Cook Huari Figurines

THE STONE ANCESTORS: IDIOMS OF IMPERIAL
A lsl lRE AND RANK AMONG HUARI FIGURINES
AnitaG. Cook
Two caches wererecoveredat Pikillacta,the largestHuari state installation(A.D. 550-1000) in the southern
Andeanhighlands;each contained40 richlygarbedvotiveturquoisefzgurines.Thefzgurinesare analyzedin terms
of their production,use, and depositionas well as their overall morphology.To the extent possible, the rank
associatedwith the costumes wornby eachfzgureis also considered.Referenceis made to Inca appareland its
potentialfor interpretingHuari officialgarments.Because the number40 also held special importancein Inca
state organizationas an administrativeunit or division,the Inca example providesconceptsof administration
vital to the interpretationof the fzgurines.A more unusual source-origin myths associated with the Chimor
Kingdom-supportsthe relationbetweenturquoisefzgurines and ancestorworship.I arguethat the stonefzgurines
embody qualities and convey conceptsthat are central to Andean political administration,and that they are
intimatelytied into the webof ancestralcultsthroughwhichkinship,hierarchy,and inheritanceweredetermined.
Pikillactaes la mas grandeinstalacionestatal huari(A.D.550-1000) ubicadaen la sierrasurperuana.En este
sitio se encontrarondos depositosde ofrenda,cada uno conteniendo40figurillas humanastalladas en turquesa,
con prendasde vestirgrabadas.El analisis de estas ofrendasse hace en terminosde los contextosde produccion,
uso y deposicionde lasfigurillas,de su morfologla,y hasta dondees posible,del rango.asociadocon las prendas
representadasen cadafigurina. No obstanteque se mencionelas prendasincas y que estas tienen buenasposi-
bilidadespara la interpretacionde las prendasoficialeshuari;el numero40 tambienhabrlatenidouna importancia
especial en la organizacionestatal inca como una unidad de division administrativa.El ejemplo inca provee
conceptosde administracioncrlticospara la interpretacionde lasfzgurillas;mientrasque un recursopoco comun-
mitos de origendel estadochimor-apoya la relacionentrefzgurillasde turquesay los cultosancestrales.Aqulse
argumentaqueestasfigurillascontienencualidadesy proyectanconceptoscentralessobrela administracionandina,
y serlan estasfzgurillaslas que estan Intimamenteligadas al complejosistema de cultos ancestrales,mediantes
los cuales se sustentaronlas relacionesde parentesco,diferenciacionsocial y la herencia.
In this paper I address the relations between political office, ritual legitimation, and aspects of
specializedproductionin an early state that lacked any known form of writing.The site of Huari
(Figure 1), an urbancenterwhose architecturalcore alone is estimated to have covered 1000-1500
ha (Isbell 1984, 1986) of agglutinatedarchitectureis the centerof one ofthe firstknown expansionist
states in the centralAndes (Isbell and Schreiber1978; Lumbreras1960, 1980; Menzel 1964, 1969;
Rowe et al. 1950). The form of administrationis not clearly understood,but a bureaucraticstate
model has been employed to explain Huari organizationin the sixth and seventh centuries A.D.
(Isbell and Schreiber 1978; Schreiber 1992). Explosive spatial expansion into areas beyond the
AyacuchoValley followed upon the urbangrowthof Huari which, as many have argued,probably
involved militaristicinvasions. These forcefulincursions into foreign territorywere accompanied
by the spread of an organizedform of state religion (Isbell and Cook 1987; Isbell and McEwan
1991; Lumbreras1960, 1974; Menzel 1964, 1969).
At present,Huariprovincialadministrationhas been demonstratedmost successfullyin consistent
and repeatedarchitecturalforms (Isbell and McEwan 1991), in settlement-patternchanges (Cook
1989, 1990; Isbell and Schreiber1978; McEwan 1984; Schreiber1992), and in shifts in local econ-
omies (Browman 1981) that have been found in areas that came under Huari control. Pottery
exhibitinga consistentrepertoireof designsis commonly associatedwith Huarisites in the Ayacucho
Basin and has served as a principal indicator of Huari presence or foreign imposition at these
Anita G. Cook, Departmentof Anthropology,The Catholic Universityof America, Washington,DC 20064
Latin American Antiquity, 3(4), 1992, pp. 341-364.

Copyright t 1992 by the Society for American Archaeology
341
This content downloaded from 129.252.86.83 on Wed, 12 Mar 2014 20:37:11 PM

All use subject to JSTOR Terms and Conditions
342 LATIN AMERICANANTIQUITY [Vol. 3, No. 4, 1992
Figure 1. Map of Peru showinglocationsof sites discussedin the text.
locations. Somewhat surprisingly,ceramics and architectureremain the principal archaeological

identifiersof the Huari polity, despite an abundanceof other rich sourcesof materialculture.
In the Andean region, where no form of writingwas used in the precolonialperiod, it is widely
acceptedthat mnemonicsplayedan essentialrole in the conservationand distributionof information
during episodes of state and empire administration.Visual images assumed greaterimportance
under these conditions, constituting and reconstitutingessential multivalent symbols that served
multiplepurposesin earlystate formation.I explorethese issues and situatethem withinthe specifics
of Andean cultureby focusing on two votive offeringsof turquoisel figurinesrecoveredin 1927 at
Pikillacta,near Cuzco, the largestHuari provincial center in the southernhighlands(Figure 1).
Three-dimensionalstone figurinesin human form have been found on the surfaceof Huari sites,
buried in caches, and in gravelots(CuestaDomingo 1985:165-169; Engel 1976:169;International
Congressof Americanists 1935; Jones 1964:Figures27-30; Larco Hoyle 1966:Figure123; Menzel
1964:61-62, 1969:51-52; Ramos and Blasco 1977; Ravines 1970:502; Valcarcel 1933). Buried
offieringsat ceremonial centers or within ritual spaces characterizeboth Inca and Huari relations
with the spiritual world (Anders l 990a; Cook 1983, 1987; Isbell and Cook 1987; Menzel 1964,
1969; Ravines 1969, 1977). The only proveniencedcaches of human stone figurines(Figure2) that
date to the Middle Horizon are those found at the site of Pikillacta(Cook 1985a, 1986).
Duringthe Middle Horizon vessels in numerousceramicoffieringsdepict a seriesof staffiedfrontal

STONE ANCESTORS
THE 343
Cook]
Figure2. Examplesof the Pikillacta figurinesfromthe Madridand Cuzcocollections:(a) CuzcoCollection

(43 mm);(b) Madrid Collection(52 mm);(c) Cuzco Collection(43 mm);(d) Madrid Collection(42 mm)*(e)
CuzcoCollection(41 mm);(t) MadridCollection(47 mm);(g) CuzcoCollection(40 mm)-(h) MadridCollection
(38 mm).Three pairs are illustrated:c and d, e andf, and g and h.
or profileanthropomorphicfiguresin relativelystandardizedposes. Parallelsare drawnfrequently

betweenthe representationalarton BolivianTiwanakustone sculptureandthe Huarifiguresdepicted
on pottery.In Huari ceramicofferings,sacredbeings on votive pottery (see Cook 1985b:Figures2
and 3, 1987:27, 30-32; Isbell and Cook 1987:21-24, 32; Menzel 1977:Figures62, 66, and 67) are
visibly associatedwith humanelite captives(Cook 1985b:Figure2, 1987:Figure32; Isbelland Cook
1987:30; Menzel 1977:Figures62, 66, and 67) wearing ear spools, elaborate collars, hats, and
plumage,with arms bound with rope behind their backs. These human captives are appendedto
thebaseof staffsheldby largerfrontalandprofilefigures.In this context,staffedfiguresarerepresented
with miniaturehumans that constitute one of the earliest examples of this association duringthe
MiddleHorizon.These cachesof offeringvessels canbe datedby stylisticseriationto MiddleHorizon
1 (or roughlyA.D. 550-700) and offera temporalframeworkfor the firstappearanceof humanson
Middle Horizon objects.2
I arguethat the Pikillactafigurinesaregarbedin robesthat expresstheirrankand convey concepts
that arecentralto earlyAndeanpoliticaladministration.As explainedbelow, a strongrelationexists

3, No. 4, 1992
[Vol.
AMERICAN ANTIQUITY
LATIN
344
context of these figurinesthat bears upon

amongthe image, the material, and the archaeologicalcults duringthe Middle Horizon (ca. A.D.
mortuarypracticesand the existence of ancestral
Huari Inca ancestorcults, and their survival into
550-1000).Thereis continuityof such practicesin royal sacred
betweenlos gentiles(the ancestors)and the
times is evident from mythic associations
colonial been worshipped over the centuries.
which have
into which they were reputedlytransformed, use, and deposition, in addition to attributes
stones
Inthis analysis, the contexts of figurineproduction, achieve
by the figurineswithin the Huari state. To
attire,provide insights into the role played
of present a descriptionand comparison
understanding ofthe politicalimportofthese figurines,I first
an
by a quantitativeanalysisof figurinesize to assesswhetherrankis expressed
figurines,followed
ofthe perceived
presentsthe analysis of rankand status as
figurineheight differences.The final section
in derived from ethnographic and eth-
throughattire or individual costume, based on information for figurineproduction.
the motives
sources, and an interpretationconcerning
nohistoric
THE PIKILLACTAFIGURINES
ranging
dressedin distinctive hats and tunics, and
A total of 80 turquoise-coloredstone figurines (Figure 3) at the site
2 caches within the same room
size from 18 to 52 mm, was recoveredin
in site is a planned architectural
in the LucreBasin. The
Pikillacta,located 27 km south of Cuzco
of period
(Figure3), similar to others built and controlled by Huari officialsduringthe early the
facility empire in
near the southern frontierof the Huari
imperial expansion. Pikillactais located
of within this frontier region, but none rival
highlands; smaller Huari sites are situated farthersouth
thesize and architecturalcomplexity of Pikillacta. of
compound and has the most complex arrangements
Pikillactais the largestruralHuari state differentrectangularenclosures that usually
architecturalforms. It is characterizedby a series of galleriesor small longitudinalrooms.
or more peripheral
contain a centralplaza borderedby one the center
within a rectangularenclosure situated near
Thefigurineswere found in a galleryroom complex sector. One large (90 x 95 m)
and architecturally
ofthe site and within the most complete inferred
smaller plaza (65 x 65 m), dominate Pikillactawithin this zone. McEwan focus
andone slightly was the main ceremonial
that the smaller plaza
fromexcavations and architecturalanalysis (44
1984:161). The room that is reportedto have containedthe figurinesthe
ofthecomplex (McEwan focused on
smaller plaza. Investigationshave not
x50 m) abuts the southwest corner of this of this type, three have been excavated at
the
structures
roomof the figurinesitself, but of the 119 reveals the presence of these
of this sector of Pikillacta
site(McEwan 1984:94). The overall plan main entrance.
and two avenues that connect this area to the
twolarge plazas for gatherings, over access to
city indicate that there was tight control
Numerouscheckpointsto the heart of the McEwan (1984:161) as the
of Pikillactais interpretedby
thisarea. For these reasons, this sector and the center of administration.
and religiousofficers
residencesof the highest-rankingpolitical one of the heavily plasteredinteriorgallery
rooms
The two figurinecollections were found within de la
in the Museo
One set of these figurines,currently
of Pikillactanext to a doorway (Figure3). Collection. The figurines
is referredto as the Cuzco
UniversidadSan Antonio Abad del Cuzco, slab that interruptedthe floor, and were associated
1 m below a stone
werefound in a layerof sand princeps,and a marine snail (Strombus)
shell (used
with a thick, pointed "bronze"bar, Spondylus not found in an upright
1933:3).The figurineswere
duringPrehispanictimes as a trumpet)(Valcarcel The figurinesof the second or MadridCollection,
sand.3
position, but ratherwere lying down on the 5 m below groundsurfaceand arrangedin
a circle
housed at the Museo de America, were situated and
thick "bronze"bar with a curved tip (Trimborn
arounda verticallypositioned, crudelymade,below a stone slab that covered a floor 3 m below the
m
Vega 1935:36-89). The cache was found 2 missing
1935:89). There originallywere 40 figurines.One piece was
surface(Trimbornand Vega the cache was discovered. Associated with
shortlyafter
upon arrivalin Spain and was lost in Peru Spondylus(2 whole valves, 8 worked pieces [some
and
the figurineswere 2 marine-snailshells,4 manufactured
and 5 worked rectangularfragments).The figurineswere
with intentionalperforations], Valcarcel 1933:7-9), a rawmaterial
Ruppert1982:69-124;
ofturquoise(see Ramos and Blasco 1977;
Huari times.
that became increasinglypopularduring

l- .. ......
*---*.- .. ..... . l O C
THE STONE ANCESTORS

Cook]
345
luL
-<D'' '
:*:-g:-:*n-:-.:.:-:-:-
::::u:::*u:::::::::::
. . * . . . * * *
. . . * . . . .
. . ^w . . 4 - . . - . .
. . . . . . - - . -
..4 .. .-.-.-.-.-.
. . . * *
* . . . .
:::::::::::
* . . . .
* . . . .
* . . . .
. . . .
. . * . .
* . . . .
. . * * . * * * .
*--- :-*- -------:-:- =S

:::: *:::::::::
. . * * . * * * .
X
.-.- .-. -.-.*.-.- .
.-.- .- - *.-.-.-.- .
. *. ..-***
. * * . * * .
. . . . *
- go :.:-.:-:-:-:
nnnml
-
s - s w I- \\
0 10 20M
I I s
Figure3. Plan of Structure34-2B at Pikillacta, wherethe two turquoisecaches reportedlywere found.The

asterisks indicate the location of each cache. Shaded areas indicate looted areas (after McEwan 1984:Figure
3-3).
An essential aspect of these objects, which became apparentduring the course of the analysis
reportedhere, is that 20 figurineswithin the Cuzco Collection are exactly the same as 20 figurines
in the MadridCollection(Figure4): they shareidenticalattire,headdresses,and facial features,and
diffieronly slightly from one another in size. Other less-frequentexamples of identical figurines
within the Cuzco or Madrid collections occur as twin (Figure 2), triplet, and quadrupletsets of
identical figurines.The two caches of figurinesare thereforeclosely related. The dual partitioning
of each group of 40 into 2 sets of 20 is determinedon the basis of comparisonsbetween the two
CUZCO COLLECTION MADRID COLLECTION
4 0 f igurines 3 9 f igurines
/ \ Matching- / \
/ \ Figurines /
20 2< o 19
40
Figure 4. Unity and duality within and between the figurine collections.

346 LATIN AMERICAN ANTIQUITY [Vol. 3, No. 4, 1992
16 17 18 19 20
21 22 23 24 25
26 27 28
Figure 5. Pikillacta figurinehats from the Cuzco Collection:I occurs eight times, 2 occurs three times, 3
occurstwice, and 18 occurstwice.
collections. This is further supportedby the spatial proximity of these offieringsto one another
within the same room at Pikillacta.
Duality is a common culturalconcept in the Andes. It is expressedin territorialmoiety divisions,
in the presenceof dual lordships in political hierarchiesand in genderdiffierences.Specificaspects
of attire that have gendersignificanceare addressedhere and althoughthey are difficultto discern,
the results suggestfurtheravenues of study.
In the analysis that follows, four variables are used in the descriptionand analysis of presumed
social rank of the figurines.These are: (1) facial features,(2) bodily ornamentation,(3) tunic and
headdressvariations (see Figures 5-8), (4) figurineheight,5and (5) the relation between figurine
statureand attire.
Facial Features
The Pikillactafigurineseach have distinctlydiffierentand disproportionatelylargefacial features.
This attributeis sharedby large stone sculpturesat Huari (Lumbreras1974:Figure177) and in the
southernAndean altiplano. This enlargementallows the producerto convey details of expression
that would otherwisebe impossible on such small figurines.The facial variations suggestthat the
producerswerenot merelycopyinga singlemodel, but wereinterestedin diffierentiating facialfeatures
and expressions.
The eyes lack irises and are the only physicaltrait that is standardized.The meticulouscaregiven
to the details of attire stands in starkcontrastto the expressionlesseyes. In this instance, the Huari
figurinesseem to embody the concept of life crystallized in stone, or as Inca lore suggests, the
transformationof the ancestorsor deceased rulersinto veneratedstones.

Cook] THE STONE ANCESTORS
347
7 8 9 10 11 12
13 14 15 16 17 18
7-
L 2
21 22 23 24
19 20
25 26 27 28 29 30
31 32 33 34 35 36
Figure 6. Pikillacta figurine tunics from the Cuzco Collection: I occurs twice, 8 occurs twice, 19 occurs twice,
and 22 occurs twice.
Bodily Ornamentation
All but one of the figurines(which appearsas a kneeling nude captive with arms bound behind
its back [Figure7:18 and Figure8:17])wearfine layeredgarmentsand headgear.In addition, several
figurinesdisplaybodily ornaments,such as ear spools, nose plugs,and necklaces.These threeaspects
of attire convey rank on the basis of analogy with non-Huari elite burials6and contemporary
traditionalcustoms of dress.
Ear spools and nose plugs are known from elite coastal Moche burialsthat immediately predate
and overlapwith the initial period ofthe Huariempire (Alva 1988; Donnan 1988). Earspools were
importantindicators of elite rank among the Inca aristocracy,while nose plugs have been found
frequentlyin burialcontextsand as ornamentson mummy bundlesin areasrangingfromthe Paracas
occupationof the south coast to the Moche on the northcoast. Earspools are found on five figurines
in the Cuzco Collection and three in the Madrid Collection;the latter are identical to three of the
five in the Cuzco Collection.
Like ear spools, the use of nose plugs was common duringprecolonialtimes, and ethnographic
examples are numerous in the tropical lowlands. Two figurineswith nose plugs are in the Cuzco
Collection, and one is identical to a figurinein the Madrid Collection. They share identical dress

Q ag 9
rX
<,%
3 4
(
1 2
10
-
1 5
11 12 13 14
, 1 9
20
17 18
16
,
25
22 23 24
21 ,
\W/3o
27 28 29
26
Figure 7. Pikillacta figurine hats from the Madrid Collection: 2 occurs six times, 3 occurs twice, and 25
occurs 3 times.
and headgearthat are stronglyreminiscentof early colonial descriptionsof lowland tropicalforest

groups.
Collarsappearon three figurines.Such ornamentationis also frequentlydepictedin paintedform
or on effigy vessels assigned to the late Nazca and Moche phases that immediately precede and
overlap with the beginningof the Middle Horizon (ca. A.D. 500-600). Jewelryconsistingof collars
or necklaceswith multiple strandsof beads was associatedwith the warrior-priestburialsof Sipan
(Alva 1988; Donnan 1988). Zuidema has also emphasized the importanceof collars in Huari art
by drawingthe followinganalogyto Inca examples:"In CentralPeru,in the XVIth century,a similar
collar was called tamtaor huacaand representedthe peoples subject to the chief. The Inca king
also had such a collar of feathers as part of his royal dress" (de Avila l991:Chapter 5; Zuidema
1972:41). This stronglysuggeststhat collars can be consideredinsignia of high rank,althoughthey
can also occuron the StaffiedFront-Facingand Profilefiguresin offieringcontexts(e.g., Menzel 1977:
Figures 62 and 67). Among the Pikillacta figurines,eight wear ear spools, three wear nose plugs,
and three wear necklaces.Two observationsare noteworthy.First, none of the threetypes of bodily
ornament are particularlycommon in the assemblage,occurringas they do on only 14 of the 79
figurines.This observationlends additionalsupportto theirinterpretationas insigniaof rank.Bodily
ornamentationassumesgreatersignificancewhen it is recognizedthat 10 of the 14 figurinespresented

Cook] THE STONE ANCESTORS 349
11 12 13 14 15
26 27 28 29 30
Figure 8. Pikillacta figurine tunics from the Madrid Collection: 4 occurs twice, 12 occurs three times, 18
occurs four times, 19 occurs twice, 23 occurs twice, and 29 occurs twice.
with insigniaof rank,particularlythose wearingear spools, occur as matched sets in the Cuzco and
Madridcollections.
Tunic and Headdress Variation

The figurinesare remarkablydiverse: only four hats and tunics occurredmore than once in the
Cuzco material, while only three hats and five tunics occurred more than once in the Madrid
Collection (Figures 5-8). The numbers in Table 1 reveal that there is a greatervariety of hats in
the MadridCollection,while tunics occur in more variationsin the Cuzco Collection.To assess the
possible significanceof this diversity, the internal variability and patternedassociations between
tunic and headdressvariationsare brieflypresented.
Figurine Tunics
Huari tunic designs known from woven textiles are repetitive and can be divided into specific
design scheme classes. However, Huari weavers manipulatedcolor in such a way as to enhance
specific design layouts. This makes it possible to subdivide design schemes on the basis of color
patterns.A series of themes and subthemesresultsfrom which rank diffierencesmay be discernible

Table 1. Comparisonof GarmentTypes in the Cuzco and

MadridCollections.
Layered
Unkus Attire Other Total
Cuzco 29(15) 11 (5) 0 40
Madrid 26 (15) 12 (5) la 39
Total 55 23 1 79
Note:Parentheses indicate the number of figurines with matches in the other
collection (e.g., 15 of the 29 Cuzco figurines wear unkusand are replicated in
the Madrid Collection).
a A kneeling figurine with hands bound behind its back with the same rope
wrapped around its neck and waist.
by analogy to later Inca examples (Zuidema 1992). Huari tapestrytunics display vertical panels
of design within which design themes are displayed. The figurinetunics do not exhibit this consis-
tency; instead, lackingcolor, they convey a remarkablenumber of vertical and horizontalpatterns
that are rarelyrepeated,unless they occur in the context of paired figurines.
There are two evident forms of attire in the combined figurinecollections. The first includes
simple short-sleevedtunics or unkus of various lengths;the second consists of layeredgarments,a
tunic covered by a shouldermantle that terminatesat the waist, and infrequentlyadditionallayers
of textiles are evident. These shouldercoverings may have been puma hides or other animal furs
that are illustratedmore clearlyin paintedceramicforms (Ravines 1977:1aminaXXVI,32). Layered
garmentsare found on about one quarterof the figurinesin each collection (Table 1). As Table 1
reveals, 70 percent(55) of the figurineswear simple tunics. The remainingfigurines(24) wear more
complex layeredgarments.The depiction of simple vs. complex figurineattireis a distinctionbased
on both morphologicaland quantitativeobservations.In the final section of this paper, I illustrate
that this distinction is also expressed in the cultural values associated with simple and complex
forms of attire worn on specificoccasions in contemporaryethnographiccontexts, and describedin
ethnohistoricaccounts.
In summary,combining both collections, there is a strongtendencyfor figurinesto be portrayed
wearinga simple tunic with or without decoration, while more than one-quarterof the figurines
wear composite garments.The same patternexists among the paired figurines.
Figurine Headgear
Valcarcel(1933) illustratesanthropomorphiceffigyvessels from numerousPrecolumbiancultures
as potentialsourcesfor the identificationof hats on figurines.The comparisonsare overgeneralized,
yet there is little doubt that turbanedhats with a sling are typical of south-coast late Paracasand
Nazca society, and turbanedhats or rectangularridged hats with chin strapsare often depicted on
north-coastMoche effigyfigures(see also Valcarcel1933:10-14, laminasVI-IX). Duringthe Middle
Horizon a four-corneredhat often embellished with tassels is known in the centralAndean region
(Frame 1990).Therearearguablythreefigurinesin the combinedcollectionsthat wearfour-cornered
caps that lack tassels (Figure5:26, Figure7:22, 31; for detailed costume descriptionssee Valcarcel
[1933]; and Ramos and Blasco [1977]).
Figurineheadgeardefies easy classificationinto types. Valcarcel(1933:11) suggested16 diffierent
forms that might easily be furthersubdivided. In the classificationthat I establishedfor this study,
33 types were isolated (23 composite types, 2 round turban types, square turbans, round hats,
feather,checkerboard,and 4 differentcap types).
In general,headgearconsists of turbanswith or without decorationand frequentlyplaced over a

THESTONE ANCESTORS 351
Cook]
Cuzco Collection Madrid Collection
12 121 r
O All 10
FigurineS O 9
rT-rr Flgurines 2 Z |-
8
:° X;/XS Z/l , 48
n_
51 53 mm
33 36 39 42 45 48 Sl 53 mm 0 18 21 24 27 30 33 36 39 42 U
0 18 21 24 27 30
Figure9. Figurinesize frequencyhistograms.
head cloth that covers the neck and shoulders.Long hair is evident in three instances(Figure5:13
and Figure 7:17, 23). Beyond generalhat form, each hat contains unique featuresthat recur only
in the context of repeatedfigurineswithin each collection, or as paired figurinesacross collections.
Earlierit was mentioned that a comparison of the Madridand Cuzco collections revealed that
20 figurinesin the MadridCollectionfound their twin in the Cuzco Collection(Figure4). Five twin
pairsof figurineswearhemisphericalhats (two of these pairs wearidenticalheadgearand garments;
Figure 2g-h; for the Cuzco Collection see Figure 5:1 and Figure 6:19; for the Madrid Collection
see Figure7:2 and Figure8:19). One pair of matchingfigurinesweara hemisphericalcap, but unlike
thosejust describedthey lack a headclothand boast prominentear spools (CuzcoCollection:Figure
5:5 and Figure6:36; MadridCollection:Figure7:6 and Figure8:30). Althoughspeculative,this last
pair with ear spools arguablycould represent the highest-rankingindividuals within the group
wearinghemisphericalhats. Four pairs of figurineswear turbansof various kinds. The remaining
10 pairswearindividualhat types. This divides the 20 pairsinto two sets of 10 pairs,which includes
10 individuallygarbedfigurinesand 10 in the patternoutlinedabove. Sets of 10 wereorganizationally
importantunder Inca rule, an observationto which I returnbelow.
Figurine Height Comparisons

Size frequencyhistogramswere assembled to examine the assumption that if size suggests an
internalhierarchythen frequencydistributionsshould be skewed toward the largestfigurinesizes
or to the right, with increasinglyfewer figurinesoccupyingthe highest positions.
A size frequencyhistogramof each collection of figurines(Figure9) reveals that two figurinesin
the Cuzco Collection fall into the two largestsize categories.This can be comparedto the hatched
areathat representsonly the size distributionof the 20 figurinesthat find their match in the Madrid
Collection. In this instance, the patternis repeated,with only one figurinein the highest position.
The histogram of the Cuzco figurinessuggests that size may be a function of rank, most clearly
expressedin the distributionof the 20 matched figurines.
In the MadridCollection, size distributionsreveal a single cluster,with one figurine,the largest,
separatedin size from the rest of the figurines.The range of measurementsdisplay a distribution
skewedto the rightwith the largestfigurineseparatedfrom the main cluster. In this case, size may
be expressiveof rankdiffierencesto be determinedbelow in comparisonsof the actual figurinesand
their respective garb. The hatched area representingthe 20 matching figurinesparallels the size
distributionfound among the 20 matching figurinesin the Cuzco Collection.

FigurineStatureand Attire
Figurineattire within each size class was examined in an effortto determinewhetherindicators
of rank are associated with the largestfigurineswithin each collection or with particulargroups.
Among the figurinesin the Cuzco Collection, the size clusters suggestsome insights concerning
rank relations in terms of attire. First there are some anomalies with respect to size alone. The
presence of pairs within the Cuzco Collection has been noted. If, however, rank is conveyed by
figurinesize we would expect identical figuresto have similardimensions. But this does not always
hold true:for example,two pairs of figureswithin the Cuzco Collectionare notablydifferentin size.
Although I have no clear explanationto account for this, the statureasymmetryis not foreign to
Andean concepts of duality that embody complementaryopposition. If, for instance, there is a
hierarchicalrelationshipbetween two people then asymmetricforms of reciprocalobligationsare
more characteristic(mink'a)than balanced exchanges(ayni). Gender cannot be ignored, and may
also account for the differencesin stature.Additionally,the smallest figurinestands apartfrom the
rest;it has one of the most detaileddesigns,featuringa long headdressand extendedpiercedearlobes,
from which it can be inferredthat the individual had worn ear spools, featuresindicative of elite
status (Figure5:28 and Figure6:34).
Although size and costume do not initially appearto be closely related in the Cuzco Collection,
closer inspection of the figurineswithin each size class reveals that each of these clustersincludes
figurinesthat convey high status. Most of the individualswith layeredclothingand jewelry are also
the largestmembersof their size group.Among the Cuzco figurines,each size groupingthen contains
higher-rankedand lower-rankedindividuals.
The Madridfigurinesare slightly more regular,with the smallest figurineswearingsimple tunics
but boasting variationsin headdresses.The largerfigurineshave both complex tunics and hats. In
the middle rangeall possible combinations occur in no visible pattern,with the exception of one
triplet,two of which have not only identicalattirebut also identical stature.Additionally,thereare
two identically paired figurineswithin the Madrid Collection which, as in the case of the Cuzco
paired figurines,are carved in differentsizes.
Paired Figurines
Paralleltrends are noticeable in both collections, such as the presence of pairs within each col-
lection, repetitionin hemisphericalhats, and increasingstaturewith associatedelite insigniawithin
each cluster.Overallthe gradualincrementsin the sizes of the figurinesare so negligiblethat major
differencesand patternsare best observed by comparingthe 20 paired figurinessharedby both the
Cuzco and Madridcollections.
Size frequency distributions of the 20 matched figurines(Figure 9) reveal strong similarities
between the two collections. The four largestfigurinesin both collections were comparedand two
found their match in the other collection. This suggests that greatersimilarity in size may exist
betweenidenticalpairedfigurines.To examine the associationbetweenthe sizes of pairedfigurines,
Spearman'sRank CorrelationCoefficient(Doran and Hodson 1975:143-145) was calculatedusing
all pairedfigurinesfrom both collections. First the size rankof all 79 figurineswas established,and
the ranks for the paired specimens were then used to calculatethe correlationcoefficient(rS= .75;
values for r lie between-1 and + 1 and arepositive when values for both variablesincreasetogether).
The result is a clear positive estimate of association in size between the rankedmatched figurines.
The size relationbetween the two collections of matchingfigurinesis thereforenot random. Size is
significantfor the paired figurines.
A final comparisonconcerns costumes worn by the matching figurines.The averageheight was
calculatedfor each pair of figurines,which were then rankedby size to evaluate whether layered
attire was associated with the largestpaired figurines.Figurinesof smaller size would likewise be
expected to reveal simpler headdressesand unkus. The comparisonsconfirm that figureswearing
layeredgarmentswere among the largestfigurines.Four figuresthat rank second, fourth, fifth, and
sixth in heightalso exhibit collarswhich, in de Avila's descriptionsof the Inca, identifiedelite status.
The similaritiesare more than coincidental.

Cook] THESTONE ANCESTORS 353
ROBESOF ROYALTY:ETHNOHISTORY,ETHNOGRAPHY,
AND THE ARCHAEOLOGICALRECORD
There is little question that Andean precolonialattire was a semanticallyrich form of expressing
social, political, and economic information.In the following passage,some of the firstimpressions
felt by the Inca upon viewing the Spaniardsare clearly conveyed.
Y que todoseran amortajados,toda la cara cubiertade lana, y que se le pareciasolo los ojos. Y en la cauesa
trayyaunas ollitas colorado,ari manco [olla sin estrenar],y suri uayta [adornode pluma de avestruz].Y que
trayyanlas pixas colgadasatraslarguicimos,decian de las espadas,y que estauanbestidos todo de plata fina.
Y que no tenia senor mayor, que todosparecianermanosen el tragey hablary conuersar,comery vestir. Y
una cara solo le parecioque tenia [GuamanPoma de Ayala 1980 (ca. 1615):383;emphasisadded].
And that they wereall shroudedlike corpses,their entire faces covered with wool, and that only their eyes
could be seen. And on their heads they wore colored pots and ornamentsof ostrich plumes. And that they
carriedtheir penises, very long, hangingbehind; this they said of the swords. And that they were dressed
completelyin fine silver.And that they did not havea greaterlord,that all seemedto be brothersin theirdress,
in speakingand conversing,eatingand dressing.And it seemed that they had one singleface [Adorno 1986.
148; emphasisadded].
The initial interestin this study was motivated by the distinctive costumes worn by the Pikillacta
figurines:each headdressand tunic detail was delicately carved so as to display both the layers of
textile clothing and their designs. I reviewed the literatureon traditionalAndean costume in an
effort to decipher elements of continuity in dress, the specific meanings of which may have been
recordedin recent times, but which might shed light on the differenttypes of apparelworn by the
figurines.
Depictions of attirein ethnohistoricsourcesand ethnographicstudies on costume and the culture
of fabricsmadeit possibleto addressthreespecificquestions.These concernedwhethercontemporary
traditionaldresspreservesprecolonialcomponentsindicativeof(a) ethnicity,(b)rank,and (c)gender.
Althoughdirect analogyis not always applicablein archaeology,inferencesconcerningthe relation
between ethnicity and attire seem plausible in the Andean region because, on the level of the
individual, weaving styles and motifs mark ethnic affiliationsand define a person'splace of origin
(Harris 1980; Medlin 1987). For instance, hats continue to identify individuals from specificareas.
Economically,we know thattextilesservedthe Incastateas culturallyvaluedcommodities.Within
communities,textileswereassociatedwith legitimationand politicalcontrol,and servedas signifiers
of rank and political office (Murra1962).
Most of our initial impressions concerningAndean peoples and their value system are derived
from the early colonial sources. There are few early images of Inca attire that can be used, but at
least one source warrantscloser attention.
A colonial-periodletter to King Felipe II of Spain, written and illustratedby Guaman Poma de
Ayala (1980 [ca. 1615]), offers invaluable images of the Inca during the immediate Postconquest
period. Although it is prematureto suggest correlationsbetween Inca class markersand those at
Huari,it is interestingto note that hemisphericalhats in the Inca empirewere worn by high-ranking
Inca nobility, warriorcaptains and soldiers, each with its appropriatestatus marker.In contrast,
among the Pikillactafigurinesonly plain hemisphericalhats are repeatedin sufficientnumbers(24
in the combined collections) to warrantcomparison with later examples; the remaining figurines
convey greateroverall variabilitythan those depicted in Guaman Poma de Ayala's illustrations.
Althoughit is difficultto determinethe degreeof precisionGuaman Poma de Ayala intended in
these drawings,they leave little doubt that rank differencesare symbolized by increasingthe dec-
orative complexityof helmets. The numberand form of Pikillactahats, and their rangefrom simple
to complex parallel the depiction of Inca rank shown in his drawings. Like their ethnographic
counterparts,describedbelow, these rank differencesare symbolized by increasingthe embellish-
ments of Inca attire and are thereforeuseful for the interpretationof the Pikillactafigurines.
Some unexploredassumptions circulateamong those scholarsinterestedin Andean costume as
markersof ethnic identity and rank.The degreeof specificityvaries from area to area but generally
hats are broadlyunderstoodto be community- or region-specificidentifiers(i.e., Wobst 1977);and

354 LATIN AMERICAN ANTIQUITY [Vol. 3, No. 4, 1992
tunics, dependingupon the occasion and the design structure,can symbolize one's place of origin
or one's rankin community life. Yet there are very few studies that explorethese specificquestions.
To disclosethe veracityofthese assumedrelationsI reviewedthe literatureand questionedcolleagues
who work with weavers and the social contexts in which textiles play an essential role.
Fieldworkon weaving and traditionalattire in the northernPotosi area of Bolivia reveals con-
tinuity in the use of cloth and hats as indicatorsof culturalidentity. Abercrombie(1986:202) notes
that a punchu(poncho)and awayu (shawl)are the outermostitems of men's and women's clothing.
In Bertonio'sdescriptionof the sucullurite in which childrenwere first clothed, the orientationof
stripeson clothing is describedas an indicatorof gender.Abercrombiealso noted this in southern
Potosi, where men's ponchos are worn over the head with stripesrunningvertically,while women's
awayusare worn over the shoulderswith colored design bands runninghorizontally(Abercrombie
1986:202).Among the figurinesthereare some with only verticalbands and otherswith exclusively
horizontalstripes, suggestinggendercategories.
Otherresearch(Zorn,personalcommunication1990)indicatesthat hats serveas regionalmarkers.
Forexample,white bell-shapedhats arefoundthroughoutthe region,howeverethnicand community
identity is markedin brim-size variations. It should be noted that brim-size differencesmay vary
as little as 3.8-5.1 cm. The differencesin brim size indicate a person'sethnicity.Within the market
place these subtle differencesare widely recognized.
Hats alone do not indicate rank or notions of community authority,however they assume im-
portance in association with particulartypes of attire worn exclusively by authority figures on
ceremonial occasions. Leadershipand the symbols of authorityare prominentlydisplayed by the
individual sponsorof community or multicommunityfestivities and are expressedin ponchos and
mantles (tunics are still worn by some Lake Titicaca groups).
The possibility of identifyingan individual's rank from his/her attire seems to depend upon a
combination of factors. In the few descriptionsavailable, male authorityfiguresare identified by
the manner in which certain textiles are folded and positioned on the body, and these forms of
attire are always worn for certain ceremonies in conjunction with specific objects as symbols of
authority(Zorn 1987:67).
Abercrombie(1986) also observed the presenceoffemale festival sponsors(e.g., on the occasion
of the Day of the Dead and Carnival)in southernPotosi. When a female sponsorsa festival she is
cargada (burdenedor loaded) with multiple textiles strung to her llEjlla.These consist of a large
numberof individual shouldermantles that expressher wealth and authority.She is literallyladen
down by textiles, and the same is true of malejilakatas (communityleaders)in this region. In this
context, we find a contemporaryexample of the values embodied in textiles that Murra (1962)
emphasizes for the earliercolonial period.
It is considerablymore difficultto discuss differencesin attire that are gender specific and that
date to the precolonialera. In fact, there has been an alarmingdisregardfor the identificationof
gender archaeologically(Gero and Conkey 1991) in the Andes, despite the excellent preservation
that typifies coastal archaeology.The Andean region certainlydoes not stand alone in this regard,
althougha few studies reveal a growingconcernwith the topic (Hocquenghemand Lyon 1980; Lyon
1979;Silverblatt1987).The archaeologicalstudiesconcernthe identificationof femalesupernaturals
in Moche and Huari depictions.
Menzel (1964:26, see also 1977:Figures122 and 123) first noted Huari gender distinctions for
supernaturalson the basis of costume and plant depictions in the two Frontal Staffedfiguresthat
are painted on an oversized offeringvessel from the south-coastalsite of Pacheco. Other studies
(Lyon 1979; Rowe 1979) addressthe identificationof gender differencesin Huari art. All sources
agreethat thereare few examplesof genderdifferentiationin Huarimaterialculture.Those discussed
pertainto the associationbetween particularattributesof costume that include depictionsof maize,
a predominantlyhighland crop. What has been given less attention is the fact that all of these
examples are from the coast, where cultivation largely depended upon the availability of water,
which was associated with the highlands and mountain deities. On the coast, issues concerning
fertilityand cultivationmay have set the stagefor more explicit referencesto female representations

Cook] THESTONE ANCESTORS
355
asseen in the offieringvessels with female figuresas tripod bases from Maymi in the Pisco Valley
(Anders1990b) and Nazca cloth figurines(Rowe 1991).
Costumeand culturalconceptsof attireshouldmakegendermore accessible,althoughthe specifics
ofthis subjectare beyond the scope of the presentdiscussion. It should be evident from the above
thatspecific forms of attire characterizeboth high-rankingmales and females in the ethnographic
recordand that these have clear historical antecedents(e.g., see also Isbell 1978:93-94, 107-108,
111-1 12;Silverblatt1987). The underlyingprinciplesgoverningelite dress are repeatedin the attire
of the Pikillactafigurines.
In summary,ethnographicand ethnohistoricexamples illustratehow cloth serves as a symbol of
communityauthority.Greateramounts of textiles expressgreaterdegreesof prestige.I suggestthat
the greaterthe authorityvested in a local leader the greaterthe textile wealth possessed and worn
on ceremonialoccasions. These notions are still visually markedin modern-dayceremoniesby the
presenceof overtly overdressedindividualswith multiple layersof cloth, in some instancesliterally
carryingtheirwealthin textiles.Given this continuity,it seems plausibleto drawan analogybetween
the generaluse and symbolism of contemporarytraditionalgarmentsand archaeologicalcontexts
that share similar configurations.By analogy, Pikillacta figurinesthat wear multiple layers of gar-
ments, in associationwith elaborateheadgearand symbols of prestigesuch as ear spools, nose plugs,
and collars, should be indicators that help differentiateindividuals of higher rank from those of
lesser rank, both within and between the two collections.
DISCUSSIONAND CONCLUSIONS
The results of this study suggest that if the analogy to contemporarydress is valid then hats
indicate ethnicity and community identity. About 75 percentof the hats within each collection are
diffierentwhile the remaining25 percentdo recurmore than once. Withineach collection of figurines
ethnicityor communityidentityis markedand expressedin variationsof headdressform and design.
Tunics with specific designs have been shown, in contemporaryand ethnohistoriccontexts, to
reveal rank and intracommunityauthority.A comparisonof the two collections reveals that rank
insigniaare more visibly displayedamong the Cuzco figurinesthan those of the MadridCollection.
To assess the significanceof this observation an independent evaluation of figurineheight was
presented,in orderto determinewhetherhat and tunic variabilitycorrespondsto particularclusters
of figurines.
For both the Madridand Cuzco collections, each size mode includes figurinesthat are believed
to representhigher-and lower-statusindividuals. The high-statusinsignia are primarilyassociated
with figurinescomprisingthe largestmembersof each mode. Additionally,the resultsconfirmthat
size was intentionallycontrolledto indicate status and to help diffierentiate individual figures.This
distinctionis even more pronouncedwhen the 20 figurinesin the Cuzco Collection identical to 20
figurinesin the MadridCollection are viewed as a separategroup. In this instance, 15 figurinepairs
wearunkus,while only five pairedfigurinesevidence layeredgarments.In fact, one of the five paired
figurineswith layeredgarmentsalso wearsear spools (CuzcoCollection:Figure5:5 and Figure6:36;
MadridCollection:Figure 7:6 and Figure 8:30).
To date, we have little additional evidence of bluish-greenstone sourceswithin areas that came
under Huari control. The quantity of unworkedand worked greenstone fragmentsfound within
certain sectors of Huari indicate that large quantitiesof this semipreciousmaterialwas broughtto
the site. In contrast,there is little evidence that turquoise,as a raw material,was cached or kept in
large quantities at Pikillacta (McEwan,personal communication 1989) because figurinesare the
only form in which turquoise objects have been found at this site. The conclusions reached in
Ruppert's( 1982)sourceanalysisof Andeanturquoiseinvites closerinspectionof sites in the vicinity
of Pikillactaas potential turquoise-processinglocations or workshops.
Stone carvingwas and remains part of the ritual sphere in the Andes (Ramiro Matos, personal
communication 1991), and as such it might betterbe conceived as a form of specializedproduction
removed from the generic class of craft specialization.One probablelocus for these "specialized"

workshopsis Huari in the Ayacucho Valley, the only site where remains of the differentstages of
production(Brewster-Wray1990:114-116) have been found: the raw material,worked but unfin-
ished pieces of greenstone,and complete figurines.Pikillactamay have served as a nodal point in
a long-distanceexchangenetworkof raw lapidarymaterialsdestined for the capitalof Huari,where
the materialswere workedand then redistributed.
It is unlikely that adequate answers can be found to questions concerningthe identificationof
who these specific figureswere intended to represent.The figurineshave only been found in two
caches of 40 objects within the confines of the site of Pikillacta. This suggests that individuals
representedin stone held special significancefor the occupants of Pikillacta.The recognitionthat
other figurineshave been found at Huari and coastal sites greatlyexpands the areal distributionto
include major areas that came under Huari control duringthe first centuryof its existence.
Pikillactawas an early Huari state installation. It is conceivable that the inhabitantsof the site
were involved not only in the business of administrationand boundarymaintenance,but in estab-
lishing this regionalcenter as a local Huari oracle (e.g., Shea 1969). The purposeof manufacturing
the figurineswas to depict individuals with positions of rank that were known to the makerswith
marksof regionalorigin.The figurinesare ideally suited to the commemorationof local Huarilords
as representativesof royal lineages and their specific foundingancestors.Several lines of evidence
supportthis interpretation.
First,the color turquoiseand various relatedshades of greenmay have held symbolic importance
duringHuari hegemony because its use is restrictedto high-qualitytextiles and lapidaryitems. In
textiles,the color turquoiseis mostly used to interruptotherwisesystematicpatternson finelywoven
tunics and largerpieces, while lapidaryobjects of varying dimensions and forms are carved with
stones of greenish-bluecolor. The color turquoiseis unknown as a pigment on ceramics.
A second line of evidence involves ethnographicand ethno-archaeologicalobservationsthat bear
upon the associations between stones, the color green, and the ancestors.A few examples suffice.
In Sonqo, near Cuzco, where CatherineAllen conductedfieldwork,All Souls Day, November 1, is
calledKawsasqanchis,when the dead visit their relatives.The word means "OurLiving"("theword
groupsthe dead with the living throughthe suffix-nchis [Allen 1988:164]").On this occasion,Allen
(1988:164) recordedhow "Rufinadescribedto Felichahow 'prettyand green'the souls were as they
hovered aroundthe rafters."
In the villageof Chuschi,duringYarqaSspiy,the ritualassociatedwith the cleaningofthe irrigation
canals,the villagersgatherin their best attirewith their hats covered with angoripa,an uppersallqa
(puna region at 33004000 m asl) plant used to symbolize the ancestors (Isbell 1978:144). The
branchesof this plant are called mallki, a generalterm with bipolar meaning, on the one hand it
refersto a (green)saplingor youngtreereadyfortransplanting,on the otherit refersto the ancestor(s).
Consequently, mallki simultaneously signifies renewal and continuation through the generations
(Isbell 1978:147).
When Allen (1988) observed that occasionallyfamilies in Sonqo kept the bones of their dead in
a wall niche of their storehouse,it was said that the bones care for or protect (khuyay)the stored
goods. In like manner, the small stone heirloom power objects called enqas or illas are described
as caringprotectorsand as kawsaqkinaor living ones (Allen 1988:59). On another occasion Allen
visited the Cuzco archaeologicalmuseum (wherethe Cuzco figurinecollectionis housed)with some
friends from Sonqo. These Sonquenos were both impressedand disturbedat the sight of so many
kawsaqkunaenclosed in case after case on public display. In this instance, the Cuzco Collection
figurineswere identifiedas "the living ones" (Allen 1988:59, personalcommunication 1991).
From an Andean perspective,the compact hardness of stones, bones, and statues implies not a lack of
animation,but a differentstate of animation-life crystallized,as it were. Hard,unusualstones (suchas illas)
and bare bones (like the skull kept for kFuyyay) are felt to be the most potent sources of energy.They are
intimatelyconnectedwith lightningand sunlight,whose power they absorband condense [Allen 1988:63].
The landscapeitself is inhabitedby the living and the "animated"dead. This relationshipbetween
the living, the deceased, and the land also existed in the countrysideon the level of everyday life.
In the village of Ocros, Hernandez Principe (1923 [1622]:50-64; Zuidema 1977-1978) reports

Cook] THE STONE ANCESTORS
357
genealogicaldata and the story of an individual who was elevated to the rank of curaca over
neighboringvillages following the constructionof an irrigationcanal that he had organized.The
curaca sacrificedhis daughterby buryingher alive in a shaft tomb on top of a mountain where the
storehouses were located to house the crops of the newly irrigatedfields. The curaca was also
presumablyburied in another shaft tomb. Zuidema has given historical value to this event by
locating the actual shaft tombs in Ocros. The fields and mountain tops were the sites of lineage
burials the extent of which was only made evident in seventeenth-centuryecclesiasticalinquiries
(MacCormack1991:132-133).
The associationbetween political rankand the degreeof ritualelaborationexpressedin past and
presentpracticesfor disposal of the dead has received considerableattention. These practicesvary
considerablyacrossthe Andes, while conversely"it seems clearthat they (the dead)wereworshipped
in a material form, and that the images and relics of the dead lords enshrinedmemory, wisdom,
power and fertility"(Harris 1982:46).
As I suggesthere, the ancestor cults describedin contemporaryethnographiesand for the early
colonial era (to which I return below) had a long history, most probably predatingthe Middle
Horizon.
A third line of evidence consists of archaeologicaldata that offer glimpses of cults of the dead.
For Huari and other Andean cultures,these relationshipsare difficultto confirmdue to the lack of
prehistoricwrittenaccounts.Ancestralcults, however,were practicedwith greatpopularityeven by
the EarlyHorizon (900-200 B.C.). Such cults are known to have existed in various Andean culture
areas and time periods. One strategyof Huari conquest that may have helped consolidate support
on the local level and which would have facilitated the establishmentof legitimate supremacyin
new areas is the institution of ancient ancestral cults which were tied to power and land rights
(Duviols 1986; MacCormack1991:131).
There is evidence that ancestorcults describedduringthe sixteenthcenturywere of much greater
antiquity, as exemplified in the customs associated with mummy-bundleconstructionand com-
plexity in Earlyand Middle Horizon coastal cultures.Huari power was intimately tied to control
over land, and colonization or penetration into coastal valleys may have been more eflectively
legitimized by the adoption or continuation of more ancient coastal ancestralcults and the burial
of their "living"dead in their newly acquiredlands. Middle Horizon coastal mummy bundleswere
treatedin accordancewith the rankof the deceased,as in the presenceof finelywoven outer tunics,
jewelry, and gold face masks, in contrastto those lackingthese symbols of wealth and prestige.
Vreeland's(1980) analysisof Middle Horizon mummy bundlesand the natureof tombs in which
they were found revealedthat the better-preservedcoastal examplesconsist of tightlyflexedburials,
which have antecedentsin earliersouth-coastalcultures.
In a recent survey (Cook 1989, 1990) of the lower Ica Valley,7Huari mummy bundle remains
were found on the surfaceof Middle Horizon sites, disinterredand ravagedby looters. While these
mummy bundlesarefound with greaterfrequencyat elite residentialHuarisites, they arealso known
from cemeterylocations (see also Rowe 1984). The deceasedis usuallyseated on one or more coils
of workedcloth stuffedwith raw cotton, wrappedin a bale or bundle, and topped with a false head
and face mask. These bundles were often dressed with a finely woven tunic that had been worn
duringthe life of the individual.This complex of traitsrevealsan emphasison the individualpersona
and his or her status in society. In the highlands, where preservationis poor, several examples
illustratethat flexedburialswere common at Huari sites, includingJargampata(Isbell 1977:20,22-
23, 28-29) and Huari proper(Lumbreras1974:160, 177).
Anotherimportantfeaturesharedby both highlandand coastal Huaricontextsis the construction
of mausoleums for elite single or multiple burials that provide access for reentry.At Willkawain
nearHuarazin the Callejonde Huaylas(Figure1), subterraneanburialhouses were recorded.Those
that were investigatedconsisted of gallerieswith capped accesses (Bennett 1939; Lumbreras1974:
118). At Jargampata,at least one flexed burial still contained a cappingstone that sealed the tomb
(Isbell 1977:29).At Huari, in the Cheqo Wasi sector, elaboratedressed-stonestructuresare found.
These are now empty (Lumbreras1974:Figure174), having been looted in prehistorictimes, but
there is speculationthat these multistoried stone-slab structuresmay have been burial chambers,

on the basis of comparisonwith a similar structureat the site of WariWillkanear Huancayo.These

structuresare covered by huge cappingstones with holes throughwhich libations could be poured.
A similar example comes from Huari in the Moraduchayoqsector, where several adjacentrooms
containedsubfloorcists sealed with largecircularcappingstones with one or more small holes. The
cists had been looted prehistoricallybut they still containedofferingsof variouskinds, mostly pottery
and some human bone fragments(Cook 1985b). In summary,the evidence that ancestorworship
was a focus of Huari ritual activity can be seen in the personificationof Middle Horizon Huari
mummy bundles, the evidence for libations and associated offerings,as well as the possibility of
burial chamberreentry.
An additional relation can be inferredfrom the archaeologicalcontext of the Pikillacta figurine
offerings.The "bronze"bar situatedin the midst of both figurinecachesis a metonym for authority.
In Huari votive-offeringart, the FrontalStaffedFigureis associatedwith a spear or staff.The bars
situated among the figurinesvividly recall versions of the Inca origin myth, in which a "golden
staff" was thrown to test for fertile soils for the foundingof the city of Cuzco.
Juan Larreahas persuasivelyarguedby analogyto Inca sourcesthat the Pikillactametal objects
are not only made of copper (ratherthan "bronze,"as they were initially reported),but that they
representthe pre-Incaicroyal scepteror yauri. Yauri means copper in Aymara, while the word is
used in Quechuato referto the Inca royal staff(Larrea1960:59-94). Larrea'sstudy is extensive and
groundedin ethnohistory.He revealsthat the yauri was an insigniaof high Inca nobility, appropriate
to and held exclusivelyby the emperorand the royal ayllus. The staffwas normallymade of copper
or silver. The Inca himself held a gold staff called a tupayauri. Although this insignia of royaltyis
not held by any of the individual Pikillactafigurines,it is found instead in their midst-a collective
symbol that possibly foreshadowstheir royal status.
A final line of evidence supportingthe interpretationof the Pikillacta figurinesderives from
ethnohistoricmaterials.Extrapolatingfurtherfrom our knowledgeof Inca religiouspractices(for a
detaileddiscussion see Brundage[1967:34-35]; Conradand Demarest [1984:96-105]; Rowe [1946:
252-297]; and Zuidema [1973]), ancestorworship played a centralrole in Inca kinship, hierarchy,
and ultimately leadershipinheritance.The mummies of deceased Inca rulers were housed in the
Temple of the Sun in Cuzco and were venerated by Inca nobles and their own lineages. Several
studiesdirectlyaddressthe pre-Incaicoriginsof these practices(Conrad1981;Conradand Demarest
1984).
The Pikillactafigurinesarguablyrepresentthe legendary40 foundingancestorsof the Huaripolity.
The number40 has specialimportancebecauseit conveyed an administrativeunit of division under
Inca rule. The city of Cuzco was organizedon the basis of a system of imaginarylines or ceques,
that emanated out of the Koricanchaor Temple of the Sun (Zuidema 1964). These were grouped
into four suyus or quarters.The Cuzco Valley was divided into these four parts,each of which was
organizedand spatiallysubdivided using territorialand water access divisions based on a decimal
system. Two of the suyus, Cuntisuyuand Collasuyu,comprisedthe southernmoiety or lower Cuzco
(Hurincuzco);Chinchaysuyuand Antisuyu made up the northernor upper moiety (Hanancuzco).
This quadripartiteand dual system was extendedspatiallybeyond the Valleyof Cuzco as the empire
grew. Outside of Cuzco, each of the huamani or provinceswas governedby a CapacApu or "great
lord" who was chargedwith 40,000 families.
One anonymous source (cited in Rostworowskide Diez Canseco 1970:162-173) referringto the
Valley of Cuzco mentions how the king and queen would initiate the agriculturalyear by ritually
plowing a field. They were followed in turn by the lords and first wives of the four suyus of the
empire, rankedas units of 100,000 families, and followed by those second in rank,the 40 lords and
their wives of units of 10,000 families (see also Zuidema 1990:68-69). Below this rank, but not
mentionedwith referenceto this ritual,were the 40 lords that were Incas by privilege,each charged
with 1,000 families(GuamanPoma de Ayala 1980 [ca.1615]). Althougha distinctionexists between
the rankedunits of administrationin the Valley of Cuzco and those of the provinces, there is no
doubtthatat the Incacapitalof Cuzco,the secondand thirdhighest-rankedgovernorswereorganized
in groups of 40. Given the presence of 40 figurinesin each of the Pikillacta collections I suggest

Cook] THESTONE ANCESTORS
359
that the origins of the Inca subdivision into 40 lords, at the highest levels of administration,has
greatantiquityand is conveyed as well in Huaripolitical organization(see also Anders 1986,1990a).
In Cuzco there were 10 territorialsocial units or ayllus called panacas and 10 nonroyal ayllus,
for a total of 20 ayllus (Sarmientode Gamboa 1942 [1572]:49-59). Urton (1990) has convincingly
arguedthat the hierarchicalrelationshipbetweenthe ayllus seems to derive from the rankingof the
four sisters and brothers who emerged from a cave in Pacariqtambo.This Inca origin myth is
describedin detail by Sarmientode Gamboa, and Urton (1990:20) notes:
Sarmiento next gives the names of eight ancestral siblings. It is important in naming the ancestors to retain
the order in which Sarmiento enumerates them, because he lists the two groups-first, that of the four brothers
and then that of the four sisters-in parallel hierarchical rankings. The brothers, Sarmiento says, were ranked
by authority, the sisters by age [Sarmiento de Gamboa 1942 (1572):49].
When the two Pikillactacollections are comparedthere are 20 matching figurines,which suggests
that these may representHuari ayllus or the mythical ancestors of the 20 highest-rankeddescent
groups. The Inca myth cited above emphasizes parallel hierarchicalrankingsfor both men and
women. This could have some bearingon futuregenderdistinctionsamong the Pikillactafigurines.
The above analogy offers furthersupport for the interpretationthat the figurinesmay represent
an analogousgroupof highly rankedmembersof Huari society. While it is possible to segregatethe
administrativeorganizationof the Valley of Cuzco from that of its provinces, the same has not
been possiblefor Huari,whose administrativestructuremust be derivedfromarchaeologicalremains
and thereforeprovides a simpler and more centralizedpicture.
The two sets of figurinesmay have been buried togetherintentionally,because their relation to
one another is more than superficial(Figure4). A new set of 40 figurinesis created when the 20
figurinesfrom the Cuzco Collectionare pairedwith their 20 twin figurinesin the MadridCollection.
These observationslend support to the interpretationthat the figurinesoffer a glimpse of Huari
political organizationand rankdifferencesthat are expressedlater in Inca administrativehierarchy
and linked to the internal rankingof the founding ancestors (for a detailed discussion see Urton
[1990:1849]).
Two final accountsextend the spatial and temporalsignificanceof turquoisefigunnes. Valcarcel
(1933) cites a passage from the Relacion del LicenciadoFelipe de Medina recordedon March 23,
1650, which concerns a visit to Huacho, just north of Chancay. In the account, an offenng was
recoveredthat included many bnghtly colored red shells (probablySpondylus)and two bivalves
each fastened shut. When they were opened a small greenstoneidol was found, which the local
inhabitantsdescnbed as their ancestor,and three small stones, said to be the firstlima bean, wheat,
and chili pepper(Valcarcel1933:5;translationby the author).
Additional legends lend further credibility to Felipe de Medina's description. These concern
accountsrelatedto the foundingof the north-coastLambayequekingdom. In 1586 Miguel Cabello
de Balboarecounteda legend associatedwith an earlydynastyof the LambayequeValley that ruled
before the emergenceof the Chimor kingdom. A translationCabello de Balboa'sstory is provided
in abbreviatedform by Donnan (1978:100, 1990:243-245). An excerpt reveals the importanceof
the passage:
The people of Lambayeque say that in times so very ancient that they do not know how to express them, a
man of much valor and quality came to that valley (from a place far to the south) on a fleet of balsa rafts.
His name was Naymlap. With him he brought many concubines, and a chief wife named Ceterni. He also
brought many people who followed him as their captain and leader. Among these people were fortyofficials
including BloweroftheShellTrumpet;
PitaZofE, Ninacola,MasteroftheLitterandThrone;Ninagintue,Royal
Cellarer(hewasin chargeof thedrinkof thatlord);FongaSigde,Preparer seashell
of the Way(hescattered
dustwherehis lordwasaboutto walk);Occhocalo, RoyalCook;XamMuchec,Stewardof theFacePaint;
Ollopcopoc, Purveyor
MasteroftheBath;andLlapchillulli, Garments....Naymlapalsobrought
ofFeathercloth
withhim a greenstoneidol namedYampallec. him, wasnamedfor him andgaveits
Thisidolrepresented
nameto thevalleyof Lambayeque [Donnan1990:243-244;emphasisadded].
Thereis disagreementconcerningthe chronologicalplacementof these events (Bennett 1939:120;
Donnan 1978:101; Means 1931:54-56; Rowe 1948:36). Some favor interpretingthe accounts as

3, No. 4, 1992
[Vol.
AMERICANANTIQUITY
LATIN
360
prefer
histoncal value (Rowe 1948), while others
myths with legendaryfiguresthat have little
ongin penod of Huari influence on the north
is also the
placethe events dunng late Moche V, which in a recent volume that concerns the northern
to
The issue is reexaminedby several authors
coast. there is no clear consensus, many of the
(Moseley and Cordy-Collins 1990). Although was an histoncal figure(Donnan 1990).8
dynasties
contnbutionspoint to the strongpossibility that Naymlap between 40
seems less relevantthan the association
question of who Naymlap may have been
The details
with whom Naymlap was identified.The
specialistsand the ancestralgreenstoneidol and to the offeringcontext of the Pikillacta
official
account
strikinglysimilar to Felipe de Medina's
are
as Huari royal lineage or ayllu ancestors.
and their potential significance tech-
figurines in a single medium the two oldest Andean
Whatis certain is that these figurinesconvey is constructed. In
so much of Andean culture
stoneworkingand the fiber arts, on which
nologies, portable mnemonic system
part of the visual and
identifiableheadgearand costumes form
essence,
the necessity for writing.
may have also precluded
that to the early
humans and sacred figurescan be assigned
EarlyHuari images that combine both be most pronounced.
practiceswould be expected to
of the Huari empire, when legitimation
years more
and commenced the annexationof new territories, was
OnceHuarihad achieved militarysuccess be expected. The legitimation process
and specific signaturesof the state might
standardized
guise of an established regime with practicesthat would ensureHuan social
continued but underthe exemplifiesthese ritualsof legitimation
reproduction and supremerule. The Huari site of Pikillacta served as signaturesof the state but
ritualssuperficially
the provinces followingannexation.The
in sacredstatus,
embody informationconcerning Huan regalpractices,the elevation of humansto that
they ancestors.It is in this capacity that suggest
I
anda unifying political cosmology of venerated to ensure rights of inheritance, domination,
ancestor worship could be appropriatedby the state ends.
sense serve administrative
andsacredlegitimacyand in this
on this paper and
Tom Zuidema for his constructivecomments
Acknowledgments. I would like to thankfigurinecollection from Pikillacta.Specialappreciationis owed to
fortelling me of the existence of the Madrid figurines.I would also like to thank Tom
Cummins
Frank Meddens for photographsof the Cuzeo Collection to the Naymlap accounts.
references
andHelaine Silvermanfor providinghelpfulat the Conferenceon the Archaeologyof Letteredand Unlettered
Earlierdraftsof this paperwere presented FiftiethAnnualMeetingof the Societyfor AmericanArchaeology
Societiesat ColgateUniversityin 1984, inat the ColumbiaUniversity.
the Arts of Africa,Oceania,and the Americas,
inDenver, and the University SeminarLatin American Summer LibraryFellowshipfrom Cornell University
ADumbartonOaks Fellowship and a this paper to fruition. This study has benefited
from thoughtful
supportedpart of the researchthat broughtWilliamConklin,Gordon McEwan,WilliamIsbell,SusanTaylor,
commentsby Joan Gero, TerrenceD'Altroy, and PrudenceRice.
TamaraBray,severalanonymousreviewers,involved in the preparationof this paper,the issues includeareas
Despite the extensive backgroundresearch the usual disclaimersapply.
ofexpertisebeyond the archaeological.Consequently
REFERENCESCITED
Abercrombie,T. A. Department
Cosmologyin Action.UnpublishedPh.D. dissertation,
1986 ThePoliticsof Sacrifice:AnAymara Chicago.
of Anthropology,University of Chicago,
Adorno, R. Austin.
in ColonialPeru. University of Texas Press,
1986 GuamanPoma, Writingand Resistance
Allen, C. Press, Washington,D.C.
1988 The Hold Life Has. SmithsonianInstitution
Alva, W. Unlooted Tomb. National Geographic174:510-550.
1988 Discoveringthe New World'sRichest
Anders,M. WariAdministrative
Inferredfrom the Planned Site of Azangaro-
1986 Dual Organizationand Calendars University Microfilms, Ann Arbor.
Strategies.Ph.D. dissertation,CornellUniversity.
de Azangaro:Notas de advertenciaparael
modelo de wari
1990a Estructuray funcionen el sitio planificado
de Lima 72:55-95.
como estado secularcentralizado.Boletln
serrano-costena en los Andes centralesduranteel Horizonte
Medio
1990b Maymi, modelo de interaccion
(nota preliminar).Boletln de Lima 71:29-31.
Bennett,W. C. Museumof Natural
AnthropologicalPapersVol. 37. American
1939 Archaeologyof the NorthCoastof Peru.
History,New York.

THESTONE ANCESTORS 361
Cook]
Benson,E.
1974 A Man and a Feline in MochicaArt. DumbartonOaks ResearchLibraryand Collection,Washington,
D.C.
Brewster-Wray, C.
1990 Moraduchayuq:An AdministrativeCompoundAt the Site Of Huari, Peru. Ph.D. dissertation,State
University of New York, Binghamton.University Microfilms,Ann Arbor.
Browman,D.
1981 New Lighton Andean Tiwanaku.AmericanScientist69:408-419.
Brundage,B. C.
1967 Lords of Cuzeo:A History and Descriptionof the Inca People in Their Final Days. University of
OklahomaPress, Norman.
Conrad,G.
1981 CulturalMaterialism,Split Inheritance,and the Expansionof Ancient PeruvianEmpires.American
Antiquity46:3-26.
Conrad,G., and A. Demarest
1984 Religion and Empire: The Dynamics of Aztec and Inca Expansion. CambridgeUniversity Press,
Cambridge.
Cook, A. G.
1983 Aspects of State Ideology in Huari and TiwanakuIconography:The CentralDeity and Sacrificer.In
Investigationsof theAndeanPast, edited by D. Sandweiss,pp. 161-185. CornellUniversityLatinAmerican
StudiesProgram,Ithaca.
1985a Small Men and Big Questions:The Role of CarvedFigurinesin the Huari State. Paperpresentedat
the 50th Annual Meetingof the Society for AmericanArchaeology,Denver.
1985b Artand Time in the Evolutionof AndeanState Expansionism.(Peruand Bolivia).Ph.D. dissertation,
State University of New York, Binghamton.University Microfilms,Ann Arbor.
1986 The Politico-ReligiousImplicationsof the Huari OfferingTradition.In Revistadialogo andino No. 4,
edited by M. A. Rivera, pp. 203-222. Departamentode Historia y Geografia,Universidadde Tarapaca,
Arica, Chile.
1987 The Middle Horizon CeramicOfferingsfrom Conchopata.NawpaPacha 22-23:49-90.
1989 Arid EcologyUrbanismin the Ica Valley of Peru. Paperpresentedat the 88th Annual Meetingof the
AmericanAnthropologicalAssociation,Washington,D.C.
1990 Late Nazea Settlementsin the Ica Valley: A Re-evaluationof Huari Presence on the South Coast.
Paper presentedat the 18th Annual Midwest Conferenceon Andean and Amazonian Archaeologyand
Ethnohistory,Chicago.
CuestaDomingo, M.
1985 Arqueologiaandina:Peru. Imprentadel Ministeriode Cultura,Madrid.
de Avila, Francisco
1991 The HuarochirlManscript.A Testamentof Ancientand ColonialAndeanReligion. Translatedby F.
Salomon and G. L. Urioste. University of Texas Press, Austin.
Donnan, C.
1978 MocheArt of Peru. Museum of CulturalHistory, University of California,Los Angeles, California.
1988 Unravelingthe Mysteryof the Warrior-Priest.National Geographic174:551-555.
1990 An Assessment of the Validity of the Naymlap Dynasty. In The NorthernDynasties:Kingshipand
Statecraftin Chimor,edited by M. Moseley and A. Cordy-Collins,DumbartonOaks ResearchLibraryand
Collection,pp. 243-274. Washington,D.C.
Doran, J. E., and F. R. Hodson
1975 Mathematicsand Computersin Archaeology.HarvardUniversity Press, Cambridge.
Duviols, P.
1986 Culturaandina y repression:Procesosy visitas de idolatrfasy hechicerfas,Cajatambo,siglo XYII.
Archivosde HistoriaAndina,vol. 5. Centrode EstudiosRuralesAndinos "Bartolomede las Casas,"Cuzeo.
Engel,F. A.
1976 An Ancient WorldPreserved,Relics and Recordsof Prehistoryin the Andes. Crown,New York.
Frame,M.
1990 AndeanFour-CorneredHats:AncientFolumes.MetropolitanMuseumofArt,NewYork.
Gero, J., and M. Conkey(editors)
1991 EngenderingArchaeology.Basil Blackwell,Oxford.
Guaman,Poma de Ayala, F.
1980 [ca. 16151 El primer nueva coronicay buen gobierno.3 vols. Edited by J. V. Murraand R. Adorno
and translatedby J. L. Urioste. 3 vols. Siglo Veintiuno, Mexico, D.F.
Harris,O.
1980 The Power of Signs:Gender, Culture,and the Wild in the Bolivian Andes. In Nature, Culture,and
Gender,edited by C. MacCormackand M. Strathern,pp. 70-94. CambridgeUniversity Press, Cambridge.
1982 The Dean and the Devils Among the Bolivian Laymi. In Death and the Regenerationof Life, edited
by M. Block and J. Parry,pp. 45-73. CambridgeUniversity Press, Cambridge.

3, No. 4, 1992
[Vol.
AMERICAN ANTIQUITY
LATIN
362
Hernandez Principe, R.
1.
1923[1622] Mitologia andina. Inca
Hocquenghem, A. M., and P. J. Lyon Nawpa Pacha 18:27-48.
Supernatural Females In Moche Iconography.
1980 A Class of Anthropomorphic
InternationalCongress of Americanists del XXVI Congreso Inter-
Larrea. Publicacion del comite organizador Madrid.
1935 Arte peruano. Coleccion Juan octubre 1935. Tipografia de Archivos,
12 de
nacional de Americanistas, Sevilla,
B. J.
Isbell,
Press, Prospect Heights, Illinois.
1978 To Defend Ourselves. Waveland
W. H. 10. University of Illinois
Isbell, Illinois Studies in Anthropology No.
1977 The Rural Foundations of Urbanism.
Press, Urbana. Andes, edited by A. Kendall,
Archaeological Projects in the Central
1984 Huari Urban Prehistory. In Current British Archaeological Reports, Oxford.
210.
pp. 95-131. BAR International Series Middle Horizon. In Andean Ar-
and State at Wari, Ayacucho, Peru, During the
1986 Emergence of City
Evans, edited by R. Matos M., S. A. Turpin, and H. H. Eling, Jr.,
chaeology: Papers in Memory of Cliffiord University of California, Los Angeles.
Institute of Archaeology,
pp. 189-200, Monograph No. XXVII.
W. H., and A. G. Cook
Isbell, Peru. Archaeology 40(4):26-
Origin of Expansionist States in Ancient
1987 Ideological Innovation and the
33.
W., and K. Schreiber
Isbell,
Antiquity 43:372-389.
1978 Was Huari a State? American
G. F. McEwan (editors)
Isbell,W. H., and and State Government. Dumbarton
Prehistoric Monumental Architecture
1991 Huari Political Organization: D.C.
Washington,
Oaks Research Library and Collection,
Jones, J.
Museum of Primitive Art, New York.
1964 Art of Empire: The Inca of Peru.
Larco Hoyle, R.
Publishing, Cleveland.
1966 Peru. Arqueologia Mundi. World
Larrea, J.
Nacional de Cordoba, Argentina.
1960 Corona incaica. Universidad
Lumbreras, L. de Letras, Universidad Nacional
Arqueologia 1(1):130-227. Facultad
1960 La cultura wari. Etnologia y
Mayor de San Marcos, Lima. Institution Press,
Peru. Translated by B. J. Meggers. Smithsonian
1974 The Peoples and Cultures of Ancient
Washington, D.C. Lima.
Peru II:9-91. Editorial Mejia Baca,
1980 E1 Imperio Wari. Historia del
Lyon,P.
Peru. Nawpa Pacha 16:95-140.
1979 Female Supernaturals in Ancient
MacCormack, S.
Incas. Representations 33:121-143.
1991 Demons, Imagination and the
McEwan, G. Occupation of Pikillacta in
of Cuzeo, Peru: The Impact of the Wari
1984 The AIiddle Horizon in the Valley Microfilms, Ann Arbor.
University of Texas, Austin. University
the Lucre Basin. Ph.D. dissertation,
Means, P. Charles Scribner's Sons, New York.
1931 Ancient Civilizations of the Andes.
Medlin, M. A. Textiles of Peru and Bolivia, edited
Woven Design. In A ndean Aesthetics:
1987 Calcha Pallay and the Uses of of Wisconsin, Madison.
of the University
by B. Femenias, pp. 60-66. Regents
Menzel, D.
Horizon. Nawpa Pacha 2:1-105.
1964 Style and Time in the Middle Pacha 6:47-114.
in Middle Horizon Epoch 2A. Nawpa
1969 New Data on the Huari Empire the Work of Max Uhle. R. H . Lowie Museum of Anthropology,
and
1977 The A rchaeology of A ncient Peru
University of California, Berkeley.
Moseley, M., and A. Cordy-Collins Oaks Research Library and
and Statecraft in Chimor. Dumbarton
1990 The Northern Dynasties: Kingship
Collection, Washington, D.C.
Murra, J. 64:710-728.
Inca State. American Anthropologist
1962 Cloth and Its Function in the
Ramos, L. J., and M. C. Blasco textiles y pequenas escultura
en el Museo de America: Orfebreria,
1977 Materiales huaris no ceramicas 4:55-108.
Prehispanicos
(figurillas de Pikillacta). Cuadernos
Ravines, R. del Peru. Nawpa Pacha 6 (1968):
Horizonte Medio en la sierra central
1969 Un deposito de ofrendas del
1946. Lima.
Peru. Edicion de Petroleos del Peru,
1970 100 anos de arqueologia en el

THE STONE ANCESTORS 363
Cookl
1977 Excavaciones en Ayapata, Huancavelica, Peru. Nawpa Pacha 15:49-100.
Rostworowski De Diez Canseco, M.
1970 Mercaderes del valle de Chincha en la epoca prehispanica: Un documento y unos comentarios. Revista
Espanola de A ntropologia A mericana 5: 135- 177.
Rowe, A. P.
1979 Textile Evidence for Huari Music. Textile Museum Journal 18:5-18.
1984 Textiles from the Nasca Valley at the Time of the Fall of the Huari Empire. In The Junius B. Bird
Conference on Andean Textiles, edited by A. Rowe, pp. 152-182. Textile Museum, Washington, D.C.
1991 Nasca Figurines and Costume. Textile Museum Journal 29-30:93-128.
Rowe, J. H.
1946 Inca Culture at the Time of the Spanish Conquest. In TheAndean Civilizations, edited by J. H. Steward,
pp. l 93-220. Handbook of South American Indians, vol.2. Bulletin No.143. Bureau of American Ethnology,
Smithsonian Institution, Washington, D.C.
1948 The Kingdom of Chimor. Acta Americana 6(1-2):26-59.
Rowe, J. H., D. Collier, and G. R. Willey
1950 Reconnaissance Notes on the Site of Huari, Near Ayacucho, Peru. American Antiquity 16: 12s137.
Ruppert, H.
1982 Zur Verbreitung und Herkunft von Turkis und Sodalith in Prakolumbischen Kulturen der Kordilleren.
Baessler archiv Beitrage zur Volkerkunde 30:69-124. Berlin.
Sarmiento de Gamboa, P.
1942 [1572] Historia de los Incas. Emece Editores, Buenos Aires.
Schreiber, K.
1992 Wari Imperialism in Middle Horizon Peru. Anthropological Papers No. 87. Museum of Anthropology,
University of Michigan, Ann Arbor.
Shea, D. E.
1969 Wari-Wilka: A Central Andean Oracle Site. Ph.D. dissertation, University of Wisconsin, Madison.
University Microfilms, Ann Arbor.
Silverblatt, I.
1987 Moon, Sun and Witches: Gender Ideologies and Class in Inca and Colonial Peru. Princeton University
Press, Princeton.
Trimborn, H., and P. F. Vega
1935 Catalogo de la exposicion arte inca. Imprenta Martosa, Madrid.
Urton, G.
1990 The History of a Myth. Texas University Press, Austin.
Valcarcel, L. E.
1933 Esculturas de Pikillajta. Revista del Museo Nacional II(1):3-27. Lima.
Vreeland, J.
1980 Mummies of Peru. In Mummies, Disease, and Ancient Cultures, edited by A. Cockburn and E. Cock-
burn, pp. 135- 174. Cambridge University Press, Cambridge.
Weigand, P. C., G. Harbottle, and E. Sayre
1977 Turquoise Sources and Source Analysis: Mesoamerica and the Southwestern U.S.A. In Exchange
Systems in Prehistory, edited by T. Earle and J. Ericson, pp. 15-32. Academic Press, New York.
Wobst, M.
1977 Stylistic Behavior and Information Exchange. In For the Director: Research Essays in Honor of James
H. Grifgin,edited by C. E. Cleland, pp. 317-342. Anthropological Papers No. 61. Museum of Anthropology,
University of Michigan, Ann Arbor.
Zorn, E.
1987 Encircling Meaning: Economics and Aesthetics in Taquile, Peru. In Andean Aesthetics: Textiles of Peru
and Bolivia, edited by B. Femenias, pp. 67-80. Regents of the University of Wisconsin, Madison.
Zuidema, T.
1964 The Ceque System of Cuzeo: The Social Organization of the Capital of the Inca. Translated by E. M.
Hooykaas. International Archives of Ethnography, supplement to vol. L. Brill, Leiden.
1972 Meaning in Nazea Art. In Goteborgs EthnograJiska Museum Arstryck, pp. 35-54. Goteborg.
1973 Kinship and Ancestor Cult in Three Peruvian Communities. Hernandez Principe's Account of 1622.
Bulletin de l'Institut Franfais d'Etudes Andines 2:16-33.
1977-1978 Shafttombs and the Inca Empire. Journal of the Steward Anthropological Society 9(1-2):133-178.
1990 Inca Civilization in Cuzvo. Translated by J-J. Decoster. University of Texas Press, Austin.
1992 Guaman Poma and the Art of Empire: Towards an Iconography of Inca Royal Dress. Transatlantic
Encounters, edited by K. Adrien and R. Adorno, pp. 151-277. University of California Press, Berkeley.
NOTES
' The term "turquoise" is generally used in two distinct ways: "[T]he narrower definition, which is a chemical
one, and a broader designation, a cultural term embracing a whole range of blue and blue-green stones" (Weigand
et al. 1977:16).

364 LATIN AMERICAN ANTIQUITY [Vol. 3, No. 4, 1992]
2 Ceramicfigurinesare not included in this discussion but they constitutean importantbody of information
that warrantsfurtherstudy.
3 A layer of fine sand sealed one of two offierings
excavatedat the site of Conchopatain 1977 (Cook 1985b,
1986, 1987; Isbell and Cook 1987). This featureadds evidence that the figurinedeposits were treatedat burial
in a similar manneras offeringswithin the Huari heartland.
4 Both Valcarcel(1933) and Trimbornand Vega (1935) note the presenceof marine-snailshell. The drawing
in Valcarcel(1933:Figure2) correspondsto a Strombusshell, referredto in Quechuaas a pututu,and which is

still used today as a musical instrumentin rituals.
5 The dataarebasedon individualfigurine descriptionsand heightmeasurementsprovidedby Valcarcel(1933)
for the Cuzeo Collectionand by Ramos and Blasco (1977) for the MadridCollection.Weightcould not be used
as a variablebecauseit has not been reportedfor all of the figurines.
6 At presentthere are no publishedHuari elite burialsthat have been systematicallyexcavated so that com-
parisons must be made with rank indicatorsthat are known from similar time periods or have documented
continuity.
7 From 1988 to 1990 I directed the Lower Ica Valley Survey (Cook 1989, 1990). The researchwas made
possiblewith an AmericanRepublicsFulbrightResearchGrant,two summergrantsfromthe CatholicUniversity

ResearchFund, and most recentlyby the H. John Heinz III CharitableTrust.The finalreportis in preparation
as are severalarticleson the findings.
8 It is instructiveto recognizethat greenstonehumanfigurineshave not been found in associationwith Moche
materialculture(with the exception of the greenstonehuman figurewith a feline describedby Benson [19741).
Evidencerecentlyreportedfrom the site of Cunturwasiin Cajamarcarevealedgreenstoneidols from an Early
Horizon Chavin-relatedcontext (Onuki,personalcommunication 1991) but these remainthe only pre-Middle
Horizon examples I have encountered.
ReceivedApril 17, 1991; acceptedJanuary 7, 1992


Cook Huari Figurines

Hochgeladen von

Dokumentinformationen

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Cook Huari Figurines

Hochgeladen von

Copyright:

Verfügbare Formate

THE STONE ANCESTORS: IDIOMS OF IMPERIAL

A lsl lRE AND RANK AMONG HUARI FIGURINES

Anita G. Cook, Departmentof Anthropology,The Catholic Universityof America, Washington,DC 20064

Latin American Antiquity, 3(4), 1992, pp. 341-364.

This content downloaded from 129.252.86.83 on Wed, 12 Mar 2014 20:37:11 PM

Figure 1. Map of Peru showinglocationsof sites discussedin the text.

locations. Somewhat surprisingly,ceramics and architectureremain the principal archaeological

This content downloaded from 129.252.86.83 on Wed, 12 Mar 2014 20:37:11 PM

Figure2. Examplesof the Pikillacta figurinesfromthe Madridand Cuzcocollections:(a) CuzcoCollection

or profileanthropomorphicfiguresin relativelystandardizedposes. Parallelsare drawnfrequently

This content downloaded from 129.252.86.83 on Wed, 12 Mar 2014 20:37:11 PM

context of these figurinesthat bears upon

This content downloaded from 129.252.86.83 on Wed, 12 Mar 2014 20:37:11 PM

THE STONE ANCESTORS

*--- :-*- -------:-:- =S

Figure3. Plan of Structure34-2B at Pikillacta, wherethe two turquoisecaches reportedlywere found.The

CUZCO COLLECTION MADRID COLLECTION

This content downloaded from 129.252.86.83 on Wed, 12 Mar 2014 20:37:11 PM

This content downloaded from 129.252.86.83 on Wed, 12 Mar 2014 20:37:11 PM

This content downloaded from 129.252.86.83 on Wed, 12 Mar 2014 20:37:11 PM

348 LATIN AMERICANANTIQUITY [Vol. 3, No. 4, 1992

and headgearthat are stronglyreminiscentof early colonial descriptionsof lowland tropicalforest

This content downloaded from 129.252.86.83 on Wed, 12 Mar 2014 20:37:11 PM

Tunic and Headdress Variation

This content downloaded from 129.252.86.83 on Wed, 12 Mar 2014 20:37:11 PM

Table 1. Comparisonof GarmentTypes in the Cuzco and

wrapped around its neck and waist.

This content downloaded from 129.252.86.83 on Wed, 12 Mar 2014 20:37:11 PM

Cuzco Collection Madrid Collection

Figure9. Figurinesize frequencyhistograms.

Figurine Height Comparisons

This content downloaded from 129.252.86.83 on Wed, 12 Mar 2014 20:37:11 PM

This content downloaded from 129.252.86.83 on Wed, 12 Mar 2014 20:37:11 PM

This content downloaded from 129.252.86.83 on Wed, 12 Mar 2014 20:37:11 PM

This content downloaded from 129.252.86.83 on Wed, 12 Mar 2014 20:37:11 PM

This content downloaded from 129.252.86.83 on Wed, 12 Mar 2014 20:37:11 PM

This content downloaded from 129.252.86.83 on Wed, 12 Mar 2014 20:37:11 PM

This content downloaded from 129.252.86.83 on Wed, 12 Mar 2014 20:37:11 PM

on the basis of comparisonwith a similar structureat the site of WariWillkanear Huancayo.These

This content downloaded from 129.252.86.83 on Wed, 12 Mar 2014 20:37:11 PM

This content downloaded from 129.252.86.83 on Wed, 12 Mar 2014 20:37:11 PM

This content downloaded from 129.252.86.83 on Wed, 12 Mar 2014 20:37:11 PM

This content downloaded from 129.252.86.83 on Wed, 12 Mar 2014 20:37:11 PM

This content downloaded from 129.252.86.83 on Wed, 12 Mar 2014 20:37:11 PM

This content downloaded from 129.252.86.83 on Wed, 12 Mar 2014 20:37:11 PM

in Valcarcel(1933:Figure2) correspondsto a Strombusshell, referredto in Quechuaas a pututu,and which is

possiblewith an AmericanRepublicsFulbrightResearchGrant,two summergrantsfromthe CatholicUniversity

ReceivedApril 17, 1991; acceptedJanuary 7, 1992

This content downloaded from 129.252.86.83 on Wed, 12 Mar 2014 20:37:11 PM

Das könnte Ihnen auch gefallen

--- :-- -------:-:- =S