14, 16; Yacovleff and Muelle 1934 inter alia) because the complex poly-
chrome images on the textiles resemble complex polychome images on Nasca
pottery from the lea-Nazca region [Note 1], while the thin-walled monochrome
71
72 Helaine Silverman
pottery in apparent association with the mummy bundles is stylistically compara-
ble to the Topara ceramic tradition of Pisco, Chincha and Cafiete (see Lanning
1960; Wallace 1986). My second goal in this paper is to define the textile tradi-
tion corresponding to Early Nasca pottery and to evaluate the competing textile
definitions in terms of the ramifications of stylistic identity and stylistic influence
for the interpretation of south coast culture history two thousand years ago.
74 Helaine Silverman
ceremonial life of the period." She argues that Early Nasca textiles extended and
developed the Necropolis tradition, "especially a preoccupation with needlework
as the dominant mode of textile expression, although I do not see the identical
production of mainstream Necropolis-style textiles at the site [Cahuachi]. Instead,
the three-dimensionally constructed needlework [cross-knit looping] took on
distinctive and dominant characteristics, elaborately decorating the edges of the
textiles" (Phipps 1989: 314). Phipps notes that "the central figures found in the
Paracas Necropolis mantles are not found in the materials coming from
Cahuachi." She concludes that Cahuachi's textiles belonged to a local style and
were locally produced. According to Phipps (1989: 3l~3l7), Nasca people
inherited, transposed, and transformed the Paracas Necropolis textile tradition.
The difference between the two textile traditions was succinctly expressed in
1954 by Junius Bird and Louisa Bellinger in their choice of title for a "catalogue
raisonne" of the Textile Museum's south coast textiles: Paracas Fabrics and
Nazca Needlework.
Because Nasca 1 pottery does not constitute adequate antecedents for the
iconographic explosion that occurred in Nasca 2 pottery (as seen on, e.g., the
Haeberli Panpipe: Figure 3.1; the Bernstein/Guggenheim Drum: Figure 3.2) and,
especially, Nasca 3, some scholars argue that Paracas Necropolis textile imagery
was transferred to the medium of Nasca ceramics (Dwyer 1971; Peters 1997).
But, Sawyer (1997: 42) attributes the poor impression of Early Nasca textiles to
the fact that no high status Nasca burials have been scientifically excavated.
Sawyer (1997) argues that extraordinarily fine and iconographically elaborate
Early Nasca needlework exists, but has long been been misidentified as Paracas
Necropolis in style and origin. Sawyer (1997) contends that Nasca 2 people made
embroideries whose subject matter is as vast as that present on the Paracas
Necropolis textiles. He says that this largely unknown corpus of elite Nasca 2 tex-
tiles provided the visual language for Nasca 3 pottery without calling into playa
transference of imagery from Block Color Paracas Necropolis textiles.
Sawyer (1997: 27) specifically argues that double-faced embroidery, the
double running stitch, and cross-knit looping were textile techniques executed
only by Nasca people. He emphasizes embroidery thereby making Early Nasca
textiles technically comparable to those of the Paracas Necropolis. Sawyer (1997:
97) also argues that the Early Nasca use of color on textiles is "sumptuous," par-
ticularly in comparison to the darker tones of Paracas Necropolis textiles, in spite
of the latter's rich polychromy.
Furthermore, in a reversal of the usual argument, Sawyer (1997: 33) says
that Early Nasca textiles of lea and Nazca evolved rapidly and influenced the
development of the more naturalistic Block Color style of Paracas Necropolis tex-
tiles. Sawyer (1997: 42) regards Paracas Necropolis Block Color textiles as more
formal and less iconographically complex than those made by Nasca people.
Specifically, he observes that "Necropolis Block Color embroideries display one
or two deity figures related to the ritual function of the deceased. In contrast, elite
Figure 3.1. The Haeberli Panpipe. Nasca 2. (photo: courtesy of Joerg Haeberli).
76 Helaine Silverman
Figure 3.2. The magnificent Nasca 2 drum originally in the Solomon R. Guggenheim Museum and
published by Sawyer (1968: fig. 403), now in the collection of David Bernstein (David Bernstein Fine
Art, New York City).
78 Helaine Silverman
Table 3.1. Similarities between EIP 1 Paracas Necropolis Textiles and
Nasca 1 Pottery.
Paracas Nec. image/theme Mummy Bundle No.1 Date Nasca I pot published in
publication figure no.
spotted cat/pampas cat 3781 Paul I990a: 7.14 EIP IB Wolfe 1981: fig. I
with and without trophy Purin 1990: fig. 035
head Purin 1990: fig. 036 right
Lapiner 1976: fig. 473
serrated serpentine figure 310/Peters 1991: 7.66 EIP IB Strong 1957: fig. 10j
human in spotted cat costume 378IPaui 1990a: plate 13 EIP IB Purin 1990: fig. 036 (see
eating trophy head fig. 035 without trophy head)
human in killer whale aspect, 89IPaui 1991b: 5.7b EIP IB/2 Lapiner 1976: fig. 469
clutching trophy head, 378IPaui 1990a: plate 21 EIP IB
with or without forehead
ornament
human holding trophy head 310/Peters 1991: 7.68 EIP IB Purin 1990: fig. 033
human wearing forehead 89IPaui 1991 b: 5.9 EIP IB/2 Lapiner 1976: fig. 466
ornament and/or mouthmask
double-headed cat-face 310IPeters 1991: 7.64 EIP IB Donnan 1992: fig. 75
serpentine creature
beans 310/Peters 1991: 7.46,7.76 EIP IB Donnan 1992: fig. 72
birds 89/Peters 199\: 7.2\ EIP IB/2 de Lavalle 1986: 120 top,
262/Peters 1991: 7.26,7.28 EIP IB 122
310/Peters 1991: 7.31,7.32 EIP IB
camelids 262/Peters 1991: 7.51 EIP IB Helsinki panpipe: Purin
1990: fig. 138
Dwyer and Dwyer (1975: 151) state that among some rich bundles from the
Paracas Necropolis, later textiles were found among the outer wrappings and ear-
lier textiles in the inner mass. The Dwyers suggest periodic renewal of offerings.
Paul (1990a: 9) counter argues that "a bundle was constructed within a short
period of time and ... its contents were roughly contemporary in origin." This dis-
agreement is important because, furthermore, Frame (1995: 8, 15) argues that
there are Nasca textiles in various of the latest Paracas Necropolis bundles. She
says that these iI'.trusive textiles are recognizably distinct from the Paracas
Necropolis repertoire because they exhibit innovative features in the border, cen-
tral band, tab, and layout of the ground cloth as well as having a new kind of
embroidery-a double-faced interlocking stem stitch which created an image
visible on both sides of the fabric. If Frame is correct, this would be a case of
actual material exchange (see below). In this regard, I am unaware of EIP 1 or EIP
2 Paracas Necropolis style textiles in Nasca 1 or Nasca 2 contexts in lea or Nazca.
I have not seen provenienced Nasca 1 polychrome slipped and incised pottery or
80 Helaine Silverman
Tombs 26 and 38 at Pena de Ocucaje in lea (see Dwyer 1971: 133) and an
embroidered fragment from the Nasca 1 Cordero Alto site in lea published by
Sawyer (1997: fig. 15). Speaking of Pena de Ocucaje, Dwyer (1971: 133) writes:
"As far as can be determined from Dawson's drawings and photos [of the Rubini
gravelots] these textiles did not differ significantly in style from contemporane-
ous specimens from Paracas. However, an exact comparison as to color, etc. was
not possible." In the absence of full description of the Ocucaje materials and
given the limited information provided by Sawyer about clear Nasca 1 textiles as
well as the lack of reported associations of other Nasca 1 pottery and textiles, to
all intents and purposes the Nasca 1 textile style is unknown.
Figure 3.3. The Helsinki panpipe (Didrichsen Art Museum). The figure wears a tunic and leggings,
a forehead ornament, elaborate recurving mouthmask and earrings; the head is inverted; above the
forehead ornament is a domed black cap topped by a rayed/undulating headdress; the notch between
a camelid head's ears fits around a projection of the cogged headdress; the figure's protruding tongue
is a serpentine figure; the figure's feet have thumbs and rest on a serrated serpentine figure; one hand
grasps a human trophy head with painted face, the other a forehead ornament. A Nasca I sherd
recovered by Strong (1957: fig. IOj) at Cahuachi shows parts of a trophy head and serpentine figure
in similar relation.
82 Helaine Silverman
Table 3.1. In addition, the Helsinki panpipe shares overall attributes with a bird
impersonator figure on Mantle 15 from Mummy Bundle 89 (Figure 3.4; Mummy
Bundle 89 is dated to EIP IB by Dwyer [1971: appendix B) and Paul [1990a: tables
5.2,5.3], but Paul [1991b: 183] dates it to EIP 2). Similar comparisons also may be
made betweeen the Helsinki panpipe figure and a more complex textile image in
Mummy Bundle 38 (Peters 1991: fig. 7.54; Mummy Bundle 38 is dated to EIP IB
by Paul [1990a: tables 5.2, 5.3], but Dwyer [1971: appendix B) dates it to Nasca 2).
On the other hand, there are Paracas Necropolis textiles dated to EIP lA
(e.g., a border fragment from Mummy Bundle 243 [Peters 1991: fig. 7.41]) and
ElP IB (e.g., mantle 24 from Mummy Bundle 310 [Paul 1990a: fig. 7.57, plate
25]) that seem to anticipate later iconographic developments on Nasca pottery
such as the Trophy Head Taster of Nasca 3 (Wolfe 1981: fig. 212) and Harpy Bird
Figure 3.4. Figure from Mantle 15 of Paracas Necropolis Mummy Bundle 89. It has in common with
the personage of Figure 3.3 the forehead ornament, mouth mask, a serpentine serrated tongue pro
jecting from a mouth in which the two teeth rows are clearly indicated, a tunic, dangling legs whose
feet point in the same direction and are thumbed, wings, and a forehead ornament grasped in one hand
in the case of the panpipe figure and in both hands on the Paracas Necropolis mantle. The panpipe
figure holds a human trophy head in the other hand. The panpipe figure's head is fully inverted; the
textile figure's head is turned sideways (redrawn from Paul 1991b: fig. 5, 6).
84 Helaine Silvennan
to Phipps (1989: 269), Haeberli (1995), and Sawyer (1997), BMT could have
come from Nazca and may be a Nasca 2 piece. Peters (1997: 867) considers BMT
to exemplify the "fuzzy interface between the Topani complex and artifacts of the
Nasca tradition in EIP 2." She says that BMT, a cross-knit looped textile, is "nearly
indistinguishable in style and iconography from the Necropolis embroideries, con-
trasting only in its emphasis on plant imagery. It represents a wide range of the late
Necropolis Block Color imagery, as if interpreted with precision and encyclopedic
knowledge by the craftsperson(s) of a different technical tradition, perhaps from
the Nazca area" (Peters 1991: 314). She continues that "[i]n general, the use of
cross-knit looping to form independent Block Color figures as a fringe-like border
around textiles is characteristic of Nasca 2 and absent from the Topara complex
burials" (Peters 1997: 867). Phipps (1989: 269) specifically argues that BMT and
GEMT "are most likely to have come from the Nazca Valley... From a techno-
logical point of view, as well as stylistic ... they conform to the aesthetic which is
originating in the Nazca Valley, and represent a transformation of the brilliant
achievements from the slightly earlier period."
BMT has an elaborate border composed of 90 intricate three-dimensional
crossed-knit looped figures attached by a band of three-dimensional flowers to a
plain brown cotton ground cloth that is decorated with 32 rayed heads (4 each in
8 rows) executed by the technique of warp-wrapping. The ground cloth heads
alternate by color (brown, yellow, green, pink, red, blue). In style, these heads are
readily comparable to King's (1965: figs. 42, 76a, b, c) "sun faces" from Ocucaje
in the lea Valley and surely represent the Oculate Being (compare to Dwyer 1979:
fig. 11, a bodiless Oculate Being from Caverna Vat Paracas dating to EH 9, also
fig. 23, a full-bodied flying Oculate Being in the Broad Line style, from Mummy
Bundle 410, dated to EH lOB [see Dwyer 1971: appendix B]). They also are com-
parable to the image on a textile published by Engel (1991: fig. 70), said to be
from Cabezas Largas at Paracas. Given the ubiquity of the "sun face" in lea, I
believe BMT's ground cloth was made in lea or by people from lea or by people
closely related to and in contact with those of lea. In their petaled form, alternat-
ing color scheme (blue, yellow, red, green) and central point of different color, the
three-dimensional flowers are reminiscent of but not identical to the embroidered
flower mantle ("Textile Specimen 3") from Paracas Necropolis Mummy Bundle
318 (de Lavalle and Lang 1983: 61), dated to EIP 2 (Paul 1990a: tables 5.2, 5.3).
It has a ground cloth embroidered with eight-petal flowers of alternating colors
and a cross-knit looped border band of three-dimensional birds, fully within the
Nasca tradition, attached to the ground by multicolored tabs sewn to the cross-
knit looped edging. That mantle, according to Frame (1995), was a Nasca 2 tex-
tile gifted to the individual in Mummy Bundle 318.
Stylistically, the three-dimensional figural border of BMT post-dates
the ground cloth (though archaistic imitation can not be excluded), based on the
obvious comparison of the 32 heads to lea Valley Oculate Being imagery and the
86 Helaine Silverman
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
88 Helaine Silverman
Table 3.3. (Continued)
GEMT ground figure no.
(following Paul 1979: 46-48)
Table 3.4. Similarities between GEMT and Early Nasca Textiles and
Nasca 1-3 Pottery.
GEMT ground figure no.
(following Paul 1979: 46-48)
31
12
13
14
B (falcon)
15
16
17
23
24
7
8
10
II
C (bird)
D (human figure)
90
Other Textiles
Helaine Silverman
91
92 Helaine Silverman
drum published by de Lavalle (1986: 135). The forehead ornaments, mouth
masks, tunics, earrings, and leggings on GEM are generic and, as such, compa-
rable to those on all the textiles and pots mentioned. But my gut feeling is that the
painted ground cloth of GEM is a Nasca 2 product while its embroidered border
is Paracas Necropolis in style.
Textile Museum 91.205 is identified as a Paracas altar (?) cloth by Bird and
Bellinger (1954: 53, 54, plates XVI-XIX); it is an example of the Nasca 2 inter-
locking warp and weft style (Rowe 1972). In my experience, the layout of the
ground cloth with repeating central anthropomorphic figures enclosed within
rhomboidal frames is unique for the south coast at this time.
In addition to the textiles discussed above, Sawyer (1979) presents three
fragments of painted textiles which he identifies as Nasca 2 (his Early Nazca
Phase A). One of these is the oft-published CMAT. Another is Textile Museum
1965.40.26a (Sawyer 1979: fig. 3). The other is American Museum of Natural
History 41.217082 (henceforthAMNH; Sawyer 1979: fig. 4). Sawyer (1979: figs.
4, 5) compares AMNH to the design on a Nasca 2 interior-decorated bowl.
Because it naturalistically depicts a condor, it also is similar to Paracas
Necropolis imagery (see Pau11990a: fig. 11.23). In overall composition and style
of painting AMNH is quite similar to other Early Nasca textiles (e.g., Lapiner
1976: fig. 476) and I accept Sawyer's Nasca 2 attribution.
Figure 3.8. The complete figure on the Nasca 2 painted cotton ceremonial cloth in the Norweb
Collection at the Cleveland Museum of Art, accession number 1940.530. (copyright permission: The
Cleveland Museum of Art, 2001).
painted medium. I agree with Sawyer's EIP 2 dating of CMAT and believe that
CMAT is a Nasca 2 product.
94 Helaine Silverman
and style of the textile suggest a Karwa provenience, not to mention the fact that
the cited location has no correspondence to actual geography. Given the reality of
the illicit Peruvian antiquities scene, I urge a degree of scholarly skeptism in
accepting proveniences for looted materials. Recent history shows that looters
both disguise the locale of a rich strike to mislead potential competitors and they
attribute their finds to a particular place so as to enhance the commercial value.
The textiles are "said to be from Cabildo" (Sawyer 1997: 45, 46). Sawyer
describes Cabildo as "a large important site." Presumably, it is in the immediate
vicinity of the eponymous contemporary farming settlement in the middle Grande
Valley. Sawyer says that "Archaeologists have not excavated at the site since most
of it is under cultivation and is covered by 10-13 feet ... of powdery alluvial soil,
making excavation difficult and hazardous. Local antiquarians [surely a euphe-
mism for antiquities dealers] state that huaqueros have found several rich Nasca
tombs at Cabildo in areas where the Nasca habitation levels have been exposed.
They believe that the site was a ceremonial center equal in importance to
Cahuachi." Given the annual flooding of the Grande River and the relative
scarcity of arable land in the drainage, I am dubious that rich tombs would have
been located in the valley bottom setting described by Sawyer. In 1989, I sur-
veyed the middle Grande Valley where the modem town of Cabildo is located. All
sites were off arable land. Eighteen had surface evidence of Nasca 2 occupations.
Most of these sites were located on the south/east side of the river. These sites are
briefly described in Table 3.5; their locations are shown in Figure 3.9.
Let us tum now to the looted textiles discussed by Sawyer (1997). Sawyer
interprets the material as a gravelot consisting of a large sampler and 27 in-
process textiles, 12 of which are embroidered and 15 of which are cross-knit
looped. Sawyer argues that "their association is highly credible, with the possible
exception of two fragments of an unfinished mantle border... that are in poor con-
dition while the rest of the items are remarkably well preserved." No pottery or
other artifacts accompanied the textiles when acquired.
On the basis of this material and textiles he relates to the cache, Sawyer
identifies an iconographically complex Nasca 2 embroidery style that he suggests
extends back to Nasca 1. This iconography includes supernatural figures, human
figures, and ordinary plants and wildlife.
The Cabildo textile no. 1 sampler (Sawyer 1997: figs. 30-33,40,41,46,47,
49) depicts iconographically complex human ritual performers/anthropomorphic
deities. It is comparable to samplers in the Textile Museum, Art Institute of
Chicago, Metropolitan Museum of Art, and Museo Nacional in Lima which are
illustrated in Sawyer's figs. 34-38, 50-55.
The exquisite Cabildo textile no. 2 in-process double-faced embroidered man-
tle border (Sawyer 1997: fig. 60) also portrays iconographically complex human
ritual performers/anthropomorphic deities in full ritual attire who are in association
with trophy heads, decapitation knives, Spondylus jewelry, gold ornaments, and
Table 3.5. Sites with Nasca 2 Occupations in the Middle Grande Valley.
Site # Brief description of site Size in ha. Nasca phases present
278 broken funerary urns; simple pits; destroyed barbacoas 1.25 Nasca 1-3
279 mound of cane-marked conical adobes and round adobes; 0.5 Nasca 1-4
perishable watlle-and-daub superstructure was burnt in
antiquity; looters holes surround mound; parts of
barbacoas and human skeletal remains scattered on surface
287 major complex of small mounds, geoglyphs, and looted 8 N 1,2,3,5,7
cemeteries
300 looters holes surrounded by remains of barbacoas and 0.75 Nasca 1-3
human bones; small trapezoidal geoglyph
301 artificially terraced natural hill; vast area of looters holes 3.0 Nasca 1-5
around hill
305 major complex of small mounds, geoglyphs and areas 43 Nasca 1-5,7
of severe looting
311 geoglyph (damaged by plowing); four circular piles of 0.4 N2-5
stone (refugios)
313 small, planned architectural unit made of fieldstones set 3 Nasca 2-7
in mud mortar; unit consists of a square-shaped, cleared
patio and rooms bordering north and south walls of patio
321 probably ceremonial architecture (visible on 1944 aerial 4.2 (bulldozed Nasca
photo) destroyed by agricultural activities; geoglyph dispersion) 2,3,5,6,7
326 looted remains of barbacoas and human bone; 3.75 Nasca 2, 3, 5, 6
scattered loaf-shaped adobes and midden deposits
502 possible potter's house and grave 0.25 Nasca 1-3
504 massively looted area 0.08 Nasca 2, 3, 5
507 massively looted area 1.0 Nasca 1-3
508 massively looted area 1.0 Nasca 1-3
509 scatter of looted tomb contents 1.0 Nasca 1,2
510 sherd scatter 0.5 Nasca 2, 3
515 artificially terraced large hill which is badly looted; 4.5 Nasca 1-4
looting has dispersed tomb material and domestic refuse
520 terraces on a hillside; terraces have cleared areas 1.8 Nasca 2, 3
measuring 3 X 2 m; terraces massively looted;
scant amount of domestic refuse visible in
looters holes
96 Helaine Silverman
d
"
Figure 3.9. All Nasca 2 sites recorded by Helaine Silverman in the middle Grande Valley. The
middle Grande Valley is a relatively large tract of arable land irrigated by the confluence of the Grande
and Ingenio rivers; it is the "bread basket" of the drainage. Topographically. it has easy access to the
lower lea Valley (around Ullujaya) due west and even greater proximity to the lower Ingenio and
lower Nazca valleys immediately to the northeast and southeast, connections of potential social and
economic importance. See Table 3.5 for site descriptions.
98 Helaine Silverman
illustrated by Sawyer also lack site provenience. One can pick and choose from
among these materials to support one's case.
DISCUSSION
Bean (1998: 115) echoes many scholars when she says that cloth and cloth-
ing are "arguably the pre-eminent material product of human ingenuity" playing
"central roles in political, social, economic, and cultural systems around the
world." Indeed, the recent identification of depictions of elaborate fiber clothing
on Upper Paleolithic "Venus" figurines (see Soffer et al. 2000) suggests that
textile technology is almost as early as the appearance of anatomically modern
humans in Europe and that, from the beginning, textiles were associated with
social differentiation, power, prestige, and value.
The Paracas Necropolis textiles continue to be exceptional because of their
quantity, quality, known provenience, and context. Looted textiles without secure
provenience but presumed to be Early Nasca in date and cultural affiliation also
are an important piece of the puzzle about the relationship and implications of
style for understanding south coast societies. It is difficult to know which evi-
dence to favor. I attempt some tentative conclusions in this section.
There has been a significant amount of scientific excavation in Nasca ceme-
teries in the lea Valley and the Rio Grande de Nazca drainage since early in the
20th century. Published data suggest that these excavations have yielded poor tex-
tile remains (i.e., iconographic complexity is rare) in comparison to the abundant
fancy textiles of the contemporary Paracas Necropolis mummy bundles. This dis-
parity must be due to the limited distribution of fancy textiles in Nasca society,
and also to the amount of time required for their production (i.e., that there were
not many fine textiles).
Other differences between Early Nasca and Paracas Necropolis textiles can
be noted. Early Nasca and Paracas Necropolis textiles differ in terms of Early
Nasca's emphasis on design elaboration through edgings in contrast to the typi-
cally whole field iconography of Paracas Necropolis textiles. Also, Paracas
Necropolis iconography was elaborated on a wide range of garment types consti-
tuting elite or ritual attire depicting the role of the individual wearing them and
belonging to that individual (Paul 1990a). In contrast, in Early Nasca society,
there appears to have been an emphasis on communal participation in rites guided
by religious practicioners who were not necessarily the owners of their role-spe-
cific costumes (unless archaeologists find the costumes in graves, as is the case
with Paracas Necropolis ritual attire and, spectacularly, with the Moche burials at
Sipan). In addition, at least some of the published fancy Early Nasca textiles were
ceremonial cloths meant to be viewed rather than worn (Bird and Bellinger 1954).
ACKNOWLEDGMENTS
I gratefully acknowledge information on CMAT provided by Karen
Thompson, Curatorial Assistant, Textiles and Islamic Art, Cleveland Museum of
Art. I thank that museum for permission to publish CMAT. I express my appreci-
ation to Maria Didrichsen and the Didrichsen Art Museum for permission to pub-
lish the beautiful Helsinki panpipe. Museum research in Peru was conducted by
the author with the assistance of a Fulbright University Affiliations Program
grant. Fieldwork in the Middle Grande valley (as part of the Ingenio Valley sur-
vey project) was conducted with a grant from National Geographic Society and
with permission from the Instituto Nacional de Cultura. These institutions are
thanked for their support.
Notes
I. Except when quoting others, Nasca written s refers to the archeological culture of the Early
Intennediate Period whereas Nazca written with z refers to the geographical region, specific river
valley, drainage, and contemporary town (see Silverman 1993: ix).
2. This little marking on each of the tabs must have meaning. I speculate that it could represent the
maize plant with pulled-back leaves (see, especially, O'Neale 1937: plate LXIIc).
3. The hyperceremoniallocus of Orefici's find supports my argument that fancy textiles were made
for use in ceremonies at Cahuachi, but not necessarily for particular individuals as their property.
4. Although some EIP lB Paracas Necropolis textiles have tabs (de Lavalle and Lang 1983: 69 [prob-
ably from EIP lB Mummy Bundle 382]) instead of the more common multicolored thread fringe
(de Lavalle and Lang 1983: 45, 46, 49, 51-53, 55, 57, 58 inter alia), their ubiquity in Early
Nasca textile art and rarity in the Paracas Necropolis textile corpus suggest that they are a Nasca
invention.
REFERENCES
Bean, Susan S., 1998, Reviews. Journal ofMaterial Culture 3 (I): 115-120.
Bennett, Wendell c., 1954, Ancient Arts ofthe Andes. The Museum of Modem Art, New York.
Chapter 4
Proto-Writing in Moche Pottery
at Cerro Mayal, Peru
MARGARET A. JACKSON
INTRODUCTION
Moche ceramic art is well known for its sensitive realism and attention to fine
detail. It is a tradition that continues to fascinate, in part, because the corpus of
imagery suggests a ready venue for cultural understanding, despite formidable
barriers of distance and time. Yet, Moche art includes a sizeable component of
images that can only be described as abstract. As scholarship surrounding Moche
imagery advances, it becomes apparent that these are not simple mimetic repre-
sentations. The images actually functioned as an integral element of a complex
communication system widespread among the valleys of the north Peruvian
coast, where the Moche held sway throughout the 1st through 9th centuries AD.
The Moche iconographic system was multivalent and multi-operational,
functioning at different levels of the social hierarchy in different ways. Different
individuals or classes of individuals in Moche society had differing levels of pic-
torial literacy. For the general populace, the images seem to have been locked
firmly into oral traditions, directly tying the overall production of cultural mem-
ory to mnemonic visualization, pictorial cueing, and patterns of rhetoric occa-
sioned by important events and accompanied by the circulation of ideologically
invested artwork. For religious specialists, the images were polyvalent, drawing
upon reservoirs of esoteric knowledge and comprising an iconic, semasiographic
system. Semasiographic systems, a term which derives from the Greek word
"semasia" (meaning), are those notational forms that use marks to convey mean-
ing in a non-verbally tied manner (see Boone and Mignolo 1994: 15; Brice 1976:
29-41; Sampson 1985: 29).
107
CERROMAYAL
Cerro Mayal provides a nearly unique opportunity to examine the ways in
which pots were fabricated and distributed throughout Moche society because it is
one of only three excavated sites known to have been involved in large-scale man-
ufacture of ritually important Moche ceramics (see also Armas et al. 1993; Bawden
1982; Uceda, Mujica and Morales, 1995). Cerro Mayal is centrally located in the
lower Chicama Valley (Figure 4.1). It is adjacent to the Moche civic-ceremonial
center of Mocollope, an area which includes an impressive sector of monumental
architecture, as well as an extensive, but mostly destroyed, zone of habitation.
Cerro Mayal was a large, nucleated workshop where the potters both lived
and worked. The site was first excavated in 1992 by Russell, Leonard and Briceno
(1994), with a second field season in 1997. Carbon-14 samples indicate that the
site was in use from roughly 550 to 880 AD (Attarian 1996: 15-16, table 2.1).
The pottery can be categorized as predominantly Moche IV, according to Larco
Hoyle's (1948) five-phase seriation, with occasional examples more approximating
Moche V in style.
There were distinct areas supporting different activities, for example, raw clay
mixing, pottery firing, waster dumps, and community hearths. The overall site is
thus divided into areas of intensive production, what appear to be production sup-
port areas, and habitation zones, strongly suggesting specialization of activities.
Botanical remains indicate that the bulk of the agricultural products consumed con-
sisted of maize and beans (Attarian 1996, 1998). Preliminary analysis of faunal
remains indicates that camelid meat was also a significant component of the diet.
The presence of staples suggests that artisans at the site were attached specialists
who received food provisions, probably from the adjacent palace at Mocollope. The
workshop's production was most likely a part of a system of embedded patronage
under the direct control of Mocollope's rulers (Russell and Jackson 2001).
The data that form the basis for the present report are a representative sam-
ple of 1474 sherds, pulled from the overall collection of 139, 240 sherds because
_ River
= Highway
• Modern town
• Archaeological site
t
o 5 10Km.
Figure 4.1. Map showing location of Cerro Mayal in the Chicama Valley, Peru.
of their iconographic content. Some 91.5% of the total collection were fragments
of undetermined ceramic forms. The remaining 11,793 sherds had some type of
diagnostic characteristic, such as rim or handle. The iconographic sample was
selected from the diagnostic group.
Incidence of motifs and vessel forms were subsequently cross-referenced by
the author to arrive at a correlation between ceramic type and image (Jackson
2000: table 2.1). Additionally, the larger diagnostic sample was analysed accord-
ing to the use-function of the product or vessel type (Russell and Jackson 2001:
fig. 6). Most imagery occurred on jars, bottles, tall flaring vases (known as flo-
reros), figurines, appliques and molds. Virtually all detailed imagery on vessels
was created using molds. The workshop's artists did not employ the fineline
painting technique to any large degree. They did, however, use slip paint for var-
ious motifs and decorative elements, conforming to well-known canons of Moche
imagery and iconographic content. Hand modeling was employed on a miniscule
number of pieces (a few spoon handles, for example).
MOLD TECHNOLOGY
In Moche society it seems that the finest pots were generally mold made,
while the common wares were handbuilt. This is something of a reversal to the
values we commonly associate with mold-made items today. A clarification of the
role of mold technology in Moche ceramic production is among the most signif-
icant results of the present analysis.
It has long been known that two-piece molds were used to produce some
types of Moche pottery; such molds essentially comprised the front and back
sides of a vessel's main chamber (see illustrations and descriptions in Donnan
1992: 60-63; Larco Hoyle 1945: 15,31). Yet, at Cerro Mayal, in addition to one-
or two-part molds, Moche artists were using complex joining techniques to
assemble sculptural compositions in an additive manner. This meant that any
given vessel might have any number of mold made elements, either stamped
directly onto it, or added through an applique process. The appliques themselves
could be formed of two-, or even three- part molds, allowing for an impressively
high degree of complex artistry.
Each mold comprised a complete iconographic package, whose message
remained self contained regardless of who was using the mold or what other
elements were selected for use. The molds could be used in a mix-and-match way,
MOLD INSCRIPTIONS
The most compelling argument in favor of a system of visual notation among
potters and related specialists comes from a body of inscriptions present on the
exteriors of many of the molds from Cerro Mayal and other sites. In the Cerro
Mayal study sample, approximately 19% of molds analyzed had some form of
exterior inscription. Of the 1271 molds and mold fragments collected in the 1992
field season, 368 were closely analyzed, yielding 69 with exterior inscriptions.
Three basic types of markings were evident: (1) those which appear to
roughly reflect the mold's interior imagery, such as eyes, nose, mouth, or feet,
incised upon the mold's exterior surface to serve as pictorial alignments assisting
the potter in positioning imagery on the finished product; (2) horizontal staight-
line incisions running perpendicular to the edges of the mold, used as register
marks to align multiple-part mold pieces, front-side to back-side or right to left;
and (3) small, sketchy pictorial drawings that bear no visible resemblance to the
imagery on the mold's interior, nor appear to serve any purpose overtly related to
the manufacture process.
I argue that none of the inscriptions can be regarded as casual or accidental.
Many refer to images and themes common to the larger Moche iconographic
repertoire, insinuating that the meaning and function of the inscriptions and the
iconographic content of their ceramic products were directly linked to a broader
symbol system. Furthermore, among the third category of molds, the conceptual
leap presented by the discrepancy between the molds' interior and exterior
imagery suggests that (1) those who were iconographically savvy thought of the
symbol system abstractly (e.g., as signs), and (2) that a Moche iconic "shorthand"
actually communicated specific information in an abstract notational form.
Pictorial Alignments
The inscriptions most readily understandable to us are those that appear to
roughly reflect the mold's interior imagery. Such things as eyes, nose, mouth,
etc., were most often crudely sketched upon the mold's exterior surface. These
markings were no doubt useful in assisting the potter to position imagery on
the ceramic object; he or she would be able to use the markings as guides.
Furthermore, in the most functionalist sense, one can imagine that the inscriptions
helped the potters to quickly differentiate one mold from another without being
forced to scrutinize each mold's interior prior to selecting it for use.
Figure 4.2. Molds whose inscriptions echo interior imagery. (a) simple figurine; (b) owl with
characteristic circle around eye; (c) figure with "banded mouth."
Figure 4.3. Mold whose inscription echoes interior imagery. (al large "Wrinkle-Face Whistler"
mold exterior; (b1modeling clay impression of mold interior.
Likewise, the applique mold shown in Figure 4.4b (one half of a two-part
mold), depicting a very realistic fox head, features exterior inscriptions that
reflect the interior imagery. But the incisions are more elaborate than needed for
simple pictorial alignment. On the mold exterior, the animal's eye is surrounded
by two concentric circles that each terminate in a stylized bird head motif, a pic-
torial addition whose meaning is unclear. Several interpretations are possible, but
all are speculative. In Moche imagery, foxes are often shown as messengers;
Figure 4.4. Molds whose inscriptions echo interior imagery. (a) "Fanged Lizard" jar neck mold;
(b) "Snarling Fox" applique mold with elaborate exterior markings.
Colonial sources identify birds as messengers from the spirit world [Note 2].
Perhaps the connection evokes both creatures' roles as emissaries. Alternatively,
the connection could have been linguistic if, for example, there were some
homophony at play and the two images juxtaposed may have constituted a rebus
device of some kind. It is even possible that the bird heads are artistic devices
116 Margaret A. Jackson
related to visual periphrasis, as was common to the ancient Chavin tradition of
terminating lines with zoomorphic heads (the kennings of Rowe 1967: 78-79)
[Note 3]. Whatever their significance, the presence of the bird head devices can
only be explained in tenus of abstraction.
In more general tenus, molds with exterior inscriptions closely keyed to their
interior imagery likely served a straight-forward technical function. It also seems
likely, however, that pictorial alignment inscriptions were linked to the larger
iconographic system by virtue of their consistent references to conventionalized
attributes of various characters.
Register Marks
The second kind of mold inscriptions are those with perpendicular straight-
line incisions leading off their edges. These doubtless served as register marks for
the purpose of aligning 2-part or multiple part molds. Still, commonly, register
marks are used by modem potters to align multiple-part mold pieces, front-side
to back-side or right to left.
Molds such as the large frog and the portrait head shown in Figure 4.5, were
clearly used in pairs, as evinced by numerous vessels in museum collections of
similar design. Each mold depicts a symmetrically divided subject; the edges of
the molds themselves show that the images were split along obvious lines of sym-
metry. The exterior of the large frog mold, Figure 4.5a, was inscribed with marks
indicating the position of the frog's main elements (eye, legs, mouth); secondary
iconographic elements (beans, com, aji peppers) were not included, suggesting
that they were not integral to the essential identification of the subject matter. The
line indicating the location of the frog's mouth runs off the leading edge of the
mold; the opposing counterpart mold (now missing) was probably inscribed with
the continuation and termination of the mouth line. Such a line served as a regis-
ter mark aiding in mold alignment.
A similar register mark is visible in Figure 4.5b, showing a mold that creates
the left side of a human face. The mold terminates in the front along the face's
axis of vertical symmetry (between the eyes and down the nose ridge) and along
the side of the head just behind the ear, suggesting that the finished portrait head
was made using at least three (or perhaps four) separate molds (two for the face
and one or two for the backside of the head). The exterior of the mold has a crude
circle corresponding to the interior location of the eye, and a horizontal line run-
ning across the bridge of the nose. The horizontal incision shows no point of ter-
mination and surely continued to the right side counterpart, forming a point of
alignment for the two mold halves.
Register marks can appear alone or in combination with pictorial align-
ments. Although examples of horizontal register marks in the Cerro Mayal
sample are relatively few, their technical purpose seems self evident.
Figure 4.5. Molds with register marks on their edges. (a) large "Frog Mold" (#181-14). insert
shows interior image; (b) portrait vessel mold (#194-15).
Pictorial Notations
Both pictorial alignment marks and register marks can be seen as directly
related to the ceramic production process, even though they may also refer to spe-
cific attributes of particular characters being depicted. A third type of marking found
on Cerro Mayal's molds is more abstractly notational. This type of inscription
Figure 4.6. (a) Rattle mold with "Rattle Player" inscription; (b) double-chambered rattle from Cerro
Mayal; (c) drawing of "Rattle Player" inscription; (d) elaborately dressed musician figure (Museo
Nacional de Antropologfa e Arqueologfa, Lima).
Figure 4.7. Rattle Players. (a) Male in elaborate dress holding double-chambered rattle (Field
Museum #100072, Chicago); (b) "Priestess Entranced," hollow figurine (Cerro Mayal); (c) Male
rattle players, small figurines (Cerro Mayal).
Figure 4.8. (a) "Birth Bowl" mold (#8570-1); (b) drawing of exterior mold inscriptions;
(c) similar, unbroken "Birth Bowl" (Lavalle 1986).
Figure 4.9. (a) Mold with "Double-chambered Rattle" inscription (#8338-1); (b) drawing of mold
#8338-1; (c) drawing of similar inscription on "Birth Bowl" mold (#8570-1).
artists had well established visual precedents for depicting this type of ritual ves-
sel in minimal form. Stirrup-spout bottles, floreros, and the like were emblematic.
They functioned as icons in their own right and as references to the spirit world,
the preparation of the dead for funeral, or the practice of making funerary offerings.
Therefore, again, the mold inscriptions do not carry a literal correspondence to that
which the mold produces, but instead refer to abstract ideological constructs. The
"ritual vessel assemblage" motif reads as specifically tied to a particular social
moment. For whatever social value they may have had, floreros and stirrup spout
vessels functioned as independent signs referring to a complex ideological cluster
that was socially widespread and well understood by the artists. As in the previous
...
~
f...
.~...
Figure 4.10. Molds with "ritual vessel assemblage" inscriptions. (a) vessel body mold (# 197-19) with
"Geometric Snake" motif interior; (b) mold (#125-4) ~
with "Female with Braids" interior imagery; (c) vessel body mold (#108-9) with low relief humans on
interior; (d) bottle mold (#138-19) with "Rolling ~
Wave" motif interior; e) jar mold (#8714-2) with tab handles, no interior imagery. =
~
~
:3.
=.
1:1
IrQ
5'
gfo
~
~
~
Figure 4.11. Bottles with painted vessel assemblages. (a) seated figure with stirrup spout bottle, paired
gourd bowls and f1orero (from Berrin 1997:
fig. 74); (b) woman and skeleton with f1orero, wrapped copper squares, stirrup spout and strap handle
bottles (from Benson 1972: fig. 6-11). ~
~
Figure 4.12. Molds with geometric "Textile Pattern" inscriptions (Cerro Mayal).
OCCUPATIONAL LITERACY
It is important to note that similar inscriptions appear on more than one
mold. If the inscriptions were only isolated examples it would be a simple thing
to dismiss them as oddities. But because there are groups of molds having essen-
tially the same inscriptions, they must be addressed in terms of their intentional
notational value. It may be that these markings represent examples of a type of
Moche iconic abbreviation that, in fact, communicated specific information
through a system of recognizable signs.
If mold inscriptions functioned within the workshop to signal specific infor-
mation (technological or other), we can assume that most of the potters working
there understood the significance of the markings. Today, this is called "occupa-
tionalliteracy," where workers of a certain kind develop and learn to read the spe-
cialized signs and signals intrinsic to their jobs (Rush et aI., 1986). Occupational
literacy is created in the workplace out of need by the specific community that
uses it. It may be common only to a single community or it may be used by
other communities engaged in similar activities. In terms of viewing Moche
EMERGENT WRITING
Can it be that these findings signal a form of nascent or emergent writing?
In an important study of how writing arose in the ancient Near East, Denise
Schmandt-Besserat challenges the idea that all forms of writing evolved from pic-
torial drawings (a developmental model which she dubs the "Pictographic
Theory") (Schmandt-Besserat 1996: 4). As late as 1974 scholars such as I. J. Gelb
were still arguing the 18th century notion that scripts such as cuneiform began as
pictography [Note 10]. Yet, Schmandt-Besserat notes that the Widespread belief
that any written script will naturally begin pictorially and gradually evolve into a
more abstract form is not consistent with the archaeological data. She proposes
instead that the antecedent of phonetic writing, at least in the ancient Near East,
was actually a counting device [Note 11].
Schmandt-Besserat ties the use of tokens (small bits of modeled clay), that
evolved to meet the needs of an expanding economy, to the rise of social structures.
Tokens originally represented the concrete counting of physical objects. They later
evolved into complex tokens with inscriptions, punctates and sub-categories. More
than a simple one-to-one object correspondence, tokens represented an entirely
new level of abstraction. The conceptual leap was to endow each token shape with
a specific meaning. Thus, the token shapes became abstract signs.
Corresponding to the increase in bureaucracy, methods of storing Near
Eastern tokens in clay "envelopes" and archives were devised. Tokens enclosed
inside envelopes were represented by impressed markings on the exterior of the
envelopes. Scribes soon realized that actual tokens were not necessary, and the
hollow envelopes were replaced by solid tablets having markings alone. These
markings became a system of their own (Archaic Cuneiform) which developed to
include not only impressed markings but more legible signs traced with a pointed
stylus. Schmandt-Besserat makes a key point in noting that the substitution of
signs for tokens was a first step toward writing. She likened the systematized
inscriptions to "picture signs" or "pictographs" because they represented pictures
of the tokens used as counters in the accounting system (Schmandt-Besserat
1996: 7-8) [Note 12]. In an analogous way, the systematized inscriptions of the
Moche potters are also representations of pictures, that is, substitutions for larger
concepts.
In the Near East, alphabetic writing resulted from bureaucratic demands
and from the invention of abstract counting. While scribes invented ever more
CONCLUSION
What, then, can we conclude when we begin to see evidence of abstract
signs and occupational literacy in workshops such as Cerro Maya!? The evidence
points to a system of pictorial notation in use among occupational specialists.
Under the circumstances, however, given the larger ritual functions of specialized
pots in Moche society, one would not expect the notational system to replace the
actual objects in production.
The question of "emergent writing" among the Moche implies that Moche
"picture signs" would have eventually evolved into an alphabet akin to other
known glottographic scripts. At the moment, based on the data from Cerro Mayal,
no solid evidence exists that they were "emerging," in an evolutionary sense, into
anything other than what I have already suggested, a specialized visual notational
Acknowledgments
Scholarship and fieldwork for this project have been supported, in part, by
the Fulbright LLE. program, the National Resource Fellowship, Edward A.
Dickson Fellowship and UCLA's Latin American Center. Special thanks go to
Dr. Glenn Russell, Banks L. Leonard, Jesus Briceno, Dr. Santiago Uceda, and the
Instituto Nacional de Cultura of Trujillo, Peru. Portions of this paper were initially
presented at the the 63rd Annual Meeting of the Society of American Archaeology
under the title, "Moche Iconography at Cerro Mayal, Peru" (Seattle, 1998).
Notes
I. The term "banded mouth" refers to the practice of depicting the lips as a continuous band encir-
cling the mouth. The term is most often seen in connection with descriptions of Chavin and
Cupisnique period artwork (see Rowe 1967; Roe 1982).
2. For more on Fox Messengers see Benson (1972:48-51) and Donnan (1978: 74-76). We can infer
that birds and owls were thought of as messengers from the spirit world from the care and redun-
dancy with which Father Carrera questions the Indians about their beliefs and idolatrous prac-
tices. As part of confession, they are specifically asked, "Ecaprecoz xllom precna pren frepi~rer,
fiaiiissap~ren, pucu, fiii, pocpoc, licaprecoz mcecha, macharec, pong, echallo?" [phonetic font
problem: the letter following the first c in macharec should be an inverted h] Translated in its most
REFERENCES
Abrams, Meyer Howard, 1993, A GLossary ofLiterary Terms. Harcourt Brace College Publishers, Fort
Worth.
Armas Asmad, Jose, Violeta Chamorro Castillo, and Gloria Jara Flores, 1993, Investigaciones
Arqueologicas en el Complejo Huaca del Sol y la Luna: Talleres Alfareros de la Sociedad Moche.
Informe de Practicas Pre-Profesionales de Arqueologfa. Facultad de Ciencias Sociales,
Universidad Nacional de La Libertad, Trujillo.
Attarian, Christopher J., 1996, Plant Foods and Ceramic Production: A Case Study of Mochica
Ceramic Production Specialists in the Chicama Valley, Peru. M.A. thesis. Department of
Anthropology University of California, Los Angeles.
Attarian, Christopher J., 1998, Diet, Status, and Affiliation in Moche Society: Ethnobotanical data
from the Chicama Valley. Paper presented at the 63rd Annual Meeting of the Society for
American Archaeology, Seattle.
Bawden, Garth, 1982, Galindo: a study in cultural transition during the Middle Horizon. In Chan
Chan: Andean Desert City, edited by Michael E. Moseley and Kent C. Day, pp. 285-320.
University of New Mexico Press, Albuquerque.
Benson, Elizabeth P., 1972, The Mochica. Praeger Publishers, New York.
Berrin, Kathleen (ed.), 1997, The Spirit of Ancient Peru. Thames and Hudson and the Fine Arts
Museum of San Francisco, New York and San Francisco
Boone, Elizabeth H. and Walter D. Mignolo, 1994, Writing Without Words. Duke University Press,
Durham.
Brice, W. c., 1976, The principles of non-phonetic writing. In Writing Without Letters, edited by
W. Haas, pp. 29--44. University of Manchester Press, Manchester.
Campana, Cristobal and Ricardo Morales, 1997, Historia de Una Deidad Mochica. A and B, S.A.,
Lima.
Carrera, Fernando de la, 1939, Arte de La Lengua yunga. edited by Ramades A. Altieri. Instituto de
Antropologfa, Tucuman. [originally 1644]
134 Margaret A. Jackson
Donnan, Christopher B., 1978, Moche Art of Peru. Museum of Cultural History, University of
California, Los Angeles.
Donnan, Christopher B.,1992, Ceramics of Ancient Peru. Fowler Museum of Cultural History, Los
Angeles.
Donnan, Christopher B. and Donna McClelland, 1999, Moche Fineline Painting. Fowler Museum of
Cultural History, Los Angeles.
Gelb, Ignacio J., 1974, A Study ofWriting. University of Chicago Press, Chicago.
Jackson, Margaret A., 2000, Notation and Narrative in Moche Iconography, Cerro Mayal, Peru. Ph.D.
dissertation. Department of Art History, University of California, Los Angeles.
Larco Hoyle, Rafael, 1945, Los Mochicas (Pre-Chimu de Uhle yEarly Chimu de Kroeber). Sociedad
Geografica Americana, Buenos Aires.
Larco Hoyle, Rafael, 1948, Cronologia Arqueol6gica del Norre del Peru. Sociedad Geognifica
Americana, Buenos Aires.
Mackey, Carol and Charles Hastings, 1977, Moche murals from the Huaca de La Luna. In Pre-
Columbian Art History: Selected Readings, edited by Alana Cordy-Collins, pp. 293-312. Peek
Publications, Palo Alto.
Peirce, Charles, 1985, Logic as semiotic: the theory of signs. In Semiotics, edited by Robert E. Innis,
pp. 4-23. Indiana University Press, Bloomington. [originally 1931]
Roe, Peter G., 1982, Cupisnique pottery: a cache from Tembladera. In Pre-Columbian Art History:
Selected Readings, edited by Alana Cordy-Collins, pp. 231-253. Peek Publications, Palo Alto.
Rojas y Silva, David Vicente de, 1981, Los tocapu: un programa de interpretaci6n. In Arte y
Arqueologia, issue 7. La Paz.
Rowe, John H., 1967, Form and meaning in Chavin art. In Peruvian Archaeology, Selected Readings,
edited by John H. Rowe 1nd Dorothy Menzel, pp. 72-103. Peek Publications, Palo Alto.
Rush, R. T., A. 1. Moe, and R. L. Storlie, 1986, Occupational Literacy Education. International
Reading Association, Newark, Delaware.
Russell, Glenn S. and Margaret A. Jackson, 2001, Patronage and political economy at Cerro Mayal,
Peru. In Arr and Political Representation in Ancient Peru, edited by Joanne Pillsbury. Center for
Advanced Study of the Visual Arts, National Gallery of Art, Washington, D.C.
Russell, Glenn, Banks L. Leonard, and Jesus Briceno R., 1994, Producci6n de cenimica a gran escala
en el valle de Chicama, Peru: el taller de Cerro Maya!. In Tecnologia y Organizaci6n de la
Producci6n de Cerdmica Prehispdnica en los Andes, edited by Izumi Shimada, pp. 201-227.
Fondo Editorial, Pontificia Universidad Cat61ica del Peru, Lima.
Russell, Glenn, Banks L. Leonard, and Jesus Briceno R., 1998, Large-scale Moche ceramic produc-
tion in the Chicama Valley, Peru: the Cerro Mayal workshop. In Ceramic Production in the
Andes: Technology. Organization. and Approaches, edited by Izumi Shimada, pp. MASCA
Research Papers in Science and Archaeology, Supplement to Volume 15. Museum Applied
Science Center for Archaeology, University of Pennsylvania Museum of Archaeology and
Anthropology, Philadelphia.
Sampson, Geoffrey, 1985, Writing Systems: A Linguistic Introduction. Stanford University Press,
Stanford.
Schmandt-Besserat, Denise, 1996, How Writing Came About. University of Texas Press, Austin.
Schumcher de Pena, Gertrud, 1991, El Vocabulario Mochica de Walter Lehmann (1929). Universidad
Nacional Mayor deSan Marcos, Lima.
Strong, William Duncan and Clifford Evans (eds.), 1952, Cultural Stratigraphy in the Viru Valley,
Northern Peru. Columbia University Studies in Archaeology and Ethnology, 4. New York.
Uceda, Santiago and Elias Mujica (eds.), 1994, Moche: Propuestas y Perspectivas. Universidad
Nacional de La Libertad-Trujillo, Instituto Frances de Estudios Andinos, Asociaci6n Peruana
para el Fomento de las Ciencias Sociales, Lima.
Chapter 5
Chachapoya Iconography and
Society at Laguna de los
C6ndores, Peru
ADRIANA VON HAGEN
INTRODUCTION
Explorers and researchers have documented distinctive Chachapoya-style
architecture and pottery across a wide swath of the northeastern Peruvian Andes.
The Chachapoya area lies sandwiched between the Marafion River to the west
and north, the Huallaga River to the east, and is bordered to the south by Pias. It
covers some 155,000 square kilometers (Figure 5.1). The land ofthe Chachapoya
encompassed a range of ecosystems: the tropical forest along the Maraiion, the
windswept jalca or high grasslands, the temperate highland valleys ideally suited
for potato cultivation and the ceja de selva, or tropical montane wet forest flank-
ing the easternmost cordillera of the Andes. In the words of an early writer, the
land of the Chachapoya was a "very rugged and wet land, all year it does nothing
but rain, and for this reason the Indians build their houses on the summits and
heights" (Primeros Agustinos 1916: 56). The name Chachapoya, in fact, may be
a corruption of the Inca (Quechua) name for the province called Chachapoyas:
sacha (tree) and puyu (cloud) which can be roughly translated as "cloud forest."
This is an apt description for much of the Chachapoya territory [Note 1].
The Chachapoya, whose territory straddled gateways (called entradas in the
Colonial literature) to the eastern lowlands, may have served as intermediaries in
exchange systems, bartering lowland forest products for highland produce and
goods. At various times in Andean prehistory the Chachapoya interacted with cul-
tures living to the west of the Maranon-as seen, for example, in pottery influenced
137
138
o
•
"'\....---.
Figure 5.1. Map of the Chachapoya cultural area (credit: Adriana von Hagen).
by the Cajamarca tradition. At other times the Chachapoya appear to have flour-
ished in relative isolation. Although the Chachapoya played a part in the greater
Andean cultural sphere, their art and architecture conveys a bold, independent spirit
that sets them apart from their neighbors.
A Chachapoya identity appears to have coalesced around AD 800 and to
have lasted through the Inca conquest. Understanding of late prehispanic
Chachapoya prehistory-the subject of this chapter-remains fragmentary
Figure 5.2. Plan of the chullpas at Laguna de los Condores (credit: Adriana von Hagen).
144
CHACHAPOYAICONOGRAPHY
Although we found close to fifty decorated gourds at the site, the imagery
and workmanship of the gourd from Chullpa 1 is unique. The gourd is missing a
few pieces and is quite fragile, but enough remains to reconstruct its iconography.
It is a bottle-shaped gourd, about 17 cm high and 14.5 cm in diameter. There is a
band, about 3 cm high, near the top of the gourd composed of a pyro-engraved
eye-like pattern, similar to the eyes of the felines in the bottom panel. (I use the
term "feline" somewhat loosely since some of the figures appear to be hybrid
animals or composites of humans and animals.) The bottom half of the gourd
is covered by a 6 cm-high band arrayed with pyro-engraved figures forming a
complex scene (Figure 5.3).
The scene is composed of five main figures, distinguished in the rollout
illustration by letters A-E, running from left to right. A distinct line separates
Figure A from Figure E. The scene appears to illustrate a narrative of some sort,
perhaps a Chachapoya myth. The protagonists in this narrative or myth are rep-
resented by the half-human, half-animal beings (shamans?) that featured in the
local religious cosmology, possibly engaged in transforming themselves into
felines, their supernatural alter egos [Note 2].
Figure A is an upside-down feline with a fanged mouth, a circle on its snout
and on its ankle, and a tail that ends in a feline head. It clutches a smaller animal.
Figure B, a goggle-eyed, anthropomorphic figure with splayed arms and wearing
earspools, follows A. Figure B's lower half-apparent when Figure B is turned
upside down-seems to be some sort of feline, judging by the rounded ears, the
fanged teeth and the zigzag markings on the tail-like appendages that emerge
from its head to become either the arms or forelegs of the lower figure or the legs
of the anthropomorphic figure above it. These zigzag markings also appear on the
tails of the other felines. These zigzag markings could also refer to those found
on snakes. Accordingly, the tails portrayed on the gourd, although they end in
feline heads, may serve as metaphoric serpents. Snake-like appendages and tails
are common themes in ancient Andean art. The earspools on Figure B resemble
wooden ones attached to carved wooden figures found at LC 1 as well as wooden
earspools that were worn by actual persons, also found at LCI.
Figure 5.3. Rollout of pyroengraved gourd from Chullpa 1 at Laguna de los C6ndores
(credit: Cecilia Nunez).
Related Imagery
Similar imagery to that on the gourd and the bamboo container appeared on
some of the textiles from the mummy bundles found in other chullpas. The vibrant,
beautifully-woven textiles from Laguna de los Condores recall Cieza de Leon's
(1959: 99) description of the skill of Chachapoya weavers: "They made fine and
highly prized clothing for the Incas, and they still make excellent garments, and
tapestry so fine and handsome that it is greatly esteemed for its quality"[Note 3].
Let us look first at an unku, or sleeveless tunic, recovered from the mummy
bundle of an adult male (Figure 5.5). He wore a bone nose ornament and his
Figure 5.4. Pyroengraved profile feline found on a hollow bamboo container from Chullpa 1 at
Laguna de los C6ndores (credit: Cecilia Nunez).
extended earlobes, which once held earspools, indicate that he was an orej6n, a
sign of status among many ancient Andean peoples, including the Incas and the
Chachapoya. The body had been wrapped in five layers of textiles. The outer
wrapping consisted of plain weave cotton cloth, dyed blue, and decorated with a
chain stitch in cotton yam arrayed in a zigzag pattern around the bundle. Like
many of the mummy bundles from LCl, a stylized, embroidered face and a cot-
ton yam braid emerging from the top of his head, topped the bundle. The next tex-
tile was plain, as was the fifth, while the third was a disassembled garment
composed of two panels, each 38cm wide and about 1.80m long. The two ends
had been basted together and wrapped around the body.
The fourth textile was an unku placed over the body with the neck slit sewn
shut over the head. It is composed of two panels, each about 39 cm wide and 1.90 m
long (97.5 cm long and 79 cm wide, as worn), sewn up the sides and the center,
leaving openings for the arms and the neck. (The neck slit was subsequently sewn
shut when the unku was placed over the body in preparation for burial.) Below the
neck slit, on either side of the unku, is a feline, 7.1 cm high, with an arching tail
embroidered in blue and red camelid fiber. The unku's 5.5 cm-wide fringed border
bears volute-like designs in interlocking tapestry and profile feline heads in
brocade. The brocaded figures on the shoulders depict seated profile felines,
Figure 5.6. Profile feline executed in brocade on an unku from Laguna de los C6ndores (credit:
Cecilia Nunez).
3.6 cm high and 7.3 cm wide, with circles on their snouts and tails that end in feline
faces similar to Figures A and C on the gourd from Chullpa 1 (Figure 5.6).
The central part of the unku is divided into two sections, one of blue
cotton with the remains of camelid fiber embroidery and another of brown cotton
decorated with two pairs of woven figures (in warp-faced plain weave with
Figure 5.7. Central figure woven on an unku from Laguna de los C6ndores (credit: Cecilia Nunez).
Figure 5.8. Tapestry fragment from Laguna de los C6ndores (credit: Cecilia Nunez).
DISCUSSION
The Chachapoya imagery I have described is both familiar and strange. Its
common Andean themes, artistic conventions and traditions include: front-facing
figures holding staffs, or in our case, the tails or necks of flanking figures; front-
facing figures with splayed arms and legs; the heads or bodies of felines or hybrid
animals shown in profile, baring prominent, interlocking canines; felines or
hybrid animals with curlicues or circles on their snouts; figures, human and ani-
mal, with circles on their ankles; metaphoric attachments, i.e., tails that resemble
snakes and end in the heads of felines; hybrid creatures that combine the fearsome
qualities of several animals, notably felines, caimans, raptors, or serpents; ear-
spools that denote rank or supernatural attributes; and finally, artistic canons such
as bilateral symmetry and anatropic organization (Kubler 1975), that is, an artis-
tic convention that reveals a new facet of an image when it is viewed upside-down
or sideways (see Burger 1992).
Many of these themes and conventions first appear in Cupisnique art, whose
culture flourished on Peru's north coast ca. 1500-600 BC, and also among coeval
societies in the upper Jequetepeque drainage, especially at the site of Kuntur Wasi
(Onuki 1997). During the Chavfn horizon (ca. 400-200 BC), a widespread reli-
gious cult inspired works of art in metal, cloth, stone and ceramics. These
displayed imagery inspired, in part, by the earlier, coastal and north highland
centers (Burger 1992). Several of these traditions and conventions recurred in
various media over the next several hundred years in the art of Moche and
Recuay, its highland contemporary (ca. 200BC-AD 700), as well as in Huari
(ca. AD 600-900).
Figure 5.9. Feathered headdress found at Laguna de los C6ndores. (credit: Adriana von Hagen)
more distant from the tropical forest than the Chachapoya area.) Some of the arti-
facts recovered from LCI are unmistakable tropical lowland imports. These
include feathered headdresses and desiccated animals not native to the ceja de
selva [Note 6]. One of the remarkably well-preserved feathered headdresses
(Figure 5.9) is festooned with what appear to be parrot feathers and has a cane
framework reminiscent of contemporary lowland headdresses (see Braun 1995:
68). The desiccated animals, tours de force of taxidermy, appear to have been
used as carrying devices of some sort, since some of them have slits behind their
necks and an attached carrying cord. The most notable example from LeI is what
appears to be a margay (Felis wiedii), so well preserved that even its whiskers are
intact. Its tail, forelegs and hind legs are stuffed with unspun cotton, and it has a
bone inserted through its septum. We have tentatively identified another, smaller
feline as either a margay or an oncilla (Felis tigrina). It was found with spindles
inside its pouch. Laguna de los C6ndores is located well above the range of the
margay, which is found below 900m in elevation (Emmons 1990). The oncilla
occurs up to 3200m in elevation (Emmons 1990).
Laguna de los C6ndores and other Chachapoya ceja de selva sites may have
served as staging areas for encounters between the Chachapoya and the Xibito
and/or Cholon to exchange tropical forest products for those of the highlands
(Schjellerup 1997). Indeed, in the 17th century, the Cholon and Xibito traded
coca for Spanish garments and iron goods, traveling eight days to Cajamarquilla
(modem Bolivar). Trade along the Huallaga in the 18th century included "salt
fish, woven pouches, beeswax, manioc meal, feathered hats, container lids, coca,
and fish lines" which were sold or traded to highland peoples (Steward and
Metraux 1948: 604). The Xibito, (also spelled Jiyito, Hibito), according to the
Notes
I. Following Lerche (1995), Schjellerup (1997) and Urton (200 I), I use Chachapoya to refer to the
people and culture that occupied the region under discussion in antiquity. Chachapoyas, on the
other hand, refers to both the modem capital of the department of Amazonas and the eponymous
modem province.
2. Although ethnographic analogies should be used with caution, especially in light of the thousands
of years that separate ancient Andean art and beliefs from those of contemporary Amazonian
peoples, it is nonetheless interesting to signal the widespread belief among aboriginal South
Americans that shamans are in fact jaguars disguised as humans, who reveal their feline alter egos
when they ingest hallucinogens (Saunders 1998). Reichel-Dolmatoff (1975: 120) points out that
the Desana people of Colombia regarded shamans as "upside-down jaguars;" that is, as reverse, or
inverted animals, not ordinary felines. Cordy-Collins (1998: 167), in her analysis of Cupisnique
and Moche feline imagery, notes: "It is probable that the Tembladera and Moche cats in this study
were represented in their strange posture, and crouched amid hallucinogenic cactus and whirling
phosphenes, to re-enforce and underscore just that message: that the animal shown in these scenes
is not an ordinary feline; rather he is a transformed, reversed jaguar-a shaman."
REFERENCES
Agustinos, Los Primeros, 1916, Relaci6n de la religi6n y ritos del Peru hecha por los primeros
religiosos Agustinos que aUi pasaron para la conversi6n de los naturales. Colecci6n de Libros
y Documentos referentes a la Historia del Peru, volume II, pp. 3-56. Madrid. [originally
1550-1557]
Braun, Barbara (ed.), 1995, Arts ofthe Amazon. Thames and Hudson, New York.
Burger, Richard L., 1992, Chavin and the Origins of Andean Civilization. Thames and Hudson,
New York.
Cieza de Leon, Pedro de., 1959, The Incas. University of Oklahoma Press, Norman. [originally 1553]
Church, Warren, 1994, Early occupations at Gran Pajaten, Peru. Andean Past 4: 281-318.
Church, Warren, 1999, Loving it to death: the Gran Pajaten predicament. The George Wright Forum
16 (4): 16-27.
Cordy-Collins, Alana, 1998, The Jaguar of the Backward Glance. In Icons of Power: Feline
Symbolism in the Americas, edited by Nicholas J. Saunders, pp. ISS-170. Routledge, London.
Emmons, Louise H., 1990, Neotropical Rainforest Mammals. University of Chicago Press, Chicago.
Espinoza, Waldemar, 1967, Los seiiorios etnicos de Chachapoyas y la alianza hispano-chacha. Revista
Historica 30: 224-333.
Garcilaso de la Vega, EI Inca, 1966, Royal Commentaries of the Incas and General History of Peru.
University of Texas Press, Austin. [originally 1604]
Grieder, Terence, 1978, The Art and Archaeology of Pashash. University of Texas Press, Austin.
Guillen, Sonia, 1998, Arqueologia de Emergencia: Inventario, Catalogaci6n y Conservacion de los
Materiales Arqueol6gicos de los Mausoleos de la Laguna de los Condores. Unpublished report
submitted to the Instituto Nacional de Cultura, Lima.
Guillen, Sonia, 1999, Evaluacion y Delimitaci6n del Sitio Arqueol6gico Llaqtacocha. Unpublished
report submitted to the Instituto Nacional de Cultura, Lima.
Hayashida, Frances M., 1999, Inka pottery manufacture in the Leche Valley, Peru. Latin American
Antiquity 10 (4): 337-353.
Isbell, William H., 1997, Mummies and Mortuary Monuments: A Postprocessual Prehistory ofCentral
Andean Social Organization. University of Texas Press, Austin.
Chapter 6
Art and Prestige Among Noble
Houses ofthe Equatorial
Andes
TERENCE GRIEDER, JAMES D. FARMER, ANTONIO CARRILLO B.,
AND BRADFORD M. JONES
INTRODUCTION
Challuabamba is located in Azuay Province in the southern highlands of Ecuador,
at an altitude of 2300m asl (Figure 6.1). Challuabamba is just west of a zone
called El Descanso where the Burgay, Deleg, and Tomebamba rivers join to form
the Paute before it plunges into the gorges that lead to the Amazon. The site is
located on the south bank of the Tomebamba River between the Salado and the
Apangora, small tributary creeks.
Uhle (1922: 108) was the first to write of the archaeological remains of
Azuay Province, including "the unexpected discovery of great treasures" (since
Uhle's time there has been a wholesale destruction of ancient tombs and their
contents). Subsequently, Collier and Murra (1943), at the end of their studies in
Cerro Namo, made an archaeological reconnaissance of Azuay. Their work was
the foundation of modem studies in the province and was the inspiration for
Bennett's (1946) excavations on the haciendas of Huancarcuchu, between
Challuabamba and El Descanso, and Monjashuaico, between El Descanso and the
mouth of the Gualaceo River. The ceramic typology established by Bennett as a
result of that fieldwork has served as a model for later investigators.
157
158
o
km
Figure 6.1. Sites in the equatorial Andean region. I. Challuabamba. 2. Valdivia. 3. Real Alto.
4. Machalilla. 5. Cotocollao. 6. Cerro Narrfo. 7. Pirincay. 8. Catamayo. 9. Cerro Naiiaiiique.
10. Bagua. II. Pacopampa.
POTTERY AT CHALLUABAMBA
The making of pottery was already ancient in the northern Andes by the time
it came into use at Challuabamba. Red-on-Cream, Red-and-Black, and Burnished
Black wares at Challuabamba were all made of the same local clays (according
to David V. Hill, a geologist and consultant in petrographic analysis of ceramic
and lithic materials), but the wares can be distinguished from each other by
the atmosphere in which they were fired. Differential firing distinguishes not
only Challuabamba's pottery, but was characteristic of the Middle Formative
Period (2200-1200BC) overall in Ecuador and indicates masterful control of a
complicated process (see Shepard 1965: 86-90,213-224).
The three named types are very generally defined based on a few features.
"Red-on-Cream" is defined by a cream body and red paint. "Red-and-Black" is
defined by a basic oxidizing fire producing a cream body with parts (usually
either the interior or exterior of the body) that are smudged to black, and with
red paint combined with the black on some areas, usually without burnishing.
,t._
Figure 6.2. Sections through Challuabamba Cut 3. Note that it shows the walls of a trench ending at the
south wall of Sector G, as in the diagram upper right.
The trench wall goes north to south from A to G, west to east across the south wall of G, then north and
east around H, and north from H to B. Subsequently
sectors H, J, K were excavated. Carbon samples: TX·9241, about 30 em east of the G-H division, on
the upper surface of Level 5, base of Level 4, at 120 cm
below datum. TX-9027, 210cm below datum point, in Sector B just below boulders (Level 2) beyond
left edge of drawing. TX-9026, in Sector A in heavy sherd
scatter 230 em below datum.
161
Red-on-Cream and Red-and-Black share a set of jar and bowl forms with very
thin walls, typically 2 nun thick even for fairly large globular jars. "Burnished
Black" is reduction-fired and burnished, either completely or in part. This type
first appeared as a deep black burnished pottery, but over time began to show oxi-
dation at the end of the firing to achieve contrasts of black and white, or overall
gray tones. Variations notwithstanding, there is a unity in the Burnished Black
category, which is defined by reduction firing, the use of burnishing for decora-
tion and, importantly, by a distinctive set of 3 to 4 nun-thick, vertical-walled, flat-
based bowl forms. The consistency of these features suggests that they were
meaningful to the designers and users of these vessels. The division of the deco-
rative designs into two groups, one for Red-on-Cream pottery, another for
Burnished Blackware, also supports the definitions of the wares.
In Figure 6.3 we see the development of these wares through time, from
about 2000 BC at the bottom to about 1200 BC at the top. At the beginning of the
sequence all three wares were present, Red-on-Cream being the largest group
(on the right), as it remained, with a group of similar forms smudged to Red-
and-Black, and a smaller group already showing the thicker walls and high bur-
nish of Burnished Black (on the left). While Red-on-Cream remained the basic
pottery ware throughout the sequence, with its own forms and decorations, the
hybrid Red-and-Black slowly disappeared as Burnished Black peaked, and at the
end Burnished Black tended to merge with Red-on-Cream, though highly bur-
nished black vessels continued to be made in succeeding periods. The history of
Red-and-Black makes clear the increasing understanding among potters of firing
as the crucial process in obtaining color in ceramics. Viewed within the history of
Andean pyrotechnology, in which metallurgy was incipient, this was an important
step [Note 2].
The best data we have on pottery history comes from about 26 cubic meters
of cultural deposits in the three adjoining sectors, G, H, and J in Cut 3. A sum-
mary of the main pottery types by levels, based on rim sherds, is presented in
Table 6.2. The percentages in Table 6.3 show that oxidized wares, particularly the
Red-on-Cream group, make up the major part of the pottery in all levels and it
was increasingly dominant in later levels. The Black-and-Red type appeared
....
~
;3
;;l
~
~
:sI...
~
~
~
,~
.,: : !aJ
· .. ·····"1 6 IJ mif\ l ~!ILJ.. ~
I \ <II ,
ch ~l,' \- 19
~! l':~\" ..
5.'~\
Figure 6.3. Challuabamba pottery. Periods are divided by horizontal dashed lines, earliest at the bottom.
Period I: ca. 2000-1800 BC. Period II: ca. 1800-
1600BC. Period III: ca. 160Q-1400BC. Period IV: ca. 1400-1 200 BC. Types are divided by diagonal
and vertical dashed lines: Burnished Black on the left,
Red-and-Black in the center, and Red-on-Cream on the right.
163
# of rims:
Levell
2
3
4
5
6
7
TOTAL
972 78 122
768 23 176
238 j I 31
584 106 65
In 50 7
100
001
2696 268 402
early, and its vessel forms show that it developed out of the oxidized type, based
on a desire for a blackware. Unpainted reduced pottery of the Burnished Black
type appeared early but in very small numbers, but gradually replaced Red-
and-Black, which nearly disappeared in Level 2, when reduction-fired wares-
Burnished Black and its gray or brownish variants-reached their maximum. In
the surface level reduction-fired types remain fairly common, and Red-and-Black
made a small recovery, both probably reflecting the greater variety of pottery forms
and decorations characteristic of that period. Challuabamba's pottery has a very
coherent history, which is reflected in the closely related technical and artistic solu-
tions it achieved. Challuabamba's potters rarely made a purely utilitarian product.
That may reflect what was an ancient tradition by the second millennium (e.g.,
Pratt 1999), and it emphasizes the role of pottery as a communication medium.
While the internal development of ceramic art at Challuabamba is important,
we wish to note that it was part of a larger development which can only be noted
here. Features in pottery from seven sites or styles within the northern Andes offer
interesting comparisons with Challuabamba and suggest interactions with it:
Valdivia, Cotocollao, Machalilla, Cerro Narrio, Catamayo, Bagua, and Cerro
Nafiafiique (Figure 6.1).
Figure 6.4. The most common Red-on-Cream rim decoration is complete in the upper left, with
two partial examples below. This diagnostic set of sherds is from Cut 3, Sector H, Level 4c.
Art and Prestige Among Noble Houses 165
found very widely in native American belief systems (e.g., Urton 1981: 68-69),
provide a baseline for discussion of symbols at Challuabamba.
Compared with the rich vocabularies of symbols found in later societies,
Challuabamba begins with a restricted set of abstract symbols. From the start dec-
orations on reduced wares were different from those on oxidized wares. Oxidized
wares, which seem to have matured first, had rim decorations on the everted rims
of globular jars. Most common was a red rectangle flanked by vertical lines and
a curl on each side, a design popular over centuries (Figure 6.4, upper left). Also
early and long-lasting, but less common, are splashes of red paint made by flick-
ing a loaded brush to leave a string of random drops on the interior of cream
Figure 6.5. Burnished Black sherds from Cut 3, Sector H, Level4b. Scale is 12cm total length.
Figure 6.6. Solid modeled head of a bat facing inward from the rim of a Red-on-Cream basin from
Cut 2, Level 2. The whole sherd is 5.4cm high.
Figure 6.7. Burnished Black effigy jar from the offering in Burial 6, Cut 3, Sector H, Level 4.
It measures 6.8 em high, 9.2 em rim diameter.
/~
oj
~ .ED
. I .to
D
.to
:
OJ
:}lJ u.c:
E
0
<l::
".'0.".
( .c'l0"l:)
::l
u
.§u
, ell)
~ "0
ll)
.>..
to
U
"'0"
to
, O'"J
ll)
.~ v'">
0.
E
to
.t..;.
0
.'2"
tl>l)
"ii
<'t":
~
\C
!=::
OIl
~
Figure 6.9. View of the site from the top of Cut 3, looking east. Level I has been cleared. Cut 2,
with similar foundation lines, is in the center in the middle distance. Cut 1 is on the terrace edge in
the distance beyond the fence-line.
Figure 6.10. Cut 4, looking east, showing earlier phase of small stones flanking wall trench,
at depth of 60-45 em. In foreground are large stones of later revetment wall.
Acknowledgments
The Challuabamba project was carried out under authorizations of the Instituto
Nacional de Patrimonio Cultural del Ecuador dated June 6, 1995, June 17, 1996,
January 22, 1997, June 14, 1999, and June 16,2000. The project was funded by
research funds of the David Bruton Jr. Centennial Professorship in Art History at
the University of Texas at Austin, by gifts for Art History faculty research from
Dana De Beauvoir, and by the University Grant-in-Aid Program for Faculty,
Office of the Vice-President for Research and Graduate Studies, Virginia
Commonwealth University. Special thanks to Francisco Chimbo and the
Jaramillo family, represented by Juan Pablo Jaramillo, for permission to carry out
excavations on their properties. We thank Jorge Marcos for consultation on the
excavation collections and review of a draft of the paper. We are grateful to Joyce
Marcus and Karen Stothert for helpful comments on a draft of the paper. Any
remaining errors are the responsibility of the authors.
Notes
I. The spelling "Challuabamba," Quichua for "fish field," follows the spelling used by the Ecuadorian
Instituto Geognlfico Militar ( the source for all geographic names used here) and traditional in the
region. Max Uhle (1922: 21ft) used that spelling in the first report on the archaeology of the site.
Jij6n y Caamafio, beginning in 1930, changed the spelling to "Chaullabamba" (e.g., 1951:
145-158). The fact that "Chaullabamba culture" as described by Jij6n y Caamaiio corresponded to
the Chorrera culture as defined by Evans and Meggers (1957) (see Bischof 1975: 15, Tellenbach
1998: 272) makes return to the earlier spelling advisable in order to associate the name with mate-
rial more directly related to the Challuabamba site.
2. It is significant that the Formative traits associated with the rapid spread of rank societies in both
Peru and Mesoamerica include a common set of ceramic vessel forms, reduction-fired and differ-
entially-burnished vessels, and many common decorative designs. The spread of these features
through the equatorial Andean region appears to be prototypical for the wider spread of Formative
culture. Lathrap (1973) proposed similar conclusions on the basis of different evidence from the
tropical lowlands.
REFERENCES
Arellano, A. Jorge, 1994, Lorna Pucara, a Formative site in Cebadas Valley, Ecuador. Research &
ExpLoration 10 (I): 118-120. National Geographic Society, Washington, D.C.
Arellano, A. Jorge, 1997, Lorna Pucara. Un asentamiento del formativo tardio en el valle de Cebadas,
sierra central del Ecuador. Fronteras de Investigaci6n I (I): 78-100.
176 Terence Grieder et al.
Bennett, Wendell C., 1946, Excavations in the Cuenca Region, Ecuador. Yale University Publications
in Anthropology, Number 35. New Haven.
Bischof, Henning, 1975, La fase Engoroy - perfodos, cronologfa y relaciones. Estudios sobre la
arqueologia del Ecuador, edited by Udo Oberem, pp. 15-39. Bonn.
Bruhns, Karen Olsen, 1987, Los tallares de cristal de roca de Pirincay, Provincia del Azuay, y el inter-
cambio entre la sierra y la costa en el Formativo Tardfo. Miscelanea Antropologica Ecuatoriana
7: 91-100.
Bruhns, Karen Olsen, lames H. Burton, and George R. Miller, 1990, Excavations at Pirincay in the
Paute Valley of Southern Ecuador, 1985-1988. Antiquity 64: 221-233.
Burger, Richard L., 1994, Chavin and the Origins of Andean Civilization. Thames and Hudson,
London.
Carmichael, Elizabeth, 1981, Dataci6n mediante el carbono 14 de muestras de carb6n de sitios de
cenimica fina provenientes del altiplano del Ecuador. Revista de Antropologia. no. 7 (Nov.)
Secci6n de Antropologfa y Arqueologfa del Nucleo del Azuay de la Casa de la Cultura
Ecuatoriana, Cuenca.
Carmichael, Elizabeth, Warwick Bray, and lohn Erickson, 1979, Informe preliminar de las investiga-
ciones arqueol6gicas en el area de Minas. Rio lubones, Ecuador. Revista de Antropologia, no. 6
(lulio). Secci6n de Antropologfa y Arqueologfa del Nucleo del Azuay de la Casa de la Cultura
Ecuatoriana, Cuenca.
Collier, Donald, and lohn C. Murra, 1943, Survey and Excavations in Southern Ecuador. Field
Museum of Natural History, Anthropological Series, 35. Chicago.
Evans, Clifford, and Betty 1. Meggers, 1957, Formative Period culture in the Guayas Basin, coastal
Ecuador. American Antiquity 22: 235-247.
Flannery, Kent V. and loyce Marcus, 1994, Early Formative Pottery of the Valley ofOaxaca. Mexico.
Memoirs of the Museum of Anthropology, Number 27. University of Michigan, Ann Arbor.
Gomis, Dominique, 1989, La alfarerfa de Chaullabamba. Catedral Salvaje, revista cultural del edito-
rial EI Conejo, circula con El Mercurio 24: 4-5 (II junio).
Grieder, Terence, J982, Origins of Pre-Columbian Art. University of Texas Press, Austin.
Grieder, Terence, Alberto Bueno Mendoza, C. Earle Smith, lr., and Robert M. Malina, 1988, La
Galgada. Peru: A Preceramic Culture in Transition. University of Texas Press, Austin.
Guffroy, lean, 1987, Laja Prihispanique. Editions Recherche sur les Civilisations, Paris.
Hammond, Norman and Karen Olsen Bruhns, 1987, The Paute Valley Project in Ecuador, 1984.
Antiquity 61: 50-56.
Helms, Mary W., 1979, Ancient Panama: Chiefs in Search of Power. University of Texas Press,
Austin.
Hocquenghem, Anne-Marie, laime 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'lnstitut Franrais d'Etudes Andines 22 (2): 443-466.
Hosler, Dorothy, 1994, The Sounds and Colors of Power. MIT Press, Cambridge.
Idrovo Urigtien, laime, 1992, El Formativo Ecuatoriano. Museo del Banco Central del Ecuador,
Cuenca.
lij6n y Caamano, Jacinto, 1951, Antropologia Prehispdnica del Ecuador, 1945. La Prensa Cat6lica,
Quito.
Langdon, E. lean, 1981, Cultural bases for trading of visions and spiritual knowledge in the
Colombian and Ecuadorian montana. In Networks of the Past: Regional Interaction in
Archaeology, edited by P. D. Francis, et al. Department of Archaeology, University of Calgary.
Lathrap, Donald W., 1973, The antiquity and importance of long-distance trade relationships in the
moist tropics of Pre-Columbian South America. World Archaeology 5 (2): 170-186.
Lechtman, Heather, 1977, Style in technology-some early thoughts. Material Culture: Styles.
Organization, and Dynamics of Technology, edited by H. Lechtman and R. S. Merrill, pp. 3-20.
1975 Proceedings of the American Ethnological Society.
Part III
Landscapes of Power
Introduction
HELAINE SILVERMAN AND WILLIAM H. ISBELL
Space is the means, medium, and outcome of human action (Soja 1989). It is
usually dealt with in terms of a scale, "place" being a small portion of space
created and emotionally experienced by each actor, "landscape" a larger unit
encompassing numerous places, and "space" the largest unit, including even the
abstraction (Hirsch 1995). Our use of "landscape" is not intended to be rigorous
about scale, but inclusive and encompassing larger and smaller spaces, under-
standing them as produced by human action and experienced by human bodies,
more and less intentionally.
Places and landscapes are manufactured by physical and cognitive alteration
of topography and built environment, but they only have meaning when experi-
enced by senses, emotion, and memory-imagined and real. Landscapes are cog-
nitively perceived and conceptually constructed by human actors. Consequently,
they have different meanings for different individuals and groups (Rodman 1992),
meanings and memories that may be contested. Thus, landscape also has the
aspect of power for it is the medium for action in the present (directed at the past,
present, or future). Therefore, agency operates in a spatial context. As theorized in
a poststructural social science, space-rather than structure-is the repository for
behavioral potentialities and constraints, the primary resource for action, and the
accumulated result in the present of past action (historical contingency).
Emphasis on landscape in the study of prehispanic Andean cultures includes
both structural and poststructural approaches with scholars often preferring struc-
turalism which assumes uniform response to landscape in terms of cultural rules.
Also, Andeanists have employed ethnographic analogies that suppose changeless
Andean culture, psychology, and political systems. In Central Andean archaeology
the exploration of power in regional landscape may be said to have begun with
Zuidema's (1964) study of the structural organization of the ritual geography of
the Inca capital of Cuzco, followed by Lathrap's (1974, 1985) and Isbell's (1978,
181
Chapter 7
The Gateways of Tiwanaku
Symbols or Passages?
JEAN-PIERRE PROTZEN AND STELLA NAIR
INTRODUCTION TO AN ARCHITECTURAL
APPRECIATION OF TIWANAKU
Tiwanaku sits an elevation of 3840 m asl in a broad, treeless valley that drains into
nearby Lake Titicaca, the highest navigable lake in the world. The site of
Tiwanaku was the center of a civilization of the same name that emerged here
around 300BC and lasted to about AD llOO when it collapsed. At its apogee
Tiwanaku had expanded its sphere of influence far beyond its immediate envi-
ronment into modem day northern Chile, southern Peru, and eastern Bolivia.
Over the last five centuries, the archaeological site of Tiwanaku has suffered
unspeakable damage inflicted by treasure hunters, hacendados, farmers, builders,
practice-shooting armies, and railroad barons. The site today consists of two dis-
tinct areas (Figure 7.1). The larger of the two, just east of the town, centers around
the huge man-made platform mound of Akapana, and comprises the main struc-
tures of Kantatayita, the Semi-subterranean Temple, Kalasasaya, Putuni, and
Kerikala. The smaller area is to the south of the town and consists of the man-
made mound of Pumapunku (literally, "gate of the puma"). There is little standing
189
FULL-SIZED GATEWAYS
The known large gateways at Tiwanaku are the aformentioned Gateway of
the Sun, Gateway of the Moon, Sandstone Gateway, and the fragments of several
gateways at Pumapunku. Most of these gateways and their fragments have been
previously documented. In 1848 Leonce Angrand, a French diplomat, made
exquisite and accurate drawings that are now housed in the Bibliotheque
Nationale in Paris (Prtimers 1993). Alphons Sttibel, a German geologist and stu-
dent of volcanoes, measured most of the gateways, gateway fragments, and many
stones in 1876-77 with astonishing precision. His work was later published with
the assistance of Max Uhle who added an historical dimension to the publication
(Sttibel and Uhle 1892). From the measured and drawn fragments of Pumapunku,
Sttibel and Uhle concluded that the fragments came from three distinct gateways.
Arthur Posnansky, an Austrian engineer and naturalized Bolivian, who spent con-
siderable time studying Tiwanaku's ruins, thought that there may have been four
gateways (Posnansky 1945 [II]: 146). The piece that Posnansky believed to
belong to the fourth gateway, Sttibel and UWe suspected to belong to Gateway I
[Note 1] (Sttibel and Uhle 1892: plate 29, figs. 29c, d).
All the above mentioned gateways share a number of common traits. They are
(or were) monolithic. That is, they were cut from a single slab of stone, including
the threshold; the actual doorway openings are set into a double-stepped recessed
frame, or chambranle [Note 2] on both sides; and the reveals of the jambs and the
doorhead are beveled, Le., they flare open to one side, such that the opening of the
actual doorway is larger on one side and smaller on the other (Figure 7.2).
The Gateway of the Sun and the Gateways of Pumapunku share additional
features. They are plain on one side, and divided into several fields by a step
molding on the other. On the plain side, the doorway is flanked by two rectangu-
lar recesses or pockets, one on either side of the doorway (the function of which
will be explained below) (Figure 7.2). The step molding that divides the other side
at about two-thirds of the height of the gateways wraps around, crowns or roofs,
the doorway head, thus defining six distinct fields: one each on either side of the
doorway below the step molding, one each above it, a field in the middle, the
~ J-----,C ~ 7.2
7.4
Figures 7.2-7.4. (Figure 7.2) Plan, Section, and Front View of Gateway III. (Figure 7.3) Back side of
Gateway ill. (Figure 7.4) T-shaped sockets suggesting the addition of building blocks to Gateway III.
Pumapunku Gateway II
We identified at least three more pieces for Gateway II than had been
recorded by Sttibel. This gateway is nearly complete, all that is missing is a piece
Pumapunku Gateway I
The piece that Posnansky believed to be a part of a fourth gateway, and that
Sttibel and Uhle suspected to belong to Gateway I, we have positively identified as
the left jamb of Gateway I, or at least of one exactly like it. On this gateway, the
step molding roofs the doorway in two steps, not three as on the previous two gate-
ways. We also noticed that on this gateway the Type 2b niches have slightly
different proportions (Figure 7.7). It is this difference that let us associate the frag-
mentary left jamb piece with Gateway I, for its Type 2b niche matches the exact
measurements of that on the right jamb. We will return to the question of this dif-
ference and its consequences later. Little of the gateway's lintel has survived, but
what is left is· enough to affirm that, like Gateway II, it had no incised frieze.
Instead, where the frieze would be, there is a large recessed area into which, again,
some equivalent ornamentation could have been set (Figure 7.8). Again, Gateway
I has T-shaped cramp sockets, recessed or plain, in similar places as Gateway III,
and therefore it, too, was in a similar context with walls extending its main plane
and walls extending perpendicular to its front side and flanking the opening.
Pumapunku Gateway IV
A stone fragment at Pumapunku that came from a door head does not fit any
known gateway. We therefore postulate the previous existence of a fourth gate-
way at Pumapunku.
Gateway of the Sun
As noted above, the Gateway of the Sun shares a wide range of features
with other gateways at Tiwanaku, but it also differs from them in some aspects.
..Ol
7.5 J
7.7 '0]
'.'j
~O~
7.6
7.8
~
l'll
~
~
's"a.
~
~
Ql
5
8
Figures 7.5-7.8. (Figure 7.5) Front view of Gateway II. (Figure 7.6) Back side of Gateway II. (Figure
7.7) Back side of Gateway 1. (Figure 7.8) '"""
Front view of Gateway I. ;S
Figures 7.9-7.10. (Figure 7.9) Back side of Gateway of the Sun. (Figure 7.10) Front view of
Gateway of the Sun.
7.13
Figures 7.11-7.13. (Figure 7.11) Frieze on the Gateway of the Sun. (Figure 7.12) Detail of central frieze
(meander center left). (Figure 7.13) Detail of
left-hand frieze (meander).
g
~
~
~
~
J
l
rJl
~
Z
:Ql
:;.
Akapana Gateway
In the summer of 1995 we recorded the remains of yet another gateway on
the summit of Akapana. Some of the pieces of this gateway were known to the
local attendants at the site but, to our knowledge, they have never been docu-
mented. Stiibel and Uhle (1892: 27) did mention the remnants of a gateway on the
mound's plateau. Whether these remnants are identical with the pieces we
recorded will probably never be known. This gateway differs from the aforemen-
tioned gateways in that it is not a monolith, but a trilithon, that is, a gateway
assembled from an enormous lintel and two jamb stones (Figure 7.14). Only frag-
ments of the lintel, and one candidate for the left jamb stone remain. A cut stone
depicted in Squier (1877: 280, center right) has all the characteristics of our jamb
stone and could be identical with it [Note 5]. The right jamb stone, no longer
existing, may have been the one illustrated by Stiibel and Uhle (1892: plate 38,
fig 22). It should be noted that Bernabe Cobo (manuscript: Libro XIII, capitulo
XIX; Cobo 1964 (II): 196) who visited Tiwanaku in 1610, and probably again in
1620, described just such a gateway: "cinquenta pies al oriente del (Acapana) ha
quedado en pie una portada grande de solas tres piedras bien labradas, a cada lado
la suya, y otra encima de ambas" [Note 6].
The gateway atop the Akapana shows a very unusual construction technique.
The underside of the lintel is cut at an angle, the exact angle of the bevel of the
door head flaring open to the inside. The top of the jamb is cut at the same bevel
or angle to receive the lintel. With this configuration, the bottom of the Type 2a
niche above the jamb had to be carved out of the beveled surface of the jamb. It
is noteworthy that the short extension of the niche's height so created is exactly
what is needed to bring the niche's dimensions in line with the established pro-
portions (Protzen and Nair 1997: 155-156). These details bespeak a remarkable
sophistication in stereotomy, or the art of stone-cutting, presuming an under-
standing, if not knowledge, of descriptive geometry.
The reader will have noticed that the jamb stone lacks the Type 2b niche. We
are not sure that the jamb stone we identified really was connected with the lin-
tel pieces. While the position and dimensions of the niche bottom and of the step
molding on the jamb match up precisely with the corresponding parts on the lin-
tel, the jamb stone's depth, or thickness, is 3 cm less than that of the lintel. It is
thus possible that the jamb stone belonged yet to another gateway with the same
features as the ones we proposed above, but with different dimensions.
Figures 7.14-7.16. (Figure 7.14) Construction of Akapana Gateway. (Figure 7.15) Comparison of
Scheme I and Scheme 2. (Figure 7.16) Gateway of the Moon,
front and back.
~
714~
7.15
7.16
']
lJ~ tL
>----
203
The gateways described thus far, in spite of their apparent identical design, fol-
low two different vertical composition schemes. In other words, the vertical layer-
ing of the various elements, the lower fields and within them the bottom and top
of Type 2b niches, the position of the step molding, and the upper fields and within
them the bottom and top of Type 2a niches, is not the same for all gateways. The
Gateway of the Sun and Gateway I follow one pattern: Scheme 1; Gateways II, III
and the Akapana Gateway follow another pattern: Scheme 2 (Figure 7.15).
The differences, as mentioned in our description of Gateway I, seem to stem
from the number of steps in the molding that wraps around the head of the door-
way, which in tum determines the fields on either side of the doorway. This obser-
vation led us to revise our earlier determination of the proportions of Type 2b
niches (Protzen and Nair 1997: 155-156). Originally, we established the propor-
tion of 1: 1.4495 based on the entire set of such niches found at the site. Although
there are variations in the dimensions of individual niches, the spread of the width-
to-height ratios of all niches measured only 0.043 standard deviations. This gave
us the confidence that we were on the right track. Re-analyzing the raw data, and
plotting only the Type 2b niches found on the gateways, we do, however, fmd two
distinct clusters with differing average proportions. The niches in the first vertical
composition scheme (Gateway I and Gateway of the Sun) have average propor-
tions of 1:1.5474, while the proportion of those in the second scheme (Gateway II,
and III) average 1:1.3431. The newly created sets have a much narrower spread of
values, with respective standard deviations of 0.007 and 0.009. Henceforth, we
will distinguish between niches of Type 2bl and 2b2 for niches on gateways of
Scheme 1 and Scheme 2, and Type 2b3 for all other niches of this kind. A similar
analysis of the gateways' openings measured on both the front and the back,
reveals that they all are in the average ratio of 1:2.1435, regardless of the vertical
composition scheme to which they belong. In spite of the noted differences, taken
together, the gateways discussed above form a single stylistic group; the two
different compositional schemes are only variations of the same theme.
Although the Gateway of the Sun and Gateway of Akapana do not have
cramp sockets on their narrow sides (they may not yet have been carved), the par-
tial Type 2a niches on either suggest that these gateways, too, were meant to be
embedded in walls extending out to the left and right that would have continued
a row of Type 2a niches. As seen above, Gateways I, II, and III similarly were
meant to be imbedded in walls extending to the left and right of them. Yet, the lat-
ter walls did not immediately continue the row of Type 2a niches above the step
molding, for if the spacing of these niches were the same or at least similar to that
on the Gateways of the Sun and of Akapana, one should find the next niches at
least partially outlined in their respective fields on the gateways themselves.
Curved Architraves
The doubly curved lintels, seen in elevation, are cut in a shallow elliptical arch;
in cross-section their backsides recede in a steep parabolic curve. The combined
curvatures produce an incredibly beautiful, but complicated surface the creation
of which would tax any stonemason's skills today. One such architrave of andesite
7.17
[ ~ l&f5Jb" L
7.20
=N
Figures 7.17-7.20. (Figure 7.17) Meander Frieze of Gateway III. (Figure 7.18) Sandstone Gateway,
front view. (Figure 7.19) Doubly curved architrave at
Kantatayita. (Figure 7.20) Anticephaloid architrave. (Drawing by Mireille Rodier).
7.18
f- --I
..u
7.19
¢....,
til
:ll::>s
~
til
~
'"C
~
:t:isl
l:l::>s
l=lo
IJ:J
[
;-
.~.,
MINIATURE GATEWAYS
Strewn about Pumapunku are not only fragments of full-sized gateways but
a large number of building stones of various shapes. While some of these stones
are still intact, most are badly mutilated. Nevertheless, working at Tiwanaku over
the last few years we succeeded in determining the relative position of many
stones and establishing their relationship to others. Some of these building stones
fit together to form both blind and open diminutive gateways.
Blind Miniature Gateways
We identified two fragments that once formed a monolithic, blind miniature
gateway, 48 cm wide and 95 cm high (Figure 7.21). From the front, these frag-
ments have the appearance of a gateway set into a double stepped, recessed cham-
branle. Just inside the gateway there is a shallow space, 17.4 cm deep, the sides
of which are adorned with a step molding at about a quarter of the space's height.
The back of the space is formed by a plain wall some 8 cm thick. One might argue
that blind gateways are not really gateways, but more closely represent niches.
But niches are unlike gateways in their basic form. Cross-sections in plan and ele-
vation reveal the difference: in the case of niches there is a smooth beveled tran-
sition from the opening to the back of the niche, whereas in the blind gateway
there is a deliberate hiatus, a distinct hint of a room into which the opening leads
that is absent in the niche.
Composite Blind Miniature Gateways
Our "discovery" of this blind miniature gateway was crucial for our under-
standing and interpretation of other stones at Pumapunku. There is a variety of
stones, executed in right-handed and left-handed versions, that feature a stepped
rabbet on one side, beveled jamb reveals, a step molding at about a quarter of their
7.23
7.21
Figures 7.21-7.23. (Figure 7.21) Blind miniature gateway. (Figure 7.22) Composite blind miniature
gateway, decorated with crosses. (Figure 7.23) Row of composite blind miniature gates.
210 Jean-Pierre Protzen and Stella Nair
height, and a straight rabbet on the opposite side with the traces of T-shaped cramp
sockets. We figured these stones to be the jamb stones of blind miniature gate-
ways. A plain stone slab fitted into the straight rabbets on the "back" side and
anchored to the left and right-sided jamb stones with cramps, crowned with a suit-
ably shaped lintel stone, of which there are several, indeed combine to form a
composite blind miniature gateway.
Some of the jamb stones, as described above, are decorated with crosses
recessed within larger crosses, with only half a cross motive at the upper end of
the stone. A lintel fragment with half a cross matching in position and dimensions
the half cross on the jamb stones, suggests that some cross-stones, too, formed
blind miniature gateways (Figure 7.22).
A number of H-stones (see Protzen and Nair 1997 for a typology of the
building stones at Pumapunku) with Type 1 and 2a niches, within which are
carved the corresponding niche icons, together with lintel stones with recessed
panels could have composed a whole row of blind miniature gateways. All the
details on the H-stones match such a reconstruction: beveled jamb reveals, step
moldings, and rabbets with the appropriate T-shaped cramp sockets for the attach-
ment of a stone backing (Figure 7.23). It should be noted that Sttibel and Uhle
(1892 [II]: 38) had anticipated this very same combination, although they had not
recorded any actual lintel stones.
Open Composite Miniature Gateways
From other stones and stone fragments we infer that there were also com-
posite open miniature gateways of both designs, plain and decorated with crosses.
Monolithic Miniature Gateways
Further fragments show the existence of monolithic miniature gateways: the
two pieces of what Posnansky called the "Little Pumapunku," and two fragments
of what we designate as Gateway A (Figures 7.24, 7.25). Both Angrand and
Sttibel found the Little Pumapunku still intact, measured, and drew it. Their
respective drawings agree fairly well, with one exception: they differ significantly
in the representation of the small window above the gateway. Angrand shows the
window to be mushroom-shaped, flush with the outside of the gateway, and set
into a rectangular recess on the inside (Prtimers 1993: 460, fig. 45). Sttibel and
Uhle (1892 [I]: plate 36-2, figs. 2, 2a) show three detail drawings of the opening.
When we tried to reconstruct the window from these drawings we were not able
to make the two sides agree; the window could not be built as drawn. Today that
window is destroyed, and key elements are missing, such that we cannot recon-
struct it with absolute certainty. Enough clues, however, are left to show that there
is no evidence whatsoever for the "platelets" shown on the inside by Sttibel, nor
for the recessed tympanum on the outside. Given the accuracy of Sttibel's other
work, we can only guess that the sketches of this particular detail got garbled
7,26
7.27
Figures 7.24-7.27. (Figure 7,24) "Little Pumapunku." (Figure 7.25) Gateway A (fragment of left
jamb on the right). (Figure 7,26) "Escritorio" type stones follow the same composition scheme.
(Figure 7,27) Gateway A and Five Niche stone.
Another Style
A lone fragment of a miniature gateway has an entirely different design or
style. The gateway's inside appears to have been entirely plane; on the outside the
actual doorway is set in the usual double stepped recessed chambranle, but that
chambranle is then triply framed in relief, the two outermost frames in quadruple
steps and the innermost in triple steps. This stone may be related to two other iso-
lated pieces, one at Pumapunku, the other inside an office of the Tiwanaku
Museum complex. Both these stones show niche icons similarly framed in three
layers of relief. The three stone fragments, the two niche icon stones and the gate-
way fragment, are unique, no others of this kind have been uncovered to this day,
and none of their features can be related to the other stones. The three stones are
not only different in design, but they vary in dimensions, material, and execution.
Compared with the finish of other stones at Pumapunku, the three stones are not
cut with the same precision as most of the others, and are of a much poorer qual-
ity andesite. Of course, different styles, materials, and execution do not necessar-
ily stand for different historical periods-they can be contemporaneous-but they
most likely would signal buildings with different functions or status. But three
stones do not make a building, and will not support a theory.
"- )J I
ur
-.......J ff
ill ~
" 1"-1 \ ~ '- l
" r-'
II "
() .. ---
y /[ .....
.. L -
'" h • ~A •
u. V /
wV
bu
ill / 1/
C(
/
'" A B C 0 E F x
7.28
7.29
Figures 7.28-7.30. (Figure 7.28) Gateway A and Gateway III. (Figure 7.29) Scaling factor relating
the miniature to the full-sized architecture. (Figure 7.30) Entrance to Putuni today (from north and
perpendicular to the gateway's axis).
Half-scale Gateways
If we have been able to establish a dimensional and compositional relation-
ship between the small Gateway A and the Escritorio type sto~es and a possible
connection between the Little Pumapunku gateway and one of the Escritorio type
stones, we have to this date not found any clues as to where this half-scale archi-
tecture stood, nor how it fit in with the full-scale architecture.
CONCLUSION
Our investigations have established that the large monolithic gateways of
Tiwanaku were not free-standing, but set, or meant to be set, into walls extending
the gateways' main plane. Fritz Buck and Javier Escalante (Escalante 1993: figs.
162 a, b, c, d) have presented reconstructions in which the Gateway of the Sun is
flanked by several copies of the Gateway of the Moon. In our view, this is not a
plausible combination because the differences in design and proportions between
the two gateways do not permit them to be put side by side. However, it is entirely
possible to assemble walls that potentially could accommodate gateways of either
Scheme I or 2. The hypothetical wall corresponding to Scheme 1 shown here is
made up of actual building stones we recorded at Pumapunku (Figures 7.31,
7.32). All the building stones used have the appropriate dimensions, and have
T-shaped clamp sockets in the right locations. Plausible as it may appear, the
proposed reconstruction is altogether speculative-we have no evidence that such
a configuration existed, nor that it was connected in this form to any gateway-
yet it is suggestive and may lead to further research.
Perhaps most surprising among our findings to date is that the openings of
the large monolithic gateways at Tiwanaku were, or were meant to be flanked by
walls perpendicular to the gateways' main plane on their front side. (Note that in
the above hypothetical reconstruction. walls are also projecting from the front
side). These walls would have constricted and channeled the approaches to the
gateways, and thus perhaps underscored the passage of the gateway proper by a
Figures 7.31-7.33. (Figure 7.31) Hypothetical reconstruction of a wall with actual stones found at
Pumapunku (back side). (Figure 7.32) Hypothetical
reconstruction of a wall with actual stones found at Pumapunku (front side). (Figure 7.33) So-called
model stone at Kanatayita.
.N...
\C
!sa,
i:l
7.32
~ 7.31
Notes
I. In this chapter, the numering of the gateways follows that of Posnansky.
2. Chambranle: "A structural feature, often ornamental, enclosing the sides and top of a doorway,
window, fireplace, or similar opening. The top piece or lintel is called the transverse and the side
pieces or jambs the ascendants" (Harris 1977: 104).
3. In another paper we have described the three basic niche types found at Tiwanaku: Type I, Type
2a and 2b. We also have established that Type 2a and 2b niches can be inscribed in rectangles with
very specific proportion: 1: 1.0394 for Type 2a, and I: 1.4495 for Type 2b (Protzen and Nair 1997).
REFERENCES
Castelnau, Francis, Comte de, 1850--59, Expedition dans les Parties Centrales de l'Amerique du Sud,
de Rio de Janeiro ii Lima, et de Lima au Para. Paris.
Chapter 8
Religious Ideology and Military
Organization in the
Iconography ofaD-Shaped
Ceremonial Precinct
at Conchopata
INTRODUCTION
Between August 1997 and January 1998, the authors conducted archaeological
excavations at Conchopata, on the edge of the modem city of Ayacucho, in the
central highlands of Peru (Figure 8.la). Research was undertaken at Conchopata
for three primary reasons. First, we hoped to learn more about the art and iconog
raphy of the Middle Horizon. Oversize Conchopata pottery decorated with repre
sentations of the Front Face Staff God, which also appears on the Gateway of the
Sun at Tiwanaku, is well known from the site and Conchopata's imagery raises
crucial questions about Andean prehistory. Second, little was known about spatial
organization at Conchopata, and particularly about the organization of craft
production. Third, and perhaps most decisive, was the history of destruction at the
site and the continuing threat of loss of archaeological information (see summary
in Isbell and Cook, this volume).
225
Airport
Terminal
Conchopata Site
Pre erved Area
b
Figure 8.1. (al Location of Conchopata with reference to Ayacllcho. Note encroachment of airport
runway; (b) Sectors of the preserved area of Conchopata.
Our 1997-98 excavations concentrated on Sector B (Figure 8.1b), adjacent
to Ayacucho's airport. We excavated some 450 sq m during the field season,
recovering approximately 2.5 tons of archaeological materials, mostly pottery
fragments that range from extremely fine to coarse domestic wares. Evidence
associated with the production of this pottery was also recovered, including
molds, smoothing and burnishing tools, paints, pestles to grind up ceramics, piles
of tempering sand, and kilns. In addition to pottery production loci, a series of
other architectural remains were uncovered including burial areas, living spaces
and, perhaps most important, a ceremonial enclosure constructed during the early
Middle Horizon.
10m
z.o---- 1997-98 Excavation
227
JJis
Ia
3m
@)
Figure 8.2. (a) Architectural remains in Sector B-III at Conchopata; (b) D-shaped ceremonial
precinct as it appears in Stratum B.
vessels. This small compound revealed simple stone and earth drainage canals
located under the wall foundations and floors.
The most exciting discovery was made in the second subsector. We found
a D-shaped building with evidence of numerous activity areas devoted to ritual.
We define this building as a ceremonial structure.
Burials were recovered in the third subsector from what were initially resi-
dential units. Upon the death of the inhabitants the floors were broken and cylin-
drical cysts were excavated into the underlying rocks, lined with flat stones and
then sealed with mud. Large stone slabs covered each tomb. The majority of the
tombs show evidence of post-depositional disturbance and typically contain only
fragmentary osteological remains and a few grave goods. The disturbance proba-
bly was caused by conquered populations who invaded Conchopata to defile
tombs and sacred areas after the site was abandoned and after the collapse of the
capital city of Huari. It is also possible, of course, that the tombs simply reflect
Huari mortuary behavior.
Finally, the fourth subsector consists of a residential unit with rectangular
rooms, mud plaster and white paint on the internal wall surface. Associated with this
compound is a kitchen and patio area where ceramic vessels were made and dried.
Small pits with remains of guinea pigs as well as burnt and unburned camelid bones
were identified in the comers and along the internal edges of some enclosures.
It appears that Conchopata was an important pottery production center and
that it also had special areas for worshiping the supernatural, houses for craft
-3m -
--3m - b
Figure 8.3. (a) D-shaped ceremonial precinct as it appears in Stratum C; (b) D-shaped ceremonial
precinct as it appears in Stratum D.
the center of the enclosure. It measured 2.2 m north to south and 4.15 m east to
west with a depth of 15 to 20 em. Ovoid hammer stones-probably for breaking
ceramics-were dispersed throughout this feature that included intentionally
smashed sherds as well as camelid bones.
The majority of the material recovered consists of fragments of elaborate
ceramic urns and necked jars. The pottery shows a superior manufacture in forms,
finish and decoration. It was here that an entire oversize modeled face was recov-
ered, along with urn fragments painted with human figures kneeling on a small
reed boat and holding a shield in one hand, and either a bow and arrows or an axe in
the other (Figure 8.5). Other human figures may be warriors or priests (Figure 8.6);
all appear to be males. Numerous geometric designs on the decorated pottery frag-
ments probably depict decorations on the clothing worn by these men (Figure 8.7).
Feature 3
This feature consisted of a ritual camelid burial. It was found close to
the southern wall of the enclosure. The burial appeared to have been sealed with
mud, now a compact soil layer located immediately over the skeletal material.
The cranium was oriented toward the north and was deposited on a hardened sur-
face of burnt soil. Postcranial remains were located on a bed of gravel deposited
over the same burned surface. Despite the poor preservation of the bones it was
clear that the camelid was immature and was probably a ritual sacrifice.
Stratum C (Figure 8.3a) was defined by darker gray soil mixed with fine
orange dirt, producing a pale pink color along the interior wall of the D-shaped
building. Toward the southern edge there was a clear change in the stratigraphy
which, although contemporaneous, is more gray and compact, as if burned. Only
a few loose stones were found in the northern half of the structure, but a feature
formed of large stones occurred in the south. This level was located 50 or 60 cm
from the wall foundation with a depth of 10 to 15 em, covering a compact plaster
floor [Note 1] with a series of new features.
Feature 4
Next to the interior northern wall of the D-shaped building, and below
Feature 3 was a lens of fine pinkish ash that covered the floor. On its surface lay
a large, globular and utilitarian vessel with a flat base and a curved everted neck,
discarded in its place of use. It has chevrons under the lip and parallel,
231
cDE
Figure 8.5. Images of warriors kneeling in reed boats on Conchopata pottery.
undulating lines on the neck with three protuberances decorated with curved
lines. It appears to be an Qkros style vessel with a form similar to those of the
Huarpa period.
Feature 5
This irregular feature was located immediately below Feature 1 in the north-
west comer of the enclosure. It had a depth of 20 em, ending at the plaster floor.
It was also associated with the lens of burned pale pink-gray soil containing
burned material resembling contemporary plastic.
This feature consisted of another concentration of medium-sized ceramic
jars and utilitarian bowls associated with a broken ceramic smoothing tool. As it
was located immediately on the floor of the structure, and the vessels were
AB
Figure 8.6. Images of standing warriors on Conchopata pottery.
restored in their entirety, it would appear that these vessels were broken inten-
tionally during ritual activities.
Feature 6
This feature containing basalt flake debitage was covered with fine sandy
soil with gray and pink patches. It consisted of an intentionally fragmented jar
located just inside the northern wall of the D-shaped structure. Its base was sit-
ting in a 39 em-wide pit, associated with a pounding stone.
This vessel corresponds to the de facto deposit of an anthropomorphic
necked jar with a globular body and conical base in the Chakipampa style, with
octopus and starfish designs and circles with dots. The field of decoration covers
the upper half of the vessel body, ending at the insertion of the neck.
Feature 7
A pit was located in the northeast comer of the enclosure near the previous
feature. It contained decorated and undecorated pottery corresponding to two or
more vessels, one in the Chakipampa style, and three stone tools, a polisher and
two pounding stones, next to a complete Chakipampa style spoon with a small
curved handle decorated with a snake.
San
J ••
K
Feature 8
This feature was identified in the northern sector of the enclosure near
Features 6 and 7. It consisted of an irregularly shaped concentration of thick-
walled, undecorated jar and urn fragments. It contained a number of vessels that
were broken intentionally and covered with a lens of fine sediments. A large peb-
ble chopper was located nearby. The feature had a depth of 8 to 10 cm.
Feature 9
This was another concentration of large ceramic fragments and camelid
bones situated on the floor of the building. Located in the northeast, it had a depth
of 15 cm. The ceramic sherds were thick-walled, corresponding to an undeco-
rated urn. Disarticulated camelid remains, flakes and blades of basalt, and what
appeared to be small blocks of burnt plastic, were also recovered along with the
same fine sediments that cover almost all the features. Approximately 50 cm to
the west, a small diatomite square (8 cm thick) was found directly on the floor.
234
Feature 10
Feature 10 also was located in the northeast section of the structure, close to
the previous feature. Consisting of a further concentration of thick-walled, undec-
orated ceramic sherds, the feature produced a large urn of poor manufacture. The
vessel represents another de facto depositional event, as it was broken intention-
ally upon the abandonment of the structure, as part of an unknown ritual. A whole
pot was placed on top of the fragments and pushed up against the wall. This pot
has a composite body of a straight everted flaring inferior half and a straight
inverted flaring superior half, with two handles and a short neck. It was undeco-
rated and in good condition.
Feature 11
Another concentration of broken pottery, this feature was located 1.2 m from
the northeastern edge of the building, close to the ring of calcareous rock near the
middle of the room. Situated beneath Feature 2, Feature 11 shared the same
general characteristics as previous features such as the covering of fine gray-pink
sediments. Whereas other concentrations, however, contained fragments of
decorated Conchopata urns, this feature presented remains of a large Chakipampa
style jar with rough finish.
Feature 12-A Possible Solar Clock
This unusual feature was found in the northeast quadrant of the D-shaped
building, but close to its center. It was a semicircular lithic object of calcareous
rock, 10 to 12 cm thick and 8 cm high, placed there when the floor was laid.
The object was broken but marks on the floor show that it was originally circular,
forming a ring about 1 m in diameter. Inside the stone circle a cylindrical rock
with a conical end was found. This object measures 60 cm long with a diameter
of 25 cm. It was discovered lying on its side within the ring. The surrounding
sediments were not compacted nor was there evidence of a floor within the
ring, so we believe that cylindrical stone was placed upright in the center of the
little circle. We furthermore suggest that the feature represents a solar clock or
sundial.
Feature 12 was also covered by a lens of fine pinkish-gray sediment, the
result of a general burning episode prior to the filling in of the D-shaped build-
ing. Excavation in the interior of the ring revealed soil varying from compact to
semi-compact, associated with a few thick-walled, decorated and undecorated urn
fragments. Further segments of the ring were also recovered in association with
small blocks of clay and diatomite. Bedrock was reached 25 to 30 cm below the
Feature 17
This was a pit 70 cm in diameter and 40 cm deep, near the northeast edge of
the building, excavated through the floor to bedrock. It contained the flexed
remains of a young camelid. The faunal remains were covered by a soil contain-
ing a fragment of Spondylus shell, a smoothing tool and fragments of undeco-
rated pottery and decorated sherds in the Chakipampa style.
Feature 18
This pit was made by breaking through the floor surface. It was located in
the southeastern sector of the enclosure and measured 90 cm across and 38 cm
deep (to the bedrock). It contained another camelid burial in very poor condition,
but perhaps secondary. The pit fill contained soil with a few domestic ceramic
fragments.
Feature 19
This is another offering pit in the central part of the D-shaped building. The
oval-shaped pit measured 1.2 m long and 75 cm deep. What appear to be adult
camelid remains were found resting on bedrock at the bottom of the pit, some
apparently burned. The pit was filled with soil containing the distal end of a pro-
jectile point and a few fragments of domestic pottery. The discovery of these
offering pits shows that the D-shaped bJlilding was a ceremonial structure where
camelid sacrifice took place.
241
The first theme corresponds to human figures found on the upper half of the
exterior of urns. The men are represented kneeling, in profile, on a reed boat hold-
ing weapons and a shield in extended arms (Figure 8.5). The theme varies in the
warriors' facial tattoos, their weapons and the designs on their shields. They have
headdresses with either rectangles or zigzags made with small rhomboids similar
to those motifs found in lowland Amazonian societies. There are two basic types
of facial tattoos: repeated step blocks or lines with zig-zags that curve around the
eye. The figures wear tunics decorated with small checkerboard squares that are
sometimes filled with little circles. In their right hand they carry a shield with
feline heads and chevrons on the top and bottom or with concentric circles. The
u.u..l.-.....l'-"'-- A
Scm
left hand holds weapons: a bow and arrows or an ax with a handle and a tassels
on top.
The second type of warrior is also represented on urns, but is found equally
on the inside surface as the outside. These profile human figures may have been
Nasca-like Motifs
We believe that there are motifs in our ceramic corpus that are Nasca-like
[Note 2]. We specifically point to a large composite vessel with the same careful
finish as the urns, but with different design motifs. Its rounded body form is dec-
orated with lobed motifs similar to those of the Chakipampa style (Figure 8.8e).
The flaring neck has various types of designs filling two panels defined by central
bands of white and black. Near the lip, two horizontal bands are present, the upper
divided into small alternating squares of color and the lower filled with chevrons.
A line divides one of the neck panels into two fields with repeating figure-8s, and
four small rectangles with stylized human heads.
There is only one panel on the other side of the vessel. It has four heads, the
lower two on small pedestals. The heads are round with circular eyes, a straight
nose and oblong mouth. A band of interlocking frets surrounds each face, and on
Acknowledgments
Our excavations were financed by the Wenner-Gren Foundation for
Anthropological Research whose support we gratefully acknowledge. The project
was conducted by a team of researchers from the Universidad Nacional de San
Cristobal de Huamanga under the direction of Principal Investigators Jose
Ochatoma Paravicino and Martha Cabrera Romero. The research team included
Ismael Mendoza, Walter Lopez, Maria Cahuana, Haydee Ccaipani, Cesar Alvarez,
and Lorenzo Huisa, all of whom are thanked.
1. Editor's note by William H. Isbell: White plaster and a soft white stone found frequently at
Conchopata have been identified as diatomaceous earth or chalk. This material was probably
processed into gypsum plaster as well as white wash for walls.
2. Editor's note by Helaine Silverman: The material that Ochatoma and Cabrera report in this section
is internally consistent. However, their identification of "Nasca-like" is not necessarily acceptable
to a Nasca specialist.
REFERENCES
Cook, Anita G., 1987, The Middle Horizon ceramic offerings from Conchopata. Nawpa Pacha
22-23: 49-90.
Cook, Anita G., 1994, Wari y Tiwanaku: Entre el Estilo y la Imagen. Fondo Editorial, Pontifica
Universidad Cat61ica del Peru, Lima.
Isbell, William H., 1987, Conchopata: Ideological innovator in Middle Horizon IA. Nawpa Pacha
22-23: 91-126.
Isbell, William H. and Anita G. Cook, 1987, Ideological origins of an Andean conquest state.
Archaeology 40 (4): 27-33.
Chapter 9
A New Perspective on
Conchopata and the Andean
Middle Horizon
WILLIAM H. ISBELL AND ANITA G. COOK
INTRODUCTION
The Middle Horizon (AD 550-1000) was a time of many cultural changes, from
northern Chile to Cajamarca (Figure 9.1). Burial practices, household organiza-
tion, residence patterns, and/or ceramic styles were transformed in many places.
New polities replaced old ones. Expansive states or empires emerged for the first
time, unifying vast numbers of formerly independent cultural groups (Schaedel
1993). An emblematic new art became conspicuously popular. Its principal figure
was a mythical person represented front face with ray appendages about the head,
who grasped a staff in each outstretched hand (Front Face Staff God in the ter-
minology of Ochatoma and Cabrera, this volume); secondary human figures
kneel in profile with a staff held before the body. These and other icons appear to
represent the ideology of a powerful new religion (Cook 1987, 1994; Isbell 1983;
Lanning 1967; Moseley 1992).
This distinctive religious art had two great centers, each with a capital city
that appears to have been a primary political and ceremonial seat during the
Middle Horizon (Figure 9.1). Lake Titicaca and the Bolivian altiplano constituted
a southern heartland, with its metropolis at Tiwanaku [Note 1]. The Ayacucho
Valley was the northern heartland, with its great city at Huari [Note 2]. Each
heartland included smaller cities, probably one-time competitors of the first order
capital that were eventually subordinated by their more powerful neighbor.
Tiwanaku's best-known second city is Lukurmata, on the shore of Lake Titicaca
249
250 William H. Isbell and Anita G. Cook
San Miguel
River
AyacuchoValley
Huanta Basin.
:uc
U
,...,'. .J
85
85
54
' ........
'\..
l.
/~~..
56
"--"';: .._..-..,
I=II
Q.
>e.
S'
~
(")
Q
~
N
'£.
~
~
;:c
~
r:r
,, ,, '.:.. J.."
I: ,:- " -== .
'-Probable Pe~imeter Wall
~t
-=::-
I.0 ZONE B
'. ~ # •
:":':":-
.":.::.::.:1.:: I
::1' I
r-f"
IlprOb;b;
Perimete
Ii Wall
II
II
n Ii 10
LEGEND
CONCHOPATA ARCHAEOLOGICAL
ZONE
=::=::::::=::=::::::: Eat.'II',1IOtlTIttl(I'lU
Uncl.IitlfdAtdl,ftolOOleatW... ~
I:
+-IOm..
...... , .
ZONES
"/···
'.-
--.
, .t
'"""'~Clt·~n;;~
-_.1'.."==.~- .
;~';
CONCHOPATA ARCHAEOLOGICAL ZONE
Figure 9.5. Map of Conchopata showing locations of each oversize ceramic offering.
AD 1000
AD 850
AD 700
AD 550
Conchopata Chronology
Vallejos Phase
B
Alarcon Phase A
B
Huisa Phase A
B
Silva Phase A
Mendosa Phase
Huamanf Phase
mSTORY OF INVESTIGATIONS AT CONCHOPATA
Conchopata is famous for its spectacular over-size pottery decorated with
polychrome depictions of the Middle Horizon mythical beings in a style that
(along with "Robles Moqo" from the Pacheco site in the Nazca Valley on the
south coast) is more like the figures on Tiwanaku incised stone sculptures than
any other art north of Lake Titicaca. Even at Huari nothing quite so similar to
Tiwanaku has been found.
The archaeological site of Conchopata was first described in 1927 by
Benedicto Flores. The first excavation at Conchopata of giant ceramic vessels
with Tiwanaku iconography was conducted in 1942 by Julio C. Tello.
Unfortunately Tello did not publish his findings, so Menzel (1964, 1968, 1977)
interviewed students and colleagues who had accompanied Tello during his field-
work to ascertain the archaeology of the site. She concluded that the offering pot-
tery had consisted of large urns, some with a band of icons that included the
front-face deity with ray appendages about the head, painted on the upper third of
the vessels. These urns had been smashed, apparently by blows directed at the
religious images, implying deliberate breakage. The fragments had been found
in simple pits, suggesting some kind of buried offering. This was the evidence
for new religion in Ayacucho, focused on the front-face deity also known at
Tiwanaku.
-~~~
Type 1 •
~~--
Type 2
~~/f[:1&n
Type 3
Oversize Ceramic Offering Type 1 (OCO Type 1). This kind of offering con-
sists of numerous giant vessels, smashed and buried in a simple pit excavated into
the ground. Two examples have been excavated in recent years, the 1977 offering
and the 2000B offering, although offerings 2000A and 1999B probably belong as
well. Both 1977 and 2000B contained oversize face-neck jars, all decorated,
although 2000B included some other shapes as well. We suspect that the offer-
ings excavated by Tello belong to this type, although they contained only urns.
Perhaps OCO Type 1 urns and OCO Type 1 face-neck jars should be differenti-
ated into OCO Type lA and OCO Type 1B respectively, but since we have secure
information only for offerings of jars, we have shown only jars in our illustration
(Figure 9.6, Type 1).
Figure 9.7. (a) Urns; (b) Face-neck jars; (c) D-shaped icon on smaller set of vessels.
Oversize Ceramic Offering Type 2 (OCO Type 2). This type consists of
numerous oversize vessels, smashed and deposited on a surface within a sizable
architectural enclosure. Two examples have been reported, 1997-98 excavated
in a D-shaped building by Ochatoma and Cabrera (this volume), and 1999A,
excavated in a rectangular patio and adjoining room by Isbell and Cook. The
1999A offering consisted of urns only, many undecorated. However, the 1997-98
offering contained both urns and face-neck jars (and other forms), so we show
both urns and jars in our illustration (Figure 9.6, Type 2).
If the offering was originally deposited in a pit, these other ceramics may also
have been in the pit as well. However, they may have been mixed into the same
strata when the area was disturbed.
The iconography of 1999B is new and unique, consisting of 7 human faces,
probably male, each distinguished by its apparel and adornments. Each face is
shown in profile with protruding tongue, one behind the other as though in pro-
cession (Figures 9.14, 9.15). A radiocarbon date of AD 680 ::!: 60 was obtained
from the deepest stratum containing examples of 1999B offering pottery.
1999C-OCO Type 4. Many fragments of smashed jars, including some face-
neck examples, were found on the floor of room EA-36 (Figure 9.16). One face-
neck vessel was especially large (Figure 9.17), but none of these jars was
particularly fine, and none has any Tiwanakoid designs. Burned wood from just
above the floor in EA-36 radiocarbon dated AD 700 ::!: 60.
1999D-OCO Type 4. Room EA-31 also contained many smashed jar frag-
ments, although none were particularly fine, or decorated with Tiwanaku-related
icons.
2000A-OCO Type 1B (7). In the western edge of the EA-I00 extension were
many smashed oversize sherds. One set was clearly contained within a pit so
these sherds were attributed to offering 2000A. Most of the pottery is plain, and
little of it has been studied. The fragments appear to belong to jars, and the only
icon identified on the sherds is the warrior also discovered by Ochatoma and
Cabrera (this volume, their Figure 8.l1A). It seems possible that originally there
were several oversize ceramic offering pits in this area, but that most if not all
were disturbed, perhaps including 1999B. Alternatively, the entire area was a
spectacular pottery dump.
2000B-OCO Type lB. This offering of oversize vessels is important for it
demonstrates the variability in Conchopata's offering. This offering was placed in
a pit, and includes several different shapes, some only regular size. There are
some 73 regular size jars that are plain or decorated only on the rim. A few were
unbroken. All the oversize vessels, about 27, come from face-neck jars, and all
were decorated (Figures 9.18, 9.19). However, these oversize face-neck jars are
not as large, as thick or as well finished as the 1977 offering jars. Furthermore,
although they are elaborately decorated, they have few colors and no Tiwanakoid
icons. All their decoration is in the "Chakipampa" style, native to the Ayacucho
Valley (Menzel 1964). The face-neck vessels were concentrated toward the bot-
tom of the 2000B offering pit, so it appears that the ceramics were deposited into
the pit in layers [Note 4]. A sample of charcoal from within the 2000B offering
pit yielded a radiocarbon date of AD 680 ± 40.
2000C-OCO Type 3. This find consisted of fragments from two incomplete
oversize urns, one elaborately decorated (Figure 9.20) and the other plain.
Numerous fragments of the two were found on the floor of room EA-78, as
though they were broken on the spot. A few more pieces were found on the floor
of adjoining room EA-79, and several pieces were recovered in the construction
trench of the modem house cutting the south end of EA-78. Other fragments may
have been destroyed by this building. About 50% of the decorated urn was easily
reassembled.
2000D-OCO Type 3. This collection may represent disturbed ceramic offer-
ings reused as fill. However, several partial urns of high quality were discovered,
including one that was nearly half complete. This is rare except in OCO Type 3
offerings, so we have assigned the find to that class. The oversize pottery came
from under a red clay floor in patio room EA-6. The most complete urn represents
the warriors, grasping weapons and shields, that kneel in reed boats (Figure 9.21).
Examples of these icons were first discovered by Ochatoma and Cabrera (this
volume, their Figure 8.5) in the D-shaped building, EA-72. We obtained a radio-
carbon date of AD 570 ± 40 from the stratum of EA-6 that contained the over-
size pottery. Several other fragments of oversize urns from below the floor in
EA-6 had Tiwanakoid designs, but more consistent with Robles Moqo than the
Conchopata style.
2001A-OCO Type 3. One smashed urn, virtually complete and decorated
with disembodied mythical profile heads (Figure 9.22), was found on the floor of
the large patio EA-98. The smallish oversize urn had been placed near a doorway,
in excavation EA-98D, on a dais of stones, smashed to pieces, and left as it fell,
except for one or two sherds that somehow disappeared. This is the most nearly
complete urn in Conchopata style ever found. It can be associated with a radio-
carbon date from a burned roof beam, lying on the same floor at the south edge
271
of the patio, in excavation EA-98(A). This date is AD 780 ± 60, but this should
date the cutting of the beam for the roof, perhaps a century or so before the patio
was abandoned with the smashed urn in one of its doorways.
2001B-OCO Type 3. Fragments of several oversize urns were found under a
possible floor or surface in EA-168. They include urn fragments similar to the
1999B, OCO Type lA offering, and several Robles Moqo pieces most like frag-
ments from the 2000D, OCO Type 3 offering. Although a few of these fragments
fit together, it was not possible to mend significant portions of any urns, so this
context may represent trash rather than offering activity.
Analyses of ceramic collections from Conchopata have only begun so many
conclusions are still far in the future. However, it seems likely that the ceramic
urns and jars in at least OCO Types 1 and 2 are incomplete. That is to say, all the
fragments of each vessel were not placed into the pit in OCO Type 1 offerings,
and all of the fragments of each vessel were not left on the floor of a large room
or courtyard in the case of OCO Type 2 offerings. We conclude that OCO Type 1
and OCO Type 2 probably were not broken on location and then simply covered
up (OCO Type 1) or abandoned (OCO Type 2). Deposition processes were more
complex and prolonged.
A second observation that can be made at this time is that sub-styles of
Conchopata oversize offering vessels can be identified on the basis of vessel shape
and decoration themes. Pottery of the 1942 offering may represent the same style
found in the 1999A offering. Interestingly, 1999A consists of pottery scattered
across the surface of a patio and room, while 1942 came from one or more pits.
Perhaps these two offering deposits represent part of the same offering ritual.
273
Within the 1997-98 OCO Type 2 offering is a warrior icon (Ochatoma and
Cabrera, this volume, their Figure 8.llA) who also appears in the 2000A OCO
Type lB offering. Again the same style of oversize pottery occurs on a floor and
in a pit. Other offering styles appear in more than one context too, but not in OCO
Type 1 and OCO Type 2 offerings. Warriors kneeling in reed boats occur in the
1997-98 OCO Type 2 offering, and in the 2000D OCO Type 3 offering. Profile
human faces appear in the 1999B OCO Type lA (?) offering, and in the 2001B
OCO Type 3 offering. Other fragments of offering styles appear here and there,
although some seem to be unique. To date, the style that characterizes the 1977
OCO Type lB offering has only been found within a few meters of the pit.
Similarly, pottery of the 2000B OCO Type lB offering is not known from other
contexts, except for a couple of face-neck jar fragments in the 1977 offering.
~...•.
~ .••.
275
276 William H. Isbell and Anita G. Cook
Two radiocarbon dates for offerings 1999B and 20ooB, located only a few
meters from one another, are virtually identical: AD 680 ± 60, AD 680 ± 40.
Neither one contains Tiwanaku characteristics, although the markedly Tiwanaku
1977 offering was found only a few meters to the south, and a couple of 2000B
face-neck jar sherds were also found in this pit. In terms of structure, the 1977
iconography is probably most like the Tiwanaku sculpture on the Gate of the Sun.
All this suggests a new and different iconographic scenario. Perhaps the first
Conchopata offering pottery appeared in the Silva A Phase, between AD 550 and
600. It may have depicted local themes, especially warriors displaying fancy
clothing, a weapon, and a shield. Chakipampa icons probably continued through
Huisa times, when the first Tiwanaku themes appeared on face-neck jars. But
Tiwanaku icons that were much more accurate in details such as divided eyes,
headdresses, and other details of the kind that were painted on the 1942, 1999A,
2000D and 2001A offering vessels may not have appeared at Conchopata until
the Alarcon Phase, after AD 850.
Of course, there are problems with this alternative chronology of offering
styles. Warriors in reed boats associated with the date AD 570 ± 40 are also
associated with Tiwanaku-related Robles Moqo pottery in possible offering con-
text 2000D. And both Tiwanaku and non-Tiwanaku styles appear in the 1997-98
offering from the D-shaped building. Perhaps the explanation is that oversize
offering pottery was kept for decades and even centuries before it was broken in
Type 4
profile
~."
Type 6
[T1ype33-
.@
Type 7
Burial Room
CONCLUSIONS
Although we have just begun analyses, already we have learned much from
the new excavations at Conchopata. It is already clear that many ideas about the
Middle Horizon that were considered firmly established must now be rethought
and reformulated. The magnitude of the revisions can only begin to be appreci-
ated. This discussion is limited to preliminary observations about oversize
ceramic offerings, pottery production, mortuary patterns, radiocarbon dates, and
architecture at Conchopata.
The most obvious and yet most problematic change in thinking about the
Middle Horizon regards the dating and chronology of the Conchopata offering
style, with its Tiwanaku art and iconography. This is essential for reevaluating the
spark that is believed to have initiated state expansionism and empire in the
Peruvian Central Andes. Is the Tiwanaku style really much later at Conchopata
than originally thought, belonging not to MH lA, but to a significantly later time?
If so, a new Tiwanaku religion could not have been the stimulus for the cultural
changes that inaugurated the Middle Horizon. In fact, Tiwanaku influence or
interaction must have occurred relatively late in the Middle Horizon, unless the
style was early and remained unchanged for centuries. Another possibility is that
there were other, earlier sources from which the distinctive iconography reached
Ayacucho. In this case, new iconography of the Middle Horizon may have
reached Tiwanaku and Conchopata/Huari at about the same time, each center
adopting slightly different expressions of the art from the beginning. In this case,
the greatest similarity between Tiwanaku and Conchopata iconography could
have been later in the Middle Horizon.
Whatever we learn about the sources of Conchopata/Huari iconography-
specifically, the front-face deity who grasps two staffs and the associated profile
figures who kneel and carry a single staff-ehronology within the Middle
Acknowledgments
Authors William H. Isbell and Anita G. Cook wish to thank the sponsors of
the Conchopata Archaeological Project, which they co-direct with Jose Ochatoma
Notes
I. Tiwanaku is the currently popular spelling for the name of the site and culture. Earlier literature
used Tihuanacu or Tiahuanaco.
2. Huari is also spelled Wari.
3. We have chosen not to calibrate our C 14 dates for this discussion. Most of the radiocarbon assays
on which the absolute chronology of the Middle Horizon was constructed were made before radio-
carbon dates were calibrated. Consequently, the current Middle Horizon chronology is in radio-
carbon years. Until we calibrate and reevaluate all the Middle Horizon dates it seems wisest to
employ and critique the chronology as it has been formulated and used, in radiocarbon years.
4. Layers of pottery in a ceramic offering pit is reminiscent of the Epoch 2A ceramic offering exca-
vated at Ayapata by Rogger Ravines (1968, 1977).
5. Mortuary behavior at Conchopata was almost certainly part of a larger domain of Huari burial
practices during the Middle Horizon. A definitive Huari mortuary discussion must consider infor-
mation from the Huari capital and other Huari settlements. However, such a comprehensive study
cannot be undertaken until information from many individual sites is available for comparison.
REFERENCES
Benavides c., Mario, 1965, Estudio de la certimica decorada de Qonchopata. Tesis para optar el
grado de Bachiller en Ciencias Sociales. Universidad Nacional San Crist6bal de Huamanga,
Ayacucho.
Benavides C., Mario, 1984, Carticter del Estado War£. Universidad Nacional Mayor San Crist6bal de
Huamanga, Ayacucho.
Benavides C., Mario, 1991, Cheqo Wasi, Huari. In Huari Administrative Structure. Prehistoric
Monumental Architecture and State Government, edited by William H. Isbell and Gordon F.
McEwan, pp. 55-69. Dumbarton Oaks Research Library and Collection, Washington, D.C.
Bennett, Wendell C., 1936, Excavations in Bolivia. Anthropological Papers ofthe American Museum
of Natural History 35 (4): 329-507.
Bennett, Wendell c., 1946, The archaeology of the Central Andes. In Handbook ofSouth American
Indians, Volume 2, The Andean Civilizations, edited by Julian H. Steward, pp. 61-147. Bulletin
143. Bureau of American Ethnology, Smithsonian Institution, Washington, D.C.
Bennett, Wendell C., 1953, Excavations at Wari, Ayacucho, Peru. Yale University Publications in
Anthropology, Number 49. New Haven.
Bermann, Marc, 1993, Continuity and change in household life at Lukurmata. In Domestic
Architecture, Ethnicity, and Complementarity in the South Central Andes, edited by Mark S.
Aldenderfer, pp. 114-135. University of Iowa Press, Iowa City.
Bermann, Marc, 1994, Lukurmata: Household Archaeology in Prehispanic Peru. Princeton University
Press, Princeton.
Cook, Anita G., 1979, The Iconography of Empire: Symbolic Communication in Seventh-Century
Peru. M.A. thesis. State University of New York at Binghamton.
Cook, Anita G., 1983, Aspects of state ideology in Huari and Tiwanaku iconography: the Central
Deity and the Sacrificer. In Investigations of the Andean Past, edited by Daniel H. Sandweiss,
pp. 161-185. Latin American Studies Program, Cornell University, Ithaca.
Cook, Anita G., 1986, Art and Time in the Evolution ofAndean State Expansionism. Ph.D. disserta-
tion. Department of Anthropology, State University of New York at Binghamton.
Cook, Anita G., 1987, The Middle Horizon ceramic offerings from Conchopata. Nawpa Pacha 22-23:
49-90.
Cook, Anita G., 1992, The stone ancestors: idioms of imperial attire and rank among Huari figurines.
Latin American Antiquity 3 (4): 341-364.
Cook, Anita G., 1994, Wari y Tiwanaku: Entre el Estilo y la Imagen. Fondo Editorial, Pontificia
Universidad Cat61ica del Peru, Lima.
Cook, Anita G., 1997, The emperor's new clothes: symbols of royalty, hierarchy and identity. Journal
ofthe Steward Anthropological Society 24 (I & 2): 85-120.
Cook, Anita G., 2oo1a, Huari D-shaped structures, sacrificial offerings, and divine rulership. In Ritual
Sacrifice in Ancient America, edited by Elizabeth P. Benson and Anita G. Cook, pp. 137-163.
University of Texas Press, Austin.
Cook, Anita G., 200lb, Los nobles ancestros de piedra: ellenguaje e 1a vestimenta y rango imperial
entre las figurillas Huaris. In Wari: Arte Precolombino Peruano, edited by L. Millones,
M. Cabrera Romero, E. Gonzalez Carre, W.H. Isbell, F. Meddens, C. Mesfa Montenegro,
Chapter 10
The Correlation Between
Geoglyphs and Subterranean
Water Resources in the Rio
Grande de Nazca Drainage
DAVID W. JOHNSON, DONALD A. PROULX, AND
STEPHEN B. MABEE
INTRODUCTION
The giant markings etched onto the Pampa de San Jose near Nazca are one of the
most enduring archaeological mysteries of ancient Peru. These geoglyphs consist
of geometric forms (triangles, trapezoids, quadrangles, single and parallel lines)
and biomorphic figures (birds, plants, animals, sea creatures). In addition, other
geoglyphs are located throughout the drainage. The majority of geoglyphs can be
attributed to the Nasca culture (ca. AD 1-700) on stylistic grounds (the biomor-
phic figures are iconographically similar to images painted on Nasca pottery), and
by the presence of broken fragments of Nasca pottery scattered on the surface of
the geoglyphs (see, e.g., Silverman 1990). However, Clarkson (1990) has argued
convincingly that some of the lineal geoglyphs date to the Middle Horizon
(AD 600-900) and Late Intermediate Period (AD 900-1476). This suggests that
geoglyph-making was a long, local tradition on the south coast of Peru.
First discovered in 1927 by Toribio Mejia Xesspe (1940), who thought they
were ancient roads, the geoglyphs received only sporadic attention until the 1940s
when American geographer and historian Paul Kosok (1965) visited the Pampa
de San Jose and happened to observe the sun setting over the end of one line on
307
Udluymarca
...
SaodacoiCbuquimarin
Figure 10.1. Regional map of the Rio Grande de Nazca drainage showing the major tributaries, modern
cities, and the location of important archaeological
sites. Sites described in this paper are labeled within a box (Modified from Silverman 1993).
Study Areas
The association among geoglyphs, subterranean water, and archaeological
sites can be demonstrated by examining the results obtained from four sites.
These sites are Cerro Aja, Orcona, Usaca and Cerro Colorado (Figure 10.1).
Cerro Aja is a mountainous ridge separating the Aja Valley from the
Socos Valley to the north and is located immediately north of the city of Nazca
(Figure 10.2). The archaeological site of Aja is situated on the southern slopes of
the hills, facing the city. It is primarily a Nasca culture site. A geoglyph system
characteristic of the upper Grande drainage is located across the hills to the north
and there are four filtration galleries paralleling the river between the Aja River
and the hills.
Orcona is located 5 kIn east of Nazca along the north side of the Aja River
directly across from Orcona filtration gallery (Figure 10.3). Orcona consists of a
north-northwest trending alluvial fill valley that enters the Aja River from the
north. There are numerous geoglyphs, a cemetery, and terraced hillsides here.
Figure 10.2. Aerial photograph of Cerro Aja showing the location of the Aja Fault, geoglyphs, habi-
tation sites, cemeteries and filtration galleries (heavy solid lines). The small arrows with the numbers
indicate the locations of the electromagnetic induction surveys conducted across the Aja Fault. Results
are shown in Figure 10.8.
Usaca is the name of a small settlement found at the head of a quebrada with
the same name (Figure 10.4). It is situated at the confluence of the Trancas and
Chauchilla rivers at a location where a natural spring seeps water into a pool. More
than thirteen archaeological sites are situated along the Quebrada Usaca, including
a huge multi-occupational cemetery covering almost 2.5 sq lan. There is surface
water only where faults cross the river valleys. There is a geoglyph system here.
314
170 >-./
. ..-"': '0 •
{
C6~
l~ -'IV •
_iii _,.- •
'-~-~ .
o "" ~. --.._~.......
... 000
28'1
Figure 10.3. Map of Orcona showing the location of the faults. spring. geoglyphs, cemeteries and
filtration gallery. Solid triangles are pointers, thin lines are linear geoglyphs (lines). Double headed
line southeast of the spring is the continuation of the Orcona Lineament across the peninsula into the
Tierras Blancas River valley.
Figure 10.4. Aerial photograph of Usaca showing the location of the Atarco and Nazca faults, water
sources, geoglyphs, and habitation sites and cemeteries. The trapezoid marking the location of the
Atarco Fault is actually located further east off the figure. The line soulh of the Alarco Fault is a geo
glyph marking the limit of the Atarco Fault/aquifer system. Ovals are cochas, the circle is a spring,
and the hexagon is the geoglyph system marking the position of the Nazca Fault. Habitation sites and
cemeteries are indicated by the numbered boxes.
that more than one archaeological culture contributed to the construction of the
geoglyphs (see Figure 10.6).
Methodology
The hypothesis being tested argues that some of the geoglyphs mark the path
of subterranean aquifers that carry water through geological faults and structures.
In order to test this hypothesis five questions must be addressed: (1) do the faults
exist; (2) do the faults intersect the valleys; (3) is there evidence that the faults
transmit water; (4) are the faults marked by geoglyphs; and, (5) are there habita
tion sites associated with the faults? These questions have been answered by
employing standard geological and hydrogeological techniques in conjunction
with customary archaeological methods. A brief outline of the methods is
described below. Fieldwork was conducted in July 1998 and September 1999.
Figure 10.5. Aerial photograph of Cerro Colorado showing the location of the regional
flexure (fold) and springs. Springs are indicated by circles.
Geologic Mapping
Faults are discontinuities in the bedrock that arise from tectonic forces caus
ing the rocks in the earth's crust to shear and move, displacing one side of the
fault relative to the other. Once the rocks have been broken, the fault plane
becomes a permanent zone of weakness that often enhances the permeability of
the rock by providing a natural conduit through which water can move.
Depending on the type of rock involved and the nature of the fluids moving along
the fault plane, some faults can become mineralized and act as a barrier to flow
forcing groundwater to seep out at the ground surface whereas other faults remain
open allowing water to flow freely along the fault plane. Faults can be a few
millimeters to several kilometers in width and range from a single fault plane to
a wide zone of highly fractured rock that contains an anastomosing, interwoven
network of permeable, interconnected fractures.
There are several clues that provide evidence for faulting in the field. These
include offset between rock types on opposite sides of the fault plane, folding of
the host rock due to drag along the fault, brecciated rock within the fault zone and
,,
\, ,
~'
, L'-- _
Figure 10.6. Detail of aerial photograph showing the configuration of the geoglyph system and
habitation site/cemetery at Cerro Colorado.
striated fault plane surfaces, highly fractured rock adjacent to the fault, and exten-
sive mineralization suggesting past or recent fluid movement.
For this study, the geology of each of the four sites was examined three
ways. First, all published geologic maps were gathered for the region and stud-
ied. Second, satellite images and aerial photographs were acquired to look for
evidence of faulting. Because faults represent weaknesses in the rock, they are
more prone to weathering. As such, they can be more easily eroded than the
surrounding country rock and, therefore, often appear as topographic depressions
or linear scars across the landscape. These features can often be detected on
topographic maps, aerial photographs, and satellite imagery particularly in arid
to semi-arid environments where vegetative cover is sparse. These features are
referred to as lineaments and very often represent faults. As a final step, traverses
were made at each site by geologists to verify the published geologic maps and
to confirm faults observed on the satellite images and aerial photographs. At each
site, the location of faults was mapped and characterized.
Archaeological Analysis
Over the course of four years, a systematic survey of archaeological sites has
been conducted in the vicinity of the city of Nazca and from Usaca on the Trancas
River, down the Nazca River to its confluence with the Grande River, and then
Geophysical Analysis
In order to verify that the faults observed in bedrock exposures and on satellite
imagery intersect the major river drainages and lie beneath the surficial materials
deposited within the valley, geophysical tools were employed, where feasible, to
locate the faults in the subsurface. In addition to locating faults, geophysical tools
were also utilized to determine the depth to bedrock below the unconsolidated sur-
ficial deposits and the depth to the water table. Several geophysical methods have
been successfully used over the last two years of fieldwork. These include: (1) seis-
mic refraction to locate the depth to bedrock beneath the surficial soil material and
to determine the depth to the water table and, (2) electromagnetic induction (EM-34)
to map the position of buried faults. Future work will include additional geophysical
methods including seismic reflection and ground penetrating radar. Both of these
techniques provide high resolution images of the subsurface stratigraphy including
the water table, depth to bedrock and bedrock discontinuities (faults).
Surveying
A field map was made at each site using a total station surveying system. The
maps include the location of any faults identified by the geologist and by the geo-
physical surveys, the position of any affiliated geoglyphs, the position of archaeo-
logical sites, and the location of wells, springs, or seeps. These maps provide the
basis for establishing any correlations between the geoglyphs and groundwater
resources. The total station was also used to su;:vey the water elevations of wells.
Water level data provided information on the direction of groundwater flow in the
valley and, at some of the sites, was used to estimate the source of the groundwater.
RESULTS
Preliminary results concerning the correlation between the geoglyphs and
water resources are described below for each of the four sites. Although data are
still being gathered and analyzed, evidence exists supporting the hypothesis. The
results that follow are characteristic of the evidence obtained at other sites
throughout the Rio Grande de Nazca drainage including La Mufia in the Palpa
Valley in the north, around Hacienda Taruga in the south, and Monte Grande in
the lower Grande Valley to the west.
CerroAja
Three faults were identified at Cerro Aja, one of which can be traced on the
ground for three kilometers into the Socos Valley to the north (Figure 10.2). The
largest fault, herein named the Aja Fault, is clearly marked by geoglyphs. Field
evidence for the faults included the presence of fault gouge, fault breccia, recrys-
tallized and strained rock along the fault plane, slickensides, and drag folds
(Figure 10.7). There is also extensive mineralization indicating fluid movement
along the faults in the past. Five geophysical transects were made across the
Aja Fault using the EM-34. Results show distinct subsurface anomalies and
were observed exactly where geologic mapping predicted the fault should be
(Figure 10.8). Accordingly, there is convincing field evidence and geophysical
data suggesting that the Aja Fault exists and can be traced into the valley.
In addition, groundwater levels from nineteen wells and three puquios were
surveyed at Cerro Aja near the intersection of the Aja Fault with the valley. Results
of the water level data collected in September 1999 indicate that groundwater flow
Figure 10.7. The fault at Cerro Aja. Note the white area of brecciated rock along the fault plane just
below the field notebook. The fault zone is approximately 2 to 3m wide. Field book is 20cm high.
... .~"".
Transect 1
~.
._.~.. "
Transect 4
Transect 2
.d.\--Vr-.-'
Transect 3
~~
\-./,.
TransectS
Observed Data
-- Modeled Data
Figure 10.8. Results of electromagnetic induction surveys (EM-34) across the Aja Fault. Note the
anomalies (downward depression in the observed data) indicating the presence of a discontinuity in
the subsurface.
is moving from the valley side walls toward the southwest. In other words, a por-
tion of the groundwater flow is entering the valley from the sides and, as it
approaches the river, the flow turns to the west and becomes part of the ground-
water flow regime beneath the river (Figure 10.9). The water elevations in wells
Figure 10.9. Aerial photograph mapping the groundwater flow at Cerro Aja. Small numbers in bold
are well locations. Large numbers are groundwater contour labels indicating the elevation of the
groundwater in meters above mean sea level. The bold lines with arrows indicate the direction of
groundwater flow with respect to the Aja Fault. Note the general direction of groundwater flow is
northeast to southwest indicating a contribution of water to the river from the valley sides.
closest to the valley side walls are at least 10m higher than the water levels in the
filtration gallery closest to the Aja River. This groundwater configuration signi-
fies a "gaining" stream scenario where groundwater from the sides of the valley
is contributing water to the main trunk of the stream. Similar results are obtained
when water level data from July 1998 are plotted; the pattern is consistent over
two successive years. These results suggest that some of the groundwater in the
river valleys is derived from bedrock.
Although all of the analyses performed on the water samples collected during
the September 1999 field trip have not been completed, we can report on the results
of the field parameters (temperature, specific conductivity and pH). Water from
wells located closest to the valley side walls are slightly warmer (26.8 degrees vs.
24.2 degrees Centigrade), exhibit higher conductivity (106 vs. 75 /-LS/cm), and have
higher pH (7.1 vs. 6.6) than the groundwater in wells adjacent to or in the bed of the
river. Although preliminary, these data suggest water from two different pathways.
Geoglyphs and archaeological sites are associated with the Aja Fault. In
1926 Julio C. Tello and Alfred L. Kroeber discovered several sites in the Aja area
(Kroeber and Collier 1998: 33, 66-fJ7). All of these were cemeteries, the largest
(Aja B) dating to Nasca Phase 3. A large Nasca and Late Intermediate Period
Figure 10.10. Aerial photograph showing the position of the large trapezoid that marks the location
of the Aja Fault. A pointer (not visible on photograph) points to the ridge in the background exactly
where the Aja Fault crosses into the Socos Valley (location marked by solid line). The geologists at
the base of the trapezoid are conducting an EM survey across the fault. The results of this survey are
shown in Transect I on Figure 10.8.
habitation site (not identifiable from Kroeber's notes and perhaps not recorded by
him in 1926) extends along the south slope of Cerro Aja (Figure 10.2), and there
is a broad agricultural zone between the river and the hills. Along the trace of the
Aja Fault there is a trapezoid with a narrow triangle extending from the trapezoid
to a line center (Figure 10.10). The trapezoid maps the trend of the fault contain-
ing the aquifer, the triangle points to the next ridge where the fault crosses into
the Socos Valley and the line center, located on a small hill, provides a vantage
point for observing the geoglyph system. Where the trace of the fault turns east-
ward into the Nazca Valley, the bend is marked by a curved trapezoid and trian-
gle with a large stone circle (Figure 10.2). Large circles have been found
throughout the drainage where aquifers tum into the valley as in this case.
Orcona
Orcona consists of a north-northwest trending alluvial fill valley that
appears on aerial photographs as a major topographic lineament extending from
the Socos Valley, across the Aja and Tierras Blancas river valleys into Cerro
Blanco (the lineament follows the valley marked Fault 3 on Figure 10.3). This
feature parallels a major thrust fault that has been mapped several kilometers to
Usaca
Two large faults cross at Usaca that are not indicated on the geological map
of the quadrangle. However, they are clearly evident on the aerial photographs
(Figure lOA) and satellite imagery and also are visible on the ground. For the pur-
poses of this chapter, they will be referred to as the Atarco and Nazca faults. The
Atarco Fault is a large unmapped fault system that crosses normal to the Trancas
River in an east-west direction. The fault enters the Quebrada Atarco (east of
Usaca) and follows it for two to three kilometers before crossing the ridge sepa-
rating the Chauchilla River and Quebrada Atarco. The fault can be seen in out-
crops on both sides of the Quebrada Atarco. It exhibits 1.5 m or more of apparent
offset in fossiliferous sandstone beds exposed in the valley walls. There is also
evidence of fault breccia within the fault plane. The Atarco fault trends 265
degrees and has a near vertical dip. Further up the Atarco Valley are fossiliferous
chalk beds that also show the same apparent offset of approximately 1.5 m.
CONCLUSION
Each of the sites we have examined exhibits all the elements predicted by the
hypothesis. Major faults and structures, many of which extend for several kilo-
meters, can be observed entering the valleys. The faults are not only observed on
the ground but, in some cases, such as Cerro Aja, their existence is supported by
geophysical evidence. Where the faults enter the valleys, groundwater can be
observed discharging directly from the faults or, in the case of Cerro Aja, ground-
water flow directions determined from water level data indicate a flow gradient
from the valley side walls towards the river. All of these observations clearly indi-
cate that a portion of the water in the river valleys has its source from faults in the
Acknowledgements
Our field research in Peru was supported in part by a Research Grant (6229-
98) from the National Geographic Society, by the H. John Heinz III Fund Grant
Program for Latin American Archaeology, by a Healey Endowment Grant (1-
60246) from the University of Massachusetts, and by Rotary International. We
wish to thank each of these sponsors for their funding of our work and their faith
in our projects.
Notes
I. Springs are places where water naturally discharges from the bedrock or sediments that are not
necessarily part of a manmade underground structure.
Chapter 11
Rock Art, Historical Memory,
and Ethnic Boundaries
A Study from the Northern Andean Highlands
TAMARA L. BRAY
As a specific and widespread fonn of cultural production, rock art has been studied
from a variety of perspectives including evolutionary (Biesele 1983; Kirkland and
Newcomb 1967), functional-adaptive (Ellis 1975; Heizer and Baumhoff 1959;
Olson 1977; Thomas 1976), fonnal art historical (Castleton and Madsen 1981;
Grant 1967), and semiotic (Berenguer and Martinez 1989; Lewis-Williams 1980,
1982, 1983, 1995; Llamazares 1989). In this chapter, I adopt the latter framework
to examine the performative and connotative aspects of a set of late prehistoric
period petroglyphs from the Pimampiro district of northern highland Ecuador.
Given the importance of context to the interpretation of meaning, I first situate
these petroglyphs within a specific cultural, geographic, and historical frame-
work. I then analyze the imagery found on the stones with respect to literal
content, internal coherence, external linkages, and symbolic associations. Relying
on ethnohistoric and archaeological data as well as local myth to establish a
referential context of social action (after Hodder 1982), I suggest that these
petroglyphs articulate aspects of territoriality, boundary maintenance, and ethnic
identity in this region of the northern Andes.
334
t
Tamara L. Bray
.....
i.· .....
i
Pimillmpiro '._
lborr.:'if ('
~),.~
W:J
Figure 11.1. Map of the Chota-Mira river valley delineating the approximate boundaries of the
Pimampiro district.
gateway to the Oriente (the term commonly used to refer to the eastern flanks and
foothills of the equatorial Andes) and an important multi-ethnic trade center
(Borja 1965; Ordonez de Cevallos 1960). Other early sources indicate that ties
between the northern Ecuadorian highlands and the Oriente comprised a com-
plexity of political, commercial, and ideological elements. Oberem (1974: 347),
for instance, citing a 16th century document, notes various instances of interzonal
marriage and comments on the political implications of such practice. Borja
(1965) discusses the various modes of tribute and exchange in this region,
highlighting the commercial aspects of the regional economic system. Ideological
linkages between the two zones are underscored in references to the use of
tropical forest paraphernalia as insignia of status (Caillavet 1983: 17) and the
Escala 1,50,000
CurvlS de Nivel con Intervalos de 100 metros
general respect accorded lowland healers and their medicinal herbs (Oberem
1974: 351).
The Pimampiro district lies at the extreme eastern end of the semi-arid
Chota-Mira river valley. This river is one of the few in Ecuador that breach the
western cordillera to empty into the sea. The warm, dry climate of this low-lying
DESCRIPTION OF PETROGLVPHS
In 1996, two large petroglyphs displaying carved images of tropical lowland
fauna (Figures 11.3, 11.4) were discovered above the modem village of
Shanshipampa in the southernmost sector of the Pimampiro district. Found in the
same general area as the stones are numerous small hemispherical mounds
(tolas), terraces of varying widths, and long linear mounds. These features occupy
an area of nearly one square kilometer above the modem community of
Shanshipampa (Figure 11.2). On the basis of the archaeological remains, clues
from the ethnohistoric record, the site's geographical location, and local infor-
mant input, this locality has tentatively been identified as the late prehistoric
period site of Chapi (Bray n.d.a.).
Building upon previous survey work conducted in the Pimampiro district
(Bray 1994, 1995a), a program of mapping and archaeological excavation was
initiated at the site of Shanshipampa by the author in 1997. During this season,
six additional petroglyphs were documented in the vicinity of Shanshipampa
though these do not manifest as complex of iconography as registered on the first
two examples (Bray 1998). Over the course of the next three years, a variety of
surface features, including tolas, burial caves, terraces, and linear mounds, were
mapped and tested by the Pimampiro Archaeological Project.
Figure 11.3. A carved stone from Shanshipampa, still in situ, which displays various curly-tailed
quadrupeds.
Figure 11.4. Line drawings of imagery found on the two largest petroglyphs from Shanshipampa.
The largest of the carved stone monuments (A: approximately 180 X 100 cm as drawn; B: approxi-
mately 150 X 75 cm as drawn) exhibits a relatively complex iconographic program. Note the presence
of the mammalian curly-tailed quadruped figure on both stones, interpreted as a simian-form, and the
pair of bicephalic creatures that occupies the center position on the larger stone.
Rock Art, Historical Memory, and Ethnic Boundaries 339
Preliminary analyses suggest that these features are variously associated
with domestic, agricultural, mortuary, and possibly metal-working activities.
Radiocarbon dates obtained from several different contexts suggest a period of
occupation for the site dating from approximately AD 900 (cal) to the early
Colonial period (Bray n.d.a., n.d.b.). The petroglyphs distributed about the site
are generally associated with these various archaeological features and earth-
works. The eight stones recorded display a range of motifs. On most, the imagery
consists of simple geometric forms including spirals (Figure 11.5), lines, and
circles. As noted above, however, the two largest monuments carry more complex
iconography (Figures 11.3, 11.4).
The compositions on the two larger stones are dominated by two principal
motifs (Figure 1104). The first figure, which is common to both monuments, is a
mammal. I interpret this figure as a simian-form based on its physical features
(prehensile tail, flexed extremities), comparisons with other zoomorphic repre-
sentations in the Carchi-Nariiio region of northern Ecuador and southern
Colombia (Duncan 1992; Echeverria 1988; Labbe 1986: 132-149; Plazas and
Echeverri 1995; Rodriguez 1992: 7-81), and ethnographic information
(Rappaport 1992: 210). Even though the figure has a somewhat beak-shaped face
on the larger stone (Figure llAa, around perimeter), there are depictions of
monkeys with extremely prognathic jaws approximating this same beak-shaped
Figure 11.5. Example of a petroglyph in the vicinity of the Shanshipampa site exhibiting a simpler
spiral motif.
Figure 11.6. Flagstone pavement at the site of La Mesa as it appeared when exposed by huaqueros
in the early 19705.
REGIONAL ICONOGRAPHY
The animal world is a natural source of visual metaphors for symbolizing
human relations and cultural knowledge (Helms 1995; Linares 1977; Morphy 1989;
Urton 1985; Willis 1990). The precolumbian art of the Carchi-Narifio region is par-
ticularly rich in representations of fauna. The zoomorphic forms in the iconography
of this region are generally portrayed quite naturalistically and are often identifi-
able, sometimes to the level of genus and species (Rodriguez 1992). A review of the
precolumbian imagery from the Carchi-Narifio region as evidenced in various
media, including pottery, rock art, and goldwork, indicates that stylized simians
were an extremely important component of the regional iconography.
In the ceramic medium, monkeys constituted a common decorative element of
Piartal materials (Figure 11.8). Piartal pottery, which combines a resist technique
with a red over-paint to produce a distinctive tri-chrome ware, dates from approxi-
mately AD 750 to AD 1250 and is associated with the proto-Pasto inhabitants of
Figure 11.7. Examples of the flagstones sculpted in bas-relief that display the bicephalic creature
and the simian-like quadruped (photos courtesy of Dr. Albert Meyers).
the Carchi-Narifio region (Labbe 1986: 132-149; Uribe 1977-78, 1992, 1995).
While less frequently portrayed in the subsequent Tuza stylistic phase
(AD 1250-1500) that is associated with the proto-historic Pasto, monkeys are still
present in the regional iconography (e.g. Labbe 1986: 168; Valdez and Veintimilla
1992: 180). Besides their appearance in Piartal and Tuza pottery, simian-forms
are also frequently portrayed in the goldwork from both of these periods
(see Plazas and Echeverri 1995: figs. 2,3; Valdez and Veintimilla 1992).
Figure 11.8. Drawings of the interior designs found on three Piartal bowls with annular ring bases
depicting images of monkeys. Stippling indicates areas of red over-paint on resist-created negative
space; white zones indicate areas of negative resist-created designs; black indicates designs created
through reduction during firing. Top: El Angel, Saville Collection, National Museum of the American
Indian, Cat. no. 3/280, diameter: 20cm. Left: EI Angel, Saville Collection, National Museum of the
American Indian, Cat. no. 3/282, diameter: 18.5 cm. Right: Provenience unknown, Museum of the
Central Bank of Ecuador, Cuenca, Cat. no. 22-1-78d, diameter: 18 cm.
Monkeys are also a common decorative element of Capuli style pottery from
the region, where they are rendered as small modelled figures on the rims of ves-
sels (e.g. Labbe 1986: 146). The Capuli style is found throughout the same sector
of the Carchi-Narifio zone and pertains to approximately the same time period as
the Piartal and Tuza phases, dating from ca. AD 800 to AD 1500 (Duncan 1992;
Uribe 1977-78, 1992). The most pronounced similarities between Capuli motifs
and the imagery of the Shanshipampa petroglyphs is found in the goldwork attrib-
uted to this phase (e.g. Plazas and Echeverri 1995: fig.7; Rodriguez 1992: 75-78;
Valdez and Veintimilla 1992: 184-187). Comparing this imagery across various
media, it is clear that there are marked similarities between the simian-forms noted
on the pottery and goldwork from the region and the images carved on the petro-
glyphs at Shanshipampa, specifically in terms of their representation in profile and
the treatment of the extremities, i.e., the head, the flexed legs, and the long tail.
Figure 11.9. Gold pectoral with decoration in repousse note the shape of the central figure's head
(Museum of the Central Bank of Ecuador, Quito; Cat. no. 3-7-59; diameter: 16cm).
Figure 11.10. Interior of a Tuza style footed bowl in the private collection of Mr. Humberto Rosero of
Tulcan. Note the headdresses of the painted anthropomorphic figures (drawing after a photo published
in Martinez 1977: 85).
CONCLUSION
The mythic traditions of the native peoples of the northern Andean region
offer some insight into the symbolic significance of the imagery observed on the
late prehistoric period petroglyphs from the Pirnampiro district. The stories
describe a fearsome double-faced or double-headed creature, explicitly snake-like
Acknowledgements
The Pirnampiro Archaeological Project has received financial support from
the National Science Foundation (Grant No. SB-9810477), the Wenner-
Gren Foundation for Anthropological Research, and Wayne State University.
REFERENCES
Alvarez Lobo, Ricardo, 1984, Estudio Etno-hist6rico del Urubamba y Alto Ucayali. Editorial San
Esteban, Salamanca.
An6nimo, 1965, Relaci6n de las cibdades y villas que hay en el distrito de la Audiencia Real que
reside en la Cibdad de San Francisco del Quito ... In Relaciones Geograficas de Indias, edited
by Marcos Jimenez de la Espada, pp. 183-189. Biblioteca de Autores Espaiioles, volume 184.
Editorial Atlas, Madrid. [originally ca.1582]
Athens, Stephen 1., 1980, El Proceso Evolutivo en las Sociedades Complejas y la Ocupaci6n del
Periodo Tardio-Cara en las Andes Septentrionales del Ecuador. Instituto Otavaleiio de
Antropologia, Otavalo, Ecuador.
Baer, G., E. Ferstl, and C. N. Dubelaar, 1984, Petroglyphs from the Urubamba and Pantiacolla rivers,
eastern Peru. Verhandl. Naturf. Ges. Basel, Band 94: 287-306.
Benson, Elizabeth P., 1992, Motifs, meanings, and myths: the iconography of Sitio Conte gold
plaques. In River ofGold, edited by Pamela Hearne and Robert Sharer, pp. 22-31. The University
Museum, University of Pennsylvania, Philadelphia.
Berenguer, Jose, 1984, Figurillas post-formativas en la sierra norte del Ecuador. Gaceta Arqueol6gica
Andina 10: 4-5.
Berenguer, Jose and Jose Luis Martinez, 1989, Camelids in the Andes: rock art, environment and
myth. In Animals in Art, edited by Howard Morphy, pp. 390-416. Unwin Hyman, London.
Biesele, Megan, 1983, Interpretation in rock art and folklore: communication systems in evolutionary
perspective. New Approaches to Southern African Rock Art. South African Archaeological
Society, Goodwin Series 4: 54-60.
Borja, Antonio, 1965, Relaci6n en suma de la doctrina e beneficio de Pimampiro y de las cosas nota-
bles que en ella hay, de la cual es beneficiado el P. Antonio Borja. In Relaciones Geograficas de
Part IV
Conclusion
Chapter 12
Issues of Cultural Production
and Reproduction
HELAINE SILVERMAN AND WILLIAM H. ISBELL
This collection of essays examines a wide range of topics from a broad sample of
prehispanic Andean societies. Each essay reflects the reality of a state of knowl-
edge in archaeological research, as well as greater goals in anthropology and art
history. All share a concern with the materialization of culture, and what ques-
tions can and must be asked of the material remains available for study. Each
essay represents another step in increasing understandings of the Andean past,
supporting one another in an ever expanding dialogue of knowledge production.
Each reader will likely focus on different issues discussed by the essays, perceiv-
ing disparate questions that may now be addressed more cogently.
The chapters on Pucara (Chavez), Tiwanaku (Protzen and Nair) and Huari
(Ochatoma and Cabrera; Isbell and Cook) reveal multiple and important parts of
a great southern highland cultural tradition, out of which Inca culture eventually
arose. They mutually reinforce one another, making it possible to ask and at least
begin to answer new questions. If early Pucara culture was dominated by a bal-
anced female/male opposition in iconography, religion, and world view, as
Chavez's research implies, what does this mean for our understanding of the
Pucara-derived religious iconography that was shared by Tiwanaku and Huari?
Especially, what can now be suggested about the staff god, and the popular iden-
tification of this prominent Middle Horizon figure with the Inca creator god
Viracocha? And how does the description of Conchopata mortuary practices pro-
vide for a better understanding of the meaning of death and ancestors in
Tiwanaku and earlier Pucara society? Our new knowledge about the Middle
Horizon helps resolve questions about Inca ancestor worship and the antiquity
and meaning of founder mummies.
357
359
REFERENCES
Bauer, Brian S., 1998, The Sacred Landscape of the Inca: The Cusco Ceque System. University of
Texas Press, Austin.
Cereceda, Veronica, 1986. The semiology of Andean textiles: the talf'!:.·s of Isluga. In
AnthropoLogicaL
History of Andean PoLities, edited by John V. Murra, Nathan Wachtel, and Jacques Revel,
pp. 149-173. Cambridge University Press, Cambridge.
Conklin, William J., 1997, Structure as meaning in Andean textile:-. Chungara 29 (I): 109-133.
Connerton, Paul, 1989, How Societies Remember. Cambridge UJ~:versity Press, Cambridge.
Issues of Cultural Production and Reproduction 363
Gosselain, Olivier P., 1999, In pots we trust. The processing of clay and symbols in sub-Saharan
Africa. Journal of Material Studies 4 (2): 205-230.
Gow, Peter, 1995, Land, people, and paper in Western Amazonia. In The Anthropology of Landscape:
Perspectives on Place and Space, edited by Eric Hirsch and Michael O'Hanlon, pp. 43-62.
Clarendon Press, Oxford.
Hendrickson, Carol, 1995, Weaving Identities: Construction of Dress and Self in a Highland
Guatemala Town. University of Texas Press, Austin.
Isbell, William H., 1991, Huari administration and the orthogonal cellular architecture horizon. In
Huari Adminislralive Slruclure. Prehisloric Monumenral ArchileclUre and Slate Governmenr,
edited by William H. Isbell and Gordon F. McEwan, pp. 293-315. Dumbarton Oaks Research
Library and Collection, Washington D.C.
Isbell, William H., 2000, What we should be studying: the "imagined community" and the "natural
community." In The Archaeology ofCommunities: A New World Perspective, edited by Marcello
A. Canuto and Jason Yeager, pp. 242-266. Routledge, London.
Joyce, Rosemary A., 2000, Heterarchy, history, and material reality: "communities" in Late Classic
Honduras. In The Archaeology of Communities: A New World Perspective, edited by Marcello
A. Canuto and Jason Yeager, pp. 143-160. Routledge, London.
Kubler, George, 1985, Style and the representation of historical time. In Sludies in Ancient American
and European Arl. The Collecled Essays of George Kubler, edited by Thomas F. Reese,
pp. 386-390. Yale University Press, New Haven.
Lawrence, Denise L. and Setha M. Low, 1990, The built environment and spatial form. Annual Review
ofAnthropology 19: 453-505.
Lightfoot, Kent G., 2001, Traditions as cultural production. In The Archaeology ofTraditions. Agency
and History Before and After Columbus, edited by Timothy R. Pauketat, pp. 237-252. University
of Florida Press, Gainesville.
Lynch, Kevin, 1988, The Image oflhe Ciry. M.LT. Press, Cambridge.
Medlin, Mary Ann, 1984, Learning to weave in Calcha, Bolivia. In The Junius B. Bird Conference
on Andean Textiles, edited by Ann Pollard Rowe, pp. 275-287. The Textile Museum,
Washington, D.C.
Moore, Jerry D., 1996a, Architeclure and Power in the Ancient Andes: The Archaeology of Public
Buildings. Cambridge University Press, Cambridge.
Moore, Jerry D., 1996b, The archaeology of plazas and the proxemics of ritual: three Andean tradi-
tions. American Anrhropologisl 98 (4): 789-802.
Murra, John v., 1989, Cloth and its function in the Inka state. In Cloth and Human Experience, edited
by Annette B. Weiner and Jane Schneider, pp. 275-302. Smithsonian Institution Press,
Washington, D.C.
Pauketat, Timothy R., 2001, A new tradition in archaeology. In The Archaeology of Tradilions, edited
by Timothy R. Pauketat, pp. 1-16. University of Florida Press, Gainsville.
Paul, Anne, 1990, Paracas Rilual Allire: Symbols of AUlhoriry in Ancienl Peru. University of
Oklahoma Press, Norman.
Rowe, John H., 1946, Inca culture at the time of the Spanish conquest. In Handbook ofSouth American
Indians. Volume 2, The Andean Civilizations, edited by Julian H. Steward, pp. 183-330. Bulletin
143. Bureau of American Ethnology, Smithsonian Institution, Washington, D.C.
Rowe, John H., 1967, What kind of a settlement was Inca Cuzco? Nawpa Pacha 5: 59-76.
Rowe, John H., 1985, La constitution inca del Cuzco. Hist6rica 9 (I): 35-73.
Salomon, Frank, 1991, Introduction. In The Huarochiri Manuscript, edited by Frank Salomon and
George Urioste, pp. 1-38. University of Texas Press, Austin.
Urton, Gary, 1997, The Social Life of Numbers. A Quechua Ontology ofNumbers and Philosophy of
Arithmetic. University of Texas Press, Austin.
Urton, Gary, 2001, A calendrical and demographic tomb text from northern Peru. Lalin American
Antiquiry 12 (2): 127-147.
Index
Bamboo containers
Chachapoyaiconography, 145,146,149
Banco de Credito del Peru
Arte y Tesoros del Peru, 8, 24
Bat, head
Challuabamba pottery, 166
Bellinger, Louisa, 74, 92
Bibliotheque Nationale, Paris, 193
Bird
attendant, floating, 266
Guanaco and Bird Motif
Feline man, Pucara ceramic art, 55-56,
57
realistic, 266
Bird, Junius, 74, 92
Bird-monkey forms, 340
Bird Priest, 3
Birth bowl, Moche pottery, 121,122
365
366
Blind miniature gateways, Tiwanaku, 193, 209
Block Color style
Brooklyn Museum Textile, 84-85
Goteborgs Etnografiska Museum Textile, 87
Paracas Necropolis textiles, 28, 75, 83, 87-
88, 97
BMT: see Brooklyn Museum Textile
Borja, Antonio, 336
Bottles, Moche
painted vessel assemblages, /25
pictorial notations, mold inscriptions, 121,
/24
Bray, Tamara L., 186, 333-354
Brooklyn Museum Textile, 83-86
Brown agouti, 149, 153n. 4
Buck, Fritz, 218
Burial: see Mortuary practices
Burnished Black pottery, 159, 161, 163, /65,
166, /67, 172-174
Index
Index
Chachapoya iconography (cont.)
pyroengraved profile feline on hollow
bamboo container, 146
related imagery, 145-149
rollout of pyroengraved gourd from
Chullpal, 144
settlements, 139-140
tombs, 142
unku, 145-148, 147, 148
Chakipampa mesa
Conchopata, Middle Horizon, 289
Challuabamba, 157-177
pottery, 159-163, 358
bat, head, 166
chief's house, 167-172
content, 172-175
cosmological symbolism, 164-167
Cuts I, 2 and 3, 169
Cut 4, /70, 171
Cut 5,171
effigy jar, 167
firing practices, 164
medium, 172-175
oxidized wares, 165
percentages of each type by level, 163
periods and types, 162
political ideology, 172-175
rims, number of, 163
stamps, seals, and ceramic cup sherds, 168
radiocarbon samples, 161
trench wall, 160
Chapi, 336-337
Chauchilla River, 324, 325
Chavez, Karen Mohr, 24, 35-36, 66-67, 357
Chavez, Sergio J., 24
Pucara ceramic art, themes, 26-27, 35-69
Chavez Ball6n, Manual, 35-36
Chavin, 149, 345
Pucara ceramic art, 37
Chav[n (Tello), 7
Chief's house
Challuabamba pottery, 167-172
Chota-Mira river, 334, 335-336
Chullpas
Laguna de los C6ndores, Peru, 141-143, 144,
150
Cieza de Leon, Pedro, 217, 221
Civic center
Conchopata, Middle Horizon, 254, 289-290
Cleveland Museum of Art Textile, 92-93
367
368
Conchopata (cont.)
oversize ceramic offering (cont.)
1999C OCO Type 4, 267
1942 oeD Type lA, 255, 262-263
OCO Type I, 260
OCO Type 2, 260, 261
OCO Type 3, 262
OCO Type 4, 262
2000A OCO Type IB, 267-268
2000B OeD Type IB, 268-269, 272, 273
2000C OCO Type 3, 269, 274
2000D OCO Type 3, 269-270, 275
200IA OCO Type 3, 270-271, 276
200IB OCO Type 3, 271-277
types of, 259, 260
urns, 259, 261
warriors in boats, 275
woman nursing spotted feline, 267
perimeter walls, 254
Conchopata Style, 251
Conklin, William 1., 93-94, 182, 192,207-208
Cook, Anita G., 184-185,239,240,249-305
Cordy-Collins, Alana, 9
Corporate styles, 10
Cotomachaco,347
Courty, Georges, 217
Crequi-Montfort, Georges de, 217
Crucifixion theme, 40
Cultural production and reproduction, 357-364
materiality, 359-362
Dawson, Lawrence, 85
D'Orbigny, Alcide Dessalines, 215
Desena people of Colombia, 152n. 2
Doig, Kauffmann, 8
Domingo Canepa of Pisco, 83
Donnan, Christopher B., 9, 10,25
Moche style, 28
Double chambered rattle, Moche pottery, 118,
119
Double-faced embroidery, 74
Drum, Nasca 2 pottery, 76
D-shaped temple, Conchopata, 184, 225-247,
361
ceramic iconography, 238-244
anthropomorphic themes, 242
feline heads, 239, 241
Front Face Staff God, 239, 246
gods and mythological beings, 239-240
Nasca-like motifs, 243-244
Index
Index
Face markings
Camelid woman, Pucara ceramic art, 44
Feline man, Pucara ceramic art, 51, 52
Face-neck jars, Conchopata, 259, 261, 271-273
Face urn, Conchopata, 268
Fanged Lizard, Moche pottery, 115
Farmer, James D., 157-177
Faults: see also specific fault
subterranean water resources, Rio Grande de
Nazca drainage, 315, 316, 319,320, 321,
322, 325-326
Feathered headdress, 151
Feline, spotted
oversize ceramic offering tradition,
Conchopata, 267
Feline attendant, floating
oversize ceramic offering tradition,
Conchopata, 265
Feline-headed snakes, 50, 54, 56
Feline heads
D-shaped temple, Conchopata, 239, 241
Feline iconography, Chachapoya, 144-145,146
Feline man, Pucara ceramic art
ceremonial/mythical component, 65-66
cream color for face, 52-53
effigy bowls, 50-52
face markings, 51, 52
facial hair, 51, 53
feline-headed snakes, 50, 54, 56
feline man pose, 64
feline motif, 56, 58, 60, 61
garments, 54
geometric designs, 59
goatee beard, 53
Guanaco and Bird Motif, 55-56, 57
hair, 51
iconology,60-67
identification of, 40, 50-54
male-headed effigy jar and bowls, 46
motifs and designs, 54-60
moustache, 53
mouth/lips, 51, 52
nose and protruding chin, 51
outflaring bowl, 47, 55
pedestal-base bowl, 47-49
political aspect, 64-65
profile, 53
related motifs and designs, 50-60
Severed Head Motif, 53, 54-55, 55-56, 61,
65,66
369
370
Gateways, Tiwanaku (cont.)
Gateway of the Sun (cont.)
location of, 215-216
half-scale gateways, 218
location of, 215-218
miniature gateways, 208-213
full-sized gateways, compared, 208-213
Pumapunku, 189, 193, 195-196,219
location of, 217-218
Sandstone Gateway, 205
signific".nce of, 191-193
Gateway~,of Pumapunku, 189, 193, 195-196,219
location of, 217-218
GEM: see Goteborgs Etnografiska Museum
GEMT: see Goteborgs Etnografiska Museum
Textile
Geoglyphs and subterranean water resources,
Rio Grande de Nazca drainage, 307-
332, 312, 314, 322-324
archaeological analysis, 317-318
Cerro Aja, 312, 313, 319-322, 329
Cerro Colorado, 313, 316, 317, 327-328
faults, 315, 316, 319,320, 321, 322, 325-326
geological mapping, 316-317
geophysical analysis, 318
hypothesis, new, 309-312
archaeological analysis, 317-318
conceptual model to explain, 328-329
development of, 309-312
geological mapping, 316-317
geophysical analysis, 318
methodology, 315-319
regional map of drainage, 311
results, 319-328
Schreiber-Lancho Rojas model, 310
study areas, 312-315
surveying, 318
testing hypothesis, 312-319
water sampling program, 319
surveying, 318
Usaca, 313, 324-327
water sampling program, 319
Geometric Snake motif, /24
Gold
Conchopata, artifacts, 287
ornaments, Popayan region of Colombia,
340
Gold pectoral, 344
Goteborgs Etnografiska Museum
other textile, 90, 91
Index
Kantatayita, 216
Kaulicke, Peter, 9
Kidder, Alfred II, 26
Kosok, Paul, 307
Index
Kroeber, A. L., 72
Kubler, George, 23
Kuntur Wasi, 149
371
372
Moche pottery (Cerro Mayal, Peru) (cont.)
pictorial notations, mold inscriptions (cont.)
face stamp molds, 126, 126-127
Geometric Snake motif, 124
rattle inscription, 123
rattle player, 118, JI9-120
register marks, mold inscriptions, 116-117
Moche style, 28-29
Moche war clubs, 128
Mold inscriptions, Moche pottery: see Moche
pottery
Monkeys, rock art, 340, 347-348
bird-monkey forms, 340
Piartal pottery, 341-344, 343
Monolithic miniature gateways, Tiwanaku,
210-212
Mortuary practices
camelid burial, D-shaped temple,
Conchopata, 229
Chachapoya iconography, burial sites, 141-
143
Conchopata, Middle Horizon, 280-288, 298
burial type 1,282,283
burial type 2, 282, 283
burial type 3, 282, 283
burial type 4, 282, 283, 284-285
burial type 5, 283, 285-287
burial type 5b, 285-287, 286
burial type 6, 283, 287
burial type 7, 283, 287-288
gold artifacts, 287
niched halls, 281
pregnant woman skeleton, 284
Moseley, Michael E., 10
Mummy Bundle, 38
Nasca 2 textiles, 83, 85
Mummy Bundle 89
Paracas Necropolis textiles, 82
Museo Nacional de Antropologia y
Arqueologia, 262
Museum fur VOkerkunde, 6
Museum of the Central Bank of Ecuador in
Quito, 344
Index
Index
Ochatoma Paravicino, Jose, 184, 225-247,
258, 261, 263
Open composite miniature gateways,
Tiwanaku, 210
Orcona, 312, 314, 322-324
Orcona Lineament, 323-324
Oriente, 334
Oversize ceramic offering tradition,
Conchopata, 259-277
bird, realistic, 266
decorated urn, 274
excavation of room EA-36, 270
face-neck jars, 259, 261, 264, 271-273
face urn, 268
floating bird attendant, 266
floating feline attendant, 265
front-face diety, 265
map, 255
1977 OCO Type IB, 261, 263, 264
1997/98 OCO Type 2, 263
1999A OCO Type 2, 263, 265, 266
1mB aco Type IA, 264, 267, 267, 268, 269
1999C aco Type 4, 267
1942 aco Type lA, 255, 262-263
OCO Type I, 260
aco Type 2, 260, 261
OCO Type 3, 262
aco Type 4, 262
2000A OCO Type lB, 267-268
2000B OCO Type lB, 268-269, 272, 273
2000c oeo Type 3, 269, 274
2000D aco Type 3, 269-270, 275
200iA aco Type 3, 270-271, 276
200lB aco Type 3, 271-277
types of, 259, 260
urns, 259, 261
warriors in boats, 275
woman nursing spotted feline, 267
Oversize urns, decorations on rims
D-shaped temple, Conchopata, 242
Owl Motif
Camelid woman, Pucara ceramic art, 45, 49, 63
373
374
Pontificia Universidad Cat61ica del Peru, 8-9
Posnansky, Arthur
full-sized gateways, 193
"Little Pumapunku," 210, 211, 213, 218, 220
Pumapunku, 196, 217
Pottery
Burnished Black, 159, 161, 163, 165, 166,
167, 172-174
Challuabamba, 159-163, 358
bat, head, 166
chief's house, 167-172
content, 172-175
cosmological symbolism, 164-167
Cuts I, 2 and 3, 169
Cut 4, 170, 171
Cut 5,171
effigy jar, 167
firing practices, 164
medium, 172-175
oxidized wares, 165
percentages of each type by level, 163
periods and types, 162
political ideology, 172-175
rims, number of, 163
stamps, seals, and ceramic cup sherds, 168
Conchopata, Middle Horizon, 277-280
firing areas, 279
pit kilns, 279-280
D-shaped temple, Conchopata: see D-shaped
temple, Conchopata
Formative Period, 30-31
Moche: see Moche pottery (Cerro Mayal,
Peru)
Nasca I pottery, 74, 78, 80-83
Piartal, monkeys, 341-344, 343
Pucara: see Pucara ceramic art
Red-and-Black, 159, 161, 163, 172
Red-on-Cream, 159, 161, 163, 164
Pozzi-Escot, Denise, 258
Proto-writing, Moche pottery, 107-135
Proulx, Donald A., 307-332
Prozten, Jean-Pierre, 182, 183-184, 189-223
Pucara ceramic art, 26-27, 35-69
analysis, 38-40
Camelid woman: see Camelid woman,
Pucara ceramic art
Chavin,37
classification, 38-40
design, defined, 39
documentation, 38-40
Index
Index
375
376
Textiles (cont.)
Goteborgs Etnografiska Museum
other textile, 90, 91
Goteborgs Etnografiska Museum Textile,
86-89
looted, 93-94
Nasca, 71-105; see also Nasca textiles
Paracas Necropolis, 28
differentiating from early Nasca, 71-105
Tiwanaku, 183, 189-223
Akapana Gateway, 201, 202, 203
anticephaloid architraves, 207-208
curved architraves, 205, 207
full-sized gateways, 193-208
miniature gateways compared, 213-215
Gateway I (Pumapunku), 195, 196, 197
Gateway II (Pumapunku), 194, 195-196, 197,
203
Gateway III, 204, 214
Gateway III (Pumapunku), 194, 204, 206,
214
Gateway IV (Pumapunku), 196
Gateway A, 210-212
Gateway of the Moon, 191, 193, 204-205
location of, 216-217
Gateway of the Sun, 191, 193, 195, 196-201,
203-204, 221
location of, 215-216
gateways, significance of, 191-193
Gateways of Pumapunku, 189, 193, 195-196
location of, 217-218
half-scale gateways, 218
introduction, 189-191
location of gateways, 215-218
map, 190
map of spheres, 250
miniature gateways, 208-213
full-sized gateways compared, 213-215
Sandstone Gateway, 205
Index
Warrior Priest, 3
Warriors kneeling in reed boats, pottery
D-shaped temple, Conchopata, 231, 297
oversize ceramic offering tradition,
Conchopata, 275
White Courtyard
Conchopata, Middle Horizon, 291-292
Winged Man Motif
Feline man, Pucara ceramic art, 59, 62
Woman nursing spotted feline
oversize ceramic offering tradition,
Conchopata, 267
Wrinkle-Face Whistler mold
Moche pottery, 112, 114
Xesspe, Mejia, 71