A MITO-STYLE STRUCTURE AT CHAVIN DE HUANTAR: DATING AND IMPLICATIONS

Daniel A. Contreras

Excavations west of themonumental core at Chavin de Hudntar, Peru in 2005 revealed a well-preserved plastered struc turewith a central circular hearth, in the style of theMito Architectural Tradition. This find challenges standard definitions of both theMito Tradition and Chavin itself I discuss thematerial remains and associated radiocarbon dates from thisfea ture, and use these new data to re-assess Chavin's involvement in interregional networks and its relationship to earlier cer emonial centers in the Central Andean highlands.

Excavaciones al oeste del nucleo monumental de Chavin de Hudntar, Peru en 2005 expusieron una estructura enlucida con unfogon circular central, en buen estado de preservacion, del estilo arquitectonico Mito. Este hallazgo permite cuestionar los conceptos existentes sobre la Tradicion Mito y Chavin en si En este articulo discuto los restos materiales yfechados radio carbonicos asociados a este elemento arquitectonico, y utilizo estos nuevos datos para reevaluar la participacion de Chavin en redes de interaccion a nivel interregional y su relacion a otros centros mas tempranos de la sierra de los Andes Centrales.

The Late Preceramic Period in the Central dition widespread in theAndes, defined by the Andes has been increasingly recognized as burning of offerings in enclosures with central a time of notable cultural complexity. The hearths. Elisabeth Bonnier (Bonnier 1987, 1988; majority of thiswork?dating toMoseley's proposal Bonnier and Rozenberg 1988) and Rosa Fung of maritime-based development of "Andean civi (Fung Pineda 1988) have also used a similar set of lization" (Moseley 1975)?has been focused on the data?as well as, in Bonnier's case, the architec coastal valleys of Peru (recently,for example, Haas tural sequence from the site of Piruru?to suggest and Creamer 2006; Shady and Leyva 2003; Shady themore restricteddefinition of what they termthe Solis et al. 2001). While thehighlands have been less Mito Tradition; I explore the distinction between studied, recognitionof theirimportance in theLate the two below. Preceramic (sometimes also called theLate Archaic) My 2005 excavations at the InitialPeriod/Early has a long history,dating to the excavations of the Horizon siteof Chavin de Huantar revealed a struc Japanese Mission toNuclear America at Kotosh ture very similar to the known KRT/Mito exam (Izumi and Sono 1963; Izumi and Terada 1972). ples fromKotosh, La Galgada, and elsewhere. This One of theprincipal tools used toexamine inter find is remarkable for two reasons. First, its asso regional contact and interaction in the Late ciated radiocarbon dates fall in thefirst half of the Preceramic?the concept of theKotosh Religious firstmillennium B.C., nearly 1,000 years later than Tradition (KRT)?was developed by Richard theLate Preceramic examples mentioned above. Burger and Lucy Salazar-Burger on the basis of Second, it has been thought thatChavin repre excavations atKotosh (Izumi and Sono 1963; Izumi sented a cultural departure from this early and and Terada 1972) and La Galgada (Bueno Men long-livedAndean tradition.This paper describes doza and Grieder 1980; Grieder and Bueno Men the recent find at Chavin and discusses its impli doza 1981) and their own work at the site of cations forour understanding of Chavin and inter Huaricoto (Burger and Salazar-Burger 1980,1985, regional interaction in thePeruvian Initial Period 1986). The KRT, they suggested, was a ritual tra and Early Horizon.

Daniel A. Contreras Archaeology Center, Stanford University, P.O. Box 20446, Stanford, CA 94309 ([email protected])

Latin American Antiquity 21(1), 2010, pp. 3-21 Copyright ?2010 by the Society forAmerican Archaeology

3

This content downloaded from 129.252.86.83 on Thu, 13 Mar 2014 23:00:42 PM All use subject to JSTOR Terms and Conditions 4LATIN AMERICAN ANTIQUITY [Vol. 21, No. 1,2010

e-z^zzbhbbhhbihh Kilometers ^^^^Bfe"'" DEM derivedfrom gtopo30 data ^IJsBflHHN gcs84 ^__Gr '"jaBBBI

Figure 1. Location of Chavm de Huantar.

Background since. Two projects are particularly salient. Exca vations directed by Luis Lumbreras and Hernan a on Set in high valley theeastern slope of theCen Amat (1966-1974) revealed previously unknown tralAndes' Cordillera Blanca (see Figure 1), the site architecturalelements, most spectacularly theCir of Chavin de Huantar is a sprawling complex of cular Plaza and theOfrendas Gallery, and began monumental stone construction, unique in the the process of establishing a radiocarbon and region in its scale and sophistication.Archaeolog ceramic chronology for the site (Lumbreras 1977, ical investigationof the site began in earnest with 1993, 2007). Burger's subsequent 1975-76 exca Julio C. Tello's visit in 1919 (Tello 1943, 1960), vations focused on the buried domestic remains continued when Wendell Bennett carried out the surrounding themonumental core of the site, and firstscientific excavations atChavin in 1939 (Ben particularly on stratigraphicallyexcavated ceram nett 1944), and has burgeoned in the near-century ics (Burger 1984, 1998).

This content downloaded from 129.252.86.83 on Thu, 13 Mar 2014 23:00:42 PM All use subject to JSTOR Terms and Conditions Contreras] A MITO-STYLE STRUCTURE AT CHAVIN DE HUANTAR

Figure 2. Areas of the site discussed in text.

While this work?and the ongoing Stanford terms?in the largerpanorama of Central Andean University Chavin de Huantar Research and Con prehistory.These various projects have led to sig servationProject (Rick 2005; Rick et al. 1998)? nificant disagreement about Chavin's chronology has much improved our understanding of Chavin, (Burger 1998, 2008; Kembel 2008; Kembel and the site is large and complex enough thatmany Rick 2004; Lumbreras 2007; Rick 2005, 2008); I new questions remain. Moreover, as data have come consider below theways in which the data pre to lightresearchers have consistently revisited the sented here articulate with this debate. These fundamental archaeological questions at Chavin: debates at Chavin are embedded within pan understanding the sociopolitical dynamics that Andean problems of chronological terminology made the site's construction possible and assess (Kaulicke 1998; Silverman 2004). For the sake of ing its place?chronologically and in systemic ease of comparison with the published syntheses

This content downloaded from 129.252.86.83 on Thu, 13 Mar 2014 23:00:42 PM All use subject to JSTOR Terms and Conditions 6 LATIN AMERICAN ANTIQUITY [Vol. 21, No. 1,2010

bench, immediately east of the hearth, is pene tratedby a 29 cm wide by 12 cm high duct (Fig ure 4, Element 6). Carbon recovered from the inner reaches of that duct suggests that itmay feed another hearth, but the limits of the excava tion did not allow exploration of thatpossibility. It is noteworthy thatanalogously located features Monumental at Kotosh were as c^^^^^-^^^ ^/^^*C^?''^^^^V^ (see Figure 5) reported niches, not ducts, butwere as much as 50 cm deep (Izumi and Terada 1972:144-164). The prominent niches in both bench and wall atKotosh and La Galgada thus apparently have no counterpart at Chavin, though furtherexcavation may of course change thatperception. If the hearth is taken to be truly central, the 3. View of ESE out of themouth of Figure Chavm, looking sunken area thehearth would have mea the canyon of the Rio Wacheqsa. Note location ofWF-07 containing relative tomonumental core. sured approximately 2m by 2 m; such a recon struction would indicate that we excavated on theKotosh Religious Tradition and theMito Tra approximately 50 percent of the central sunken dition, as well as theoriginal excavations at themost area (what Bonnier terms the pericaust). Unfortu prominent sites considered, I here employ Rowe's nately,with only thenorthern extreme of the struc (1967) chronological terminology.This is a prac tureproperly defined, precisely estimating the size tical measure only; addressing the problems of of the entire structure is impossible. However, if chronological terminology in theCentral Andes is we posit relative symmetry around the hearth, a beyond the scope of thispaper. roughly4 m x 4 m chamber is reasonable (see Fig My 2005 excavations, carried out as part of the ure 5). Stanford Project, focused on the periphery of the The upper portion of the northern facade, pre site's monumental core (see Figures 2 and 3; also sent on the east side of the entry, is damaged, but Contreras 2007). These resulted in the discovery, what remains suggests a facade at least 75 cm high. in the area known as theWest Field, of a structure The wall may have been higher, or supported a less that closely matches the architectural forms permanent superstructure,but no evidence of either described by Burger and Salazar-Burger (1980) was found. It is also possible thatthe structurewas and Bonnier (1997) as diagnostic of theKotosh unroofed, but the vulnerability of exposed plaster Religious Tradition and theMito Tradition, respec during the pronounced rainy season in the region tively. suggests that a roof would have been necessary (unless the structurehad a use lifeof only one dry season before sealed; the scale of Excavation of theMito-style Structure being given investmentin theconstruction of the structure,this Excavated features of the structureinclude several seems unlikely). salient elements (see Figure 4): entryand exterior The only exception to the spare, undecorated step (Element 1), floor (pericaust) (2), lateralbench appearance of the chamber was the remnant of a (epicaust) (3), central circular hearth (4), and par sculptural pillar affixed to the northern facade, tial facade (5). These elements combine to form a flanking theentry to theeast (Figure 4, Element 7). small rectilinear structure,accessed from thenorth Although more thanhalf of thepillar was destroyed by ascending at least one step toenter, whose focus prior to the sealing of the chamber,what remained was a small perfectlycircular hearth 39 cm indiam was sufficient to indicate an irregular cylinder, eter and 11 cm deep in the center. embedded in the chamber floor and extending at While the circular hearth does not have any least as high as thepreserved portion of facade. The ventilation duct, as prominently featured in the remnant portion was sufficient to suggest a form well-known examples fromKotosh, the face of the that bowed outward just above the base before

This content downloaded from 129.252.86.83 on Thu, 13 Mar 2014 23:00:42 PM All use subject to JSTOR Terms and Conditions Contreras] A MITO-STYLE STRUCTURE AT CHAVIN DE HUANTAR

?

O

0

Figure 4. Elements ofWF-07 Mito-style structure: entry and exterior step (1), floor/pericaust (2), lateral bench/epicaust (3), central circular hearth (4), partical facade (5), duct (6), remnant sculptural pillar (7), and construction detail (8).

was tapering to a neck -45 cm above the floor; above painted ormolded decoration found.The orig the neck thepillar flared out to form an irregular inal color of the plaster was also difficult to dis form too damaged to reconstruct.1The emphasis cern, as it appears to have been affected by the on entryways thatBonnier (1997) ascribes to the grey-green soil of the sealing fill.White or yellow Mito architectural form is germane here; also of white are themost likelycandidates; theonly color interestis the step entry,which argues thatthe struc variation observed was in and around the hearth, turewas raised relative to its surroundings. Inves where the plaster was heat-altered to a brick-red. tigationsof theexposure provided by thecut of the The lightcolor of theplaster distinguished itfrom Rio Wacheqsa, approximately 5 m to the north, the darker clay matrix of the substructure,but that suggest thatcontemporary ground surfacewas at light color seems to have been a characteristic of most 1.5m lower, and an associated floor proba the plaster itself rather than of any applied slip bly sat only 70 cm below the level of the structure (unlike La Galgada and Kotosh?see Bonnier (see Figure 6). 1997:137). The construction technique of the structure Cultural material was recovered fromonly two appears to be uniform;where damage to the plas areas: the circular hearth and the inner reaches of terhas exposed substructure,selected angular rock the duct. In both contexts thematerials appeared had been placed in a mud mortar, towhich plaster to be primary depositions, while the remainder of had been applied and carefully smoothed (Figure theexcavated area gave every indication of having or 4, Element 8). The only exception, apparently,was been carefully cleaned before its interment, per the remnant of the sculptural pillar affixed to the haps kept clean during its use life (see discussion facade, flanking theentry, which had no stone sub of depositional context below). structureand consisted entirelyof a prepared mud The inner reaches of the duct (40-50 cm from con covered with plaster. Overall, the plaster was its opening in the face of the lateral bench) as a remarkably well preserved, but no evidence of tained abundant carbon, as well fragment of

This content downloaded from 129.252.86.83 on Thu, 13 Mar 2014 23:00:42 PM All use subject to JSTOR Terms and Conditions 8 LATINAMERICAN ANTIQUITY [Vol. 21, No. 1,2010

MHo StructureSchematic WF-07/07A Reconstruction (adaptedfrom Izumi and Terada1972:163 and Bonnier1997:138; based on theTemple of theCrossed Hands)

Figure 5. Schematic drawings of an idealized Mito Structure (left) and a hypothetically reconstructed WF-07 structure (right).

an anthracitemirror and a solitary ceramic sherd bration curve (BronkRamsey 2001,2006; McCor (unfortunately,like the two other sherds recovered mac et al. 2004), the samples from thehearth itself from thehearth, a small?approximately 2 x 2 cm combine for a tighterrange, centered on 800 cal ?nondiagnostic plainware body sherd). B.C. The hearth dates, moreover, seem from their The circular hearthwas filled with in situ car context tomark the terminusof occupation, sug bon and ash deposits without internalstratification. gesting that the structure itselfdates back some The deposit also included two small, nondiagnos what further.Moreover, itbears repeating that the tic ceramic sherds, bone remnants too heat WF-07 Mito-style structureis not a basal deposit? damaged to be much more than powder, and a architectural evidence from the cut of the Rio quantity of fragmentary obsidian (a mixture of Wacheqsa to the north demonstrates the presence smallwaste flakes and shatter; 129 fragments total of earlier construction stratigraphicallyprior to the ing 22 g). Analysis of macrobotanical and phy structure(see Figure 6). Further, theKotosh exam tolith remains from thehearth deposit is ongoing. ple of multiple superposed and juxtaposed Mito Two fragments of wood charcoal (samples AA structures suggests that this chamber may be part 69446 and AA 69447) from this hearth deposit of a largercomplex ratherthan a stand-alone struc yielded radiocarbon dates thatwhen calibrated fall ture.Given the limitedareal exposure thusfar pos roughly between -850-750 B.C. (the 2-sigma sible at Chavin, assessment of this prospect ranges are 841-540 B.C. and 911-772 BC); see unfortunatelymust remain speculative. Figure 7).2 Four carbon have thus far been dated samples Context and Process by theUniversity ofArizona AMS Laboratory, after pretreatmentby Dr. Herbert Haas at RC Consul The 2005 excavation thatexposed thenortheastern tants. All samples were charcoal, from well quadrant of theplastered chamber described above documented and well-sealed contexts; two from was located approximately 300 m west of themon the contents of the hearth and two from distinct umental core. As the investigatorsdid not antici strata in theoverlying deposits. The results are both pate the exposure of such a structure, time and internally regular and consistent with the strati funding did not allow a broad areal excavation. graphic relationships observed during excavation Nevertheless, theportion excavated has important (see Figure 7).While thedates from theupper strata implications for the early development of the cer encompass a broad range (800-400 cal B.C.) due emonial center of Chavin de Huantar and its rela to the flatness of the relevant portion of the cali tionship to other early highland centers. I here

This content downloaded from 129.252.86.83 on Thu, 13 Mar 2014 23:00:42 PM All use subject to JSTOR Terms and Conditions Contreras] A MITO-STYLE STRUCTURE AT CHAVIN DE HUANTAR 9

road construction, suggesting a structure or struc Modernground surface(eroding slope) tures in theWest Field rather than simple mega lithic terraces; this leadsDiessl todescribe a "West Temple" (Diessl 2004:510-516). The apparently substantialChavm-period architecture in this sec tor has mostly been buried by a combination of slope processes, subsequent occupation, and a major aluvion (debris flow) in 1945 (Diessl 2004; Indacochea G. and IbericoM. 1947; Turner et al. 1999).3 The combination of slope erosion and thismod ernmining of the aluvion deposit meant thatour excavation rapidly exposed archaeological strata in spite of the substantial post-occupation deposition. The shallow overburden covered a series of loose fills of small angular rock, alternating with thin lenses of clayey soil indicating periods of rapid, apparently intentionaldeposition of fill alternating with brief periods of exposure and natural deposi tion from slope erosion. These strata, judging by the small amounts of cultural material included, appear to represent a post-Chavin attempt to level the area, perhaps also with consideration for ade quate drainage. If therewas a northern retaining wall, itwas either dismantled over theyears or did not survive the southward of the river.To 1m cutting Modernground the east the laterfills, at least,were contained by a surface(base of cut) cdH-v/F-io/!OA ProWe constructedwall of small boul DrownbyFrankiin Romero and roughly alternating ftrtriooQuifonez, 8 Sep! 2006 ders and stacks of chinking cobbles. This wall (Ele ment 1 inFigure 8) was both differentin style than Figure 6. Profile drawing of WF-10/10A?note relative level ofWF-07 hearth. the Chavin-period walls excavated below (Ele ments 2 and 3 inFigure 8) and in a differentarchi tectural orientation. describe the location of thisMito-style structure Below the series of loose rock fills two terraces within the site and itsstratigraphic context and bur were excavated. Both were well-built, consisting ial processes, before turningto those implications. of selected angular stones set in clay mortar to cre core The area west of the monumental of ate walls with flat faces. Some suggestion of plas as Chavin?here referred to theWest Field?has tering remained on the upper terrace, in the form been recognized since at least JulioC. Tello's vis of a very fine, homogenous, and distinctly colored its to the site (beginning 1919) as containing sediment associated with some parts of thewall Chavin-era construction. Two megalithic walls base. The terraces are parallel, running roughly (apparently terrace facades), constructed of east-west (and architectural east-west, in the archi quartzite blocks in a style similar to that of the tecturalalignment of themonument). Between the are visible on area structures in themonumental core, lower terrace and the upper?an approxi the surface,as is one canal draining northward into mately 2 m wide?a fill of angular rock had been con theRio Wacheqsa (see Figure 3). Until the carefully laid, the large rocks making up the fill structionof the road that currently separates the chinked with smaller stones at the surface to cre monumental core from theWest Field in the 1970s, ate a flat substrate for the packed-earth floor that were these east-west terraces also associated with spanned the area between the two terraces. Both a north-southwall thatwas largelydestroyed by the terraces extended into the excavation's western pro

This content downloaded from 129.252.86.83 on Thu, 13 Mar 2014 23:00:42 PM All use subject to JSTOR Terms and Conditions 10 LATINAMERICAN ANTIQUITY [Vol. 21, No. 1,2010

Radiocarbon Samples, Unit WF-07/07A

F5 jWF-07 Charcoal fromupper deposits* WF-07F11L2

WF-07AF17L34 Charcoal fromhearfi WF-07AF17-ts4 ' 1200 600 CalendarDate (BC)

[FrompCfAO) [To(BOAD) p WF-OTfnU WF-Q7APT7L3A WF-Q7APT7L3B

Figure 7. Radiocarbon dates discussed in the text.

file; to the east, as far as the presence of the large, intactand clean, was sealed by a massive deposi late retainingwall above allowed excavation, the tion of completely sterile fill.As much as 90 cm lower terrace continued into the eastern profile thick,with a uniformly level upper surface (thick while the upper was destroyed (see Figure 8). ness thusdepending on thedepth below surface of In association with thebase of the lower terrace structuralfeatures), thefill consisted of a fine grey were a series of compact, material-culture-rich green soilmatrix with abundant but neitheraligned deposits, including some fragmentarypacked-earth nor sorted clasts of angular rock, ranging from surfaces. The ceramics recovered from these small rocks (~3 cm on long axis) to boulders (as deposits and from the two surfaces above (that much as 50 cm on long axis). There was a notice between the two terraces and that atop the upper able trend toward larger inclusions in thefill to the terrace) were consistently and exclusively north, perhaps reflecting some care in placement Janabarriu-style (see Figure 9 for a sample of the of fill within the chamber as opposed to outside it associated decorated ceramics). Two charcoal sam (the excavation straddled the apparent northern ples recovered from these deposits yielded radio facade of the structure).No culturalmaterial what carbon dates thatwhen calibrated fallbetween -800 soever was recovered from the approximately 3.6 and400 B.C. (2-sigmaranges of761-107 B.C. and m3 of fill excavated, and no internaldistinctions 796-417 BC; see Figure 7).4 within thefill were detected, suggesting intentional These lowermost, compact deposits were and careful preparation as well as deposition in a directly atop the sterilefill thatsealed theplastered single fill event. structure,with no natural deposits intervening to The basal portion of the sealing fillwas free of suggest any period of disuse or abandonment the large angular rocks found throughout the rest between the deposition of the fill that seals the of the fill, ensuring that in deposition no damage Mito-style structureand theconstruction of the ter was done to theplastered surfaces,and the surfaces races (see Figure 10).5The temporal proximity of themselveswere swept clean. This suggests a con these deposits to the sealed structurebelow is con certed effortto preserve the structure intactwhen firmed by the radiocarbon dates. itwas buried. The only cultural material and/or The structureitself, described above, was only debris recovered in excavation was from thehearth partially exposed. Even such a limited exposure, and from deep in the duct.While the hearth was however, revealed a plastered chamber consisting apparently leftundisturbed, and simplycovered, the of an entry, split-level floor, and circular central duct was carefully packed with the sealing fill to a ~ hearth,with an associated duct and remnantpillar. depth of 50 cm?approximately an arm's reach. At the end of itsuse-life, that structure,largely The culturalmaterial recovered from theduct was

This content downloaded from 129.252.86.83 on Thu, 13 Mar 2014 23:00:42 PM All use subject to JSTOR Terms and Conditions Contreras] A MITO-STYLE STRUCTURE AT CHAVIN DE HUANTAR 11

CdH-WF-07-F10

Figure 9. Sample of Janabarriu-style ceramics from upper deposits.

theconstruction of themonumental core ofChavin as occurring between approximately 1000-600 B.C. (Kembel and Rick 2004; Rick 2005) brack ets the dates from the structure.While Figure 8. Unit WF-07/07A, with major architectural ele Mito-style ments labeled. These include a late wall (1), two terrace Kembel and Rick remain cautious about the facades (2 and 3), and theMito-style structure (4). absolute dating of the specific architecturalphases within theirwell-anchored relative sequence, the structure within notmixed in thatfill, but deposited deeper in the Mito-style apparently falls clearly duct; itremains unclear whether thatmaterial had theBlack andWhite Stage (see Kembel 2001:312). also been pushed into the duct or if instead ithad The clear stratigraphicprecedence of the structure inrelation to in the entered through another, yet-unexcavated, open Janabarriu-styleceramics upper even ing at theduct's eastern terminus. layers also indicates contemporaneity?or structure This intentional,prepared filling of the structure, temporalpriority?for theMito-style with to at although itwas not followed by construction of a regard the latter ceramic phase Chavin de such new temple atop the old, recalls the "temple Huantar. The upper layers?those containing entombment" described byMatsuzawa at Kotosh, Janabarriu-style ceramics?yielded two floor where superposition of Mito structureswas the associated radiocarbon dates, both falling within norm (Izumi and Terada 1972:176; see also Onuki 820-410 cal B.C. (with an overwhelming proba see 1993, 1999). Burger and Salazar-Burger (Burger bility of pre-490 cal B.C.; Figure 7). In terms and Salazar-Burger 1985:116) also describe simi of absolute dates, thus, these strataalso fallwithin lardeposits atHuaricoto. Kembel's Black and White Phase and Burger's Urabarriu Phase (Burger 1992:231; Kembel and Location the Spatial Temporal of Mito-style 2001:312), while the associated ceramics are dis Structure within Chavi'n de Hudntar tinctivelyJanabarriu-style (Janabarriu is the lastof The structure is linked to the ceremonial core of Burger's three ceramic phases, dating 390-200 Chavin?the well-known monumental architecture B.C. [Burger 1984]). more at the site?by several features.Contemporaneity The stratigraphic evidence also allows is themost obvious: Kembel and Rick's dating of subtle inferences.The careful cleaning (ormainte

This content downloaded from 129.252.86.83 on Thu, 13 Mar 2014 23:00:42 PM All use subject to JSTOR Terms and Conditions 12 LATIN AMERICAN ANTIQUITY [Vol. 21, No. 1,2010

Figure 10. Cut drawing of Unit WF-07/07A.

nance) of the structurefloor prior to thedeposition and isolated. If theMito-style structureis taken to of the sealing fill, coupled with the apparently rel have significant ceremonial importance given the atively undisturbed ash and carbon deposits left in centrality of such architecture at other sites, the the hearth, suggests no interval of abandonment significant construction investment it represents, prior to thedeliberate intermentof the space. Fur and its careful interment,this might imply a dif thermore, the occupation surfaces above that seal fuse rather than concentric model of spatial sig ing fill sit directly atop it,without any deposits nificance within the site.6That is, archaeologists suggesting an interval of abandonment between have tacitlyassumed that theLanzon, theCircular the sealing of the structureand the subsequent occu Plaza, or the Portada?all architectural elements pation. The tightclustering of theradiocarbon dates within themonumental core?represent the site's from above and below the fill supports this inter conceptual focus. The discovery of such an appar pretation. ently noteworthy element as aMito structurewell A final piece of architectural evidence is also outside themonumental core implies, though, that suggestive: the orientation of the structure,judg theremay be multiple foci within the site. ing from the exposed lateral bench and northern However, another possibilitymust also be con facade, is close (-21.3? east of north) to the 13.65? sidered: given thedensity and scale of construction east of northwhich serves as architectural north in features in theWest Field, itmay be necessary to themonumental phases (Rick et al. 1998:194). It reconceptualize themonumental core itself,includ is also noteworthy that the structureopens north, ing the largely buriedWest Field constructions as as did theoldest portion (now thenortheast corner part of the site center. Such a model would leave of StructureA) of the stone architecture in the site theMito-style structurestill peripheral, but not nec core (Rick et al. 1998:194); subsequent construc essarily marginal: its location at thewesternmost tion phases shifted the site orientation 90? to the extreme of the site, adjacent to theRio Wacheqsa, east. places iton themost direct route to theCordillera Also of interestis the structure's location rela Blanca, theCallejon de Huaylas, and the coast to tive to both themonumental core and the sur thewest. This places itastride the prime pathway rounding landscape. Situated approximately 270 m into themonument, and also puts it in a position west of theLanzon, the structureappears, given the fromwhich itdirectly addresses thenot-too-distant modern layoutof the site, tobe somewhatmarginal apus (see Reinhard 1985) and faces both theallure

This content downloaded from 129.252.86.83 on Thu, 13 Mar 2014 23:00:42 PM All use subject to JSTOR Terms and Conditions Contreras] A MITO-STYLE STRUCTURE AT CHAVIN DE HUANTAR 13

Bah faSec aw de las - iBsj^^Q^f^Smw^B^dBtl^^^^^^^^^^^^^^ Pampa Llamas ^0X4^H^^^^^^|^|Cp{sH^^^^^^^^^^^^^^^^^^|

Zi Siteswith MT architecture Bonnier (from 1997) ^^^^^IflBGR^^S^^SHi^^ll^^^HH Siteswith KRT architecture that lack MT features ^^^^^^^BMj^y vwmF^^CySS&St^^ fl it AdditionalMT I sites_| IKfiMfeiE)!

Figure 11. Sites with Mito-style architecture in the Central Andes.

and risk of theQuebrada Wacheqsa, which regu Andean Context larlyprovided lifegivingwater but could erupt in The Mission's excavations at Kotosh catastrophicflood or debris flow. The water of the Japanese found a series of quadrangular plastered structures Wacheqsa River is today (and was perhaps in pre with floors, niches, and central hearths history)used to irrigatethe agricultural basin of the split-level (Izumi and Terada 1972). The multiple structures Mosna Valley, and was also apparently a key ele were in some cases contemporary and in others ment of ritual practice inChavin's ceremonial core in the lattercase earlier structureshad (see Burger 1992:141-143; Contreras and Keefer superposed; been and interred. Subse 2009; Lumbreras et al. 1976). carefully deliberately

This content downloaded from 129.252.86.83 on Thu, 13 Mar 2014 23:00:42 PM All use subject to JSTOR Terms and Conditions LATINAMERICAN ANTIQUITY [Vol. 21, No. 1,2010

quent excavations atLa Galgada byAlberto Bueno (Bonnier 1997). Similarly, Fung, noting common Mendoza and Terence Grieder (Bueno Mendoza alities in the ritual architectureof Kotosh (and the 2004; Bueno Mendoza and Grieder 1980; Grieder associated sites of Shillacoto andWairajirca, also and Bueno Mendoza 1981, 1985; Grieder et al. in theHuallaga drainage) and La Galgada, argued 1988) and Huaricoto by Richard Burger and Lucy that thesewere representativeof a highland tradi Salazar-Burger (Burger and Salazar-Burger 1980, tion of the later Preceramic Period, which she 1985) revealed examples of similar structures termed theMito Tradition (Fung Pineda 1988:73). widely removed fromKotosh (see Figure 11). While Fung's discussion remained general, Bon The construction and use at these sites of nier focused on the construction of a split-level enclosed structureswith central hearths and other floorwith associated hearth, considering both ele potentially diagnostic architectural features has ments, ratherthan just thehearth, tobe critical, and been the subject of definitional debate, as the arguing thatthe Kotosh Religious Tradition was too Kotosh Religious Tradition proposed by Burger broadly defined to be useful (Bonnier 1987:358; and Salazar-Burger (1980) has been the subject of 1997:122).7 She identified several key definitional both revision (Bonnier 1997) and rejection elements: quadrangular chamber, central hearth, (Siveroni 2006). The excavation of the structure split-level floor, niches, and use of plaster (Bon described above, as a result, necessitates engage nier 1997:137; see Figure 4), arguing that, ment with two overlapping definitions of ritual Los templos que se edifican durante esta practice(s) and associated architecture thatdate to segunda mitad del tercermilenio [B.C.] theLate Preceramic, as early as 2500 B.C. These responden a canones formales bien definidos are theKotosh Religious Tradition proposed by que permiten identificaruna tradicionde arqui Burger and Salazar-Burger (Burger and Salazar tectura liturgica... Esta corriente arquitectonica Burger 1980, 1985, 1986) and theMito Tradition y religiosa, la llamamos 'TradicionMito' en suggested by Bonnier (1987) and Fung (1988) and referenciaal sitio de Kotosh, fase eponima de elaborated by Bonnier (Bonnier 1987,1988,1997, Mito, donde se descubrieron sus huellas por 2007; Bonnier and Rozenberg 1988). primera vez [Bonnier 1988:40].8 The Kotosh Religious Tradition was defined as a ritual practice focused on the burning of offer Using thisdefinition, Bonnier thenchronologically ings ina central hearth (Burger and Salazar-Burger delimited theMito Tradition as exclusive to the 1980), observed from theLate Preceramic (most Late Preceramic. notably atKotosh, Shillacoto, Huacaloma, La Gal The implications of the two definitions are dis gada, andHuaricoto) throughthe Early Horizon (at tinct.Where Burger and Salazar-Burger envision a Huaricoto). Such practice, Burger and Salazar cultural substratumof ritual traditionwhose mate Burger argued, was the reflection of religious rial form varies depending on sociopolitical cir beliefs thatwere fundamentalenough toprehistoric cumstance, Bonnier does notmake explicit claims Andean peoples that they crosscut differences in of theMito Tradition reflecting particularities of political, economic, and social organization, find sociopolitical structure.9Instead, focusing her def ing expression in communities with distinct forms initionmore narrowly, she prefers to use theMito of sociopolitical organization. That very diversity Tradition as a tool for examining cultural interac of expression, Burger and Salazar-Burger laterpro tion.The recurrence of the architectural forms of posed, could be read as a reflectionof the varying theMito Tradition, in otherwords, evidences not forms of sociopolitical organization of the com the scattered florescence of local variations on a munities in question (specifically Kotosh, Shilla common, generalized Andean religious heritage coto, and Huaricoto, in their 1986 study). but ratherdistinct instances of a specific and elab some same With reference to of the sites,Bon orate ritual practice whose presence is suggestive nier, following her work at Piruru, distinguished of cultural interactionbetween the sites inquestion. between theKotosh Religious Tradition and what The limitedchronological range defined by the rel she termed theMito Tradition, defined as a cere evant sites reinforced this coherence; Bonnier monial traditioninvolving a particular architectural describes theMito Tradition as encompassing the form intowhich the sacred hearth is incorporated period 2500-1800 B.C. (1997:122). In contrast,

This content downloaded from 129.252.86.83 on Thu, 13 Mar 2014 23:00:42 PM All use subject to JSTOR Terms and Conditions Contreras] A MITO-STYLE STRUCTURE AT CHAVIN DE HUANTAR 15

theKotosh Religious Tradition, appropriately for patterningof Mito Tradition structuresimpossible. its broader definition, spans the period from the However, ifanything, sampling has been more thor Late Preceramic into theEarly Horizon, approxi ough on thecoast, where sitesof theperiod aremore mately 2500-200 B.C. (Burger and Salazar-Burger accessible and less apt to be buried. The associa 1980). tionof theMito Tradition with thehighlands, thus, That theMito Tradition might be a particular should be seen as provisional but reasonably robust. instance of thegeneralized ritual traditionof burn in a centralhearth?that is, theKotosh ingofferings Implications Religious Tradition as defined by Burger and Salazar-Burger?is not specifically addressed but All of theelements thatBonnier defines as integral seems implicit in thedefinition. Moore has recently to theMito Tradition?with, arguably, the excep suggested as much, writing, "the concepts [of the tionof theniches?are present in theChavin exam Kotosh Religious Tradition and theMito Tradition] ple. Moreover, theWF-07 structure seems to be cover distinct but overlapping observations" relatedmost closely to those fromKotosh. Though (Moore 2005:108). As the structureexcavated at it lacks the niches so characteristic there and the Chavin fits the definition of theMito Tradition, I hearthdoes not featurea ventilation duct (however, employ thatterm here. However, itshould be under such a duct is not among thearchitectural features stood thatthis Chavin example also fallswithin the thatBonnier identifiesas key to theMito Tradition), more broadly defined Kotosh Religious Tradition. it is rectilinearrather than subrectangular,formally Structures identified as belonging to theMito constructed, and was ceremonially interred(Bon Tradition have previously been described at six nier 1987:Fig.3). As such, it seems definitively to sites in theCentral Andes (Bonnier 1997:122) (see belong to theMito Tradition, despite its anom Figure 11). Bonnier's criteriawere based primar alously latedate. Itspresence also reinforcesBurger ilyon analysis of her own work at the siteof Piruru, and Salazar-Burger's (1980) contention that the in theUpper Marafion, and the well-published Kotosh Religious Tradition persisted later than the examples fromKotosh; shemakes reference also Initial Period. Although the structure at Chavin to thepublished examples fromShillacoto, Waira bears littledirect resemblance to the late example jirca, La Galgada and Huaricoto (Bonnier 1987, they cite fromHuaricoto, the two are generally 1997). While Bonnier restrictsher discussion to comparable indate and both fit comfortablywithin highland sites, coastal examples of related archi theKotosh Religious Tradition's broad definition. tecturehave also been published, includingmost Two sortsof implications follow from this iden saliently those in theCasma Valley, at Caral in the tificationof aMito-style structureat Chavin: those Supe Valley, and at El Silencio in theSanta Valley. forChavin itself and those for theMito Tradition Pozorski and Pozorski describe a number of "ven ingeneral. Beginning with the latter,the Mito-style tilatedhearth structures"in theLate Preceramic and structureat Chavin should be treatedwith some Initial Period sites of theCasma Valley (atHuay caution; while the relatively late dates are con - case nuna, Bahia Seca, Pampa de las Llamas Moxeke, vincing, it is also the that theyneed not imply and Taukachi- Konkan (Pozorski and Pozorski direct continuitywith theMito Tradition. How 1996), and mention also a similar structureexca ever, given the adherence of theWF-07 structure vated byTellenbach atMontegrande in theJequete to the architectural canons of theMito Tradition, peque Valley (Tellenbach 1986).10While these and more specifically its resemblance to theMito appear tobe examples of theKotosh Religious Tra structuresfrom therelatively nearby siteof Kotosh, dition but not theMito Tradition, the two structures a tentative assertion of continuity is warranted. described at Caral (Shady and Machacuay 2003; While archaism or revival are certainly possible, means Shady et al. 2003) conformmuch more closely to these are simply alternativemodels of the the otherwise highland Mito Tradition model, as of transmissionof canons of ritualarchitecture, and do the structures at El Silencio (Montoya Vera do not obviate the need to explain the link. at 2007). Certainly the small sample of excavated The intermentof theknown Mito structures Late Preceramic/Initial Period ceremonial centers Kotosh by 1500 B.C. ifnot earliermakes the infer makes any definitive statementabout the regional ence of some other link?whether direct, allowing

This content downloaded from 129.252.86.83 on Thu, 13 Mar 2014 23:00:42 PM All use subject to JSTOR Terms and Conditions LATIN AMERICAN ANTIQUITY [Vol. 21, No. 1,2010

continuity,or indirect,allowing lateremulation? overlapping but not coterminous?definitions. If essential. The absence of explicit evidence of any theMito Tradition existswithin thebroader spatial such link?either at Kotosh, Chavfn, or and temporal aegis of theKotosh Religious Tradi elsewhere?in thecorpus of archaeological data is tion, this begs the question of what is behind the by no means evidence of its absence; as Kaulicke empirical distinction based on diagnostic ritual (among others) has noted, our picture of theCen architecture.I adopt the theoreticalposition thatthe tralAndean highlands during thisperiod of devel specific architectural forms associated with the oping sociopolitical complexity is Mito Tradition reflecta specific suite of associated underdetermined, reflecting a relatively sparse ritualpractices. As theKotosh Religious Tradition known record (Kaulicke 1998,1999). Nevertheless, is defined as a very generalized architectural form this is a substantial gap, to be either explained or (the central hearth) associated with a broad "reli filled with still-missing data. Given the inclusion gious ideology" (Burger and Salazar-Burger of theWF-07 structurein the corpus ofMito struc 1980:27), it is tobe expected that itmight be long tures from theCentral Andes, one of the key dic lasting and widespread. For themore narrowly tums about theMito Tradition needs reexamining. defined practices of theMito Tradition to be simi Bonnier's assertion, based on the data then avail larlypersistent and widely distributed ismore sur able, was that, "It seems that theMito tradition prising. The link of the specifically defined ritual ends at thebeginnings of the Initial Period" (Bon architecture associated with theMito Tradition to nier 1997:143). The new data fromChavfn, how a particular ritual practice thus ties theirdistribu ever,demonstrate itspersistence all theway through tion not simply to a common cultural background thebeginnings of theEarly Horizon and its coex but to specific interconnections.This is not to sug istencewith thenew ritual forms thendeveloping. gest thatthe Mito Tradition is coterminous spatially This reinforcesBurger and Salazar-Burger's obser and temporallywith some particular cultural entity, vations thatdespite itsprimary association with the but rather that the evidence of theMito Tradition Late Preceramic, theKotosh Religious Tradition? at various sites throughout theCentral Andes has a broader set towhich examples of theMito Tra specific implications for early interregional inter dition must by definition also belong?persisted action in the region. The cultural mechanisms into theEarly Horizon. They note, however, that, behind those apparent interactions certainlymerit "outside of Huaricoto, traces of this tradition furtherinvestigation. become more sketchy after the Late Preceramic One of the specific implications is the linkmen period" (Burger and Salazar-Burger 1980:30). tioned above between Chavin and Kotosh. Itbears While these laterexamples ofKotosh Religious re-emphasizing thatwe do not yet have any evi Tradition structures?i.e., those at Huaricoto?are dence thatmight determine whether this link is closer toChavfn spatially and temporally, theyare direct or indirect.However, this is not thefirst piece similar to the structuredescribed here only in the of evidence to link Chavin to Kotosh. Previous broadest terms (Burger and Salazar-Burger examples have focused on both material culture 1980:26). Many of thehearth structuresat Huari (Izumi 1971; Kano 1979; Lumbreras 1989) and coto belong to thebroader Kotosh Religious Tra broad cultural patterns (e.g., Lumbreras 1989:22, dition but lack the specific architecturalfeatures that 1993:353; Tello 1942:Lam.VII). More recently, would associate themwith theMito Tradition. Sim Kembel has argued that the earliest ritual architec ilarly, the potentialMito structureat Huacaloma turein Chavm's ceremonial corewas affiliatedwith (Terada and Onuki 1982:Color Plate 1) also appar the Kotosh Religious Tradition as manifest at ently dates to the late Initial Period, but does not Kotosh and La Galgada, concluding, "The archi feature the specific structuralfeatures thatBonnier tecture at Chavin does appear to incorporate this defines as integralto theMito Tradition. Shillacoto, [rectangular stone chambers] and other features of in theHuallaga Valley, is also dated to the second Kotosh Religious Tradition, beginning in its earli millennium B.C. (Izumi et al. 1972), if imprecisely. est phases" (Kembel 2001:227). She goes on to The cumulative effectof this array of examples describe the later elaboration of the ceremonial is to reinforce the concepts of theMito Tradition complex at Chavin, however, as a sequence "of and theKotosh Religious Tradition as nested? transitionas well as synthesis,beginning with local

This content downloaded from 129.252.86.83 on Thu, 13 Mar 2014 23:00:42 PM All use subject to JSTOR Terms and Conditions Contreras] A MITO-STYLE STRUCTURE AT CHAVIN DE HUANTAR 17

formsbased in theKotosh-Mito traditions,and tran Field suggest thepossibility of a concurrent conti sitioning to the incorporationof coastal forms that nuity between Late Preceramic Kotosh and Early reach theirmost formal in the final monumental Horizon Chavin, spanning approximately 1,000 stage,while still incorporatingaspects of the local years. Such long-term stability, interestingly forms" (Kembel 2001:230; see also Kembel 2008). enough, is precisely one of the characteristics If the architectureof the ceremonial core began as Burger ascribes to the Initial Period on the coast something akin to the architectureof theMito Tra (Burger 1992). The period is, however, sparsely dition, inother words, by the time thecomplex was documented in theCentral Highlands, leaving the fully elaborated ithad largelymoved on?though nature of thecontinuity between Kotosh and Chavin not, theMito structuredescribed here suggests, specifically, and between theMito Tradition and abandoned those roots. This ties Chavfn more ritual practice at Chavin generally, unclear. tightlyto highland antecedents,11rendering unten The presence of a functioningMito structurein able Burger and Salazar-Burger's suggestion (rea the early part of thefirst millennium B.C. also has sonable though itwas based on the data then implications for our understanding of ritual prac available) that,"Chavfn de Huantar had very little tice and its sociopolitical underpinnings atChavm. in common with the ceremonial centers of the This find cannot be understood in isolation from Kotosh Religious Tradition" (Burger and Salazar the rest of the site?indeed, the coexistence of the Burger 1980:32). Burger has also suggested that Mito Tradition with theceremonial practices of the contrast between Kotosh and Chavfn is an impor site core is perhaps themost significantfinding. tantfeature of what marks theEarly Horizon as dis This suggests not simplyMito Tradition roots for tinctfrom earlier Peruvian prehistory,writing that, Chavin (as argued by Kembel (2001:226-227)), "The site of Chavin de Huantar...represents a rup but also Chavm's adoption of the specific ritual turewith thehighland Kotosh Religious Tradition" architectureof theMito Tradition. Such a syncretic (Burger 1989:52; see also, however, Burger's note approach is consistentwith themodels ofChavm's thatHuaricoto, linked by itsceramic style toEarly function proposed by Lumbreras (e.g., 1989), Horizon Chavfn, was participating in the KRT Burger (e.g., 1992), and Rick (2005; see also Kem [Burger 1993:66]). bel and Rick 2004), though these diverge substan Instead, the evidence presented here suggests tially in other particulars. thatChavfn, both in itsorigins and its florescence, seems tohave maintained ties to or at least Kotosh, Conclusions to the traditionsfor which that site serves as con ceptual archetype. The presence of a Mito struc The excavations described here demonstrate the turecontemporary with themonumental core thus presence of aMito-style structureat Chavin, and substantially reinforcesLumbreras's argument for the persistence of theMito Tradition through locatingChavfn's origins inKotosh, encapsulated approximately 800 cal. B.C. at the site.As such, in his description of "una tradition [Kotosh-Mito] they argue for an interaction?whose character que no fue abandonada sinomas bien continuada remains to be defined?between Chavin and the por los constructores de Chavfn" (Lumbreras highland centers of theLate Preceramic generally, 1993:354; see also Lumbreras 1989:88-90).12 and perhaps theHuallaga Valley specifically.More Assessment of the role and number of such struc over, both the close association of thisMito-style tures at Chavfn, and evaluation of the degree to structurewith well-dated Janabarriu-styleceram which theChavfn example follows the pattern of ics and the correlation of the radiocarbon dates temple burial and reconstruction, await further from the structureitself with Kembel's architectural excavation. chronology suggest the close affiliationof this rit While itremains true that themonumental core ual space with theplazas and galleries of themon of Chavfn seems more linked to the coastal tradi umental core. tions of mound-and-plaza construction that date The radiocarbon dates (Figure 7) from thehearth back as faras theLate Preceramic than to such high itself and from the overlying deposits associated land centers as Kotosh (Burger 1989; Williams with Janabarriu-styleceramics (Figure 9) reinforce 1985, interalia), these new data from theWest the need to revisit Chavm's chronology, though

This content downloaded from 129.252.86.83 on Thu, 13 Mar 2014 23:00:42 PM All use subject to JSTOR Terms and Conditions LATIN AMERICAN ANTIQUITY [Vol. 21, No. 1,2010

theyare not sufficientin themselves to construct a tradition (see Rick 2005) and Burger's now tradi new temporal framework.While these data do not tionalmodel of Chavin as a center of ideological constitutean argument forKembel and Rick's pro synthesis (see Burger 1992:Ch.7), althoughBurger posed revision in particular (Kembel 2001:Table placed such activity several centuries later in time. 7.2; Rick 2005:73), they are compatible with it. It is also wholly compatible with Rick's sugges Most importantlythey do emphasize the need for tion (2005) that the canny (re-)use of extant, reexamination of the conflicting chronologies pre respected traditionswas central toChavm's inno sentedby Burger (1981,1984), Lumbreras (1989), vative sociopolitical maneuvering. The contrasting and Rick and Kembel (Kembel and Rick 2004), models apparently owe more to divergent theoret suggesting a significantly earlier appearance for ical explanations for themotivations of prehistoric Janabarriu-styleceramics than the 390 B.C. pro peoples than to disagreements about the specific posed by Burger. historical trajectory at Chavin. The resolution of The relatively late (with respect to theMito Tra disagreements over the timing of that trajectory dition) dates and the close association of theMito remains a necessary focus for future research, as structurewith Chavfn have broad implications for the dates presented here highlight. Central Andean prehistory.The longevity of the Theoretical disagreements over the underpin Mito Tradition, with the addition of thisevidence, nings of prehistoric behavior notwithstanding,cer is extended by a millennium, to 2500-800 cal tainlysuch persistence of theMito Tradition implies B.C.13 This chronological revision has theoretical broad cultural continuity in theCentral Andes. It is as well as typological consequences. The persis also testament to thepotential fornew data?even tence into theEarly Horizon of a traditionthat dates at a sitewith a long historyof fruitfulresearch?to back into theLate Preceramic reinforces the char provoke new questions and re-open old ones.More acterization of the Initial Period as ideologically over, italso suggests a specific threadof continuity (and, by extension, politically) stable.At the same from theLate Preceramic through theEarly Hori time, its longevity also demonstrates that theMito zon. That architectural thread also explicitly links Tradition was not replaced by themound-and ritualpractice in theCentral Andean sierra to appar sunken-plaza ritual tradition,but that the twowere ently similar practice on the coast, connecting the able to coexist. Burger and Salazar-Burger suggest Central Andes spatially as well as temporally. that this coexistence occurred at an inter-site level

(see Burger and Salazar-Burger 1980:32), and these Acknowledgments. All work here described was carried out under the of Dr. John Rick's Chavm de new data demonstrate such compatibility at an auspices Huantar Research and Conservation I owe him tremendous intra-sitelevel. Similar juxtaposition is also in evi Project. thanks for that, but he should not be held accountable for the dence at Caral, where themassive Templo Mayor conclusions drawn here. My fieldwork was funded primarily houses a smallMito structure other con (among by the National Science Foundation (Doctoral Dissertation on structions) its summit (Shady andMachacuay Improvement Grant #0532350), with additional support from a Stanford School of Humanities and Sciences 2003), perhaps suggesting thatpublic, visible elite Graduate Research Opportunity Grant, an Amherst College ritual activity took place within only a fewmeters Memorial Fellowship, and the Lynford Family Foundation. of private and hidden activity. Pozorski and Funding for follow-up work in 2006 was provided by the Pozorski also note the coexistence of divergent Explorers Club Exploration Fund and a Lewis and Clark forms of ritual architecture at the coastal site of Field Scholar Grant from the American Philosophical I was of Huaynuna (1990), as does Montoya Vera (2007) at Society. also, course, supported in the field by col El Silencio. leagues, students, and locals too numerous to name. Particularly deserving of mention for their help are Maria This compatibility suggests the possibility of Mendoza, Christian Mesia, Parker van Valkenburgh, and simultaneous coexistence of long-term stability Veronica Castro Neves; the excavation itself would have (representedby thepersistence of theMito Tradi been impossible without the expert work of JuanMunoz and tion) and innovation and dynamism in ritualprac Christian Cotrino. The Stanford Archaeological project was possible thanks to the approval and support of the national tice (represented by the development of the (Lima) and regional (Huaraz) branches of Peru's Instituto monumental core of Chavfn). Such coexistence Nacional de Cultura. For comments on this text I particularly both Rick's articulatedmodel supports recently of thank John Rick, Nicole Slovak, and Helaine Silverman, as as an as Chavfn instance of created and manipulated well Elisabeth Bonnier and three anonymous reviewers.

This content downloaded from 129.252.86.83 on Thu, 13 Mar 2014 23:00:42 PM All use subject to JSTOR Terms and Conditions Contreras] A MITO-STYLE STRUCTURE AT CHAVIN DE HUANTAR 19

References Cited Papers inMemory of Clifford Evans, edited by Ramiro Matos M, Solvieg A. Turpin, and Herbert H. Eling, Jr., Bennett, Wendell Clark pp. 65-82. Monograph of the Institute of Archaeology, 1944 The North Highlands of Peru: Excavations in the University of California, Los Angeles XXVII. Institute of Callejon de Huaylas and at Chavfn de Huantar. Anthro Archaeology, University of California, Los Angeles, Los pological papers of theAmerican Museum ofNatural His Angeles. tory,New York. Contreras, Daniel A. Bonnier, Elisabeth 2007 Sociopolitical and geomorphologic dynamism at 1987 Arquitectura Preceramica en la Cordillera de los Chavin de Huantar, Peru. Ph.D. Dissertation, Stanford Uni Andes, Piruru frente a la diversidad de los datos. Anthro versity. pologica VI(6):335-36l. Contreras, Daniel A., and David K. Keefer 1988 Acerca el surgimiento de la arquitectura en la sierra 2009 Implications of the Fluvial History of theWacheqsa andina. InArquitectura yArqueologia: pasado yfuturo de River forHydrologic Engineering andWater Use at Chavin la construccion en el Peru, edited by Victor Rangel Flo de Huantar, Peru. Geoarchaeology 24(5):589-618. res, pp. 35-^4-9.Universidad de Chiclayo, Chiclayo. Diessl, Wilhelm 1997 Preceramic Architecture in theAndes: The Mito Tra 2004 Huantar, San Marcos, Chavin: Sitios Arqueoldgicos dition. In Archaeologica Peruana 2, edited by Elizabeth en la Sierra de Ancash. Instituto Cultural Runa, Lima. Bonnier and Henning Bischof, pp. 120-144. Sociedad Fung Pineda, R. Arqueologica Peruano-Alemana, Mannheim. 1988 The Late Preceramic and Initial Period In Peruvian 2007 Arquitectura preceramica en los Andes: la tradicion Prehistory: An Overview of Pre-lnca and Inca Society, Mito. Coleccion Biblioteca Andina de Bolsillo n? 27. Insti edited by R. W. Keatinge, pp. 67-96. Cambridge Univer tute Frances de Estudio Andinos, Lima. sity Press, Cambridge. Bonnier, Elizabeth, and Catherine Rozenberg Grieder, Terence, and Alberto Bueno Mendoza 1988 Del santuario al caserfo: Acerca de la neolitizacion en 1981 La Galgada: Peru Before Pottery.Archaeology:44-51. la Cordillera de losAndes Centrales. Boletin del Instituto 1985 Ceremonial Architecture atLa Galgada InEarly Cer Frances de Estudios Andinos XVII(2): 23?10. emonial Architecture in theAndes: A Conference at Dum Bronk Ramsey, C. barton Oaks, 8th to 10th October 1982, edited by 2001 Development of the Radiocarbon Program OxCal,. Christopher Donnan, pp. 93-110. Dumbarton Oaks Radiocarbon 43(2):355-363. Research Library and Collection, Washington, DC 2006 OxCal v4beta4. Grieder, Terence, Alberto Bueno Mendoza, C. Earle Smith Jr. Bueno Mendoza, Alberto and Robert M. Malina 2004 Desarollo Arqueologico al Norte del Callejon de 1988 La Galgada, Peru: a Preceramic Culture in Transi Huaylas: La Galgada, Tumshukaiko y Pashash. InArque tion.University of Texas Press. ologia de la Sierra de Ancash, edited by Bebel Ibarra Haas, Jonathan, andWinifred Creamer Ascencios, pp. 51-82. Instituto Cultural Runa, Lima. 2006 Crucible ofAndean Civilization: The Peruvian Coast Bueno Mendoza, Alberto, and Terence Grieder from 3000 to 1800 BC. CurrentAnthropology 47:745-77'5. 1980 La Galgada: nuevo clave para la arqueologia andina. Indacochea G., Angel J., and Mariano Iberico M. Espacio 9(9):48-55. 1947 Alu vionamiento de Chavm de Huantar el 17 de Enero Burger, Richard L. de 1945. Boletin de la Sociedad Geologica del Peru 1981 The Radiocarbon Evidence for theTemporal Priority XX:21-28. of Chavm de Huantar. American Antiquity 46:592-602. Izumi, Seiichi 1984 The Prehistoric Occupation of Chavm de Huantar, 1971 The Development of the Formative Culture in the Peru. University of California Publications inAnthropol Ceja deMontana of theCentral Andes: A Viewpoint Based ogy. University of California Press, Berkeley. on theMaterials from theKotosh Site. InDumbarton Oaks 1989 An Overview of Peruvian Archaeology (1976-1986). Conference on Chavm, edited by E. Benson, pp. 49-72. Annual Review ofAnthropology 18:37-69. Dumbarton Oaks Research Library and Collection, Wash 1992 Chavm and theorigins ofAndean civilization. Thames ington,DC. and Hudson, Ltd., London. Izumi, Seiichi, Pedro J.Cuculiza, and Chiaki Kano 1993 The Chavm Horizon: Stylistic Chimera or Socioeco 1972 Excavations at Shillacoto, Huanuco, Peru. Bulletin nomic Metamorphosis? In Latin American Horizons, No. 3. The University Museum, The University of Tokyo, edited by D. S. Rice, pp. 41-82. Dumbarton Oaks, Wash Tokyo. ington,D.C. Izumi, Seiichi, and Toshihiko Sono 1998 Excavaciones en Chavin de Huantar. Pontificia Uni 1963 Andes 2: Excavations at Kotosh, Peru. Kadokawa versidad Catolica del Peru, Lima. Publishing Co., Tokyo. 2008 Chavm de Huantar and Its Sphere of Influence. In Izumi, Seiichi, and Kazuo Terada on Handbook of South American Archaeology, edited by 1972 Excavations at Kotosh, Peru: A Report the Third Helaine Silverman and William H. Isbell, pp. 681-703. and Fourth Expeditions. University of Tokyo Press, Tokyo. Springer, New York. Kano, Chiaki in Pre Burger, Richard L., and Lucy Salazar-Burger 1979 The Origins of Chavin Culture. Studies 1980 Ritual and Religion at Huaricoto. Archaeology Nov Columbian Art and Archaeology 22. Dumbarton Oaks, Dec 1980:26-32. Washington, D.C. 1985 The Early Ceremonial Center of Huaricoto. In Early Kaulicke, Peter Ceremonial Architecture in theAndes, edited by Christo 1998 Reflexiones finales: problemas y perspectivas. Boletin pher Donnan, pp. 111-138. Dumbarton Oaks Research de Arqueologia PUCP 2:353-368. en el Library and Collection, Washington, D.C. 1999 Los Estudios Del Period Arcaico Peru: Logros, 1986 Early Organizational Diversity in thePeruvian High Problemas y Propuestas Boletin de Arqueologia PUCP lands: Huaricoto and Kotosh. In Andean Archaeology: No. 3.

This content downloaded from 129.252.86.83 on Thu, 13 Mar 2014 23:00:42 PM All use subject to JSTOR Terms and Conditions LATIN AMERICAN ANTIQUITY [Vol. 21, No. 1,2010

Kembel, Silvia Rodriguez tecture and Culture, edited byWilliam J.Conklin and Jef 2001 Architectural Sequence and Chronology at Chavm de freyQuilter, pp. 3-34. Cotsen Institute,Los Angeles. Huantar, Peru. Ph.D. Dissertation, Stanford University. Rick, JohnW, Silvia Rodriguez Kembel, Rosa Mendoza Rick, 2008 The Architecture at theMonumental Center of Chavm and JohnA. Kembel de Huantar: Sequence, Transformations, and Chronology. 1998 La arquitectura del complejo ceremonial de Chavm In Chavm: Art, Architecture, and Culture, edited by de Huantar: documentation tridimensional y sus impli William J.Conklin and JeffreyQuilter, pp. 35-84. Cotsen cancias. In Perspectivas Regionales del Periodo Forma Institute,Washington, D.C. tive) en el Peru, edited by Peter Kaulicke, pp. 181-214. Kembel, Silvia Rodriguez and JohnW. Rick Boletin de Arqueologia PUCP, Vol. 2. Pontificia Univer 2004 Building Authority at Chavm de Huantar: Models of sidad Catolica del Peru, Fondo Editorial, Lima. Social Organization and Development in the Initial Period Rowe, JohnH. and Early Horizon. In Andean Archaeology, edited by 1967 Stages and Periods inArchaeological Interpretation. Helaine Silverman, pp. 51-76. Blackwell Publishing, In Peruvian Archaeology: Selected Readings, edited by Maiden, MA. John H. Rowe and Dorothy Menzel. Peek Publications, Lumbreras, Luis G., Chacho Gonzalez and Bernardo Lieataer. Palo Alto. 1976 Acerca de lafuncion del sistema hidraulico de Chavm. Shady, Ruth, and Carlos Leyva Museo de Arqueologia y Antropologia, Lima. 2003 La Ciudad Sagrada de Caral-Supe: Los origenes de Lumbreras, Luis Guillermo la civilizacion andina y laformacion del Estado pristino 1977 Excavaciones en el Templo Antiguo de Chavm (Sec en el antiguo Peru. InstitutoNational de Cultura, Lima. tor R); Informe de la Sexta Campana. Nawpa Pacha Shady, Ruth, and Marco Machacuay 15:1-38. 2003 El Altar del Fuego Sagrado del Templo Mayor de la 1989 Chavm de Huantar en elNacimiento de la Civilizacion Ciudad Sagrada de Caral-Supe. In La Ciudad Sagrada de Andina. InstitutoAndino de Estudios Arqueologicos, Lima. Caral-Supe: Los origenes de la civilizacion andina y la 1993 Chavm de Huantar: Excavaciones en la Galena de formacion del Estado pristino en el antiguo Peru, edited las Ofrendas. Verlag Philipp von Zabern, Mainz. by Ruth Shady and Carlos Leyva, pp. 169-185. Instituto 2007 Chavm: Excavaciones Arqueologicas. 2 vols. Uni National de Cultura, Lima. versidad Alas Peruanas, Lima. Shady, Ruth, Marco Machacuay, and Sonia Lopez McCormac, EG., A.G. Hogg, P.G. Blackwell, C.E. Buck, T.F.G. 2003 Recuperando la historia del Altar del Fuego Sagrado. Higham, and P.J. Reimer In La Ciudad Sagrada de Caral-Supe: Los origenes de la 2004 SHCal04 Southern Hemisphere calibration, 0-11.0 cal civilizacion andina y laformacion del Estado pristino en kyr BP. Radiocarbon 46(3): 1087-192. el antiguo Peru, edited by Ruth Shady and Carlos Leyva. Montoya Vera, Maria InstitutoNational de Cultura, Lima. 2007 Arquitectura de la "Tradition Mito" en el valle medio Shady Solfs, Ruth, Jonathan Haas, andWinifred Creamer del Santa: sitio "El Silencio". Bulletin de lTnstitutFrancais 2001 Dating Caral, a Preceramic Site in the Supe Valley on d'Etudes Andines 36(2): 199-220. theCentral Coast of Peru. Science 292:723-726. Moore, JerryD. Silverman, Helaine 2005 Cultural Landscapes in theAncient Andes. Univer 2004 Introduction: Space and Time in the Central Andes. sity Press of Florida, Gainsville. InAndean Archaeology, edited by Helaine Silverman, pp. Moseley, Michael E. 1-15. Blackwell Publishing, Maiden, MA. 1975 The Maritime Foundations of Andean Civilization. Siveroni, Viviana Cummings Publishing Company, Menlo Park, CA. 2006 Mi casa es tu templo: Una vision alternativa de la Onuki, Yoshio arquitectura de la Tradicion Kotosh. Arqueologia y 1993 Las actividades ceremoniales tempranas en laCuenca SociedadXl': 121-148. del Alto Huallaga y algunos problemas generales. In El Tellenbach, Michael mundo ceremonial andino, edited by Luis Millones and 1986 Las excavaciones en el asentamiento formativo de Yoshio Onuki, pp. 69-96. Senri Ethnological Studies, vol. Montegrande, Valle de Jequetepeque en el norte del Peru. 37. National Museum of Ethnology, Osaka. Materialien zurAllgemeinen und Vergleichenden Archae 1999 El Periodo Arcaico en Huanuco y el Concepto del ologie. Verlag C.H. Beck, Munich. Arcaico. Boletin de Arqueologia PUCP 3:325?333. Tello, Julio C. Pozorski, Thomas, and Shelia Pozorski 1942 Origen y desarrollo de las civilizaciones prehistori 1990 Huaynuna, a Late Cotton Preceramic Site on theNorth cas andinas. InActas y trabajos cientificos del XXVII Con Coast of Peru. Journal ofField Archaeology 17(1): 17-26. greso Internacional de Americanistas, Lima 1939, pp. 1996 Ventilated Hearth Structures in the Casma Valley, 589-720. vol. Tomo I, Lima. Peru. Latin American Antiquity 7(4):341-353. 1943 Discovery of theChavm Culture in Peru. American Reinhard, Johan Antiquity 9:135-160. 1985 Chavm and Tiahuanaco: A New Look at Two Andean 1960 Chavin: Cultura Matriz de la Civilizacion Andina. Ceremonial Centers. National Geographic Research Imprenta de laUniversidad de San Marcos, Lima. Review l(3):395-422. Terada, Kazuo, and Yoshio Onuki Rick, JohnW. 1982 Excavations at Huacaloma in theCajamarca Valley, 2005 The Evolution of Authority and Power at Chavm de Peru, 1979. University of Tokyo Press, Tokyo. Huantar, Peru. InFoundations ofPower in thePrehispanic Turner, Robert J.W.,Rosemary J.Knight, and JohnRick Andes, edited by Kevin J.Vaughn, Dennis Ogburn, and 1999 Geological landscape of the pre-Inca archaeological Christina A. Conlee. Archeological Papers of theAmeri site at Chavm de Huantar, Peru. Current Research 1999 can Anthropological Association Number 14. Arlington, D; Geological Survey of CanadaAl-56. Virginia. Williams, Carlos 2008 Context, Construction and Ritual in theDevelopment 1985 A Scheme for theEarly Monumental Architecture of ofAuthority at Chavm de Huantar. In Chavin: Art, Archi theCentral Coast of Peru. In Early Ceremonial Architec

This content downloaded from 129.252.86.83 on Thu, 13 Mar 2014 23:00:42 PM All use subject to JSTOR Terms and Conditions Contreras] A MITO-STYLE STRUCTURE AT CHAVIN DE HUANTAR 21

ture in theAndes: A Conference at Dumbarton Oaks, 26 8. "The temples thatwere built during this second half of and 27 October edited 1982, by Christopher Donnan, pp. the thirdmillennium B.C. respond to formal, well-defined 227-240. Dumbarton Oaks Research and Collec Library canons that permit the identification of a tradition of ritual tion,Washington, D.C. architecture... We call this architectural and religious current the 'Mito Tradition' in reference to the site of Kotosh and the Mito where its traces were discovered for Notes eponymous phase, the first time" [translation by author]. 9. Lumbreras has however, that this ritual 1. Bonnier mentions that structures ER-28 and "proba argued, early architecture in theAndean sierra itselfwas of a that bly" ER-20 at Kotosh were "framed with a clay molding" quality reflects the existence of and centralized control of (Bonnier 1997:138), which may represent an analogous hierarchy labor Onuki, with reference architectural element (see also Izumi and Terada 1972:160). (Lumbreras 1989:89); specifi = toKotosh, 2. 2644 ? 45 (AA69446; wood charcoal; 813C -23.2%0) cally agrees (Onuki 1999:328). = 10. Pozorski and Pozorski do not consider the and 2712 ? 42 (AA69447; wood charcoal; 813C -23.6%c). explicitly Mito/Kotosh Tradition distinction, but group the 3. Aluvion is theAndean term for a catastrophic debris Religious structures describe in the Kotosh Tradition. flow. In 1945 the site of Chavm de Huantar, and part of the they Religious This choice is as these of circular ven adjacent town, was buried by such an event, the result of a appropriate, examples tilated hearths in square or circular rooms seem to be more landslide into a glacial lake high up in theWacheqsa River related to theKotosh Tradition than to the drainage. The resulting debris flow?a slurry of rock, earth, Religious generally Mito Tradition their on the hearth is ice, and water estimated at 900,000 m3?descended the specifically; emphasis not the elaboration of the construction hous canyon of theWacheqsa at an estimated 30+ km/hr, and accompanied by the hearth thatBonnier as to buried Chavm under up to 4 m of sediment (estimates from ing clearly distinguishes integral theMito Tradition. Indacochea and Iberico 1947). The thickness of the deposit 11. Though theMito-style structures at Caral suggest per varies widely throughout the site, with the amount of a more Mito Tradition. deposited material and the size of entrained debris dependent haps geographically widespread 12. "a tradition [Kotosh-Mito] that was not abandoned, on the energy of the flow in any given area. = but rather continued, Chavm's builders" [translation 4. 2506 ? 43 (AA69448; wood charcoal; 813C -23.9%o) by by = author]. and 2567 ? 42 (AA69449; wood charcoal; 813C -22.\%o). 13. As theMito-style structure at Chavm may also be 5. Any discussion of the excavation context must also take described as to the broader of theKotosh into account the fact that our work did not reach sterile soil? belonging category Tradition, its existence also reinforces and as attested by the architecture visible in the cut of the Religious Burger Salazar-Burger's contention?based on the later hearths at Wacheqsa River (see Figure 6). Future work may be able to Huaricoto (1985:118)?that the Kotosh Religious Tradition expand the stratigraphic exposure by working outside the persisted into the Early Horizon. boundaries of the chamber proper, thus avoiding the ethical imperative of not destroying the intactMito chamber archi tecture. However, any future excavation will be tightly con strained by the steep slope to the south and the river cut to the north.

6. Ranging conceptually further afield, it also prompts questions (as yet unanswerable) about the possibility of het erarchical rather than strictly hierarchical organization. 7. In addition, Siveroni (2006) has recently criticized the Submitted: October 2006; Accepted: August 6, 2008; very identification of structures with central hearths as ritual Revised: 2008. architecture. September 15,

This content downloaded from 129.252.86.83 on Thu, 13 Mar 2014 23:00:42 PM All use subject to JSTOR Terms and Conditions