Advances in

Titicaca Basin

Edited by
Charles Stanish,
Amanda B. Cohen, and
Mark S. Aldenderfer

Cotsen Institute of Archaeology at UCLA

Los Angeles, ~ a l i f o r n i a
EDITORIAL BOARD OF THE COTSENLYSTI'TCTTEOF ARCHAEOLOGY AT UCLA:
Jeanne E. Arnold, Christopher Donan, John Papadopoulos, James Sackett,
Julia L. J. Sanchez, and Charles Stanish

The Cotsen Institute of Archaeolog at UCLA

Charles Stanish, Director
Julia L. J. Sanchez, Assistant Director and Director of Publications
Shauna Mecartea, Publications Assistant

Edited by Joe Abbott

Production by Leyba Associates

Library of Congress Cataloging-in-Publication Data

Advances in Titicaca Basin archaeology / edited by Charles Stanish, Amanda B. Cohen, and Mark S. Aldenderfer.
v. cm. - monograph / Cotsen Institute of Archaeology, University of California, Los Angeles ; 54-)
Includes bibliographical references.
ISBN 1-931745-15-3 (alk. paper)
1. Indians of South America-Titicaca Lake Region (Peru and Bolivia)-Antiquities. 2. Tiwanaku culture-Titicaca Lake
Region (Peru and Bolivia) 3. Excavations (Archaeology)-Titicaca Lake Region (Peru and Bolivia) 4. Tticaca Lake Region
(Peru and Bolivia)-Antiquities. I. Stanish, Charles, 1956- 11. Cohen, Amanda B. 111. Aldenderfer, mark S. IV. Monograph
(Cotsen Institute of Archaeolog!. at UCLA) ; 54-
F3319.1.TS7A382005
984'.1201-4~22
2004028970

ISBN 1-931745-15-3 (paper)

ISBN 1-931745-19-6 (cloth)

Copyright O 2005, Regents of the University of California

All rights reserved.
Printed in the USA.
 Inca Ceremonial Sites in the
Southwest Titicaca Basin
Elizabeth Arkush

T AWANTINSUYU, as the Incas called their empire, was important to the Incas, both politically as
grew in perhaps a hundred years (ca. A D 1430- the home of the rich, populous, and powerful
1532) to encompass a huge territory of numer- Lupaca and Colla ethnic groups, and religiously,
ous ecological zones and peoples with diverse because it was the doorway to the famous Inca
customs, languages, economies, and political in- pilgrimage center on the Islands of the Sun and
stitutions. The Incas relied on religious ideology Moon. Analysis of the style, size, and placement
as one important element of imperial control of ceremonial sites in the Lupaca region suggests
over this vast and varied area. Ethnohistoric that Inca administrators did not mandate cere-
documents describe a concerted Inca policy of monial site construction merely as a wholesale
religious incorporation of the provinces (e.g., imposition of Inca ideology, but took many other
Cobo 1979:191 [1653: Bk. 12 Ch. 231, 1990 [1653]; factors into account, including previous non-
MacCormack 1991:98-118; Rowe 1946:293-314, Inca traditions of worship. Furthermore, it is
1982; Valera 1950:145).The subject people's local likely that some sites were constructed and
divinities, or huacas, were assimilated into Inca modified at least partly by local workers with-
state control, and subjects were gathered to en- out Inca supervision. This general picture of in-
gage in Inca rituals at pilgrimage centers or state clusion and accommodation contrasts with more
festivals at Cuzco. Inca state ritual was also rigid class exclusion at the sanctuary on the Is-
brought to the provinces and was performed in land of the Sun itself. These little-known sites
sun temples built at provincial centers, at local and their relation to the Island of the Sun sanctu-
festivals, and at special state ceremonies (such as ary give us a window into the i n n e r mechanisms
the capacocha, or sacrifice ceremony) that were of outwardly monolithic, legitimizing ideolo-
performed away from the center. Nevertheless, gies. In practice, in the Titicaca Basin, as perhaps
the ethnohistoric record gives us an incomplete everywhere, ideology was shaped and contested
and Cuzco-centric view of the way religion by countless agents of greater and lesser power. .
worked on the ground in the empire. A close ex-
amination of the archaeological record can illu-
minate the ways in which religious ideology in
RELIGION IN THE INCA E MPIRE
Tawantinsuyu interacted with, rather than sup- The question of the degree of Inca religious con-
planted or ignored, the preexisting cosmologies, trol may be viewed as a subset of a more general
ritual practices, and shrines of its new provinces. debate on the impact of Lnca conquest on local
This chapter looks at the archaeological populations. To some scholars, the Inca policies
manifestations of religion, ideology, and ritual in of indirect rule meant that little changed at the
an Inca province by compiling the results of sur- local level after the Inca conquest (Murra 1980
face survey and incorporating previous research [1956]). Others have emphasized the intrusive
on Inca period ceremonial sites in the south- policy of labor extraction used by the Incas (for
western Lake Titicaca Basin of Peru. This region example, Julien 1988; Stanish 1997). In fact, the
impact of Inca conquest in Tawantinsuyu 12 Ch. 231; Valera 1950:145).The Incas also fre-
ranged from complete reorganization of some quently enhanced aboriginal religious sites and
subject provinces to the loosest control of others consulted important local huacas (Rowe 1946:
(Bauer 1992; D'Altroy 1992; Dillehay 1977; 302; MacCormack 1991:141-159).
Gonzcilez 1983; LeVine 1985; Menzel 1959; Neth- Our archaeological knowledge of Late Hori-
erly 1978; Pease 1982; Salomon 1986a, 1986b; zon ceremonial sites outside the Inca heartland
Stanish 2001b). The amount of control exerted in is incomplete (see, for instance, Van de Guchte
any one place was affected by the length of time 1990:406, illustration 2). Some provincial cere-
the Incas controlled it, the desirability or use of monial sites, such as Sayhuite in Apuriinac and
the region in question, the sociopolitical com- Vilcashuamhn in Ayacucl-ro, are intrusive and
plexity of polities already present in the region, pure lnca in style; others, like Pachacamac or
the threat that these polities posed to lnca rule, Wari Wilka (Shea 1969),are Inca additions to im-
and strategic considerations. portant pre-Inca shrines, built with significant
For religion, the question is whether Inca re- adaptations of local, non-Inca styles and niateri-
ligious ideology was imposed wholesale in an als. Rarely, there was actual destruction of indig-
attempt to comm~~nicate Inca dominance to con- enous huncas (Nielsei-r and Walker 1999). In
quered peoples 01; rather, evolved to accommo- addition, numerous modest, rural ceremoi-rial
date the practices and needs of ne\tr subject sites exist-a few will be described below-but
populations. Both possibilities are supported by they are little known, giving the erroneous im-
different lines of evidence, and i t is certainly pression that where ceremonial sites were built
possible that different religious policies were in the provinces, they were major sanctuaries di-
utilized in different regions or time periods by rectly designed and controlled by the lnca elite.
the state. It is debatable whether ii-rtrusive Inca cere-
Most contact period documentary sources on monial installations were placed strictly to intlu-
lnca religion tend to portray a uniform imposi- ence the conquered masses. For instance, a
tion of official Inca religion everywhere in the proliferation of high-altitude mountain shrines
empire, although some accounts of the extirpa- in Argentina and Chile, a region where the Incas
tion of idolatry emphasize the importance of lo- invested comparatively little in economic or po-
cal nonelite shrines and portable huacas in the litical infrastructure, shows considerable effort
daily lives of ordinary folk (e.g., Arriaga 1968 and expense poured into ceremonial sites that
[1621]).The Huarochiri manuscript documents a were not very visible or accessible to local popu-
remarkably detailed and, from a Cuzco perspec- lations (Beorchia Nigris 1973, 1985; McEwan and
tive, unortl-rodox provincial mythology that Van de Guchte 1992; Reinhard 1992; Schobinger
must have incorporated many pre-Inca elements and Constanza Ceruti 2001). Class exclusivity is
(Avila 1991 [ca. 15981). The Huarochiri manu- clearly apparent in Inca religious practices in
script also vividly attests to the expedient Inca general and at the Island of the Sun sanctuary in .
policy of incorporating regionally important particular. Elaborate regulations controlling ac-
shrines into their origin mythology: "In the high- cess to the island sanctuary reproduced and re-
lands, they say, the Incas worshiped the sun as flected social divisions. Overall, Inca religious
the object of their adoration from Titi Caca, say- sites did more, and sometimes less, than signal
ing, 'It is he who made us Inca!' From the low- lnca dominance.
lands, they worshiped Pacha Camac, saying, 'It Factors affecting Inca ideological strategy
is he who made us Inca!"' (Avila 1991:111, sec. may have included the political and economic
276 [ca. 1598:Ch. 221). According to the chroni- investment in the region, local religious prac-
clers, the major portable huacas of conquered na- tice and the prestige of local religious sites, and,
riones were brought to Cuzco and tended to by apparently, military considerations (Hyslop
rotating colonists from the huacas' homelands, 1990:189). The ideological incorporation of the
kept by the Incas both as hostages and as poten- provinces was not a n event, but an ongoing
tially powerful guests (Cobo 1979:191 [1653: Bk. process, as Cobo noted:
 CHAPTER 14: INCA CEREMONIAL SITES IN THE SOUTHWEST TITICACA BASIN 211

. . .from the beginning of their empire chiefs of the region (Julien 1983:36-38). Chulpa
the Incas were not always steadfast in burial towers continued to be built. Monumental
their religion, nor did they maintain the examples with beautifully finished stone ma-
same opinions and worship the same sonry in the Late Horizon may have signaled the
gods.. ..They were prompted to make strength and security of a hereditary elite newly
such changes because they realized that confirmed in its position of power within the
this way they improved their control Inca hierarchy.
over the kingdom and kept it more sub- Along with the resettlement of native peo-
servient [Cobo 1990:5 (1653:Bk. 13 Ch. ple, a distinct demographic spike is observable
1)). in the Late Horizon from Stanish et al.'s (1997)
survey of the southwest basin, best explained as
the movement of large numbers of mitima colo-
THE TITICACA BASIN
nists into the Titicaca Basin. Ample textual evi-
UNDER INCA RULE
dence also exists for mitimas in the region,
The traditional date for the expansion of the In- destined for service at the Island of the Sun or
cas into the Titicaca Basin is AD 1450 (Rowe for specialized production workshops (Ramos
1944:65). The conquest of the Titicaca Basin was Gavilan 1988:84 [1621:Bk.1 Ch. 121; Murra 1978:
recounted by Cieza de Leon (1985:130-136 [1553: 418; Julien 1983:75; Espinoza Soriano 1987: 248,
Bk. 2 Chs. 41431) and Cobo (1979:140 [1653:Bk. 253; Stanish 1997:204).
12 Ch. 131). While they differ in some details, The most extraordinary act of Inca control in
both accounts state that the Lupacas and the the Titicaca Basin was the establishment of an
Collas were engaged in a war at the time of Inca elaborate sanctuary complex and pilgrimage
contact. The Collas were eventually vanquished, center on the Islands of the Sun and Moon (Titi-
and the Lupacas welcomed the Incas as allies. caca and Coati Islands).' Here, according to Inca
The Collas later rebelled several times under the n~ythology,the sun first arose, and here the Incas
yoke of the Incas (Cobo 1979:143, 153 [1653:Bk. were created by the god Viracocha. Titicaca Is-
12 Chs. 14, 161). land appears to have been used as a religious
The Inca incorporation of the Titicaca Basin center in the Tiwanaku period (Bauer and Stan-
caused dramatic changes on almost every level. ish 2001; Seddon, Chapter 9, this volume), and
The most obvious change that is visible archaeo- Tiwanaku-style archaisms at the Inca sanctuary,
logically is the massive resettlement of the Titi- in both ceramics and architecture, attest to an
caca Basin and the introduction of large Inca attempt to link their control of the island to
numbers of rr~ifinlas,or colonists from other sub- ancient tradition (Gasparini and Margolies 1980:
ject areas (Stanish 1997). The hilltop forts, or 13, 262; Julien 1993). Inca emperors regularly
pukaras, were abandoned, along with those habi- visited the sanctuary, reaffirming its importance
tation areas situated to take advantage of non- in imperial religious ideology (Cobo 1979:141,.
residential pukaras. Most of the inhabitants of 144, 154 [1653:Bk. 12 Chs. 13, 14, 161).
the region were resettled in intrusive, nucleated While hosting the highest echelons of Inca
centers along two main Inca roads, one on each society, the island sanctuary was also a pilgrim-
side of the lake (Hyslop 1984). age center, and sources state that pilgrims came
These new settlements served both as ca- to it from every corner of the empire (Cobo
beceras, or administrative centers (Diez de San 1990:95-96 [1653:Bk. 13 Ch. 181; Ramos Gavilan
Miguel 1964 [1567]; Toledo 1975 [1575]), and as 1988:41, 164 [1621:Bk.l Ch. 4, 261). After passing
tambos, or way stations along the road. Among through a number of checkpoints and purifica-
them, the most important were the new capitals tions, pilgrims arrived at the Sacred Rock on the
of the Lupacas and Collas: Chucuito and Hatun- Island of the Sun, where they participated in
colla, respectively. The cabeceras were adminis- seasonal festivals, obtained oracles, gave rich of-
tered by local lords (k~irakas)in a hierarchical ferings of gold, silver, shells, feathers, and fine
system. It is likely, although not certain, that cunzbi cloth, and observed sacrifices of children,
these kurakas were descended from the pre-Inca llamas, and guinea pigs (Ramos Gavil6n 1988
[1621]). Pilgrims then passed on to a temple on to have been as troublesome as the Collas under
the Island of the moon. Inca rule (but see Cieza 1985:155 [1553:Bk 2 Ch.
liegulations at the sanctuary compartmen- 531). Stanish (2000) argues that in practice they
talized people into distinct groups by geograph- held an apparently privileged position in the
ical or ethnic origin and by social status (Ramos Titicaca Basin in the Late Horizon, perhaps de-
Gavilhn 1988:153 [1621:Bk. 1 Ch. 241). Access to riving from their initial negotiation of the peace
the Sacred Rock itself was highly restricted. A with the Inca emperor (Cieza de Le611 1985135-
wall and a series of three gates about 250 m 136 [1553:Bk. 2 Ch. 431). According to Cobo, the
away from the Sacred Rock marked the closest Lupncas were favored for their loyalty b!. Topa
point to which non-Inca pilgrims could come. Inca and his son Guayna Capac (Cobo 1979:144,
There, they were permitted to watch rituals at 154 [1653:Bk. 12 Chs. 14, 16]), and the Lupacas
the Rock and left their offerings at the gate were considered "Incaized" relative to other
(Cobo 1990:96 [1653:Bk. 13 Ch.181; Ramos subject peoples (Hyslop 1977a:160).
Gavilan 1988:87, 94 [1621:Bk. 1 Chs. 12, 13)). In fact, both the Collas and the Lupacas, as
Some local, nonelite groups were not permitted well as the other naciones of the Titicaca Basin,
to visit the island at all (Lizarraga 1987:187 were remarkably "Incaized" in their material
[1605:Ch. 861, Ramos Gavildn 1988:150, 176 culture. Locally . -produced ceramics from the
[1621:Bk. 1 Chs. 24, 291). In brief, the picture Late Horizon, including- nonelite ware, drew
from the documents and from the archaeological heavily on Cuzco-Inca ceramic forms and de-
evidence is one of a highly structured environ- signs, and the local ware was technologically
ment in which each pilgrim's position was pre- and stylistically closer to Cuzco-Inca ware than
cisely assigned based on geography and social it was to indigenous Late Intermediate period
status. pottery (Stanish 1991; Julicn 1983). Houses, gen-
Brian Bauer and Charles Stanish's (2001) erally circular in the Late Intermediate period,
work on the Inca pilgrimage center at the Island became rectangular in the Late Horizon, follo~v-
of the Sun and Moon, as well as my reconnais- ing Inca tradition. Inca-style cut-stone masonry
sance project in 1998 on the southwest margins was occasionally adopted for edifices that d o not
of the lake, attest to a spurt of activity in the con- seem to have been executed under Inca supervi-
struction of new ceremonial sites in the Inca ye- sion, and were therefore probably constructed
riod in the southern and western Titicaca Basin. under local initiative; examples include the Inca
This remodeling of the religious landscape of the Uyu in Chucuito, capital of the Lupacas (treated
region went hand in hand with the massive al- below), and chulpas with a square footprint or
teration of the demographic, political, and eco- faced in fine Cuzco-style masonry, which were
nomic landscape. probably constructed and used by the local
Despite these Inca measures of control, or kurakas rather than by Inca administrators re-
perhaps because of them, the inhabitants of siding in the Collao (Hyslop 1977b:160; Julien
Collasuyu developed a reputation for rebellious- 1983:254). It seems that the local non-Inca elite,
ness (Molina "el Almagrista" 1968 [1552]:75). whether Colla or Lupaca, adopted an Inca archi-
The Collas rebelled at least once under the Inca tectural style for their most symbolic construc-
yoke and were harshly punished for it (Cobo tions, and even reworked the tombs of their
1979:143,153 [1653:Bk.12 Chs. 14, 161).Catherine ancestors to fit a new era in which Inca style was
Julien notes that one of the insubordinate Colla synonymous with prestige and dominance.
lords named himself "Pachacuti Inca" (Julien The interpretation of stylistic adoption is a]-
1983:258). ways problematic, and never more so than in the
In contrast to the fractious Collas, the Lu- context of political domination. However, it is
pacas followed a pattern of negotiation with worth noting that stylistic emulation in material
Inca power. Although the kuraka of the Lupacas culture need not be seen as signifying political
declared himself the son of the Sun and rebelled allegiance. Among the many Inca-style ceremo-
under Spanish rule in 1538 (Sitio del Cuzco 1934 nial sites discussed in this paper, at least three
[1539]:121),the Lupacas are not clearly reported appear to be local emulations of Inca ceremonial
 CHAPTER 14: INCA CEREMONIAL SITES IN THE SOUTHWEST TITICACA BASIN 213

forms. It is difficult to conclude exactly how The placement of sites is also telling. Sites
these sites functioned in the context of local adap- near the Inca road2 and near major towns and
tation to, co-option by, and resistance to Inca rule. administrative centers could in theory have been
However, we must remember that they may not accessible to all, including local residents (elite
have signified wholesale participation in the Inca and nonelite), Inca administrators, and pilgrims
cosmology or in Inca imperial ideology. traveling the Inca road from the Cuzco region to
the Island of the Sun. Sites far away from the
road and from major towns are less likely to
have been visited by pilgrims and Incas. Sites
How are we to interpret subsidiary ceremonial near Copacabana, the stopping-point for pil-
sites in the southwest Titicaca Basin and their grims before embarking for the Island of the Sun
relationship to the Island of the Sun pilgrimage and a major religious center in itself, were pre-
center? I argue that overall patterns in their sumably tied to the island cult. Highly visible,
style, placement, and accessibility should tell us large, and accessible sites could have served the
something about the way religion, as expressed aaditional purpose of signaling ideology or im-
in r~ligioussites, was used and negotiated be- perial control to a wide audience. Less visible,
tween Incas and local inhabitants. The style of small, or out-of-the-way sites would not have
carving at a ceremonial site indicates whether it been well suited to this propagandistic function.
was built under Inca supervision and at the or- Given that there were many new ceremonial
der of Inca administrators, or by workers (pre- sites built in the Late Horizon and that Inca ad-
sumably local) with a hazy idea of Inca ministrators were clearly interested in the re-
ceremonial style. Inca ceremonial sites are clas- gion's religious landscape, I propose to use the
sically distinguished by carved rocks, in partic- above criteria of style and placement to distin-
ular carved bedrock (as opposed to monoliths guish between two possible models of Inca reli-
or portable stones), as well as by uncarved boul- gious policy. In the first model, the Incas used
ders enclosed by masonry walls or incorporated ceremonial sites to disseminate imperial ideology
into larger sites (Bauer 1998; Hyslop 1990:Ch. 4; and mark the landscape as Inca territory. They at-
Van de Guchte 1990). They may also include tempted to incorporate the region's population
fountains, nonutilitarian canals, and carved into Inca state religion, but did not accommodate
channels for the manipulation of liquid offer- preexisting sacred sites or ceremonial styles and
ings, as well as structures with exceptionally practices. For example, Bauer and Stanish (2001)
fine stone masonry. Local religious forms and interpret ideology and its manifestations in cere-
styles were different from those executed under monial construction at the Island of the Sun as a
Inca artistic canons. Tiwanaku and pre-Tiwan- tool for supporting or legitimizing the rule of the
aku ceremonial constructions include sunken Incas over subject peoples. Smaller Inca ceremo-
courts, monoliths, and artificial mounds. It is nial sites could have functioned in the same way;
clear from the documents (Arriaga 1968 indoctrinating local participants into an im-
[1621]:79; Ramos Gavilan 1988:196-197 [1621: posed religious ideology in which the domi-
Bk. 1 Ch. 321) that many Tiwanaku or Formative nance of the Incas was naturalized. Maarten Van
carved stelae were revered in the early Colonial de Guchte uses this interpretation in his analysis
period and probably the Late Horizon. They of Inca carved rock sites: "By carving rocks, the
may also have been held sacred in the Late In- Incas effectively molded their world. The pat-
termediate period (LIP). In the LIP, huge chulpa terns on the rocks succinctly and directly helped
cemeteries such as Sillustani probably consti- to replicate icons of Inca ideology. As such the
tuted the major ceremonial centers (Hyslop carved rocksserved a purpose, similar to textiles
1977a:153). This stylistic distinction between in Andean society, as vehicles for the dissemina-
Inca-built and locally built sites is also sugges- tion of an Andean catechism" (Van de Guchte
tive of who may have used the sites. Contact pe- 1990:50). Likewise, DeMarrais et al. (1996:29)
riod textual references occasionally identify the propose that Inca roads, tambos, fortresses, and
groups who used a particular site. storehouses were symbolic as well as functional,
constituting "a landscape and architecture of tive centers, especially those that took advan-
power." Most of the small ceremonial sites in the tage of preexisting sacred places. Those sites,
southwest Titicaca Basin feature rocks carved in likewise, might not be particularly visible or ac-
more-or-less typical Inca style. Such rocks cessible.
would have indicated an Inca presence as clearly A subsidiary question is whether Inca-style
to the local inhabitants of the Titicaca Basin in sites in the region were intended or used as rit-
the Late Horizon as they do to the modern ob- ual stations on the pilgrimage route from Cuzco
server. Thus, the subsidiary ceremonial sites in to the Island of the Sun. If they were, they
the Collao could have been designed primarily should have been on the road or near tambos
to mark the landscape with symbols of Inca rule. (way stations), and either Inca in style or related
Such a policy should have resulted in iln- to the pilgrimage cult in other ways. In either
pressi\.e, intrusi\.e sites close to major towns and scenario, huacas that were locally built and used
roads, for the best visibility and the widest audi- could also have served the religious needs of the
ence. These sites would have been mandated population.
and designed bv Inca administrators and con-
structed under close Inca supervision, and thus
the st!.le of these sites should be similar to that
of sites in the Cuzco area, following the Inca sty- In July and August of 1998, I completed a recon-
listic canons outlined above. The sites could naissance of the southwest lake region under the
have been visited by Inca, pilgrim, Aymara elite, aegis of Prograrna Collasuyu, a multi-year col-
and nonelite alike, but the Incas might have laborative research project in the Titicaca Basin
wished to retain some control over their use, in directed by Charles Stanish (UCLA) and Ed-
keeping with a religious policy of imposition, mundo d e la Vega (Universidad Nacional Tec-
rather thm accommodation. Therefore, finding nica del Altiplano, Puno). My goal was to
evidence of associated elite habitation areas, and document several inadequately publislwd Inca
even ~ 7 access
t restriction, similar to the pattern period ceremonial sites in the southwest lake re-
on the Island of the Sun, would not be surpris- gion and to find new ones. Because of the dis-
ing. However, these last indices should not be persed and highly visible nature of Inca
considered essential to the model, especially as ceremonial sites, and the enormous size of the
access could ha\,e been restricted to sites while potential study area, I chose a nonsystematic re-
leaving no observable trace on the landscape. connaissance methodology. This permitted me
In the second model, imperial ideology to cover the largest possible geographical area,
might have been promoted through ceremonial target the known sites, sample areas of high
sites, but with significant accommodation to lo- probability, and therefore optimize data recov-
cal traditions and participation by local actors. ery from a relatively short season. The area of
Ceremonial sites could have been part of an on- study was the approach to the Island of the Sun
going process of negotiation and accommoda- on the southwest side of the lake, from Chucuito
tion between local and imperial actors, in which south, including the peninsula of Copacabana
imperial ideology was expanded and altered to (Figure 14.1). No attempt at full coverage of this
fit local needs. Here, more varied characteristics large area was made. Rather, the aim was to
would be expected than in the other models. characterize the general nature of the Inca reli-
Some sites could have been built by local work- gious landscape in that area. Thus, the list of
ers on Inca orders but without direct Inca super- sites given here is in no way exhaustive.
vision. Thus, some sites might be stylistically The reconnaissance methodology was based
"Inca," while others could be hybrid Inca-local on a combination of strategies: (1) the survey of
creations. In the latter category, we might also selected places, such as large rock outcrops and
see later additions by local workers to Inca cere- hilltops, that were deemed likely candidates for
monial sites, and vice versa. Some sites might be Inca ceremonial sites, (2) the use of local infor-
situated at pre-Inca huacas. Not all sites would mants, and (3) the investigation of places with
be located close to major towns and administra- toponyms in Quechua or names that indicated
 CHAPTER 14: INCA CEREMONIAL SITES IN THE SOUTHWEST TITICACA
BASIN 215

4\ TO CUZCO

Pucara
*'

Nicasio,

Juliaca .
Caracoto . BOLIVIA
Hatuncoll
$*,.
1
PERU

t N : Inca man
: (as surveyed by Hyslop 1984)

I Acora.

Korko *
Tmnbos and towns at
mid-16th century
I
I Inca ceremonial sites
I
F IGURE 14.1. The Lake Titicaca Basin, with Latc Horizon towns, ceremonial sites, and the road

the presence of sites (e.g., "Inca Pukara"). The many previously unrecorded features of these
reconnaissance was not restricted to sites along known sites were recorded. Chulpa sites that
the probable Inca road, but an inevitable bias to- lacked other Late Horizon ceremonial features
ward that strip of land arose from the simple fact are excluded from the catalogue below.
that the modern road lies close to the probable Sites or site sectors were considered to be
Inca road in most places (Hyslop 1984). In about "ceremonial" based on the presence of carved
two and one-half weeks of on-the-ground recon- rocks, ritual canals, or other clearly nonutilitar-
naissance, three new ceremonial sites were ian constructions. Sites were considered Late
found, and one large site-Kenko (Tres Venta- Horizon if they showed stylistic similarities in
nas), previously thought to be pre-Inca-was carving or masonry to known Inca sites (prefera-
shown to be an important Inca ceremonial ten- bly in the Inca heartland) or were associated
ter. In addition, all previously known Late Hori- with single-component habitation sites datable
zon ceremonial sites in the region as well as with Late Horizon pottery. The ceramic se-
several known chulpa sites were visited, and quence used for the analysis was refined by
 * CcWmonial sitcs
0 hlodcrn towns

FIGLIIE14.2. The Coyacabana Peninsula, with Late Horizon ceremonial sites

Stanish et al. (1997:40-49) for the survey of the Two known sites outside of the reconnaissance
Juli-Pomata region. Late Horizon pottery in the area, Pucara and Amantani, are briefly men-
study area includes bowl and aryballoid forms tioned in the catalogue because of their rele-
in Cuzco Inca, Local Inca, Chucuito, Sillustani, vance to the topic. The sites are listed in order of
and Pacajes types (for Cuzco Inca, see Rowe their location, moving away from the Island of
1944; for the region, see Julien 1993; for Local the Sun through the Copacabana Peninsula and
Inca, see Julien 1983; for Chucuito, see M. northward along the west side of the lake, in re-
Tschopik 1946; for Sillustani, see M. Tschopik verse pilgrimage order (Figures 14.1 and 14.2).
1946, Revilla Becerra and Uriarte Paniagua 1985,
and Stanish 1991; for Pacajes, see R y d h 1957, Intinkala and Orcohawira
Albarracin-Jordan and Mathews 1990, and "Intinkala" ("stone of the sun") lends its name to
Mathews 1992a). the collection of carved boulders in which it is
The catalogue of sites lists all known Late found. These boulders are located about 300 m
Horizon ceremonial sites from the study area, east of the modern town of Copacabana. The
excluding the Islands of the Sun and Moon. De- rocks are carved with flat altars, niches, and ca-
tailed descriptions of the new sites from the re- nals, with a quality of carving comparable t~
connaissance are included, as are somewhat Cuzco work (Figure 14.3). The site and its most
briefer treatments of previously published sites. prominent rock, called the "Seat of the Inca" or
 CHAPTER 14: INCA CEREMONIAL SITES IN THE SOUTHWEST TITICACA BASIN 217

F IG U RE 14.3. Intinkala

"Seat of the Sun," have been noted by numerous sector in Late Horizon Copacabana that sur-
travelers and archaeologists (for example, Ban- vives, and the rites related to the pilgrimage at
delier 1910; Hyslop 1990; Mantilla 1972; Rivera Copacabana that are mentioned in the docu-
Sundt 197th; Squier 1877~1;Trimborn 1967; and ments (e.g., Ramos Gavilin 1988:171 [1621:Ch.
Wiener 1880). Hermann Trimborn (1967:19-23) 281) may well have taken place at Intinkala and
gives a precise map of the Seat of the Inca. Orcohawira.
Cleaning and excavation by INAR (Instituto Na- Both carved-rock sites are so consistent with
cional de Arqueologia) in 1975 uncovered walls Inca stylistic canons that there can be no doubt
built directly into the rocks, traces of a paved they were executed on Inca orders and under Inca
floor, and a system of drainage canals (Rivera supervision. However, early Colonial documents
Sundt 1978.~76).Orcohawira (also called Rio suggest that an important local huaca, the idol of .
Macho or Rio Fuerte), first documented in 1968 Copacabana, may have been found at Intinkala or
(Mantilla 1972), consists of three stones finely Orcohawira (Sanz 1886). Ramos Gavilcin gives a
carved with "seats" and lies 200 m to the east of detailed account of where it was found:
Intinkala (Figure 14.4). These "seats" are ori- Among the Idols that were found in this
ented in the general direction of Intinkala and place, the principal and most famous
are of the same quality and style. INAR reported among the Yunguyos was the Copaca-
very fine, decorated Late Horizon ceramics on bana Idol. In the time of the priests of
the surface at both Intinkala and Orcohawira my order, certain Spaniards, looking for
(Rivera Sundt 1978:75-6). Indeed, the whole area some treasure, had the place dug out
between the sites and for about 100 m on all where the Idol was reputed to be locat-
sides has a dense scatter of fine Inca ceramics. ed, and found it, and nearby, also found
The Copacabana Peninsula was the stopping two huge rocks. One had the name
point for pilgrims just before they embarked for Ticonipa, and the other Guacocho. They
the Island of the Sun. This is the only ceremonial were worshiped by the Yunguyos, who,
 being poor, had no riches in this, their for example, K. Chiivez 1988; S. Chiivez and K.
principal sanctmry. Their continual nf- Chivez 1975). Stone sculptures of human figures
ferings were of livestock, chicha, and are practically unknown from the Late Horizon,
other things, because the silver and gold ~vhilethe continued worship of Formative or Ti-
which they managed to find, they of- wanaku monoliths in the early Colonial era is
ferred to the principal temples of the Sun ivell documented for the Lake Titicaca area (e.g.,
and the Moon. This Copacabana Idol Arriaga 1968 [1621]:79).
was in the same town [of Copacabana],
as you go to Tiquina. It was of fine blue H o r c a del Inca
stone, and it had nothing more than a On the hill of Kesanani just south of Copacabana
human face, disembodied of feet or is the misnamed "Horca del Inca," which, ac-
hands.. .this Id01 faced towards the tem- cording to the findings of Rivera Sundt (1984),
ple of the Sun, as if to signify that from functioned as an astronomical device rather than
there came its well-being.. . [Ramos
a gallows. A stone cross-beam is set between
Gavilin 1988:191 (1621:Ch 32), author's
two crags, and two small holes are drilled in out-
translation].
crops nearby (Figure 14.5). On the June solstice
Ramos Gavilin's account places the idol on the rising sun casts its light through the northern
the route east out of Copacabana toward hoIe onto the cross-beam. On the September
Tiquina, which corresponds well to the location equinox, a crag casts a shadow on the cross-
of Intinkala and Orcohawira. The two great sa- beam (Rivera Sundt 1984). The Island of the Sun
cred rocks near the idol may well have been the can be seen from the Horca, and taking into ac-
rocks at Intinkala.3 The descriptions of the idol count its proximity to Copacabana, it is likely
are far from detailed-according to Calancha that rituals in connection with the Island of the
(1972:1.139 [1639:Bk. 1 Ch. 31)) it had the head of Sun pilgrimage cuIt were performed here. While
a man and the body of a fish-but it was most Rivera Sundt (1984:98) offers the possibility that
likely a Formative or Tiwanaku monolith (see, the Horca was a pre-Inca construction, the ob-
 CHAPTER 14: INCA CEREMONIAL SITES IN THE SOUTHWESTTITIUU BASIN 219

FIGURE 14.5. The Horca del lnka

servatory can be safely assigned to the Late Ho-

rizon based on its association with nearby Inca- The Baiio del Inca
style stairs and walls. One trace of a wall, with The "Baiio del Inca" ("bath of the Inca") is lo-
some blocks still in place, cuts across the space cated in the former hacienda Kusijata, 2 km
directly in front of the Horca (Figure 14.6). Al- northeast of Copacabana. Noted and drawn by
though this wall could not have been very high Squier in 1877 (1877a:325), it is a large open cyl-
or it would have blocked the trajectory of the inder (the cavity measures 1.2 x 0.6 m) carved
sun's rays on the solstice, it probably served as a out of a single piece of rock, highly polished on
boundary separating sacred from profane space, the surface, and sunk into a stone-paved plat-
and may have indicated a social division as well, form. Two semicircular gaps in its edge served
such as that between officiators and the public to fill it with water from canals extending out
(Rivera Sundt 1984:98). from the bath.
 FIGURE14.6. Traces of a wall at the Horcn del Inka

Besides this Copacabana Idol, the Yun-

Copacclti is a carved-rock site on top of a steep, guyos had another which they called
rocky hill about 4 km south of Copacabana, first Copacati. The hill where it was located
described by Maks Portugal Zamora (1977:299, took its name from this same Idol, which
see also Rivera Sundt 1978:81). The bedrock is was on the way out of the town. It was
finely carved in flat stepped shelves ("seats" and made of stone, an evil figure completely
curled round with snakes; the people
"altars") at several points along the ridge-top in
would resort to it in times of drought,
unmistakable Cuzco-Inca style (Figure 14.7). On
asking it for water for their crops. Padre
the northwest flank of the ridge (a steep slope Almeyda, who had charge of the curacy
makes the main carvings almost inaccessible) before the missionaries of my sacred or-
are two additional worked areas first noted in der came into it, heard of this Idol and
this reconnaissance. One is a triangular "seat" had it brought to the town, and when it
carved in a rock far down on the side of the hill; was placed in the plaza in the presence
the other consists of two rectangular "seats" in of many people, a snake was seen to de-
the steep rock higher up the hill (Figure 14.8). As tach itself from the Idol and go around it.
will be shown, a pattern of carved rocks scat- The priest seeing this gave them to un-
tered away from the main ceremonial sector is derstand that it was the devil, and that
rather common for the small Inca sites described they should be ashamed to have held
here. No Late Horizon habitation site was found such a vile' creature for a god.. ..The
in the immediate area. master of the chapel, Don Ger6nimo
As at Intinkala, the Inca carvings at Copacati Carvacochachi, seventy-eight years of
were apparently associated with a pre-Inca idol. age, told us he had seen the snake uncoil
According to Ramos Gavildn: itself from the discarded Idol, which was
CHAPTER
14: INCA CEREMONIAL SITES IN THE SOUTHWEST TITICACA BASIN 221

FICLJRI-:14.7. Carved rocks at Copacati

FIGURE14.8. Carved rocks on the northwest flank of Copacati

 thrown in the lake, and the snake was the Inca road as well. The site could easily have
beaten and stoned to death [1988:19&197 been visited by pilgrims en route from Yunguyu
(1621:Bk.1 Ch. 32), author's translation]. to Copacabana, or alternatively it could have
Portugal Zamora (1977:301) notes correctly been visited as a short subsidiary pilgrimage
that the description of the idol corresponds to from Copacabana (like the Horca del Inca). The
Formative and Tiwanaku period stone mono- walking time from Copacabana to the top of Co-
liths. (A Formative date is most likely, consider- pacati is about one hour using the modern road.
ing the snake iconography.) At Copacati, as at Although it does not take long to reach Copa-
Intinkala, an elaborate, Cuzco-style Inca cut- cati, the trip does involve making a detour up a
stone site, probably connected with the island rather steep hill and knowing where the site is
cult in some way, seems to have featured an an- located-in other words, taking a miniature pil-
cient stone monolith still revered by the local grimage in itself. The Island of the Sun is not vis-
people of Yunguyu in the early Colonial period. ible from Copacati. Clearly, placing the site on
The Incas may have placed their carvings at Co- top of the hill, or near an already sacred site, was
pacati to take advantage of a preexisting huaca, more important to the Incas than making it
or they may have moved the huaca here, which highly visible and accessible.
fits with the pattern of Inca reuse of pre-Inca sa-
cred sites such as Pucara, Tiwanaku, and the Carved Stone a t C e r r o J u a n a
sanctuary itself on the Island of the Sun. Alter- On the mountain of Cerro Juana (Koana), lo-
natively, the idol's local worshippers may have cated southeast of Copacabana and northeast
moved it to this Inca cut-stone site. Either possi- from Copacati, is an unusual carved block,
bility suggests a striking degree of accommoda- which was also described by Portugal Zamora
tion between the two religious traditions. (1977:307,323, Fig. 16). The block is 4.3 x 3.3 m in
One interesting aspect of Rainos Gaviltin's size, and its top surface is covered with rectan-
account is that in both cases, the monoliths, gular depressions and canals. While the block
rather than the Inca carved rocks, were the main probably dates to the Late Horizon, it is stylisti-
focus of worship by local inhabitants in the early cally much less "Inca" than the other ceremonial
Colonial period (and the focus of destruction by sites on the Copacabana Peninsula, and displays
the Spanish clerics). Pre-Inca monoliths may very atypical features, including several short
have proved to have more enduring symbolic canals arranged in a radiating pattern. In view of
significance for the inhabitants of the southern its relatively inaccessible location and style, i t is
basin, although Ramos Gavilin's account may unlikely that this site was an "official" ritual sta-
simply reflect the Spanish concern over the wor- tion. It may have been a local imitation of Inca
ship of representative (especially anthropomor- ceremonial expression.
phic) idols, as opposed to the abstract, geometric
carvings of the Incas. Another intriguing element Playa C h a t u m a and Nearby Sites
of the account is that the idols were apparently Playa Chatuma is located on the flank of a low
worshipped by the nonelite. The Yunguyus held ridge at the lakeshore, about 10 km southeast of
them sacred, and in the case of the Copacabana Pomata. It consists of carvings on a stretch of
idol, these included Yunguyus too poor to make soft, sloping limestone exposed above the beach
valuable offerings." This post-Conquest reality and is about 400 m wide and 100 m from top to
may not have reflected Inca period conditions, of bottom. The rock is carved extensiveIy with thin
course, but it raises the question of for whom vertical channels and other carvings (Figure
these explicitly Inca-style stone carvings were in- 14.9). A network of more deeply carved channels
tended. The scenario of local nonelite worship- is located on the western edge of the site, toward
ping at Inca sites contrasts markedly with elite the top of the exposed bedrock. There is a short,
restriction at the Island of the Sun. shallowly carved "staircase" next to this network
Copacati is close to the modern road from of channels. This staircase, which is similar to the
Yunguyu to Copacabana, and probably was near Inca stairs at Kenko (Tres Ventanas) and at the
 CF-LWER SITES IN T H E SOUTHWEST TITICACA BASIN
14: INCA CEREMONL~L 223

FrcunE 14.9. Playa Chatumn

"Inca's Chair" at Bebedero, securely dates the The site of Ckackachipata, recorded by Stan-
site to the Late Horizon on stylistic grounds (Fig- ish et al. (1997:90),was probably the closest asso-
ure 14.10). The rock nearby displays sets of shal- ciated habitation area to the Playa Chatuma
low transverse cuts, a repetition of the stair motif blocks. Ckackachipata is found on the peninsula
too tiny for any human climber. The western por- just next to Playa Chatuma (Figure 14.12) and
tion of the site also has eight vertical channels has a Late Horizon component with fairly mod-
running straight down the rock face toward the est local ceramics (Stanish et al. 1997). The Island
shore, each one dotted with shallow hollows, of the Sun is visible from Ckackachipata, though
some converging or diverging on their course. not from Playa Chatuma itself.
On the east side of the site, beyond a natural Other nearby Lnca features may be related to
fault, plain channels without hollows predomi- Playa Chatuma (Figure 14.12). In the bay to the
nate. The most recognizable elements of Inca cer- east of the Ckackachipata Peninsula, a large,
emonial style-geometric planes and niches- square block of stone with a straight groove
are absent at Playa Chatuma. carved into one side rests in the water just off the
On the beach below are two identical stone beach. This is almost certainly from the Late Ho-
blocks, lying about 50 m apart, and finely carved rizon, but its relationship to Playa Chatuma can-
with an inu usual step-motif (Figure 14.11). An not be determined. Pukara Chatuma is a hill with
undecorated, but finely made block, about 1.0 x a Late Intermediate period and Late Horizon
0.5 m in dimension, is also found on the beach to presence, as evidenced by pirkara walls and an
the east. The exposed beach below the canals Inca-style square chulpa with fine Inca sherds,
yielded a total of only four diagnostic sherds, all respectively (Stanish et al. 1997:95).On top of the
dating to the Late Horizon. However, abundant hill, Stanish et al, noted several carved rocks
ground stones and grinding surfaces of hard (1997:95-96). These include an Inca-style carved
sandstone and granite are present. These were rock with a rectangular depression, a large, finely
polished to a surface suggestive of plant pro- shaped ashlar block near what may be a slab-cist
cessing. tomb, a rock with a single straight groove about
FlCL'llE 14.10. Close-up of Playa C h a t u m a

FIGURE 14.11. Carved rock on the beach at Playa

Chatuma
 CHAPTER 14: I N C A CEREMONIAL SITES IN T H E SOUTHWEST TITICACA
B AS IN 225

I* Carved rocks or bedrock

+I
F IG U RE 14.12. Playa Chatuma and surrounding area. After Stanish et al. 1997:86

10 cm long, and another rock carved with an un- water or liquid ritually. Canals and basins are
usual concentric circle design (Figure 14.13). This features of most of the Inca ceremonial sites in
last motif is characteristic of neither Inca nor lo- the study area.6 The forking canals at Playa Cha-
cal carving, but the rock may be Late Horizon in tuma (for instance, the network of canals at the
view of its location near the other rocks at this "stair") are particularly interesting because, like
site.5 These dispersed ceremonial features scat- other cases of Inca diverging charnels, they may
tered across the landscape near Playa Chatuma, have been used for divination. Forking-chamel
and the relatively disorganized, asymmetrical yacchas are found at some of the most important
layout of the channels (compare, for instance, Sa- Inca ceremonial carved-rock sites: Kenko (near
maipata in Bolivia, near the eastern edge of the Cuzco), Samaipata, Sayhuite, Ingapirca, Santa
Inca Empire), suggest that the site may not have Apolonia (Cajamarca), and Vilcashuamin (Van
been organized by a master plan. de Guchte 1990:146). The forking channels at
Hyslop (1984) was unable to find the Inca Playa Chatuma are far more crudely carved than
road south of Pomata, but it is unlikely that it those listed above, but may have fulfilled the .
passed closer to Playa Chatuma than the mod- same role.
ern road does because of the hilly topography of
this area and the swamp land just to the west. Playa Chinchin Jalave
The site is accessible by foot and requires only Playa Chinchin Jalave is located on the lake-
about a half-hour's walk to reach it from the shore north of San Bartolorn4 Hill, just west of
modern road leading to Yunguyu. It probably the Choquesuyo Peninsula, about 5 km north-
required no more effort to reach it in the Inca pe- east of the Inca (and modern) town of Juli. Al-
riod. The site could easily have been along a though the site is not on the Inca road surveyed
route from Pomata to Yunguyu and the Copaca- by Hyslop (1984), it is close enough to Juli to
bana Peninsula beyond. have been easily reachable by residents. The site
The canals at Playa Chatuma and Chinchin is on a thin strip of beach at the foot of a steep
Jalave (described below) relate these sites to nu- cliff, about 100 m from the lake edge. It consists
merous others across Tawantinsuyu that used of several boulders of chalky white limestone,
 FIGURE 14.13. Carved rocks at Pukara Chatuma

carved with short canals and basins, which are on the beach, which can be reached by a ravine
now heavily eroded (Figures 14.14 and 14.15). and a footpath, but is otherwise difficult to access.
One boulder displays seven channels, but most Playa Chinchin Jalave displays obvious sim-
have two or three. As at Playa Chatuma, the ilarities with Playa Chatuma, including the lake-
geometric shapes most typical of the Inca carv- side location of both sites, the carved channels
ing style are absent. These boulders are close to present at both sites, and the limestone used at
the level of the lake and would have been inun- both sites. Although the lack of ceramics and the
dated since they were carved, contributing to atypical carving style would make this site diffi-
their surface degradation. Not surprisingly, no cult to date in isolation, it can be reasonably
pottery is present in this area. The north end of placed in the Late Horizon based on its parallels
the Island of the Sun is visible from the site. to Playa Chatuma, its association with the
On a point just west of the beach on the cliff nearby Late Horizon site of Chinchin Jalave, and
above the lake is the small Late Horizon habita- its connection tp water ritual.
tion site of Chinchin Jalave, noted by Stanish et al. The carving here and at Playa Chatuma does
(1997:63). Plainwares abound, and there are sev- not follow the typical Inca stylistic canon evoked
eral collapsed circular structures that may be at other sites in the region, despite the fact that
fallen chulpas. Those who lived at this site may both sites are relatively close to the Island of the
have been able to control access to the boulders Sun and conveniently located for the pilgrim.
 CHAPTER14: INCA CEREMONIAL S ITE S IN T H E SOUTHWEST TITICACA
BASIN 227

14.14. Car\wi rocks .it I'laya Chinchin J a l w e

FIGUKI~

Unlike Intinkala or Copacati, these sites d o not

have landscapes that have been stamped with
clear statements of Inca power. They may have Two examples of Inca ceremonial rock carving
been fashioned without direct supervision from are found at the Bebedero sandstone outcrop,
Cuzco, either as a secondhand directive or a lo- which is located about 8 km north of Juli and
cally motivated imitation. just west of the modern road. Here the modern
road is probably very close to the Inca road and
CUT STONE NEAR may overlie it (Hyslop 1984:123). "The Inca's
Chair," first described by Squier in 1877 (1877a:
SILLUMOCCO-HUAQUINA 350)) is a set of carvings in classic Cuzco-Inca '
This isolated stone is located on the northeast style on a section of the outcrop near the road
flank of Mt. Sapacollo near Juli, not far from the (Figure 14.16). The carvings consist of several
important Upper Formative and Tiwanaku pe- shallow planes or "seats," a stairway leading to
riod site of Sillumocco-Huaquina (de la Vega, the top of the rock, and the thin vertical channel
Chapter 8, this volume). A branch of the Inca after which the outcrop is named (bebnirro, or
road surveyed by Hyslop (1984:123) runs near it. drinking-trough). The rock has been capped
It is a large (1.5 m wide) boulder that has been with a tower in recent times.
smoothed and carved in an abstract, gently The vertical channel on the front of the Inca's
curved shape. It is perhaps significant that this Chair tie this'site, along with others in the area,
rock is found on the east branch of the Inca road, to a typical Inca pattern of libation. The quality
on the lakeside branch, rather than on the more of work at this site and the similarity of its
direct branch into Juli to the west side of Mt. Sa- carving to that in the Cuzco heartland make the
paqollo. site stand out among more modest ceremonial
FIGLRE14.15. Carved stone at Playa Chinchin Jalave

FIGURE14.16. The "Inka's Chair" at Bebedero

 C U ~ P T E14:
R INCA C EREMONIAL SITES IN THE SOUTHWEST TITICACA
BASIN 229

F I G U R E 14.17. Altareni

installations in the arca. The location of the site impressive cerernoni'11 carving ivith more mod-
near the probable Inca road makes it highly est elaboration nearby.
likely that it was associatcd with a pilgrimage to While the carving style of the "Inca's
thc Island of the Sun 7 In other words, it is ex- Chair" is unmistakably Inca, the Altarani carv-
actly what we would expect to see in a pilgrim- ing is more difficult to date. The terraces and
age station symbolizing Inca power on the plain a t Bebedero yield artifacts a n d ceramics
provincial landscape. of all periods from Late Archaic to Inca (Stan-
Thc impressive carving of Altarani (Figure ish et al. 1997:61). Stanish and colleagues argue
14-17), about 2 km away, was first pubIished by that the Altarani carving is Late Horizon, based
Alberto Cuent,\s (1928) and more fully described on its association with the Inca's Chair (1977:
by Hyslop (1976:352) and Stanish et al. (1997:61). 62). Hyslop (1976) attributed the site to the Al-
It consists of a rock ocltcrop into which a 7.0-m tiplano (Late Intermediate) period, calling the
tall and 8.0-n~ wide vertical plane has been T-shaped carving a "chulpa facade." This hy-
carved. This ccntral section is ocltlined by two pothesis is difficcilt to support, given that no
grooves on the sides and an unfinished "lintel" comparable examples of Altiplano rock carving
on top. In the middle of it is carved a blind door- are known to exist. Hyslop's alternative sug-
way (1.9 x 1.1 m) in a rough T-shape. Two gestion that the carving is Tiwanaku, based on
smaller planes flank this central section, bring- the niche's T-shape (Hyslop 1977a:162), is pos-
ing the total width of the carving to 14 m. The sible. While T-shapes are not unknown in Inca
1998 reconnaissance found one previously un- carving and architecture, they are uncommon."
published feature at Altarani: a low rock outcrop The cupules nearby are analogous to scatters of
to the east (with a good view of the Altarani ccipules noted by Trimborn at known Inca
carving) displays two abstract designs of small carved rock sites such as Samaipata a n d Lacat-
holes, or cupules (Figure 14.18), similar to the ambo, near Cochabamba in Bolivia (Trimborn
cupules at Kenko (Tres Ventanas). As a t Copacati 1967:26). In sum, a Late Horizon date is most
and Playa Chatuma, Altarani shows a pattern of likely.
 13.18. Cupulcs at Altarani
FIGURE

Kc11ko (TITS\'en m l a s ) the level of the plains. The masonry is extl-emely

L.ocal rcsicients call the site Kenko. Here I add fine for the region and very regular. Indiviciual
the name "Tres Ventanas" to distinguish it from blocks vary little in size, ranging from 35 to 45
the site of Kenko near Cuzco? It was briefly de- cm across, and most are square or nearly square
scribed by M. Tschopik (1946:8) and Hyslop (Figure 14.21). The walkways are reached by
(1976:348). As one of the more impressive and passing through the fields behind the front
unusual Inca ceremonial sites in the province, it ridge. Here, a series of four rectangular depres-
deserves further attention. This large (4 to 6 ha) sions are canled into the bedrock. About 11 m
site is located about 1 km from the lakeshore, on north of this carving, a set of Inca steps carved in
the eastern edge of a range of hills east of the the rock leads through a gap in the ridge to its
plains between Acora and Ilave. The site in- northeast face, where the main walkway begins.
cludes a ceremonial sector, consisting of thin Another flight of steps is carved on the face of
walkways and niches constructed on a cliff face, the cliff at this point, but does not lead anywhere
and three distinct Inca period habitation sectors: (Figure 14.23). These steps could have been in-
one behind, a second in the fields in front, and a tended to connect with a higher walkway, which
third near the entrance to the walkways (Figures stops some 45 m from this stair. The walkways
14.19, 14.20, and 14.21). The cliff is part of a geo- are inaccessible except by this route.
logical formation consisting of a series of sheer There are four distinct walkways, the larg-
sandstone ridges running northwest to south- est being about 150 m long and 7 m wide at its
east, enfolding thin strips of steep terraced land widest point. This main walkway has three
between them. The cliff face holding the walk- raised sections or platforms, each progressively
ways is the most northeastern of these ridges increasing in length. The last platform is
and has an unimpeded view of the lake. reached by ascending a series of four small
The walkways (Figure 14.22) are formed and rises, each 0.5 to 1 m in height. Three large rect-
retained by walls of fine, Inca-style coursed ma- angular niches are carved into the rock face at
sonry clinging to the cliff face, about 10 m above various points along the walkways. One of
CWER 14: INCA CEREMONIAL SITES IN T H E SOUTHWEST TITICACA
BASIN 231

FIGURE
14.19. Kenko (Tres Ventanas)

F IGURE 14.20. Sketch map of Kenko (Tres Ventanas) v map and contour lines are
not to scale.)
 FIGURE
14.21. Kenko (Tres Ventnnas) ceremonial sector

them is in an incomplete stage of carving, and strips of land lying between ridges to the south
in addition, two groups of hollows on the rock of the site. Areas 2 and 3 feature a slab-cist tomb
face indicate the beginning stage of two more and a modest chulpa, respectively. Areas 2 and 3
unfinished niches (Figure 14.24). No ceramic or have a light scatter of mostly undecorated Late
lithic artifacts were found on the walkways. Horizon ceramics and possible Late Intermedi-
Locally produced ceramics from domestic ate period ceramics. Area 4 is located in the
occupations and stone agricultural tools are fields that lie below the cliff face to the north,
found on the surface of several sectors close to where a moderate scatter of Inca and early Colo-
the niches and walkways (Figure 14.20). Area 1 nial ceramics can be found in a 2- to 3-ha area
consists of about 1.5 ha of fields behind the ridge (Chokasuyu types as defined in Stanish et al.
and to the east of the ceremonial sector, just be- 1997:49; Pakajes Tardio in Albarracin-Jordan and
yond the stair leading to the walkways. This Mathews 1990). All four areas have stone agri-
area displays a dense scatter of fine Inca ceram- cultural tools, and these tools are particularly
ics. Areas 2 and 3 consist of the thin terraced common in Area 4.
 CHAPTER 14: INCA CEREMONIAL SITES IN THE SOUTHWEST
TITICACA BASIN 233

\
FIGURE 14.22. Masonry retaining the walkways at Kenko (Tres Ventanas)

F IGURE 14.23. Carved stairs on the cliff face at Kenko (Tres Ventanas)
 FIGURE14.24. Nichrs at Kenko (Tres Ventmas), in \,nrying stages of completion

Although Kenko is far more impressive and exists, ;l stone gateway is reported to have stood
more stylistically "Inca" than some of the other at the gap where the Inca stairs give access to the
sites listed here, Kenko, like the other sites, has a ceremonial sector.I0 It may have marked a point
pattern of modest subsidiary carvings dispersed at which access was restricted, as well as demar-
around the site. A natural gap in the rock ridge cating the border between sacred and profane
behind Area 3 allows access to a much wider space.
trough in the landscape. At this gap, on the In the overaIl pattern of Inca ceremonial
south side of the ridge, are ten small depres- sites in the region, Kenko uniquely juxtaposes a
sions, or cupules, pecked in a pattern on the rock significant labor investment and Inca stylistic
(Figure 14.25). On the opposite side of the canons with a very remote location. It currently
trough, against the ridge that rises to the south, a requires a detour of about 15 km from the mod-
thin, straight groove or canal carved into the ern road running from Acora to Ilave, although
rock is almost certainly Late Horizon in date. It the location of the corresponding section of the
might have carried water from a spring uphill to Inca road is difficult to ascertain." Why was this
the east. Finally, a shallow semicircular "seat" location chosen? The rock ridges on which the
has been carved in a boulder about 2 km north- site was constructed are spectacular and un-
west of Kenko, at the base of the cliffs facing the usual, and geologically similar to the Bebedero
lake (Figure 14.26). outcrop, which constitutes another sacred site.
It may be significant that the main ceremo- This may explain why the area, if not the site it-
nial sector has only one entrance: the stairway self, is still ritually important to local residents.
from Area 1. Decorated fineware is common in Although the Island of the Sun is not visible
Area 1. It is possible that this was an elite area from the cliff at Kenko, the site does have a mag-
and that access to the site was restricted, or at nificent view of the lake from a fairly easy-to-
least monitored, by the elite. Though it no longer reach setting. Given that many of the ceremonial
CHAPTER 14: INCA CEREMONIAL SITES IN THE SOUTHWEST TITICACA B ASIN 235

FIGURE14.25. Cupules at Kenko (Tres Ventanas)

FIGURE 14.26. Cawed stone to the west of Kenko

sites in the area are on the lakeside, this may ated structure had windows (Stanish 2001b).
have been a factor in the decision to build at The structure was therefore not an ushnu, or cer-
Kenko. Perhaps more importantly, the cliff face emonial platform.
and any activities that may have taken place on Marion Tschopik excavated the structure in
it are clearly visible from the fields below. Kenko the 1940s, finding Colonial era fieldstone walls,
is the perfect location for ritual performed pub- ceramics, and faunal remains, but no prehistoric
licly, but with restricted participa tion-a pattern stratigraphy (Hyslop 1984:129-130). Hyslop re-
identical to what the historical documents ii-tdi- lates that she was told of another structure
cate occurred at the Island of the Sun. called the Kurinuyu to the east of the Inca Uyu,
but she did not see it, and there is no observable
The Inca Uyu at Chucuito trace of this structure today. The upper row of
Late Horizon Chucuito is a very large, intrusive stones at the Inca Uyu were found and replaccd
site under and to the east of the present-day by Oroinpelio Vidal in the 1960s)who continued
town of Chucuito (Hyslop 1976:415; Stanish excxra tions at the site.
2001b; M. Tschopik 1946). I t is known from eth- Although Chucuito was probably a rest stop
nohistoric accounts to have been the cabecera, or on the pilgrimage to the Island of the Sun from
administrative center, of the Lupacas. It was also Cuzco, i t would be premature to assume that the
a tambo on the Inca road from Cuzco (Diez de Inca Uyufs primary function was tied to the pil-
San Miguel 1964 [1567]; Vaca d e Castro 1908 grimage. Its location in the center of the Lupaca
[1543]:437) and, at around 80 ha in size, was the cabecera of Chucuito and its quasi-Inca stone-
largest known Inca settlement in the Titicaca Ba- work suggest that i t is the construction of local
sin (Stanish 2003). lords unsiipervised by architects from Cuzco. Its
An intriguing structure in the plaza, known role in ritual may have had more to do with the
as the Inca Uyu, probably fulfilled a ceremonial expression of the authority of the Lupaca kura-
function. The Inca Uyu is a walled enclosure kas than with Inca-Lupaca relations.
measuring about 10 x 10 ni and consisting of two
courses of massive stonework (Figure 14.27). Al- SI'TESOUTSI~)~',
T H E SURVEYAREA
though the masonry was clearly modeled on
Cuzco-style polygonal stonework, curved rather Hatuncolla
than right-angle joinings betray it as a poor imi- Hatuncolla was the provincial administrative
tation.I2 The shape of other stones uncovered in center corresponding to the Colla ethnic group,
the plaza indicate that either this or an associ- just as Chucuito was the administrative center

FIGURE14.27. The Inka Uyu at Chucuito

 CHAPTER 14: INCA CEREMONIAL SITES IN T H E S O UTHWEST TITICACA
BASIN 237

F I G U RE 14.28. The Pachatata sunken court on Amantani Island. The upper wall is a recent addition.

for the Lupacas. As at Chucuito, it is likely there The date of the ceremonial constructions at
was at one time an Inca religious or ceremonial Amantani is difficult to determine. Although
center of some kind at Hatuncolla. Cieza d e rectangular sunken courts are common to For-
Le6n (1984:361 [1553:Ch. 1021) mentions a sun mative and Tiwanaku period ceremonial sites,
temple at Hatuncolla, among other Inca edifices. Pacha ta tats a typical fieldstone masonry, triple
No whole structure remains, but Julien (1983:90) stair, and noncardinal orientation make assign-
suggests that the fine Cuzco masonry reused in ing it to the Tiwanaku period with certainty dif-
more recent buildings-masonry including one ficult (Stanish 2003). The inclusion of two large
whole doorway-may be from this temple. (uncarved) boulders in the circular temple of Pa-
chamama (see Spahni 1971:222, pl. 3) suggests
Amantani Island an Inca period date. There is little to no surface
Amantani Island's two major ceremonial con- Tiwanaku pottery on the island, although Niles
structions, Pachamama and Pachatata, are prob- (1987b, 1988) reports some pottery looted from a
lematic, but the site is too important to leave out grave on the island, pottery that included a frag:
of this discussion. Both constructions are sunken ment of an incense-burner that is almost cer-
courts on low mountains. Pachatata (also called tainly Tiwanaku. Formative pottery, including
Aylicancha) is rectangular and Pachamama is Pucara polychrome incised, is apparent on the
circular, making it very unusual within the Titi- surface, though in small quantities (see also Kid-
caca Basin (Figure 14.28).'"0th have been re- der 1943:16; Spahni 1971:15).On the other hand,
stored and modified to an unknown extent in Inca ceramics litter the island, and the terraces
recent times, and continue to be used for rituals near the ceremonial structures have dense scat-
and offerings (Niles 1987b, 1988; Spahni 1971; ters of fine Inca ware. The Pachatata sunken
see also Kidder 1943:116 Stanish 2003:188, and court may therefore be a Formative or Tiwanaku
Viisquez 1940). The courts appear to have been court restored and modified in Inca times. Al-
in use in the early Colonial period, for Martin d e though such a construction would be unusual, it
Murila reports a famous huaca at Amantani would have fit with a program of Tiwanaku sty-
(Murfia 1946:21&7 [1590:Bk. 3 Ch. 211). listic references apparent in the sanctuaries on
the Islands of the Sun and Moon. Jean-Christian and rather than being placed for inaxirnurn visi-
Spalmi (1971:219) also reports carved seats or bility and use in the Late Horizon, their place-
shelves on the soutliwest shore of the island (fac- ment may have been intended as architectural
ing the peninsul'i of Capachica) that may possi- statements of Inca control over formerly impor-
bly be Inca period in date. If the cerenionial tant political centers (Arkush 2001).
structures and carvings at Amantani were in use
in tlie Inca period, the inaccessibility of the is- Vilcanota a n d Beyond
land makcs it very unlikcly that t h y were di- At the La Iiaya Pass, tlie entrance to the Titicaca
rcctly related to the pilgrimage. Basin fsom tlie Cuzco region was marked by the
Inca temple and tainbo of Vilcanota (lieinhard
I'ucai a 1995). Placed as it was clirectly on the road, this
While I'ucara is far to the north of the study area, facility may well have been designed as a pil-
it deserves to be mentioned for its ceremonial grimage station for high-ranking travelers. To
role in the Late Horizon. A11 important Forma- the north of the pass, another series of shrincs
tive center, it was restored and modified in Inca both large and small traced the route from
times. A tcrrace wall was rcmodclcd ivith trape- Cuzco. This series includes Raqchi, Urcos, Ti-
~ o i d a niclics,
l ancl 1' stairway ivith finely dressed pon, and huacas along the crqric lines of Colla-
corner stones \vas built into Pucara walls suyu, the southeast quarter of the empire (Bauer
(1Ylieeler and hlujica 1981:jS-9). Fine Inca pot- 1998).
tery of both local and Cuzco manufacture can be
found at the site (M'lieelcr and Mujica 1981:58-
70), and Pucara is listed as a tainbo in early Span-
ish colonial documt.nts (Guaman Poma cie Ayala The proliferation of small cci-cnionial sitcs along
1980 [1633]:10O(~[1C)91/1101]; Vaca de Castro the southwest shores of Lake Titicaca-as well
1908 [lS43]:437) P~~cai-a appcars in Inca mytliol- as thc justly more famous Island of the Sun sanc-
ogy as a sitc at Iilcli Virac'ocha, in 111s voyage of tuai-y-amounteel to something of a religious
crca tion, turned clisobcclicnt subjects to stone or florescence in the Late Horizon. While the region
called down a rain of fire upon them (Molina of does not have nearly the density of cercmonial
Cuzco 1947 [1584]):26). The association of Pu- sites as the area near Cuzco, the region displays
cara with Viracocha's journey outward from the a surprising variety of small sites available for
sacred center of Titicaca makes tlie site a good ritual. These are mostly carved rock sites, and
candidate for a ceremonial station on the pil- they clearly refer to Inca, rather than indigenous,
grimage, which traced Viracoch'~'~ course in re- ceremonial styles, yet they are highly varied in
verse back to the Islands of the Sun and Moon. location, size, level of labor investment, and
technical skill.
Other Shrines in the Northern Titicaca The most obvious pattern that emerges from
Basin this group of subsidiary sites is the very lack of a
In a more recent survey of Colla fortified sites pattern. Out of fifteen sites,'%ix would have re-
(pukaras) in the northwest basin, I found a num- quired a minor detour from the probable Inca
ber of modest Inca administrative or ceremonial road, and four sites (Cerro Juana, Kenko and the
structures located on hilltops that had formerly carved stone to the west of it, and Amantani Is-
been used by the Collas as defended settle- land) would have required a major detour. Nine
ments. While these data will be described else- sites are on the lakeshore or within view of the
where, it is worth noting that at least one of lake, even though the Island of the Sun is not
these Inca complexes displayed unusually elab- visible from most of them. However, the pattern
orate architecture, including double-jambed may simply reflect the predominance of lakeside
niches (which are very rare in the Titicaca Basin), settlement in the Late Horizon. A cluster of
and probably had ceremonial functions. This Cuzco-style sites around Copacabana contrasts
and similar Inca structures were located in rela- with sites outside the Copacabana Peninsula,
tively inaccessible places far from the Inca roads, which vary wideIy in style. Size and labor in-
 CHAPTER 14: I NCA CEREMONLAL BASIN
SITES I N THE SOUTHWEST TITICACA 239

vestment do not obviously correlate with Cuzco- area is perhaps notable for the lack of an overall
style sites or decorated pottery. Several sites "master plan." Unfinished features and addi-
have a "dispersed" site plan including minor tions at some sites imply that there may have
elaborations or additions, and two (Altarani and been an ongoing process of site construction and
Kenko) show clearly unfinished carving. Two of modification, rather than a single monumental
the most stylistically "authentic" Inca religious building program. I suggest that these sites were
sites, Intinkala-Orcohawira and Copaca ti, are as- not all conceived and constructed at one time by
sociated with indigenous and possibly nonelite one group. They were not part of an imperial
huacas. The Inca use of Amantani and Pucara program designed to stamp the landscape with
also took advantage of preexisting ceremonial symbols of Inca power in the most visible and
structures that may well have held religious sig- accessible way.
nificance for local inhabitants in the LIP. Mean- The second scenario, in which Inca religious
while, in the Late Horizon, the massive chulpa policy allowed input from local traditions and
cemeteries probably provided the main ceremo- participation by local actors, is best supported
nial sites that were not associated with the Inca by the reconnaissance data. Most telling in this
state in the minds of local residents. regard are those sites that appear to have been
While it is difficult to determine exactly what placed to take advantage of pre-Inca huacas, in-
was going on in the region in the Late Horizon to dicating a mutual accommodation between Inca
produce this heterogeneous mix of sites, compar- and local religious traditions. These instances of
ison with the expectations of the models of Inca reuse fit with the revival of a Tiwanaku sacred
policy outlined above can rule out some possibil- center on the Island of the Sun and the reuse of
ities. Tiwanaku itself as a sacred site (Vranich et al.
The first model, in lvhich Inca administra- 2002). They also fit \\.ith the lnca reuse of other
tors used Cuzco-style ceremonial sites to mark sacred sites such as Pachacamac. However, re-
the region as Inca territory and impose imperial use of sacred sites on the small scale represented
ideology on the Aymara-speaking peoples with- here is surprising, and points to finer-grained re-
out accommodation, is untenable in the face of ligious policies that integrated smaller, less elite
the data. The styles, locations, and overall pat- communities and their sites of worship into the
tern of the sites all argue against this scenario. Inca cosmology.
Several sites feature a mix of styles. Playa Cha- The location and style of these sites also in-
tuma, for instance, has a Cuzco-Inca style "stair" dicate a surprising degree of local participation
carving, canals which refer to but d o not repro- in their construction and use. Four sites (Cerro
duce Cuzco-Inca canons, and the nearby carved Juana, Playa Chatuma, Playa Chinchin Jalave,
rocks on Pukara Chatuma, whose styles range and the lnca Uyu) are so stylistically atypical
from typical Inca to highly unusual and innova- that they appear to have been constructed as lo-
tive. Kenko (Tres Ventanas) displays Cuzco-Inca cal imitations, or at least fashioned without Inca.
style stairs, masonry, and niches, but an innova- design or supervision. It is interesting that such
tive site plan. Furthermore, the locations of the modest, unsupervised sites had a place in the
ceremonial sites are not what would be expected pattern of lakeside religious installations. Stylis-
from a propagandistic building program. Sev- tic innovation and elaboration at certain Inca-
eral sites, such as Playa Chinchin Jalave and style sites also point to a construction process
even Copacati, are not very visible or accessible. that may have been open to local input. Clearly,
Many are not close to areas of dense population, there was considerable room for innovation and
and in consequence are ill-suited as vehicles for local participation in the fashioning and use of
the dissemination of imperial ideology to local Inca-style ceremonial sites, suggesting a greater
people. Kenko, in particular, has only a small as- role for Aymara residents than was permitted,
sociated habitation sector. Furthermore, there according to the ethnohistorical documents, at
are no large towns nearby comparable to the ca- the island sanctuary. Although access may have
beceras on the Urcosuyu road. In fact, the Late been restricted without leaving visible traces on
Horizon pattern of ceremonial sites in the study the ground, we may guess that small, remote
sites such as Cerro Juana and the carved rock religious autonomy. Meanwhile, their participa-
west of Kenko were probably not visited prima- tion in this religious framework helped to rein-
rily by the elite. force conceptions of social hierarchy in which
The sites described here constitute a marked the doininance of the Incas was naturalized. The
contrast with the Island of the Sun. Tight controls cosmology of the Incas readily adopted local be-
on access to the island sanctuary served to insti- liefs, embraced locally sacred places, and accom-
tutionalize a highly stratified social hierarchy modated local innovation, while retaining its
one that placed Incas at the top. In the southwest essential tenet, the divine solar origin of the
basin outside of the island, there was apparently conquering Incas. A recursive relationship for
a far looser control m7erthe sacred. Here, a range ideological influence resulted. Inca ideology, re-
of larger and smaller sites, sites that appear more ligious practices, and sacred sites were continu-
and less "lnca," and more and less controlled ally altered and reshaped by 11011-Incas and the
sites extended the cline of the sacred achieved by nonelite, as well as by their rulers.
the island sanctuary, allowing the incorporation Inca religious accommodation contrasts
and participation of small local comn~unitiesin sharply with the forcible and exclusive imposi-
the rituals and beliefs of the Late Horizon. tion oi Spanish Catholicism that was to succeed
This picture oi Inca religious policy exemph- it (MacCormack 1991). Was there a downside to
fies an interesting gradation of control over ide- an imperial policy of toleration of existing local
ology and religious sites from the center to the or regional huacas? While Inca religious accom-
periphery. The spectrum of ideological control modation probably helped to reconcile subject
bears similarities to the continuum from territo- populations to Inca rule, i t may also have weak-
rial (direct) to hegemonic (indirect) political and ened tlic empire by allowing subjects to continue
economic control argued for the Inca by D'AI- a strong tradition of identification with smaller
troy (1992). Inca political control was, of course, ethnic or regional groups. By the LIP, Andean
centered , ~ tCuzco, while major c~drninistrati\e ethnic consciousness was anchored in huacas
centcrs in the provinces such as Huanuco and p c ~ r i i l t ~(mytlwlc>giz~d
s ancestors and origin
Pampa served as nodes in the political hierarch\^. places fixed in the landscape), providing a con-
Outside of these centers and away from the ma- crete and inalienable sense of ethnicity or group-
jor roads and forts (especially on the coast), lnca hood. There is evidence that the Inca rulers were
poIitical control was looser and more flexible, attempting to weaken regional/ethnic identifica-
accommodating local hierarchies and sometimes tion with many of their other policies. These poli-
leaving few traces on the landscape (D'Altroy ties, enumerated by Rowe (1982), include the
1992; Hyslop 1990; Morris and Thompson 1985). imposition of forcible resettlement or temporary
Inca ideological power may well have followed labor service, the standardization of the arts and
a similar pattern of a highly controlled center technologies, and the spread of Quechua and
(Cuzco), subsidiary centers in the provinces some elements of Inca religion. Yet the Incas
(e.g., the Island of the Sun), and a hinterland in themselves arose out of a traditional Andean con-
which significant compromises were made with ceptualization of self and grouphood and partici-
local actors and local tradition. The result was a pated fully in it, as evidenced by their own
flexible approach to ideological governance, one landscape-based sense of ethnicity (involving de-
that would have been more economical to the scent from a handful of ancestors originating at
rulers while being more palatable to the ruled. Pacariqtambo; see Urton 1990). They could
A balance between local and imperial needs hardly have done otherwise; but this existing
in religious practice could only have been ideological base may have hindered rather than
achieved if it provided acceptable benefits to helped them in-the task of consolidation and uni-
both sides. Aymara people were given, or de- fication. The historical documents show that at
manded, license to participate in a prestigious the time of Spanish contact the empire was still
religion whose value was defined by the exclu- thought of as a collection of ethnic units, an-
sive, ancient island sanctuary, all the while re- chored to place and genealogy by their respective
taining their local huacas and a degree of pacarinas and local huacas. When these groups
 CHAPTER14: INCACEREMONIAL SITES IN THE SOUTHWEST TITICACA
BASIN 241

splintered off, as they did in the Huascar- Cuzco's plaza (it is apparently not the Kenko
Atahualpa civil war and at the time of the Span- now known to us): " m a piedra que tiene
ish conquest, they did so as whole units. The labrados dos circulos concentricos" a. M.
parts of which the body politic was composed, Blanco 1974 [1834]:181, cited in Van d e Guchte
tenuously connected by a few empire-wide insti- 1990:150, note 14).
tutions, came apart easily as soon as the head fell. 6. Liquid offerings of water, chicha,or blood, con-
stituted an important feature of worship at the
lsland of the Sun, according to contact period
texts (Cobo 1990:97 [1653:Bk. 13 Ch. 181;
1. Bauer and Stanish (2001) and Stanish and Ramos Gavilin 1988:116, 149 [1621:Bk. 1 Chs.
Bauer (2004) provide a detailed treatment of 17, 24]), and ritual canals on the islands attest
the archaeology and history of the islands, and to them (Bandelier 1910:198, 221; Bauer and
Adolph Bandelier (1910), the first archaeologist Stanish 2001).
to systematically work on the islands, remains 7. Squier (1877a:350) gives the "tradition" that
an excellent source. Also, Ephraim Squier the Inca's Chair "was the 'resting-place of the
(1877a) and other nineteenth-century natural- Inca,' in his journeys or pilgrimages, where the
ists wrote a number of useful descriptions of people came to do him homage, bringing
the remains on the Islands of the Sun and chicha for his delectation and that of his atten-
Moon. dants." Dubious as his ethnohistoric informa-
2. The Inca road from Cuzco split into two at the tion may be, it is worth noting that the site was
northern end of Lake Titicaca at Ayaviri, with specifically connected with the pilgrimages of
one road tracing each side of the lake. They the Inca even at this late date. Alberto Cuentas
rejoined to the south at Caracollo. Access to (1928) noted that some informants believed
Copacabana and the lsland of the Sun was this was where the Incas and Lupacas first cel-
from the western or Urcosuyu branch. As the ebrated their alliance by libating chicha. Thus,
Urcosuyu road continued south, a side road in its name and in legends surrounding it,
passed through Yunguyu and on to Copaca- Bebedero was strongly associated with Incas,
bana (Hyslop 1984; see also Cieza de Le6n libation, and chicha.
1984:361,364-5 [1553:Bk.1 Chs. 102,1041; Gua- 8. Examples are found on the southeast side of
man Poma de Ayah (1980 [1613]; Lizarraga Suchuna (Rodadero) Hill by Sacsayhuaman
1987:185 [1605:Bk.1 Ch. 851, and Vaca de Cas- and north of Suchuna near a large circular
tro (1908 [1543]). depression, at the carved-rock complex known
3. Such was the opinion of Fray Rafael Sanz as Kusilluchayoc or the "Templo de 10s
(1886), compiler of an early edition of Ran~os Monos" near Kenko (Cuzco), and in Tiwan-
Gavikin, who associated the idol of Copa- aku-style niches at the cave site of Choque-
cabana with the "restos de graderias" at Copa- quilla. It is worth noting also that the Incas.
cabana, as he did the idol of Copacati with the deliberately used Tiwanaku stylistic canons in
Inca site of the same name. the architecture of the temple on the lsland of
4. Ramos Gavilsn recounts that the aboriginal the Moon nearby, as well as in ceramics used
inhabitants of Copacabana and the Island of on the Island of the Sun (Julien 1993).
the Sun were relocated to Yunguyo by the 9. In both Quechua and Aymara, kenko means
Incas, and replaced by mitima colonists who twisting, sinuous, or zigzag (Bertonio 1956
enjoyed a special prestige as attendants to the [1612]:295). At Kenko near Cuzco, the term
temples (1988:84 [1621:Bk. 1 Ch. 121). He uses refers to the zigzag paccha. Here, it may refer
the term Ymguyo to refer to non-mitima inhab- to the zigzag walkways on the cliffside.
itants of the Copacabana Peninsula. 10. A local informant gives this report, which
5. One possible analogy is found in the descrip- matches M. Tschopik's (1946:s) observation:
tion of priest Jos6 Mario Blanco in 1834 of a "A well dressed stone wall and doorway have
rock he calls "Qqenco" or "Ccasana," near been erected across a break in the escarpment."
11. Hyslop (1984) found no trace of the Inca road 13. One circular structure a t Sillustani may be
between Kacha Kacha B, a chulpa site 3 km \.iewed as somewhat similar to the one at
southeast of Acora, and the town of Juli. The Pachamama.
location of the Inca town of Ilave, listed as a 14 The difficulty of determining what to call a dis-
tambo in early colonial records (Vaca d e Castro tinct "site" is considerable, since so many of the
1908 [1543]), is similarly ambiguous (Hyslop sites described here show a pattern of small,
19S4:123; Stmish 2003). widely dispersed cerenionial features. For the
12. One unusual stylistic feature of the masonry at p ~ ~ r p o s of
c s this tally, features within an arbi-
tlie Inca Ulru, a pattern of "tails" from the trary mdius of 1 km are considered a single rit-
upper coursc that fit into the cur\wl join of L I ~ area
I or "site." For instance, Intinkala and
blocks on the lower course, has an intriguing Orcohawira are grouped as one site, while the
pirallel at Ollantaytambo. Although much Icss "Inca's Chair" and Altarani are considered sep-
accentirated, masonry on tlie Wall of the Unfin- arate. Sitcs on the Islands of the Sun and Moon
ished Gate and o n some terrace walls at Ollan- are not included in thc tally; ncithcr is tlie Inca
t,i!,talnbo has a similar tailcd form, termcd occupation at Tiwanaku. Thc statistical liniita-
"scu t i form" masonry by Harthe-Tcrri. (1965: tions of this sample of fifteen are obvious.
158; scc also Protzen 1993:82, Figure 3.13). The fiftccn sites: considered here arc
There may be a possible connection between lntinkala and Orcohawira; the Baiio dcl Inca;
these two isolatcd instances, for according to the Horca dcl Inca; Copacati; the Cerro Juana
Sarmiento rle Gamboa, Colla captives werc stone; Playa Chatuma and associated features;
brought to build Ollantaytanibo by Pachacuti I'laya Chinchin Jalavc; the Sillumocco stone;
(Sarmicnto de Gamboa 1988:112 [1572:Ch. 401). Altarani; the "Inca's Chair"; Kenko; the carved
A Coll'i ayllu at Ollantaytambo is also attcc.ted stone west of Kcnko; the Inca Uyu at Chuci~ito;
in 1' land dispute document from 1560 (Protzen Alnantani Island's sunken courts; and Inca
1903:269, notc 1). additions at Pucara.