Society for American Archaeology

Mortuary Preferences: A Wari Culture Case Study from Middle Horizon Peru
Author(s): William H. Isbell
Source: Latin American Antiquity, Vol. 15, No. 1 (Mar., 2004), pp. 3-32
Published by: Society for American Archaeology
Stable URL: http://www.jstor.org/stable/4141562
Accessed: 22/09/2008 12:14
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=sam.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the
scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that
promotes the discovery and use of these resources. For more information about JSTOR, please contact support@jstor.org.

Society for American Archaeology is collaborating with JSTOR to digitize, preserve and extend access to Latin
American Antiquity.

http://www.jstor.org

ARTILES

MORTUARY PREFERENCES: A WARI1CULTURE CASE STUDY FROM

MIDDLE HORIZON PERU
William H. Isbell

Mortuarypractices reveal a great deal about the social organization of prehistoric cultures and their landscape of places.
However, tombs are favored targetsfor looters, making it difficult to determine original burial practices. Verylittle was
knownabout Wariburial during the Middle Horizon (A.D. 500-1000), even though Wariwas an imperial, early BronzeAge
culture with a spectacular urbancapital in highland Peru. Excavations at the secondary Waricity of Conchopataproduced
remains of more than 200 individuals,from disturbedand undisturbedcontexts. These burials as well as informationfrom
other sites permit an initial description of ideal patterns of Warimortuarybehavior. Theforms abstracted reveal graves
rangingfrom poor and ordinary citizens to royal potentates, supporting inferences of hierarchical political organization.
It is also clear that the living accessed graves of importantpeople frequently, implying some form of ancestor worship.
However,unlike the later Inkas, Wariancestors were venerated in their tombs, located deep within residential compounds
and palaces.
El estudiode las prdcticasfunerariases invalorablepara el conocimientode las culturasprehist6ricasy los pueblos antiguos.
Desgraciadamente,las tumbasson tambie'nel blancofavorito de los saqueadores,por lo que resultadificil en muchoscasos
interpretarlas prdcticas originales. Pese a la importanciade una cultura como Wari,un imperiode la Edad del Bronce que
tuvo una espectacular capital urbana en la sierra del Pert, conocemos muypoco respecto a sus prdcticas funerarias. Las
recientes excavaciones en la ciudad secundaria wari de Conchopatahan permitido recuperarrestos humanos,en contextos
funerarios disturbadosy no disturbados,correspondientesa ma'sde 200 individuos.Estos entierrosy la informacidndisponible
de otros sitios waris (incluyendoal centro urbanode Huari) hacen posible plantear una descripcio'ninicial de patrones ideales de la conductafunerariawari duranteel HorizonteMedio (500-1000 d.C.). Lasformas interpretadasrevelantumbasque
correspondentantoa ciudadanospobres y ordinarioscomo a gobernantesreales.Ademds,las tumbasde las personas importantespresentanevidencias de haber sido abiertas con frecuencia luego del entierro,implicandoalgunaforma de culto a los
ancestros.

studiesof tombsandmor- into builtenvironmentsof the past (Bradley1989,

Archaeological
tuaryremainshavebeencriticalforunder-

standingthe prehistoricpast since at least

Sir LeonardWoolley's (1934) discovery of the
Royal Cemeteriesof Ur. In the 1970s, grave and
cemeteryanalysisbecamemorerigorousandsystematicwiththe methodologicalinnovationsassociatedwith processualarchaeology(Brown 1971;
Goldstein 1980, 1981; Saxe 1970; Tainter1978).
Postprocessual archaeology and the study of
ancientlandscapesoffera potentialforevenbroader
understandingsfrommortuarystudies,examining
places of the dead as spatialmetaphorsinscribed

1998;Cannon1989,2002; Carr1995;ParkerPearson 1982, 1993, 2002; Silverman2002; Thomas

1996).These landscapesof deathweredesignedto
communicate,so archaeologistswouldnotbe doing
theirjobs if they rejected the hermeneuticchallenge to readandinterpretthem.However,meaningful understandingsdepend on archaeologists
determininghow ancientpeopleintendedburialto
be conducted.Thisis usuallymoredifficultto determine thanimagined.
Graves,andespeciallytheintermentsof importantindividuals,arealmostalwaystargetsof plun-

William H. Isbell a Departmentof Anthropology,State Universityof New Yorkat Binghamton,Binghamton,NY

13902-6000
LatinAmericanAntiquity,15(1), 2004, pp. 3-32
Copyright@2004 by the Society for AmericanArchaeology
3

LATINAMERICAN
ANTIQUITY

der and destruction.Tombs were loci of power

withintheirsocial arenas,makingthemtargetsof
aggression. They frequently contain significant
wealth, attractinglooters. Furthermore,mortuary
behaviormaynotrepresentan event,buta process,
consistingof a sequenceof acts over an extended
periodof time. How can the archaeologistdifferentiate the opening of a grave to add a newly
deceased member of the family, to remove an
ancestor'sbones, or to participatein some activity
fromrobbinga gravefor its wealth?At leastin part
becauseof this problem,thereareno generalsyntheses of mortuary practices for prehispanic
Andeanculturessuchas highlandChavin,Recuay,
Pucara,Tiwanaku,or Wari.
Archaeologists'discussionsof landscapesof the
deadmustbe basedon intendedconditionsof interment.Butthe archaeologicalrecordpresentssnapshots of complexprocesses,some intendedby the
mournersbut othersresultingfrom looting, construction,erosion,etc., frozen as confusing material contexts. An inventory of popular burial
patternsmustemphasizethe originalideals.While
this obscuresvariationand inferencesaboutindividual agency, in the long run the abstractionof
ideal patternsor normsseeks to recognizeculturally relevantdistinctions,on the basis of which
organizational structure may be inferred, and
observedrangesof behaviorcan be morecogently
discussed.To abstractintendedor idealpatternsan
archaeologistmust work qualitatively,evaluating
as manymortuarycontextsfrom the same culture
andtime periodas possible. Effects from destructive processessuch as lootingmustbe evaluatedin
opposition to impacts from intended mortuary
processesthatmayhavegone on overa long period
of time, such as refurbishinggrave goods. These
effects must be distinguished from differences
intendedto express status,class, gender,age, or
other socially relevant variables. No explicit
methodology exists to assure success, although
large,carefullyexcavatedsamplesareessential.
In the archaeologicalstudy of Warimortuary
behavior,it was impossibleto move directlyfrom
excavationdatato prehistoricactivity.Information
was confusing and contradictory,in large part
becauseso manymortuarycontextsweredisturbed.
Waseverypitandchamberwitha few humanbones
a tombthathadbeen looted?Orhadhumanbones
been trophiesor amuletsthatwere depositedhere

[Vol.15, No. 1,2004

and there,and not exclusively in tombs?Was any

disturbancea result of looting, or had mortuary
practices been a prolonged process involving
reopeninga grave several times? Was secondary
buriala Waripracticeor did bones become disarticulatedby otherpost-intermentprocesses?It was
only throughcomparisonof many cases thatpatternsbeganto emerge.Unfortunately,information
hasbeenpoorlyrecordedformanyyears;therefore
comparativedata were not accumulatingquickly.
ArchaeologistsdiscoveringdisturbedWariburials
paidlittleattention,for they appearedto offeronly
insignificantscrapsof informationaboutthe past.
More recently,it has become clear thateven disturbedremainsare valuablefor comparativepurposes, when carefullydescribed.
Archaeologistsengaged in inferringpast culturalpatternsmustavoidexcessiveinfluencesfrom
theory and expectations in their comparative
abstractionof ideals and norms. If we employ
favoredtheoreticalconvictionsoranalogiesto help
inferintendedburialformsandmortuaryprocesses
and thengo on to use the sametheoryto infercultural meanings,our results become overly laden
with theoretical conviction (Isbell 1995; Wylie
1992a, 1992b).Forexample,JalhDulanto(2002)
describesscatteredhumanbones and theirspatial
contextsfor a firstmillenniumB.C. settlementon
Peru'scentralcoast thatimply an ideal involving
processingof ancestors'remainsin a mannerquite
foreignto anythingknowninAndeanethnohistory.
However,his convictionsaboutcontinuityin Inka
ancestorworshipandmortuarypracticeslead him
to emphasizesimilaritiesto ethnohistorical
descriptions while de-emphasizingdifferences.The outcome is preferredpatternsmoresimilarto those of
the Inkathanwarrantedby the actualdata.
Intended Patterns of Death at Conchopata
This studyis possiblebecauseof recentexcavations
at Huari's secondarycity of Conchopata.2They
have revealedthe remainsof more than200 individuals from burialcontexts of the Wariculture.
Conchopatais one of many Wari capitals, secondarycities, provincialcenters,andcommunities
(Figure 1) that were spread across the Central
Andes during the Middle Horizon (A.D.
550-1000). Most archaeologistsinterpretHuarias
thecapitalof a vastimperialstateof the samename

Isbell]

A WARICULTURE
MORTUARY
PREFERENCES:
CASESTUDYFROMMIDDLE
HORIZON
PERU

IKalki n Palacio

PampaGrande4?
Yarnobabanba
Jos
San
de
Mar
\ j-'k
cac
ampa
Moche

Honcn

mp,

stHuari

Sphere
ariWifica

tnnc
i

v,

opata
aJargampata
opataaPikillactacha
Huaro

May-mis>
XJinolca
Pacheco
el
ac

del

I.

t
a

Iwawi

Tiw anak
Bauf

ro

,--I ------

Lukuro-Uyu

aniKal~amarca

Nno

t
W

e WariCenters
A

Wari Influenced Centers

TiwanakuCenters

Tiwanaku

phere
SanPedro

deAtcama

500 Km.

Figure 1. Central Andes showing Middle Horizon centers including the capital, Huari, provincial Wari cities and other
contemporary capitals.

LATINAMERICAN
ANTIQUITY

withits networkof centersdocumentingits administrativestructure(Isbell 1983, 1985, 1991, 1997b;

Isbell and Schreiber 1978; Lumbreras 1974a,
1974b, 1985; McEwan 1991, 1996, 1998;
Schreiber1991, 1992). But agreementis not universal.AlternativepositionsconsiderHuariandthe
Middle Horizon a confederationof lineages (J.
Topic 1986, 1991, 1994; T. Topic 1991; J. Topic
andT. Topic 1992, 2001), or a mosaic of independentcities engagedin intensivecommerce(Shady
1982, 1988;ShadyandRuiz 1979).Of course,new
of thelandscapeof deathduringthe
understandings
MiddleHorizoncan help resolve this debate.
Conchopatais locatedin the southernendof the
AyacuchoValley,about10km fromthecapitalcity
of Huari(Figure2). It has a long historyof occupation,but duringthe centurieswhen Huaridominatedmuch of Peru,Conchopatawas the second
city of the imperialheartlandandthe largesturban
centerin theAyacuchoValley'ssouthernsettlement
enclave.Todayits ruinsareoverrunby the modem
city of Ayacucho,resultingin the destructionof
mostof the ancientarchaeologicalzone (Figure2).
Originallythesettlementcoveredatleast20 ha,and
possibly as much as 40 ha. Presently,only about
threeha remain,probablythe focus of the original
civic center.All of our new informationaboutthe
dead comes from this tiny portionof the old city
(Isbell 2001a). However, this well-documented
sample of some 200 individualsis probablythe
largest collection of archaeologicallyexcavated
burialsfrom the Wariheartland.All come from a
denselyurbanizedareaof moreor less continuous
buildings,plazas, and patios (Figure3). At some
time, mostof this survivingportionof Conchopata
may have been enclosed by a perimeterwall, of
whicha northwestanda southeastcomerhavebeen
preserved.Be thatas it may,Conchopatawas long
recognizedas a communityof pottersbecauselarge
tools were disnumbersof ceramicmanufacturing
coveredatthe site (Pozzi-Escot1985, 1991;PozziEscotet. al 1994, 1998).However,once we learned
to recognizemortuaryarchitectureandhow it varied with status,it became clear thatthe surviving
portion of Conchopata contained tombs that
includedelaborateandwealthyexamples.The site
couldnothavebeen a townof craftspeopleof more
or less middle status. Rather,it appearsto have
been a landscapeof palace compoundsoccupied
by lowly servants,middle-levelcitizens, wealthy

[Vol.15, No. 1, 2004

elites, andprobablyeven pettykings or governors.

Architecture,stratigraphy,ceramic styles, and
radiocarbondatesrevealfive phasesof occupation
at Conchopata.During the Huamaniphase (240
B.C.-A.D. 300) we know that Conchopatawas
occupied,butlittleculturalmaterialcanbe assigned
to this time. During the Mendosa phase (A.D.
300-550), Huarpaand Curz Pata pottery styles
were in use. Severalgraveswere discoveredin the
north-central
portionof survivingConchopata,but
burials
these
representa distinctpatternof interment. Modest tombs appearto have been located
in an open areawith no architecture,close enough
to one anotherto imply a cemetery.Bodies were
flexed and placed in simple pits or cavities in the
bedrock,frequentlyaccompaniedby one or more
ceramicvessels, andprobablyby perishableitems
as well. Anothergrave,reportedlydiscoveredby
earthmoverswhile leveling the landingstripseveral hundredmeters southeastof our excavation
area,containedCurzPatapottery,so it alsobelonged
to the Huamaniphase. But it is reportedto have
been a bottle-shapedshaft tomb with a skeleton
extendedon its back (Lumbreras1974:112a). No
otherbottle-shapedshafttombs or extendedburials areknownat Conchopata.
The Silva phase (A.D. 550-700) initiatedthe
MiddleHorizonatConchopataandis characterized
by oversize Conchopata-styleceramicsas well as
Chakipampaand Ocros pottery.Less-fancy potteryusuallydesignatedHuamangawas also in use.
There is a great deal of evidence for large-scale
building at Conchopata,althoughmany of these
early buildingswere disturbedby later activities.
The remainsdocumenta significantchangein the
landscapeof the deadbetweenthe lateEarlyIntermediateperiodandMiddleHorizontimesthatcontinuedthroughthe Silva phase as well as the next
two phasesat Conchopata.Humanbodieswere no
longer placed in open cemeteries but below the
floorsof roomsandpatios.Theseroomsandcourts
were partsof extensive buildingcompoundsand
because, as discussed below, at least some of the
burialswerereveredandgiven offeringslong after
death,groupsof descendentsmusthaveresidedin,
and expectedto remainin chargeof, the residentialcompoundof theirancestors.TheMiddleHorizon landscape of the dead constructed a new
association between large building compounds,
ancestors,and a social group that was probably

Isbell]

MORTUARY
PREFERENCES:
A WARICULTURE
CASESTUDYFROMMIDDLE
HORIZON
PERU

THnawasin.
Hu ta

ghway to Huanta
San Miguel
River

BaB.s
Azangain

04

River
T
ablapm
Twacochba
im
hrao
pap *4.
Huamanga
BaaBasin
MyuyuOrqo4

MilitaryBase

Huari

eroChuru

Jargampata

Conchopata, "

._
Nawinpukyu

AyacuchoValleyho

Moder

Ayacucho
urban

Ayaccho~lle

Q)

L~9

ConchopataArchaeologicalZone
0Civic
Center
SuburbanPeriphery
Edge of Mesa
Figure 2. Map of Ayacucho Valley and the Conchopata archaeological zone.

ANTIQUITY
LATINAMERICAN

based on descent. Conchopata'slargestarchitecturalcomplexesseem to havepalacesoccupiedby

rulersor governors.
Itis probablethatthemortuarycomplexin room
EA-203 belongs to the Silva phase (Figure3), but
it was excavatedyearsago and,apparentlybecause
of severe looting, it was neverdescribedin print.
This tomb complex belongs to Type5a of the following proposedtypology andit could be the earliest "mortuary
room"atConchopata,representing
the firstelite gravecomplexconstructedunderthe
floorsof a palace
TheHuisaphase(A.D. 700-850) was themajor
occupationat the Conchopatasite. Oversize-Conchopatapotterycontinuedin use, butprobablydisappearedbefore the end of this time. Huamanga,
Chikipampa,and Ocros potterystyles were very
popular.Huisa is the phase to which the majority
of the burialsemployedin this analysisappearto
belong, althoughit seems thatthe most elaborate
tombs continuedin use though the final Alarcon
phase (A.D. 850-1000). Duringthatphase, there
is no evidence for constructionor occupationof
palacesexcept for the tombsthatwere still in use,
or perhapsbeing reused.However,Alarconphase
rooms nearbyhave simple tombs that are consistent with the proposedtypology. Huamangapottery was popularin Alarcontimes, but occasional
pieces of VifiaqueandAtarcostyle ceramicsalso
appear.
MortuaryremainsfromConchopata'sfinalthree
phases seem very similar,at least on the basis of
currentdata,so descriptionsfrom all threephases
were combined.Along with less detailedinformation as well as restudyof undescribedgravesfrom
formerexcavations,theyprovidethedataon which
thefollowingpreferentialpatternsarebased.Many
of the tombs sufferedsignificantdisturbance,but
some were intact.However,even damagedtombs
furnishedvaluableinformation.
Conchopata's Middle Horizon mortuary
remainsappearto fallintosevenpreferentialgroups
or idealtypesof interment,describedbelow.I omit
at Conchopata,in which
one type of "non-burial"
human remains were deliberatelydefleshed and
disarticulated before they were eventually
depositedon thefloorsof templebuildings.Norwill
I exploreinfantandchildburialsexceptwhenthey
co-occurwith adultburials.3I will make comparisons withmortuarycontextsfromHuariandother

[Vol.15, No. 1, 2004

Middle Horizoncities in the centralhighlandsto

fill out Wari'sculturalrecord,and to confirmits
mortuaryideals.
WariBurial Type1-Individual Interment
Thisformof burialconsistsof a singlebodyplaced
in a small pit excavatedinto the groundand coveredwithearth(Figure4). Sometimesthegravewas
cappedwith a flatstone or two, andoccasionallya
few flatstoneswereusedto line the sides of thepit.
Bodies appear to have been tightly flexed and
placedin the graveeitherseated,on the back,or on
one side.Tracesof textilesandcordagesuggestthat
at least some bodies were wrappedin cloth and
bound with rope. Examples appearto have been
locatedin patios,courts,andnarrowrooms.Except
when a stone slab was used to cap the pit, thereis
no evidencethatthesegravelocationsweremarked.
Occasionally,Type 1 gravesincludea ceramicvessel, a stone bead, or some other object, but typically, imperishablegravefurnishingsare absent.
WariBurial Type2-Multiple Interments
Undisturbedmultiple intermentswere found in
ArchitecturalEnclosureEA-65 and EA-151 (Figure3). Both were probablyopen patioareasrather
thanroofedchambers.Like individualinterments,
multipleintermentsconsistof unlinedpitscovered
by soil, andperhapsa stoneor two, with few or no
gravefurnishings,andthe flexedremainsof two to
fouror fivebodiesof adultsandsub-adultchildren.
It is apparentthatburialscould be addedto these
gravesas time passed,so it seems likely thatType
1 intermentsturnedinto Type2. Like Type 1 individual graves,Type 2 graves show little evidence
for markingof theirlocations.However,theywere
reopenedfor subsequentburials,so people of the
communitymust have rememberedthe locations
of thegraves.Perhapsthereweremarkersthathave
now disappeared.
Future bioarchaeological study will show
andmovedabout
whetherbonesfounddisarticulated
in Type2 graveswere movedsimplyto accommodatethe additionof morebodies,or whethersome
more elaborateactivitieswere involved.It may be
thatMultipleIntermentgravescontainedmembers
of the samefamilyor socialgroup.
WariBurial Type3-Cist Interment
This importantclass of Middle Horizon graves

Isbell]

MORTUARY
A WARICULTURE
HORIZON
PERU
PREFERENCES:
CASESTUDYFROMMIDDLE

CONCHOPATA
ARCHAEOLOGICAL
ZONE
10 meter grid

fla

11

ii

.................

....I

41

J<

lPlaza.

.....

C~h~w-3-

____lr

................Pi
....l...................
.............
....... ......
... ... . .. ---.
-..

II

ff

rAY

~~

.....41
I

"

# ............................................
I

ZL:I

S~--it

__

_ _ _

_7

1//

ii7i

,Z--

1-Y

--

-_

__

_ _ _

Figure 3. Map of Conchopata's civic center architectural remains.

(Figure4) is poorlyknownat Conchopata,for only

one undisturbedexamplewas discoveredin room
EA-205 (Figure 3). It contained an older adult
femalewithtwo ceramicvessels, butthe gravewas
only partially stone lined with a clay top as its
markerratherthana rockwitha perforationthrough
it. A small hole reachedthe grave,passingthough
a wallto theadjoiningroom.A secondlootedexample was foundin the PinkPlaza(Figure3), incompletelycappedby severalflatstonesandcontaining
the partialremainsof a single individual,a tupu,4
anda distinctivepolychromeceramicsherd.However,because cist intermentsare frequentat other
sites in theAyacuchoValley,the type deservessignificantattention.
Cist tombs probablywere markedgravesconsisting of cylindricalpits, fully or partiallystonelined, about60 to 90 cm in diameterand 60 cm to
1 m deep. They are known throughoutAyacucho
and many were sealed with a large, flat, circular
stone or by several smaller slabs of rock. Some-

times thereis a notchin one side of the cover,or a

hole about 10 cm in diameterpecked throughthe
middle of a single stone lid. Sometimesthereis a
small niche in the wall of the cist, or a grooverunning down one side. Where grooves have been
found, they appearto align with the notch in the
lid. Cist tombs probablyrepresenta more lavish
version of Type 1 and Type 2 graves. Their distinctivelids servedto markthe gravelocationand
probablyalso facilitatedreopening the tombs. I
suspectthatType3 tombsweredesignedto receive
successiveburialsovera periodof time.Whenthey
havea notchor hole throughthe lid, this musthave
been intendedfor communicationwith the dead.
I have named the notch or hole throughWari
gravelids "ttoco,"from the old Quechuatermfor
window or passage. These holes appearto have
been used for makingofferingsto the dead,probably consistingof smallluxuryitems such as shell
and stone beads. Type 3 Cist Intermentsare very
similarto the primaryand secondaryburialcham-

10

LATINAMERICAN
ANTIQUITY

Type1

[Vol.15, No. 1, 2004

Type 6

Type3
profile
prfilei

plan

1~~z~~
plan

Type 7

Type 2

Type 4

proflane

profile
p cifile

plan

Figure 4. Illustrations of Wari Burial Types 1, 2, 3, 4, 6, 7.

bers of mortuaryrooms that I classify below as

Type 5a, demonstratingunity in grave forms at
ConchopataandotherWarisettlements.However,
thereis no evidence for ttoco in Type 1 and Type
2 Warigraves.And only the more elaborateburials that have ttoco also have evidence for introducing small luxury items into the grave as
offerings.
At otherMiddle HorizonAyacuchosites, Type
3 cist intermentsappearto occurin isolationor in
cemetery groupings, in buildings, and in open
places.Theymaycontaintheremainsof one orseveral individuals, but often contain incomplete
assortmentsof human bones. Grave furnishings
were occasionally included, but rarely are the
objectsnumerousor of significantvalue.

residentialareasof largercompounds.Perhapsthis
kindof tombshouldbe recognizedas anothervariant of the mortuaryroom, which I have classified
below as the Type5 burial,an issue to be resolved
by furtherstudyof Warimortuarypractices.
Bedrock cavity tombs have different shapes,
probablybecausethecontoursweredeterminedby
cracksin the rockthatmadeit easierto removethe
stone. Most, but not all, the bedrockcavity tombs
discoveredat Conchopatawere looted.All appear
to havecontainedtheremainsof morethanone person, andsignificantnumbersof pots as well as other
offerings.Onebedrockcavityintermentwas found
intactbelow Conchopata'sroomEA-31 (Figure5).
To constructthe tomb, earthand then stone had
been cut away to produce a broad shaft-like
entrance,with two burialchambersin the deepest
WariBurial Type4-Bedrock CavityInterment northernpartof the excavation,one to the northBedrockcavityburialemployeddeep tombsexca- east andone to the northwest.A ttoco about15 cm
vatedintothebedrockunderlyingConchopata(Fig- in diameterthathadbeen cut throughthe bedrock
ures4, 5, and6). They appearto havebeen marked at the northwestedge of the tomb shaftappearsto
by raisedbench-likestructuresthatoftenhadttoco have servedboth burialchambers.
holes in them.They were locatedunderthe floors
The northwestchamberwas open, havingbeen
of buildingsthatwereprobablyroofedroomsin the looted, and containedmany fragmentsof human

Isbell]

PERU
MORTUARY
A WARICULTURE
HORIZON
PREFERENCES:
CASESTUDYFROMMIDDLE

11

Figure 5. This bedrock cavity tomb was cut though the floor of room EA-31, and a ttoco was also opened through the
rock to its left. All photographs are by William Isbell.

bones as well as pieces of brokenpottery,but the

northeastchamberremainedclosedbehinda rough
stone wall. It containedseveralindividualswhose
bones were almosttotally consumedby chemical
actionwithinthe sealedenvironmentof the grave.
Two adults were tightly flexed. One, associated
with tupus, has skeletal features diagnostic of a
female.The otheradultwas associatedwithhalf of
an archer'sbow and anotherwooden object that
may be a reworkedbow stave. Based on the bow,
it seems probablethat the individualwas male,
althoughsex determinationfrom the bones themselves was impossible.Intherearof thegravewere
additionalbones in extremelypoor conditionthat
mayrepresentearlierintermentsin the samegrave.
A jar in the grave contains a humanfetus and a
radiocarbonsamplefromvegetablefiberbindings
about one of the cadavers produced a terminal
Huisaphase date.
The most impressiveunlootedbedrockcavity
intermentwas discoveredduringour2000 season.
The grave opening was found in room EA-105,
partiallycoveredby a bench-likeconstructionthat

had a circularhole in the top suggestiveof a ttoco,

except thatit did not penetrateinto the tomb (Figure 6). Small luxury objects of turquoise and
Spondylusshell were foundin this hole.
The flooraroundthetombentrancewas covered
with sherdsfrom largejars, but therewas no lid,
only earthand rocks in the mouth.A small plain
pot with constrictedopeningwas also foundat the
entranceinto this bedrockcavity tomb.Below the
roughly80-cm-diametertombmouthwas a spherical cavity almost2 m wide and about1.6 m deep,
excavatedinto the bedrock.The graveyielded 27
ceramicvessels, includingseveralminiaturepots
that seem to imitate oversize offeringurns, small
objectsof greenstone,numerouscoppertupus,and
the remainsof 15 individuals.Osteologicalexamination documented two fetuses in jars, three
infants,partof a child,ajuvenile,one malebetween
23 and 27 years of age, and six adultfemales of
various ages, as well as a seventh skeleton too
incompleteto be sexed,butprobablyalso an adult
female (Tung 2003; Tung and Cook 2002). The
male was placedin the bottomof the graveseated

12

LATIN AMERICAN ANTIQUITY

[Vol. 15, No. 1, 2004

Figure 6. The bedrock chamber tomb in room EA-105 had large pots and other objects just inside its entrance. Its bench
and ttoco can be seen just behind the tomb opening.

Isbell]

PERU
HORIZON
MORTUARY
PREFERENCES:
A WARICULTURE
CASESTUDYFROMMIDDLE

on crossedsticks of wood thatmay have been the

frameworkof a stool or mortuarysupport.I suspect thatthis was the primaryburialof the group,
probablya husbandaccompaniedby polygynous
wives and deceased infants.If my inferencesare
correct,it seems likely thatthis gravebelongedto
a nobleman,for the numberof wives seems too
largefor a commoner.
The skeletonof a pregnantwoman was found
just inside the tomb opening. It was completely
articulatedas thoughuntouchedsincethebodyhad
been placedinto the tomb.It appearsto havebeen
addedafterotherburialsimmediatelybelow,which
were disturbedandpartiallydisarticulated.Disturbance of these skeletons was consistentwith the
intrusionof the final female body when the other
bodies still had connective tissue holding their
bones together,but when theirremainswere delicate enoughto permitpartsof the skeletonto separate from one another. This is a convincing
demonstrationthatWaritombs were reopenedby
intentionto addindividualsandit seems likelythat
it occurredmanytimes.We can also concludethat
bones were removedwhen the tomb was opened,
for some of the skeletonsin thisunlootedgraveare
incomplete.So Wariburialwas a process, not an
event. The last woman addedto the gravein EA105 was about45 years of age and was pregnant,
but she probablyalso was a wife of the young man
interredearlierat the bottomof the grave.
AnotherType4 BedrockCavityIntermentwas
in EA-40, and disturbedexamples were found in
roomsEA-9 andEA-64.A uniquecase in a larger,
probablyopen patiocame from EA-6.
WariBurial Type5a and Type5b-Mortuary
Room Interment
Thiskindof burialis named"Mortuary
RoomInterment"becausetombsoccupyso muchof the space
within a room that it is difficult to imagine any
otheractivityexceptburialandburialritualwithin
the enclosed androofedarea(Figures7, 8, and9).
In some cases a secondroomandevena thirdroom
appearto havebeen partof the mortuarycomplex,
although these secondarymortuaryrooms were
probablynot filled with tombs. However,looting
has usually disturbedthe original conditions so
cannotbe precise.At
severelythatinterpretations
least six examples of mortuaryrooms are known
at Conchopata.They are room EA-138, with its

13

neighborEA-110 that were both looted severely.

A second, and perhapsthe largestmortuaryroom
complex, consisted of EA-38 (Figures 3 and 8),
probablycombinedwith EA-44, andperhapsEA31. Mortuaryrooms thatappearto have included
only one architecturalspace are EA-39, EA-150,
EA-153, and EA-203 in the westernpartof Conchopata,acrossthehighwayfromourexcavations.
Mortuaryrooms EA-38 (Figure 8) and EA-150
(Figure9) arethe bestpreservedandprovidemuch
of the informationnecessaryfor identifyingBurial Type5a andType5b, respectively.
Type5a mortuaryrooms(Figures7 and8) contain severalcircularor rectangularstone-linedcist
tombs and skeletalremainsfrom numerousindividuals. Mortuaryrooms of Type 5a all probably
containedseveralcist chambers,but one appears
to have been the principalcist, which also may
have been the firsttomb in the room. The principalcist orburialchamberwas eithercircularorrectangular, and apparently could have two and
perhapsmorechambers.Itwas sealedwitha heavy
capstonepiercedby a notchor hole, the ttoco.All
examples probablycontainedthe remainsof severalindividuals,althoughnonehasbeendiscovered
intact. Over the capstone a small offering house
somewhatless than 1 m tall was built,containing
an altarchamber.The offeringhouse had a flattop
and a small trapezoidalentrancein one side. The
floor of the offeringhouse was the gravelid, with
the ttoco providinga tiny passage from the altar
chamberof theofferinghouseintotheburialchamber that containedhumanbodies. It seems likely
that ttoco were usually sealed with stone plugs
shapedmuch like champagnecorks.The offering
house was constructedon the heavy stone lid, so
once the littlebuildingwas in place it would have
beenimpossibleto re-openthecist withoutdestroying the altarchamberwalls. Consequently,constructionof the offeringhouse terminatedthe use
of the principalcist, and probablyinitiatedexcavationof, andburialin, secondarycists withinthe
mortuaryroom. There likely was both a chronologicalorderanda hierarchyamongtheinterments
in multi-cistburialmortuaryrooms.
In some mortuaryrooms,additionalcist tombs
were excavated throughthe floor almost everywherethatwaspossible.Sometimespartitionswere
constructedarounda seeminglysecondarycist, or
set of cists, creatinga subsidiaryoffering house

14

LATINAMERICAN
ANTIQUITY

Type 5a

[Vol.15, No. 1, 2004

Type5b

PrimaryBurialChamber
WithOfferingHouseConstructed
On BurialChamberLid

BuralRoom
Chambereringouse
tWith
Doorway

over

Ttocoe

LRoomRo
locoiiiiN

pranance T

Figure 7. Illustrations of Wari Burial Type 5a and Type 5b.

Figure 8. Mortuary Room EA-38 is an example of Wari Burial Type 5a. The offering house, now lacking a roof, is constructed over a massive lid of the primary burial chamber. Its ttoco notch is visible at the top edge of the stone. Secondary
cist tombs were located around the primary burial chamber.

Isbell]

A WARICULTURE
MORTUARY
PREFERENCES:
CASESTUDYFROMMIDDLE
HORIZON
PERU

15

Figure 9. Mortuary Room EA-150 is an example of Wari Burial Type 5b. The offering house, with roof intact, is constructed over a chamber that was entered from the side, where its rectangular lid, now broken, has collapsed into the
void. The ttoco notch is located inside the offering house.

with an altarchamber.Occasionally,an adjacent

roomseemstohave beenpartof themortuarycomplex, having its own cist tombs excavatedinto its
floor,and walls thatmay have been partsof offerlootinghouseswithaltarchambers.Unfortunately,
ing hasmadeit difficultto determinecriticaldetails
of constructionchronology,but what does seem
clear is that in Type 5a mortuaryrooms, the constructionof an offeringhouse overa tombsignaled
its importance.It also meantthatthe tombwas difficult if not impossible to re-open to insert additionalburialsor to removeany remains.
Mortuaryrooms of Type 5b representan elaborationon Type 5a that could be enteredand reentered, without disturbing the offering house.
These tombs had a separateentranceto one side,
sealedby a flatstone(Figures7 and9). A largerectangularburialchamberwas constructedbelow the
floor of the mortuaryroom andcappedwith stone
slabs at aboutthe same level as the floor.A ttoco
was constructedbetweenthestonelintelsatone end
of thechamber,andanentrancethatcouldbe sealed
witha single stoneslabwas placedatthe otherend.
An offering house with altar chamberwas built
overthettoco,coveringabout70 percentof thebur-

ial chamber,butleavingthe entranceandcovering

stoneexposed.Thiskindof mortuaryunitcouldbe
re-openedrepeatedly,while the offeringhouseand
ttoco remainedundisturbed.
All the mortuaryrooms discovered at Conchopatawere looted,butgold artifactswere found
in mortuaryroomsEA-138 andEA-150.Thisis the
only gold discoveredin our excavationsat Conchopata,so thereseems little doubtthatmortuary
rooms were the pinnacles of the local interment
hierarchy.Onlythemostpowerfulandwealthyresidents could affordso much luxury.Study of the
skeletal remainsfrom mortuaryrooms is still in
progress,and,of course,all were disturbed.However, preliminaryevidence indicatesa significant
of femaleskeletons,consistentwith
preponderance
a high-statuspalaceareawherea kingandhis noble
kinfolkwere attendedby numerouswives, concubines, and servingwomen.
WariBurial Type6-Wall Interment
This typeof intermentemployeda chambercutout
of, constructedwithin,or attachedto, a thickwall
(Figure4). Wedidnotdiscoveranywall interments
duringour excavationsat Conchopata,but Lum-

16

LATINAMERICAN
ANTIQUITY

breras(1974a:180-181) reportedone examplecontainingtwo individualsduringhis investigationsat

the site. It is possible thatit representsa late addition to a Silva phase wall at Conchopataand that
wall burialis a late Warifeature.Numerouswall
burialshave been reportedfrom Huari(Gonzalez
Carr6and BragayracDaivila 1996) and Middle
Horizon Batan Urqo (Zapata1997), so although
burialType6 does not seem to havebeen verypopularin thecivic centerof Conchopata,it was apparently a significanttype of Wariinterment.
WariBurial Type7-Communal or Sacrificial
GroupBurial
One exampleof a mass grave,probablya groupof
sacrificial victims,5 is reportedfor Conchopata.
Thisuniqueexamplecontainedfive youngfemales
coveredby a stone moundor cairn(Figure4). All
appearto have been buriedat the same moment.
This intermentwas discoveredin 1977 about 1 m
northwestof a ceramicofferingof oversize faceneckjars(Isbell1987;IsbellandCook 1987,2002).
It is likely thatthe womenparticipatedin the same
event in which the giant face-neck jars were
smashedandburied(Cook 1987, 1994).
Wari Tombs at Other Settlements
I doubt that this descriptivediscussion,based on
Conchopataburials, exhausts the range of Wari
mortuary practices. Expansion, revision, and
reevaluationswill surely be requiredas we learn
moreaboutMiddleHorizonmortuarylandscapes.
However,it is clearthattheseidealtombtypeshave
equivalentsat otherAyacuchosettlements,as well
as at moredistantWaricommunities.
Tombs of Type 1, individual interment, are
describedfor the Warisites of Jargampata(Isbell
1977:29) and Azangaro(Anders 1986:619-620).
Similargraves existed at the Rio Pampassite of
TaqsaUrqo,butwere destroyedby roadconstruction. However,I suspect that many examples of
WariType 1 burialshave gone unrecognized,and
perhapseven unreported,because they contained
no stylisticallydatableobjects.
Type2 multipleintermentsare as pooras Type
1 gravesandareprobablyalso under-reported.
One
described
Schreiber
(1992:249-250),
example,
by
is a gravecontainingtwoindividualsatJincamocco.
This tomb was a little fancier than Conchopata

[Vol.15, No. 1, 2004

examples,for the pit was partiallystone lined and

coveredwith simpleslabsof rock,butno offerings
were included.
Type3 cist intermentburialshavebeenreported
for Aqo Wayqo (Ochatoma and Cabrera2001:
83-96), whereatleastone containedpottery,tupus,
and other furnishings.However,these examples
had no ttoco.A similarburialwas found at Nawimpuquio, with a stone-lined double chamber
(Cabrera1998) that is somewhatlargerand more
elaboratethanmostcistinterments,althoughit also
lacked a ttoco. Perhapsthis representsa subclass
of cist burials consisting of a stone-lined vault
insteadof just a pit cut into earth.
A Type4 bedrockchambertombis foundat the
planned architectural complex of Azangaro
(Anders 1986:617-619). Accounts of what seem
to be Warisites in the Rio Pampas,southof Ayacucho, suggestthatbedrockchambertombsmayexist
thereas well.
Type5a mortuaryroomintermentis alsoknown
at Huari(Figure10). My studentsandI excavated
an examplein theMoraduchayuqarea(Isbellet al.
1991:34-36 andFigure 18). The presenceof only
scatteredfragmentsof humanbone madeus reluctantto identifythe chambersas mortuaryin funcwe now have
tionwithoutthekindof corroboration
from comparisons with Conchopata.At Moraduchayuq,two rooms, each 5.3 m long and 2 m
wide, connect througha doorway.The northend
of the innerroom is raisedabout20 cm with the
primarycist locatedon this bench andcoveredby
a heavy circularstone with two ttoco. Remainsof
an offeringhouse arerepresentedby wall bases on
two sides of the lid. In this mortuaryroom, there
aretwo morelargecists with lids andttoco,andin
the neighboringroom,four cists, one with lid and
ttoco still in place.All the tombshad been looted,
anda greatquantityof finepotteryof MiddleHorizon EpochIB, all severelybroken,was foundscattered about.The potterywas mostly open vessel
forconforms,suchas bowlsandcups,appropriate
sumingfood and drink(Cook 1994).
A secondexampleof a Type5a mortuaryroom
is from the Cuzco Middle Horizon site of Batan
Urqu(Figure11). Partof a largerWaricommunity
knownas Huaro,theBatanUrqucomplexmightbe
describedas a cemeterybuildingcontainingvarious mortuaryrooms(Zapata1997).Most similarto
theConchopataexamplesis theprimaryburial(Zap-

Isbell]

MORTUARY
A WARICULTURE
CASESTUDYFROMMIDDLE
PREFERENCES:
HORIZON
PERU

17

N
Cist3
Pit

Raisedstep
in floor

PitA

135

(1

Pit C

2m

Stonelid withttoco
coveringcist
nexcavated
134
RaisedstepV
in floor

234

3acy

161

MoraduchayuqCompound, Huari

Figure10. Huari'sMoraduchayuqCompoundshowinga burialroomof Type5a. (Redrawnfrom Isbellet al. 1991:

Figures6 and 18)

18

LATINAMERICAN
ANTIQUITY

ata1997:Figures33, 34) consistingof a hugestonelined cist with heavy rock lid and central ttoco.
Remainsof a smallrectangularofferinghouse surmountthe lid, andmanyothercist tombsandmortuaryrooms are locatedclose by. Therecan be no
question that the Batan Urqu mortuarycomplex
representsan orderof magnitudeor two grander
than anything at Conchopata,but the mortuary
behavioris clearlythatof a Type5a Wariburial.
I believe that ConchopataType 5b mortuary
roomsareformallysimilarto numerousexamples
from Huari, but the Huari tombs have been so
severely damagedthat most are difficultto conceptualize in their original form. Called "cheqo
wasi" (stone house), they are megalithicchamber
complexes,often of two or even threefloor levels
(Figures12 and 13).No one has attemptedto determine theiroriginalforms, althoughwe have severaldescriptionsof thelootedarchitectural
remains
(Benavides 1984, 1991; Bennett 1953; Gonzailez
Carr6and BragayracDaivila 1996; P6rez 1999,
2001a, 2001b). Based on these discussions, my
own researchat Huari,and the new Conchopata
comparisons,I conclude that the majorityof the
megalithic chambers were enclosed within the
roughstonewallsof architectural
compounds.They
were re-openable mausoleums similar to Conchopata'sType5b mortuaryrooms.
Type5 mortuaryroomsdescribedfor Huarican
encloseone largechambercomplexorseveralsmall
chambers,probablyrangingfromtwoto five.Small
and simple cheqo wasi probablywere enteredby
removingthelid (Figure12).Morecomplexexamroomor complexof
ples consist of a subterranean
rooms enteredfrom one side througha crawlway,
perhapsalso coveredby a heavystone(Figure12).
The upperlevel is often a room,or room complex
thatmay have been closed except for ttoco. Other
ttoco connect the upperchamberswith the lower
chambers.In form, Huari's cheqo wasi are like
Type5a and5b mortuaryroomsfromConchopata,
except thatthey are much grander.I proposethat
these megalithictombs be recognizedas another
subclass,Type 5c (Figure12)
All knownType5c mortuaryroomsfromHuari
werelooted,probablymanytimes,beginningin the
distantpast.Inearlypostconquesttimestheyserved
as quarriesfor constructionstone,furnishinghuge
expertlyworkedashlarsthat could be re-cut into
mill stones,waterconduits,andotherstoneobjects

[Vol.15, No. 1, 2004

used to constructthe colonial city of Huamanga

(now Ayacucho).But excavationsin and around
themhaverevealedmanyhumanremainsin thedisturbedcontexts.As ourunderstandings
grow,there
seems little doubt that the chambers were elite
tombs.
If the BatanUrqoType 5a mortuaryrooms are
granderthanConchopata'sby an orderor two of
magnitude,some of Huari'slargercheqo wasi are
greaterthanConchopata'sType5b mortuaryrooms
by half a dozenordersof magnitude.Huari'scheqo
wasi must have been tombs for kings or nobles
whose statuswas a full social level abovethe fanciest tombsdiscoveredat Conchopata.
Megalithicstonechambersof Type5c arecommon at Huari,but are very rareif they exist at all
outsidethecapitalcity.Onlyone examplehasbeen
reported.In southernAyacucho,morethan100 km
fromHuari,Schreiber(1992:154-155) reports"at
least one (andpossiblymorethanthree)semi-subchambersbuiltof largeslabsof cutstone."
terranean
This settlementappearsto have been quite small
but locatednearthe entranceinto a valley thathad
architeca sizablecomplexof Wariadministrative
ture and extensive agriculturalterracing.Perhaps
it becamethe estateof a Huarimonarchwhose relatives were eventually buriedthere, but excavations arerequiredto confirmtheexistenceof these
ruralcheqo wasi, much less infer their meanings
in the vastWarilandscapeof death.
Uncommonat Conchopata,Type 6 wall interment was frequentat Huariand at BatanUrquin
Cuzco (Zapata1997). Wallgravesareonly found
in very thickwalls, which arerareat Conchopata,
at least in the civic centerwhere our excavations
have been concentrated.
Type6 wall intermentrequiresadditionalinvestigationin thefuture.BurialsfromtheVegachayoq
Moqo sectorof Huari(Figure14) aredescribedby
Vera Tiesler Blos (1996). Most of the human
remainswere looted fromtombswithina massive
wall that was built across a courtyardwhen the
functionof thearchitectural
complexchangedfrom
palace, to mortuarymonument,to popularcemetery(see Isbell200 ib). It is nowclearthatthismassive wall, more than 2 m thick, had many large
niches, one containinga collection of secondary
burials (Bragayrac 1991), as well as numerous
chambers for wall interments. These were not
nichesbutcryptsforprimaryburialsthatwereprob-

Isbell]

MORTUARY
A WARICULTURE
PREFERENCES:
CASESTUDYFROMMIDDLE
HORIZON
PERU

19

0600

--offering

--'
......

Remainsof
house
andaltarchamber
constructedoverprimary
tomb

Mortuary
rooms

'

10m

...

....I

Figure 11. Batan Urqu, Cuzco, mortuary building with burials of Types 5a and 6. (Redrawn from Zapata 1997: Figures
5 and 34).

ably sealed except when occasionally reopened.

Some wereprobablyintrudedintothe wall afterits
construction, while others appear to have been
shapedas the wall was built (Pdrez1999; Tiesler
Blos 1996). I suspect that the large quantity of
humanremainsfoundalong the edge of this same
wall-and attributedto post-MiddleHorizonmor-

tuaryactivityby Tiesler-are actuallyHuariburials pulled from theirwall chambersand scattered

aboutthe foundationareaby looters.
At Batan Urqu in Cuzco, Zapata (1997)
describesa largerectangularbuilding,poorlypreserved, but originallyabout 33 m by 89 m, with
partsof its perimeterwall standingalmost1 m high

LATINAMERICAN
ANTIQUITY

20

Type5c HuariMortuaryRoom
with two smallchambers

[Vol.15, No. 1, 2004

Type5c HuariMortuaryRoom
with one largechambercomplex

aPlan

r
Entrance

Chamber2

r2

Profile
7m.Profile

Figure 12. Huari's cheqo wasi or megalithic mortuary rooms of Type 5c. (Redrawn from GonzAilezCarre and Bragayrac
Divila 1996: 20 and from PNrez2001a: Figure 32).

to 20 m in diameterandhalf as deep thatwere partially filled with huge rocks, includingfragmehts

of finely workedashlars,curiouslyshapedstones
thatlooked like conduitsfor aqueducts,andcircularslabsresemblingmill stones.Nearbywas a long
subterranean
hall filled with humanremains.
Although extremely damaged, clearing and
excavationby IsmaelPerez (1999, 2001a, 2001b)
in 1997 has finallyrevealedenoughof the ancient
architectureat Monjachayoqto get a sense of its
originalform. Monjachayoqconsisted of four or
five subterraneanlevels of constructionwith the
deepestreaching 10 m or more below the ground
(Figure 15). On the surfacethereappearsto have
been a perimeterwall, a "D-"shapedtemplebuilding, a large structure,and maybe a streetor corridor.Underthis, and apparentlybelow the original
Type 8 Royal Interment
groundlevel, was a complex of four halls, end to
The Monjachayoqarea of Huari is also named end, of well-maderoughstonemasonrywith mas"canterdn"(Bennett 1953:19) or stone quarryin sive cut stoneslabsfor theroof andthe floor.At the
Spanish.Before the 1970s it had gaping holes 15 southend, the hall complex passed over a deeper,

and about 1.3 m thick. Along the interiorbottom

of the west wall, he foundType6 wall burialchambers of variousforms, from rectangularto semicircularto elongated,usuallycontainingdisturbed
bones of several individuals, adults as well as
infantsand children(Figure11).
Basedon thesereportsit appearsthatwall intermentsrepresentyet anotherkindof Warigravethat
was probablyopened and reopenedfor the addition,andperhapstheremoval,of humanbodiesand
defleshed bones, respectively. Few offerings or
gravefurnishingshavebeen foundwith wall interments.Perhapsthisis becauseso manywerelooted,
but moreprobably,it is becausethey were similar
in statusto Type 3 cist interments.

Isbell]

MORTUARYPREFERENCES: A WARI CULTURE CASE STUDY FROM MIDDLEHORIZON PERU

21

Figure 13. This mortuary room at Huari contains a medium sized cheqo wasi, or megalithic tomb of Type 5c.

second subterranean
level of architecture(Figures
15, 16, and 17).
level conMonjachayoq'ssecond subterranean
tains 21 cells constructedof ashlarsin combination with rough stonework(Figures 16 and 17).
Thisconstructionwas disclosedby cleaningone of
Monjachayoq'sgapingholes of loose stone,revealing a surfaceexposedby lootersand subsequently
worked by colonial stonecutterswho converted
ancientashlarsintomillstones,waterconduits,and
otheritems requisitionedby Spanisharchitectsin
the new city of Huamanga.In fact, the 21 chambers areexposedbecausemassivecoveringstones
were removed, along with several levels of constructionabovethem.Pdrez(1999) foundstonesin
theprocessof beingre-cut,alongwithanexhausted
iron chisel of the colonial masons.
Huari's subterraneanmegalithic complex of
cells must have been opened and looted, perhaps
in prehispanictimes.Duringthe colonialeraSpanish contractorsbegan quarryingstone from Mon-

jachayoq,recuttingits originalconstructionblocks
fornew requirements
in thecolonialcapitalof Huamanga.In spiteof this destruction,therecanbe little questionthatthe complexof 21 cells represents
a mortuarygroup,of subsidiaryburialchambers,
or perhaps offering houses built above an even
granderprimarymortuarychamber.
Under the complex of 21 cells is a thirdbasement level, accessible only by a shaft. It is a hall
whose plan resembles a llama viewed in profile
(Figures 16 and 18). Pdrez (1999) observedthat
entry was at the mouth of the symbolic animal.
And, at the tip of the llama'stail a still-deeperelement was constructed,that mightbe considereda
fourthundergroundlevel. This is a circularchamber, lined with roughstonework,3.7 to 4 m deep,
reaching 1.2 m in diameterat the bottom, with a
flat-stonelid that once sealed it. It looks remarkably like a primaryburialcist froma Type5a Wari
mortuaryroom,as well as theprimaryburialchamber in the BatanUrqumortuaryroom.

[Vol.15, No. 1, 2004

LATINAMERICAN
ANTIQUITY

22

:I

WallTombs

SmeNiches

lo
s
..6 niches

surface

St

wall tombs

SProbable Walls

walls
ecorated

icstonesd

"D" ShapedhbuildinsFloorlevel
FirsdinTe

er r
TW aceFfPill

of stones

.
Floors

canal

Floor

esu

Fo

Moqosector.
Figure14.Mapof Huari'sVegachayoq

So

Isbell]

MORTUARY
PREFERENCES:
A WARICULTURE
CASESTUDYFROMMIDDLE
PERU
HORIZON

--------- -----W--

-I

---------------

23

-------------

Vegachayoq
Sector

Moqo.-..y

Entrance
SPossible

Sectorarchitectural
Sashlars

Vg

First
subterranean
hallswallslevel

meters

Monjachayoq
ashlarUncleares

Cleared
in1973
Second subterranean
level cells
Third subterraneanlevel
llama-shaped hall

Surface walls
--

Figure 15. Map of Huari's Vegachayoq Moqo and Monjachayoq sectors. For details of subterranean levels in
Monjachayoq Sector, see Figure 16.

24

[Vol.15, No. 1, 2004

LATINAMERICAN
ANTIQUITY

Originalentrance
to llama-shapedhall tomb

D3

'd?

1
~c~fV
3

~a7~s7G3

17~=3

5
llama shaped

Vto

subterranean
level
Cells of second
1
2
0

Cylindrical

cyst-tombof fourth
subterraneanlevel hall
Llama-shaped
tomb of third

-????-metersmeters

loters'
hall tombentrance

subterraneanlevel

Figure 16. Map of Huari's Monjachayoq Sector showing the second, third, and fourth subterranean levels.

Nothing of the original contents of Monjachayoq'shugeunderground

complexis lefttoday.
Many humanbones were removedfrom the first
basementhallsin 1977.The21 chambersandllama
gallery of the second and third basements were
excavatedmore recently,but they containedonly
secondaryfill, withoccasionalfragmentsof human
bones, pot sherds,and stone tools. Even the lid of
the deepestcist hadbeenremovedandnothingwas
foundwithin.Of course,fragmentaryanddisturbed
humanremainswere scatteredthroughoutthe fill
of this impressivecomplex, confirmingits mortuary function.
The form, size, and impressiveconstructionof
the Monjachayoqmortuarycomplex place it on a
par with royal burialplatformsfrom Peru's great
northcoastalcity of ChanChan(see Conrad1982).
I feel securein identifyingthe Monjachayoqsubterraneanbuildingcomplex as a royalWaritomb,
eventhough,as atChanChan,regalbodiesandtheir
offeringsdisappearedcenturiesago. Curiously,the
Huarisepulchreis virtuallythe inverseof Chimu's

royalburialplatforms-a "royalcatacomb."Itrepresents the supremehierarchicallevel in Wari's

landscapeof death.The Monjachayoqtomb ma?
be listed as a WariType 8 SubterraneanChamber
Complex Interment, probably representing an
emperorwho ruledHuariand all its possessions.
Withinthe Huarisite, I do not thinkthatType8
interment is unique to Monjachayoq. Near the
northeastcorner of Huari's architecturalcore is
anothergreathole, filled with brokenblocks, ashlars, and stones, that is also called canterrn. I
believe thatexcavationswill revealanothermegalithic subterranean tomb complex of a Huari
emperor,also looted long ago, andquarriedfor its
fineworkedstones.Perhapsa new excavationcampaign will revealan unlootedroyaltomb at Huari.
Wari's Landscape of the Dead
Waripeopleinscribedrespectfor,andengagement
with, the dead into the built environmentsof their
cities andtowns.At Conchopatathey creatednew

Isbell]

MORTUARYPREFERENCES: A WARI CULTURE CASE STUDY FROM MIDDLEHORIZON PERU

25

Figure 17. Wari Type 8 Royal Tombs are represented by the megalithic subterranean complex at Monjachayoq, Huari.
The second subterranean level consists of 21 cells that probably served as secondary tombs and offering chambers.

26

LATINAMERICAN
ANTIQUITY

[Vol.15, No. 1, 2004

Figure 18. The third subterranean level of Huari's Monjachayoq royal tomb was probably the primary burial chamber.
It consists of a narrow hall whose plan resembles a llama in profile. Located 6 to 8 m below the original ground surface,
the looted chamber now contains construction rubble and fragments of human bone.

kinds of buildingswhere the living veneratedthe

dead,who were interredbelow the floors.Some of
these tombs were modest,otherswere substantial
anda few werepretentious.Themostpowerfulresidents created mortuaryrooms for their bodies,
wheretheywouldbe visitedby generationsof their
descendents,at least some of whom would eventuallybe addedto the same complex of tombs.
Conchopatahas half a dozen mortuaryrooms
with tombsthathavegreatlids, ttoco,andoffering
housesfillingthe entirespace.This mortuarylandscape affirmsthatConchopatawas not just a city
of craftspeople,but of elites and nobles, occupying palaces and commandingresourcessufficient
to constructimpressivetombsandprovisionthem
with wealth thatincludedgold. But the poor condition of Middle Horizontombs made it impossible to describeWari mortuarybehaviordirectly.
This has been achievedonly by abstractingideal
or preferredpatternsfrom a multitudeof graves,

manydisturbedbuta few intact,fromConchopata

and relatedsettlements,includingthe Huaricapital itself. The resultingtypologyof ideal mortuary
classes is remarkablycomplex andhierarchical.It
suggestsso manyinferencesthatonly a few can be
discussedhere.
Wari's Middle Horizon landscape of death
linked ancestorsand descendentswith a house or
palace.This surelypromotedthe formalizationof
royallineagesordynastiesknownin manycultures
as "greathouses."Wariintermentemphasizedstatus differenceand social inequalityin its spatial
metaphors.Type 1 and Type 2 intermentswere
small, unmarked,and lackingin materialobjects.
Type 2 multiple intermentsmay graduallyhave
becamemorepopular,almostreplacingindividual
interments.Kin ties, or whateverformedthe basis
for mortuarygrouping,became emphasizedeven
more as multiple intermentsof the Middle Horizon replacedindividualgraves and cemeteriesof

Isbell]

PERU
HORIZON
A WARICULTURE
MORTUARY
PREFERENCES:
CASESTUDYFROMMIDDLE

theEarlyIntermediateperiodMendosaphase.Perhaps in the new urbanmilieu, new principlesof

affiliationwere exploredfor creatingnew kindsof
relationships(Smith2003).
Ttocoopeningsintotombsbecamepopularduring the Middle Horizon, implying an increased
desire to maintaincontact with ancestors.However,Type 1 and Type2 intermentshave no ttoco
and containno luxurygoods. It appearsthatlowstatusindividualswereburiedtogether,in affiliated
groupings,butthey did not become reveredancestors.
WariType 3 cists, as well as Type6 wall interments, are a step higher in the social landscape.
Type3 gravessometimes,butnotalways,hadttoco
openings,whileType6 seemsnotto havehadthem.
I suggest thatthese burialsrepresenttypical residents of Waricities, neitherpowerfulnor impoverished.
Type 4 bedrockchamberintermentsappearto
havebeentheburialplacesof minornobles,atleast
at Conchopata.They had ttoco openingsand contained many grave goods. Type 4 burialsare frequentat Conchopata,implying thatthe surviving
portionof thatcity was a palacecompound,orcomplex of associatedpalacecompoundsoccupiedin
large partby elites. Bedrockchambertombs that
were not disturbed appear to have held family
groups,and at least some examplesarebest interpreted as the polygynous family of a man with
manywives. In fact, female remainsconsiderably
outnumberthe males in ourConchopataosteological sample,a factthatI ascribeto the seraglio-like
natureof the palatialsector we have investigated
at Conchopata.
BurialType 5a and 5b mortuaryrooms representthepinnacleof theintermenthierarchyatConchopata.Theyhavettocoopenings,combinedwith
anofferinghousewithaltarchamber.Thesegraves
contained gold and other objects of wealth,
althoughnonehas beendiscoveredunlooted.As in
bedrockchambertombs, mortuaryrooms contain
a predominanceof female skeletons, seeming to
confirmtheimportanceof polygyny,andtheimportanceof womenandtheirlaborforthesmallernumber of elite men. I proposethatthe personsburied
in Conchopata'smortuaryrooms were rulersand
theirclose family members,probablypetty kings
or curaca, to use an Andeanterm.The discovery
of similarbutmoremagnificentmortuaryroomsat

27

BatanUrqo in Cuzco implies Warikings of similarnoblerankin thedistantcity of Huaro,butjudging by the graves, Batan Urqo's kings were
probablywealthierthan Conchopata'srulersin a
regionalscale of powerandaffluence.Type5a and
5b burialsseem to representa fourthlevel of social
statusin ancientWariculture,perhapsrulersof secondarycities and governorsof provincialterritories.
Mortuaryroomsof Type5a and5b weretheapex
of the funeraryhierarchyat Conchopataand at
Huaro/BatanUrqu, but they were modest when
comparedwith Huari's cheqo wasi-megalithic
chambertombs-but placed in mortuaryrooms
similarto those of Conchopata.This demonstrates
that the fourth-levelcuracas of Conchopatawere
significantlyout-rankedby more powerfulnobles
at Huari,who could build trulymagnificentmausoleums.Furthermore,
Type5c megalithicmortuaryroomsappearto havebeenlimitedto Huari,and
perhaps one provincial site in the south, where
some Huariprince may have establisheda royal
villa orcountryestate.Consequently,Type5c burials must representa fifth hierarchicallevel of status and wealth in Waricultureand society. Their
limitationto the capitalcity impliescentralization
of political power, with deceased nobles being
buriedonly in the greatcity. Wari'slandscapeof
deathproclaimsHuari'suniquehierarchicalposition, contradictinginterpretationsof the Middle
Horizonthatarguefor equivalentcities or confederationsof lineages.
Supremepowerand wealthin Wari'smortuary
landscapeis representedby Type8 royalinterment,
a sixth level in the power hierarchy.Still poorly
known, these tomb complexes were vast and
impressive.Furtherresearchwill probablyprove
thattheywerethe tombsof Huari'semperors.And
they significantlysurpassall othergravesof Middle Horizon date anywhere within the Wari
sphere-Pachacamac, Cuzco, Huamachuco,
Nasca,or Moquegua.Theironly appropriate
place
was Huariitself, wheretheyprobablydefinedcentrality,for therearehintsthatsocial relationswith
these dead emperorsneverended, and that social
memorywas constructedaroundtheirtombs.
Wari's dead, or perhaps more correctly, the
higher-statusdead,werein continuedrelationships
withtheliving.Offeringsof some sort,butcertainly
including small luxury objects, were introduced

28

LATINAMERICAN
ANTIQUITY

intotombsthoughttocoopenings.Offeringhouses
with altarchambers,built over the tombsof Type
5 as well as Type 8 royal tombs, may have containedmany otherkinds of gifts. This shows that
progenitorswere objectsof adoration,andthatthe
peopleof Waripracticedancestorworshipof some
sort.
In the sixteenth and seventeenth centuries,
Andean peoples practicedreligions that emphasizedancestorworship(Doyle 1988;Duviols 1988;
Isbell 1997a; Salomon 1995). Corpsesof important lineage founders and political leaders were
mummifiedbecausetheirbodieswereholy objects
of publicworship.The cadaverwas carefullypreserved,even body exuvia-fingernail cuttingsand
trimmedhair.Some mummiesresided in special
mortuarytowns,othersremainedin theirhomesand
palaces, and, at least some of the time, dead Inka
kings sattogetherin a greathallwithinCuzco'ssun
temple(MacCormack1991).Founders'mummies
anddeceasedkings were publicfigures.They participatedin feasts, traveledabout,and were availableforconsultation.Theydemandedandreceived
fine clothing,foods, and otheritems of conspicuous display,andwitnessingtheirenjoymentof these
gifts seems to have been an essentialpartof worship by theirdescendents.
It would be attractiveto imaginesimilarmummies populatingWari'slandscapeof death,butthis
seems unlikely.Warimortuaryfacilities were not
designedto preservemummifiedflesh. Underthe
floorsin the ground,Warideadwere soon reduced
to bones. Furthermore,some of the bones, but not
mummifiedcadavers,were removedwhile other
partsof the body remainedin the graves.Apparently, Wari ancestorswere deliberatelydismembered, somethingthat would have horrifiedInka
worshippers.
Manyof the higherstatusWaritombswereeasily openedand sealed again,but it seems unlikely
thatthey containedfounders'mummieswho were
broughtout for public worship.The entrancesof
these tombswouldhavemadeit difficultto extract
andreplacewhole mummies.But the evidencefor
Type 5a tombs is even more indicative.Theirprimary cists were impossible to re-open once an
offering house and altarchamberhad been constructedover the lid.
Principalcists of Type5a mortuaryroomscontainedimportantancestors,but it is impossibleto

[Vol.15, No. 1, 2004

imagine Inka-style mummies trapped in these

tombs, beyond the reach of their descendents.
While therewere importantdevelopmentsin Type
5 tombs that appear to document significant
changesin treatmentsandmeaningsof deadancestorsthoughthe MiddleHorizon,Waridescendents
who employedType5a mortuaryroomscontented
themselveswith communicatingwith theirprincipal ancestorthrougha ttoco
At Conchopata,andapparentlyatHuarias well,
elaboratemortuaryrooms were located far from
publicareas.They were intendedfor privateceremonies, not public display.In fact, built environments of deathimply that admissionto mortuary
roomswas limitedandexclusive.Perhapsentrance
instatedpowerthathad to be controlled.Adjacent
facilitiesdo not includecourtyardsor plazaslarge
enough for the assemblyof many people. We do
not yet fully understandhow the Waridead were
incorporatedin granderritualswheresocial memory was constructed,but currentinformationsuggests the possibility that defleshed and
disarticulatedbones of deceased ancestorscould
have been objects of displayin public landscapes
of death.
Unfortunately,the image of Inka-stylemummies is excessively powerfulin Andeanarchaeology, becoming an untested assumption for
earliermortuaryremains(see Kaulicke
interpreting
Inka
ancestormummieswere kept in open
2000).
and
broughtout to participatein ceresepulchres
monialactivitiesof the living,in manycases as the
focus of adoration.TheresaandJohnTopic(1984;
see also Isbell 1997a:204-208)reportedthe possibilityof Inka-likemummiesfroma lateEarlyIntermediateperiod/MiddleHorizonmortuarybuilding
at CerroAmaruin Huamachuco,althoughthe context was disturbedand requiredsignificantinterpretativeinference. Also on the basis of highly
disturbedhumanremainsI arguedthatJargampata,
a ruralMiddle Horizon installation25 km from
Huari,may have includeda room within its residentialquarterswheremummieswere kept (Isbell
1997a:187). But new mortuaryinformationfrom
Conchopatashowsthatthedeceasedwereaccessed
though ttoco openings, and that removalof comin publicritualswould
pletebodiesforparticipation
have been difficult or impossible. GordonMcEwan's (1998) inference that Inka-like ancestor
mummies were the principalreligious objects of

Isbell]

PERU
HORIZON
MORTUARY
PREFERENCES:
A WARICULTURE
CASESTUDYFROMMIDDLE

29

Wari'sregionaladministrativecenterat Pikillacta
now seems very unlikely. Quoted in a recent
National Geographic Magazine article (Morell
2002:123),McEwanstatedthatPikillactawas used
as a mummystoragedepotwhereWarileadersheld
capturedancestormummieshostageto insurepolitical compliance from their living descendents.
Withoutmaterialevidencein supportof this assertion, andin light of inconsistenciesbetweenMiddle HorizonAyacuchomortuaryfacilitiesandthose
associatedwith Inkapublic display of mummies,
suchancestorbundlesseem unlikelyin Pikillacta's
landscapeof death.If ancestormummiesexisted
at Pikillacta,they were partof the cultureof the
conqueredpeoples of Cuzco.
Therecan be no doubtthat this study of Wari
mortuarylandscape is preliminary.Much more
informationmust be collected and compared.As
dataincrease,so will the refinementof ideal types
of Wari mortuary practices, as well as actual
cases-the occasionalintacttomb-providing better understandingof variability and individual
strategiesin the treatmentof Waridead. But even
in preliminaryform,this typology of Warimortuarypreferencesfurnishesa tool for inferringsocial
andpoliticalhierarchyduringtheMiddleHorizon,
whileit createsa new understanding
of Wari'slandof
death.
scape

the Universityof NorthCarolina(TiffinyTung),the University

of California-Berkeley(Bill Whiteheadand Matt Seyre), La
UniversidadNacional de San Marcos (PatriciaMayta), and
fromArgentina(MabelMamaniand SilvanaRosenfeld),who
have contributedto the researchby directingexcavationcrews
and/or participatingin analyses. Special recognition is due
AlbertoCarbajalA., my friend,our projectadministrator,
and
an insightful archaeologist. I want to thank Dr. Luis
Lumbreras,currentdirectorof Peru's InstitutoNacional de
Cultura,for his encouragement.Throughthe yearsunequaled
supporthas come from Dr. EnriqueGonzalez Carr6,in his
posts as Rectorof the UniversidadNacional San Crist6balde
Huamanga,Directorof the Museo Nacional de Arqueologfa,
andDirectorof Peru'sInstitutoNacionalde Cultura.Gonzalez
had a storagefacility added to the archaeologylaboratoryof
the Universidad Nacional San Cristobal de Huamanga
expressly for materialsexcavated at Conchopata.I want to
thank the directors of Ayacucho's Instituto Nacional de
Cultura,Ulyses Lareya, Teresa Carrasco,MarianoBenites,
and Severino Castillo for their help, friendship,and advice.
CesarAlverez,directorof Ayacucho'sarchaeologicalmuseum
Hipolito Unanue, also contributedto the success of the
ConchopataArchaeologicalProject,as did manyothers.I also
wish to thankmy university,the StateUniversityof New York
at Binghamton,and my colleagues in anthropology,for support and encouragement.Dr. Cook's Catholic University of
America, and the UniversidadNacional San Crist6bal de
Huamangahave also contributedgenerously.Finally,my wife
Judy Siggins has been a source of continualhelp and inspiration. She has managedour home and family life duringmy
long researchabsences, she read and edited my proposals,
reports,andmanuscripts,andshe is the best consultantanyone
could hope for.

Acknowledgements:The ConchopataArchaeologicalProject
is directedby Dr.WilliamH. Isbell, Dr.Anita G. Cook, M.A.
Jose OchatomaP., and Lic. MarthaCabreraR. de Ochatoma.
It is administeredby AlbertoCarbajal.Special recognitionis
due the sponsors, and particularlythe National Geographic
Society that has been the primarypatronsince 1998. Initial
supportwas a grantfrom WennerGrenin 1997 to Ochatoma.
Additionalfundinghas come from the CurtissT. and MaryG.
Brennan Foundation, Dumbarton Oaks, and the Heinz
Foundation.Excavationswere conductedin 1997, 1998, 1999,
2000, 2001-02 and2003 withpermissiongrantedto Professor
Ochatoma(1997-1998) and Dr. Isbell (1999, 2000, 2001-02,
2003). I wish to thankthe co-directors,and also the archaeology students,especially those from the State University of
New York-Binghamton (Catherine Bencic, Juan Carlos
Blacker, Juan Leoni, Greg Ketteman,Mike Calaway,Marc
Lichtenfeld,Ariela Zycherman,Kris Mearish,Amy Groleau,
Dan Eisenberg, Brian Finucane, and Meridith Davis), La
UniversidadNacional San Crist6balde Huamanga(Lorenzo
Huisa, Carlos Mancilla, Ismael Mendosa, Maximo Lopez,
Teresa Limalla, Irela Vallejo, Alina Alvarado, and Edgar
Alarc6n),the CatholicUniversityof America(BarbaraWolff,
Nikki Slovak, David Crowley, Teresa Carmona, and Eric
Schmidt), La Pontifica Universidad Cat6lica del
Perdt
(Gonzalo Rodriguez, Manuel Lizarraga, and Antonio
Gamonal),the Universityof Pittsburgh(CharleneMilliken),

References Cited
Anders,MarthaBiggar
1986 Dual OrganizationandCalendarsFromthe Planned
Site of Azaingaro:
WariAdministrativeStrategies.Unpublished Ph.D. dissertation,Departmentof Anthropology,
CornellUniversity,Ithaca.
BenavidesC., Mario
1984 Cardcter del estado War(.Universidad Nacional
MayorSan Crist6balde Huamanga,Ayacucho.
1991 Cheqo Wasi, Huari.In Huari AdministrativeStrucArchitectureandStateGovture:PrehistoricMonumental
ernment, edited by William H. Isbell and Gordon F.
McEwan,pp.55-69. DumbartonOaks,Washington,D.C.
Bennett,WendellC.
1953 Excavationsat Wari,Ayacucho,Peru.YaleUniversity
Publications in Anthropology No. 49. Yale University
Press,New Haven.
Bradley,Richard
1989 RuinedBuildings,RunedStones:Enclosures,Tombs
and NaturalPlaces in the Neolithic of South-WestEngland. WorldArchaeology30:13-22.
1998 The SignificanceofMonuments;On the Shapingof
HumanExperiencein Neolithicand BronzeAge Europe.
Routledge,London.
BragayracDavila, Enrique
1991 ArchaeologicalExcavationsin theVegachayoqMoqo
Sectorof Huari.In HuariAdministrativeStructure:Pre-

30

LATINAMERICAN
ANTIQUITY

[Vol.15, No. 1, 2004

imo LopezQuispe,pp. 9-27. Laboratoriode Arqueologia,

historicMonumentalArchitectureand StateGovernment,
Facultadde Ciencias Sociales, UniversidadNacional de
editedby WilliamH. Isbell and GordonF McEwan,pp.
San Cristobalde Huamanga,Ayacucho.
71-80. DumbartonOaks,Washington,D.C.
Isbell,WilliamH.
Brown,James.(editor)
1977 TheRuralFoundationfor Urbanism.Illinois Studies
1971 Approachesto the Social Dimensions of Mortuary
in Anthropology No. 10. University of Illinois Press,
Practices. Memoirsof the Society for AmericanArchaeUrbana.
ology No. 25. Washington,D.C.
1983 SharedIdeologyandParallelPoliticalDevelopment:
CabreraRomero,Martha
HuariandTiwanaku.InInvestigationsoftheAndeanPast:
1998 Evaluacidnarqueoldgicaen el complejoturisticode
Papers of the First Annual Northeast Conference on
Nawimpuquio.Informeal InstitutoNacional de Cultura
AndeanArchaeologyand Ethnohistory,editedby Daniel
del Peri, Ayacucho.
H. Sandweiss,pp. 186-208. CornellLatinAmericanStudCannon,Aubery
ies Program,Ithaca,New York.
1989 The HistoricalDimension in MortuaryExpressions
1985 El origendel estadoen el Vallede Ayacucho.Revista
of Status and Sentiment. Current Anthropology
Andina3(1):57-106.
30:437-458.
1987 Conchopata:IdeologicalInnovatorin Middle Hori2002 SpatialNarrativesof Death,Memory,andTranscenzon IA. Nawpa Pacha 22-23:91-126.
dence.In TheSpaceandPlace ofDeath, editedby Helaine
1991 HuariAdministrationand the OrthogonalCellular
Silvermanand David B. Small, pp. 191-199. AnthropoArchitectureHorizon.In HuariAdministrativeStructure:
logical Papersof the AmericanAnthropologicalAssociaPrehistoricMonumentalArchitectureand State Governtion, No. 11. Arlington,Virginia.
ment,editedby WilliamH. IsbellandGordonE McEwan,
Carr,Christopher
1995 MortuaryPractices:TheirSocial,Philosophical-Relipp. 293-315. DumbartonOaks,Washington,D.C.
andPhysicalDeterminants.Journal
1995 "AsYou Like It":Or,How Do We Know the Prehisgious, Circumstantial
SocitoricPast?InJournalof theStewardAnthropological
of ArchaeologicalMethodand Theory2(2):105-200.
etv (Special issue "Current Research in Andean
Conrad,GeoffreyW.
1982 The BurialPlatformsof ChanChan:Some Social and
Antiquity"),editedby Ari ZighelboimandCarolBarnes,
PoliticalImplications.In ChanChan:AndeanDesertCity,
pp. 1-12. Universityof Illinois,Urbana.
1997a Mummies and Mortuary Monuments: A Postedited by Michael E. Moseley and Kent C. Day, pp.
87-117. School of AmericanResearchand Universityof
processualPrehistoryofAndeanSocialOrganization.UniNew Mexico Press,Albuquerque.
versityof TexasPress,Austin.
1997b ReconstructingHuari:A CulturalChronologyfrom
Cook,AnitaG.
theCapitalCity.InEmergenceand ChangeinEarlyUrban
1987 The Middle HorizonCeramicOfferingsfrom ConSocieties, edited by Linda Manzanilla, pp. 181-227.
chopata.NawpaPacha 22-23:49-90.
PlenumPress,New YorkandLondon.
1994 Wariy Tiwanaku:Entreel estilo y la imagen.Pontifica UniversidadCatolica,Lima.
2001a Repensandoel HorizonteMedio: El caso de Conchopata,Ayacucho,Peri. InBoletindeArqueologiaPUCP
Doyle, MaryEileen
No. 4, 2000 Huari y Tiwanaku:Modelos vs. evidencias,
1988 TheAncestor Cultand Burial Ritual in Seventeenth
and Eighteenth-Century
CentralPeru.Ph.D. dissertation,
primeraparte, edited by Peter Kaulickeand William H.
Isbell, pp. 9-68. Departamentode Humanidades,EspeUniversityof California,Los Angeles. UniversityMicrocialidad de Arqueologifa,PontifficaUniversidadCat61lica
films,Ann Arbor.
del Perd,Lima.
Dulanto,Jalh
2001b Huariy Tiahuanaco,arquitectura,identidady reli2002 TheArchaeologicalStudyofAncestorCultPractices:
The Case of PampaChica,a LateInitialPeriodandEarly
gion. In Los dioses del antiguo Peru, Vol.II, edited by
HorizonSite on the CentralCoast of Peru.In TheSpace
KrzysztofMakowski,pp. 1-37. ColeccidnArtey Tesoros
del Peru,Banco de Creditodel Peru,Lima.
and Place of Death, edited by Helaine Silverman and
David B. Small, pp. 97-117. AnthropologicalPapersof
2002 Reflexionesfinales.InBoletin deArqueologiaPUCP
No. 5, 2001 Huari y Tiwanaku:Modelos vs. evidencias,
theAmericanAnthropological
Association,No. 11. Arlington,Virginia.
segundaparte, edited by PeterKaulickeandWilliamH.
de Humanidades,EspeIsbell,pp.455-479. Departamento
Duviols, Pierre
cialidadde Arqueologifa,PontificaUniversidadCat61lica
1988 CulturaAndina
y repression:Procesosy visitasde idoldel Perni,Lima.
atriasy hechiceriasCaxatambo,sigloXVII.Centrode EstuandLindaSpickard
dios RuralesAndinos"Bartolomede Las Casas,"Cuzco.
Isbell,WilliamH.,ChristineBrewster-Wray,
1991 Architectureand Spatial Organizationat Huari.In
Goldstein,Lynne
HuariAdministrativeStructure:PrehistoricMonumental
1980 MissippianMortuaryPractices:A Case Studyof Two
Architectureand StateGovernment,editedby WilliamH.
Cemeteriesin the LowerIllinois Valley.ScientificPapers
No. 4. NorthwesternUniversityArchaeologicalProgram,
Isbell and Gordon E McEwan, pp. 19-53. Dumbarton
Oaks,Washington,D.C.
Evanston,Illinois.
1981 One-DimensionalArchaeology and Multi-Dimen- Isbell,WilliamH., andAnitaG. Cook
1987 Ideological Origins of an Andean Conquest State.
sionalPeople:SpatialOrganizationandMortuaryAnalysis. In The Archaeology of Death, edited by Robert
Archaeology40:27-33.
2002 A New Perspectiveon Conchopataand the Andean
Chapman,Ian Kinnes and Klavs Randsborg,pp. 53-69.
Vol.II:Art,LandMiddleHorizon.InAndeanArchaeology
CambridgeUniversityPress,Cambridge.
GonzalezCarr6,Enrique,andEnriqueBragayracDfvila
scape and Society, edited by Helaine Silverman and
WilliamH. Isbell,pp.249-305. KluwerAcademic/Plenum
1996 El areaceremonialen la ciudadde Wari:Una hipotePublishers,New YorkandLondon.
sis. In El TemploMayoren la ciudad de Wari,edited by
EnriqueGonzalezCarrY,EnriqueBragayracDivila, Cir- Isbell,WilliamH., and KatharinaJ. Schreiber
1978 WasHuaria State?AmericanAntiquity43:372-389.
ilo VivancoPomacanchari,VeraTiesler Blos, and Max-

Isbell]

PERU
HORIZON
A WARICULTURE
CASESTUDYFROMMIDDLE
MORTUARY
PREFERENCES:

Kaulicke,Peter
2000 Memoriay muerteen el Peri antiguo.PontificiaUnidel Peru,Lima.
versidadCat61lica
Lumbreras,Luis Guillermo
1974a Lasfundacionesde Huamanga.EditorialNuevaEducation,Lima.
1974b The Peoples and Culturesof AncientPeru. SmithsonianInstitution,Washington,D.C.
1985 El ImperioWari.In Historiadel Perui,TomoII. Juan
MejiaBaca, Lima.
MacCormack,Sabine
1991 Religionin theAndes:VisionandImaginationin Early
ColonialPeru.PrincetonUniversityPress,Princeton.
McEwan,GordonF
1991 Investigationsat thePikillactaSite:A ProvincialWari
Centerin the Valley of Cuzco. In Huari Administrative
Structure:PrehistoricMonumentalArchitectureandState
Government,edited by William H. Isbell and GordonF.
McEwan, pp. 93-120. DumbartonOaks, Washington,
D.C.
1996 ArchaeologicalInvestigationsat Pikillacta,a Wari
Site in Peru.Journalof FieldArchaeology23:169-186.
1998 The Functionof Niched Halls in WariArchitecture.
LatinAmericanAntiquity9:68-86.
Morell,Virginia
2002 Empires Across the Andes. National Geographic
Magazine201,6, June:106-129.
OchatomaParavicino,Jose, andMarthaCabreraRomero
2001 PobladosruralesHuari:UnavisidndesdeAqoWayqo.
CANOAsociados SAC, Lima.
ParkerPearson,Michael
1982 MortuaryPractices,Society and Ideology:An Ethnoarchaeological Study. In Symbolic and Structural
Archaeology,edited by Ian Hodder,pp. 99-113. CambridgeUniversityPress,Cambridge.
1993 The Powerful Dead: ArchaeologicalRelationships
Between the Living and the Dead. CambridgeArchaeological Journal3:141-151.
2002 PlacingthePhysicalandtheIncorporealDead:Stonehenge and Changing Concepts of Ancestral Space in
NeolithicBritain.In TheSpaceandPlace ofDeath, edited
by Helaine Silvermanand David B. Small, pp. 145-160.
AnthropologicalPapersof theAmericanAnthropological
AssociationNo. 11.Arlington,Virginia.
PerezCalder6n,Ismael
1999 Huari: Misteriosa ciudad de piedra. Facultadadde
CienciasSociales,UniversidadNacionalSanCrist6balde
Huamanga,Ayacucho.
2001a Estructurasmegaliticasfunerariasen el complejo
Wari. In Boletin de Arqueologia PUCP No. 4 Wariy
Modelosvs.Evidencia,editedby PeterKaulicke
Tiwanaku:
andWilliamH. Isbell,pp. 505-547. PontificaUniversidad
Cat6licadel Peru,Lima.
2001b Investigacionesen la periferiadel complejo Huari.
In XII CongresoPeruanodel Hombrey la CulturaAndina, TomoII, editedby Ismael Perez,WalterAguilarand
MedardoPurizaga,pp. 246-270. UniversidadNacional
de San Crist6balde Huamanga,Ayacucho.
Pozzi-EscotB., Denise
1985 Conchopata:Un pobladode especialistasduranteel
HorizonteMedio. Boletin del InstitutoFrancdsde EstudiosAndinos 14 (No. 3 y 4):115-129.
1991 Conchopata:A Community of Potters. In Huari
AdministrativeStructure:PrehistoricMonumentalArchitectureand StateGovernment,editedby WilliamH. Isbell
and GordonE McEwan, pp. 81-92. DumbartonOaks,
Washington,D.C.

31

Pozzi-EscotB., Denise,MarleneAlarc6nG., andCiriloVivanco

P.
1994 Cerimica Wariy su tecnologiade producci6n:Una
visi6n desde Ayacucho.In Tecnologiay organizacidnde
la produccidnde cerdmicaPrehispdnicaen los Andes,
editedby IzumiShimada,pp. 269-294. PontificaUniversidadCat6licadel Peri, Lima.
1998 WariCeramicsandProductionTechnology:TheView
fromAyacucho.In AndeanCeramics:Technology,Organization,and Approaches,edited by Izumi Shimada,pp.
253-281. MASCA ResearchPapers,UniversityMuseum
of Archaeologyand Anthropology,Universityof Pennsylvania,Philadelphia.
Salomon,Frank
AndeanAncestorShrines
1995 TheBeautifulGrandparents:
and MortuaryRitualas Seen ThroughColonialRecords.
In Tombsfor the Living: Andean MortuaryPractices,
edited by Thomas Dillehay, pp. 315-353. Dumbarton
Oaks,Washington,D.C.
Saxe, Arthur
1970 Social DimensionsofMortuaryPractices.Ph.D. dissertation, Departmentof Anthropology, University of
Michigan.UniversityMicrofilms,AnnArbor.
Schreiber,Katharina
1991 Jincamocco:A HuariAdministrativeCenterin the
SouthCentralHighlandsof Peru.In HuariAdministrative
ArchitectureandState
Structure:PrehistoricMonumental
Government,edited by William H. Isbell and GordonF.
McEwan, pp. 199-213. DumbartonOaks, Washington,
D.C.
1992 WariImperialismin Middle Horizon Peru. AnthropologicalPapersof the Museumof AnthropologyNo. 87.
Universityof Michigan,Ann Arbor.
Shady,Ruth
1982 LaculturaNieveriay la interacci6nsocialen el mundo
Andinoen la 6pocaHuari.Arqueoldgicas(MuseoNacional
de Antropologiay Arqueologia,Lima) 19:5-108.
1988 La 6poca Huaricomo iteracci6nde las sociedades
regionales.RevistaAndina6(1):67-99.
Shady,Ruth,andArturoRuiz
1979 EvidenceforInterzonalRelationshipsDuringtheMiddle Horizonon the North-CentralCoast of Peru.American Antiquity44:676-684.
Silverman,Helaine
2002 Introduction:The Space and Place of Death. In The
Space and Place of Death, edited by Helaine Silverman
andDavid B. Small,pp. 1-11. AnthropologicalPapersof
theAmericanAnthropologicalAssociationNo 11.Arlington, Virginia.
Smith,MonicaL.
2003 Introduction:The Social Constructionof Ancient
Cities.In TheSocial ConstructionofAncientCities,edited
by Monica L. Smith, pp. 1-36. SmithsonianInstitution,
Washington,D.C.
Tainter,James
1978 MortuaryPracticesandtheStudyof PrehistoricSocial
Systems.InAdvancesinArchaeologicalMethodand TheoryVol.1, editedby MichaelB. Schiffer,pp. 105-141.Academic Press,New York.
Thomas,Julian
1996 Time, Cultureand Identity.Routledge, London &
New York.
TieslerBlos, Vera
1996 Los entierrosdel sitioWari:Estudiode unapoblaci6n
Prehispinica.In El TemploMayor en la ciudadde Wari,
edited by EnriqueGonzalez Carr6,EnriqueBragayrac
Davila, CiriloVivancoPomacanchari,VeraTieslerBlos,

32

[Vol.15, No. 1,2004]

LATINAMERICAN
ANTIQUITY

andMaximoLopez Quispe,pp. 111-135. Laboratoriode

Arqueologifa,Facultadde CienciasSociales, Universidad
Nacionalde San Cristobalde Huamanga,Ayacucho.
Topic,JohnR.
1986 A Sequenceof MonumentalArchitecturefrom Huamachuco.In Perspectiveson AndeanPrehistoryand Protohistory,edited by Daniel H. Sandweiss and D. Peter
Kvietok,pp.63-83. LatinAmericanStudiesProgram,Cornell University,Ithaca.
1991 Huari and Huamachuco.In Huari Administrative
Structure:PrehistoricMonumentalArchitectureandState
Government,edited by William H. Isbell and GordonE
McEwan, pp. 141-164. DumbartonOaks, Washington,
D.C.
1994 El izamientodel gallardeteen Huamachuco.In En el
nombredel Seiior:Shamanes,demonios,y curanderosdel
nortedel Perui,editedby Luis Millonesand Moises LemAuslij, pp. 102-127. BibliotecaPeruanade Psicoanailisis.
tralisS.A., Lima.
Topic,Theresa
1991 The Middle Horizon in Northern Peru. In Huari
AdministrativeStructure:PrehistoricMonumentalArchitectureandState Government,editedby WilliamH. Isbell
and GordonE McEwan,pp. 233-246. DumbartonOaks,
Washington,D.C.
Topic,JohnR., andTheresaLangTopic
1992 The Rise and Decline of CerroAmaru:An Andean
ShrineDuringthe EarlyIntermediatePeriodand Middle
Horizon.InAncientImages,AncientThought:TheArchaeology of Ideology,edited by A. S. Goldsmith,S. Garvie,
D. Selin, andJ. Smith,pp. 167-180. Universityof Calgary
ArchaeologicalAssociation,Calgary.
2001 Hacia la comprensi6ndel fen6menohuari:Una perapectiveNortefia.InBoletinde ArqueologiaPUCP No. 4,
2000 Huariy Tiwanaku:Modelosvs. evidencias,primera
parte, editedby PeterKaulickeandWilliamH. Isbell,pp.
102-127. Departamento
de Humanidades,
Especialidadde
Arqueologfa, Pontifica UniversidadCat61licadel Peril.
Lima.
Topic,TheresaLange,andJohnR. Topic
1984 HuamachucoArchaeological Project:Preliminary
Reporton theThirdSeason,June-August1983.TrentUniversity OccasionalPapersin AnthropologyNo. 1, Peterborough,Ontario.
Tung,TiffinyA.
2003 BioarchaeologicalAnalysis of WariTrophyHeads:
Evidencefrom Conchopata,Peru.Paperpresentedat the
72nd Annual Meeting of the American Association of
PhysicalAnthropology,Tempe,Arizona.
Tung,Tiffiny,andAnitaCook
2002 IntermediateElites andTheirRole in WariImperialism as IdentifiedThroughBioarchaeologicaland MortuaryAnalysis.Paperpresentedat the 67thAnnualMeeting
of the Society for AmericanArchaeology,Denver.
Woolley,C. Leonard
1934 Ur Excavations.Vol.2: TheRoyal Cemetery.British
Museum and University Museum of the University of
Pennsylvania,Londonand Philadelphia.
Wylie,Alison
1992a On "HeavilyDecomposing Red Herrings":ScientificMethodinArchaeologyandtheLadeningof Evidence

withTheory.In Metaarchaeology:ReflectionsbyArchaeologists and Philosophers,edited by LesterEmbree,pp.

260-288. KluwerAcademic,Dordrecht.
1992b The Interplayof EvidentialConstraintsand Political Interests:RecentArchaeologicalResearchon Gender.
AmericanAntiquity57:15-34.
Zapata,Julinho
1997 Arquitecturay contextos funerariosWarien Batan
Urqu,Cusco. In La muerteen el antiguoPerdi,editedby
PeterKaulicke,pp. 165-206. BoletinArqueologicaPUCP
del Peru,Lima.
Vol. 1. PontificaUniversidadCat61lica

Notes
1. "Huari"is also spelled "Wari."This name refersto the
archaeologicalruinsof a greatcity in Peru'scentralhighland
AyacuchoValley. It also refers to the art style and archaeological culturethat probablyoriginatedin the city, and was
spreadacross much of the CentralAndes duringthe Middle
Horizon (A.D. 550 to 1000). To reducethe confusion, I have
proposed(Isbell 2002) that "Huari"be used for the city and
its contents, while "Wari"be employed for the broadlydiffused cultureand its distinctiveart found outside the capital
city. I follow that practicein this article.
2. I wish to recognize the co-directors,project administrator,sponsors,and otherparticipantsand contributorsto the
Conchopata Archaeological Project. Please see
"Acknowledgments"at the end of the article. Special thanks
are due Dr. Tiffiny Tung for her painstakinganalyses of the
Conchopataskeletal remains, and the preliminaryinformation presentedhere. Bioarchaeologicalstudy of these materials is continuing.
3. This discussiondeals with the burialof adultsandjuvenile children.Except where they were placed in what appear
to have been family tombs, the burialof fetuses and infants,
as well as young children, was significantly different from
burial for adults and youths. This probablyexpressed practices appropriatefor differentage grades.Complete analysis
of Conchopataburial practices, including the intermentof
children,will be presentedin the future.
4. A tupu is a long pin with flat head ethnohistorically
used by women to fasten a wrap-aroundgarmentover their
shoulders.
5. ChallengingIsbell and Cook's originalconclusionthat
the women were sacrificialvictims, recentre-examinationof
the bones by TiffinyTungfailed to detect evidence of violent
death. Of course, strangulation,poison, and othertechniques
for killing would leave no detectableevidence, especially on
poorly preservedbones, as these are. But the conclusion that
the women were sacrificedrequiresmore examinationin the
future.

SubmittedJanuary31, 2003; AcceptedOctober3, 2003;

RevisedNovember18, 2003.