Andean Past

Volume 10 Article 14

4-15-2012

Research Reports Andean Past 10Jason L. TooheyUniversity of Wyoming, jtoohey@uwyo.edu

Monica BarnesAmerican Museum of Natural History, monica@andeanpast.org

Catherine Gaithercat.gait@gmail.com

Robert A. Benfer Jr.University of Missouri, benferr@missouri.edu

Daniel E. Sheadeceased

See next page for additional authors

Follow this and additional works at: https://digitalcommons.library.umaine.edu/andean_past

This Research Reports is brought to you for free and open access by DigitalCommons@UMaine. It has been accepted for inclusion in Andean Past byan authorized administrator of DigitalCommons@UMaine. For more information, please contact um.library.technical.services@maine.edu.

Recommended CitationToohey, Jason L.; Barnes, Monica; Gaither, Catherine; Benfer, Robert A. Jr.; Shea, Daniel E.; Matsumoto, Yuichi; Nesbitt, Jason S.;Palacios J., Denesy; Chicoine, David; Rojas, Carol; Uribe, Mauricio; Urbina A., Simon; Adan, Leonor; Vidal M., Estefania; and Salazar,Julian (2012) "Research Reports Andean Past 10," Andean Past: Vol. 10 , Article 14.Available at: https://digitalcommons.library.umaine.edu/andean_past/vol10/iss1/14

Research Reports Andean Past 10

AuthorsJason L. Toohey, Monica Barnes, Catherine Gaither, Robert A. Benfer Jr., Daniel E. Shea, Yuichi Matsumoto,Jason S. Nesbitt, Denesy Palacios J., David Chicoine, Carol Rojas, Mauricio Uribe, Simon Urbina A., LeonorAdan, Estefania Vidal M., and Julian Salazar

This research reports is available in Andean Past: https://digitalcommons.library.umaine.edu/andean_past/vol10/iss1/14

RESEARCH REPORTS ON ANDEAN ARCHAEOLOGY

PERU

Recent Excavations at the Late IntermediatePeriod Village of Yanaorco in the CajamarcaHighlands

Jason L. Toohey (University of Wyoming,e-mail: jtoohey2@uwyo.edu) excavated at thesite of Yanaorco in the Cajamarca region ofnorthern Peru in 2003 and 2004 (InstitutoNacional de Cultura permit number C/138-2003/DREPH) with the support of a FulbrightHays Dissertation Research Abroad Grant.

Radiocarbon dates and ceramic chronologyplace the occupation of Yanaorco within theLate Intermediate Period (LIP; A.D. 1000-A.D.1465; Toohey 2009) when it was a major popu-lation center (Sachun 1986). It stands at ap-proximately 3550 m.a.s.l. and its occupantspracticed agro-pastoralism, taking advantage ofboth the upper Quechua ecological zone below3500 meters and the Suni zone herding re-sources to its southeast (above approximately3500 m.a.s.l.; Pulgar Vidal 1981).

Architectural and excavation data indicatethat Yanaorco was substantially fortified andarchitecturally complex. Its population wassocially and politically stratified. My resultssupport a view of the Cajamarca LIP as a dy-namic time when populations shifted residencefrom lower elevations to high elevation commu-nities that were often fortified. Several archaeol-ogists have characterized Yanaorco as an impor-tant Cajamarca population center during theLIP (Julien 1988, 1993; Ravines 1968, n.d.;Sachun 1986; Urteaga 1960). Surface survey ofthe site had also placed the occupation in theLIP (Martell 2002). Research at Yanaorco is thefirst community and household level investiga-tion of an LIP fortified village in the Cajamarcaregion.

Yanaorco

Yanaorco is on the continental divide nearthe headwaters of the Jequetepeque River. Itsinhabitants were part of the Cajamarca culture.The town is on a defensible ridge, with naturaldrops of approximately 500 meters on its north,west, and south sides. To further improve thenatural defenses of the site, its inhabitantsconstructed two sets of fortifications on itsgently sloping eastern edge. Each set includesthree large walls. The largest retaining wallsstand up to three or four meters high and aretwo meters thick (Figure 1). Each wall has a dryditch on its exterior side. Many of these wallsalso have abutting parapets along their interiorfaces. These works indicate that the locationoccupied by Yanaorco was important and con-tested.

Figure 1: Fortification walls preserved to a height of 3 to4 meters. People are, from viewer’s left to right, JimmyBouroncle, Wilder León Ascurre, and Mariano Esquivel.

ANDEAN PAST 10 (2012): 263-302.

ANDEAN PAST 10 (2012) - 264

In addition to occupying a defensive loca-tion, Yanaorco is situated strategically to over-see traffic along the primary route in bothprehistoric and modern times between the coastand the highland Cajamarca Basin. Yanaorcooverlooks the Gavilán Pass into the basin by 500meters. Survey uncovered the remains of aprehispanic trail connecting the pass to one ofthe village’s two entrances, physically tying thecommunity to the pass. Leadership within thecommunity would thus have been in a positionto closely monitor, and possibly control, trafficthrough the pass.

The architectural organization of Yanaorcois complex and indicates that the original com-munity was rapidly augmented at some pointduring its occupation. The site contains twoprincipal architectural zones. The older sector ison the western end of the ridge upon which thesite sits. This sector was ringed with three largefortification walls. At a later date, the commu-nity grew with the construction of a newersector of adjoined rectangular rooms to thesoutheast, outside the original fortifications.Ultimately this new set of rooms was fortifiedwith three additional walls and ditches, thelongest wall connecting to one of the originalwalls and effectively completing the fortificationof the entire community.

Within these fortifications are a series ofdistinct architectural zones integrated andconnected by ramps and staircases. At least twoareas of domestic terraces exist, broken gener-ally into well-constructed elite terraces alongthe south slope of the village, and less welldefined non-elite terraces on the northern slope.Three low platforms stand along the ridge line,dividing these domestic spaces. Although thereare rooms on the summit of the largest of theseplatforms, the presence of multiple lootedmasonry tombs on all of the platforms impliesboth mortuary and domestic functions, possiblycreating a visible symbol of collective commu-

nity identity for residents of Yanaorco. Non-domestic, public areas of architecture are alsopresent, including a sector of possibly ceremo-nial spatial features. These include a large, U-shaped plaza (Architectural Unit 18A) closelyassociated with a set of two small rooms charac-terized by extremely limited and controlledaccess (Figure 2). Excavation in one of theserooms indicated the presence of a low benchalong one wall and multiple niches recessed intothe walls. Excavation in the open plaza revealeda level activity surface and several small servingvessels. In the future, excavation will include anarea of four possible storage rooms, spatiallyassociated with a zone of elite architecturalspaces. All of these integrated architecturalspaces indicate that the community may havebeen partially planned when first constructedand that it did not grow haphazardly over time.

Figure 2: Plan of Architectural Unit 18.

265 - Research Reports

Figure 3: LIP Cajamarca Semicursive and Cajamarca

Black and Orange sherds.

Figure 4: Late Horizon Amoshulca Black Geometric sherds.

Ceramic Assemblages at Yanaorco

I analyzed approximately 34,300 ceramicsherds from both surface and excavated con-texts. These include a wide range of vessel formsand 18 distinct stylistic types, both decoratedfine wares and domestic, utilitarian wares(Toohey in press). Types diagnostic of the LIPinclude Cajamarca Semicursive (CSC) andCajamarca Black and Orange (CB&O; Figure3). The Late Horizon, Amoshulca Black Geo-metric (ABG) style is also present in high fre-quencies (Figure 4).

Although the entire surface of the commu-nity was surveyed and all diagnostic ceramicsanalyzed, no direct evidence for ceramic produc-tion was uncovered. Indirect evidence includedseveral small smoothing stones. It is likely thatmany of the vessels utilized in the communitywere made there, and future investigations willcontinue to seek direct evidence of production.There is little evidence for exotic, fine wareceramics being imported into the communityfrom other regions. However, a single exoticblack ware rim sherd (Figure 5) was recoveredfrom an excavated context, suggesting somelevel of coastal-highland interaction (Donnan1990; Sandweiss 1995:161). This sherd wasdecorated with a press-molded design, a methodthat was not used in local Cajamarca ceramicproduction.

Figure 5: Exotic blackware ceramic sherd with press-molded design.

Unusual Artifacts at Yanaorco

Excavation and surface collection at Yana-orco also revealed a number of less commonartifact types. Excavation within elite domesticroom 22A (not shown in Figure 2, see Toohey 2009:276) uncovered four copper artifactsincluding copper tweezers and a tupu (stick-pin)that had been rolled or folded in on itself beforeentering the archaeological record. Inspectionrevealed that it was still in working conditionwhen it was bent, indicating that it was purpose-fully rendered unusable before being deposited(Figure 6). This pin was originally 4.1 centime-ters long. Finally, several small quartz crystals

ANDEAN PAST 10 (2012) - 266

were recovered from the same room. Interviewswith local landowners revealed the presence oftwo ceramic whistles that had been collectedfrom an area of the site now under cultivation.These whistles contained modeled human faces.

Figure 6: Folded tupu.

Social and Economic Stratification at Yanaorco

Architectural survey and variability inexcavated materials indicate the presence of atleast two social strata living at Yanaorco duringthe LIP. There is clear evidence for two distinctdomestic zones at the site, marked by differencesin room size, organizational complexity, and wallconstruction quality. Elite domestic spacesinclude both interior, roofed rooms and un-roofed patios. Terraces are connected by well-defined staircases, and walls are made of facedand cut stone. Non-elite domestic spaces areless well preserved due to the lower quality oftheir walls and consequent poorer structuralpreservation.

While all households at Yanaorco probablyengaged in agro-pastoral subsistence activities,including camelid herding, guinea pig raising,and corn and potato production, spatial analysisindicates variation in foodways and depositionpatterns between elite and non-elite contexts(Toohey 2009). Elite households like thoserepresented at room 22A seem to have had

access to camelid resources more frequently, andin larger amounts. Excavated faunal remainsfrom this room include large limb bones with cutmarks. Many fragments also exhibit evidence ofburning or roasting. On the other hand, verysmall, friable bone fragments dominate thefaunal evidence from non-elite spaces. Theseremains may indicate the practice of boneboiling in order to extract the maximum amountof nutrition from each available bone, implyinga very conservative cuisine practiced by non-elite households. This pattern of conservativefoodways in non-elite spaces versus the capacityin elite spaces to discard large, partially roastedbones is evidence for patterned variation infaunal use within the community.

Conclusions

The general settlement shift in LIP Caja-marca from lower elevations into often fortifiedand defensible higher elevation villages is impor-tant. This movement in population has twoprimary implications for life in the Cajamarcaregion during the centuries after the collapse ofWari influence. The shift to higher elevations,in the case of Yanaorco to an elevation of 3550m.a.s.l., placed many communities at the bound-ary between the Quechua and Suni ecozones.This shift may indicate an increased focus onpastoralism during the LIP. At the same time,the settlement shift to more defensible loca-tions, in many cases into substantially fortifiedcommunities, suggests the threat of armedconflict between communities during the LIP.

Excavation at Yanaorco indicates a dynamiccommunity characterized by social and eco-nomic differentiation. Interaction betweenCajamarca villages may have involved bothexchange of fine ceramics and other resources,and armed conflict during the LIP. This investi-gation seems to show that interaction wasprimarily local and that little material exchange

267 - Research Reports

occurred between highland communities likeYanaorco and coastal polities.

There are plans for future research at thesite. These will include excavation of possiblestorage rooms associated with elite spaces.Because very little is known about mortuaryactivity, treatment, and practice in the Caja-marca region during the LIP, future research willalso focus on unlooted tombs and will continueto investigate the nature of community organi-zation, foodways, and militarism in fortifiedCajamarca communities.

REFERENCES CITED

Donnan, Christopher B.1990 An Assessment of the Validity of the Naymlap

Dynasty. In The Northern Dynasties: Kingship andStatecraft in Chimor, edited by Michael E.Moseley and Alana Cordy-Collins, pp. 243-274.Washington, D.C.: Dumbarton Oaks ResearchLibrary and Collection.

Julien, Daniel G.1988 Ancient Cuismancu: Settlement and Cultural

Dynamics in the Cajamarca Region of the NorthHighlands of Peru, 200 B.C.-A.D. 1532. Ph.D.Dissertation, University of Texas.

1993 Late Pre-Inkaic Ethnic Groups in Highland Peru:An Archaeological-Ethnohistorical Model of thePolitical Geography of the Cajamarca Region. Latin American Antiquity 4(3):246-273.

Martell Castillo, Nelly E.2002 Yanaorco: Un asentamiento Cajamarca; un caso de

control administrativo en el Valle de Cajamarca.Master’s thesis (licenciatura) Universidad Na-cional de Trujillo, Perú.

Pulgar Vidal, Javier1981 Geografía del Perú: Las ocho regiones naturales del

Perú. Eighth edition. Lima: Editorial UniversoS.A.

Ravines, Rogger H.n.d. Cajamarca hacia 1532. One-page manuscript in

the library of the Instituto Nacional de Cultura(now the Ministerio de Cultura), Cajamarca,Perú.

1968 Arqueología del Valle de Cajamarca. Cajamarca,Perú: Universidad Nacional Técnica de Caja-marca.

Sachun, Jorge1986 Patrones de asentamiento en el proceso cultural

prehispánico del valle de Cajamarca (primera

aproximación). Materiales para la Arqueología deCajamarca 3. Trujillo, Perú: Editorial Sud-américa.

Sandweiss, Daniel H.1995 Life in Ancient Túcume. In Pyramids of Tucume:

The Quest for Peru’s Forgotten City, edited byThor Heyerdahl, Daniel H. Sandweiss, andAlfredo Narvaez, pp. 142-168. New York:Thames and Hudson.

Toohey, Jason L.2009 Community Organization, Militarism, and Ethno-

genesis in the Late Prehistoric Northern Highlandsof Peru. Ph.D. Dissertation, University of Califor-nia Santa Barbara.

in press Ceramic Assemblages and Spatial Patterning atYanaorco: A Late Intermediate Period Village inthe Cajamarca Highlands of Northern Peru.Ñawpa Pacha 32:1.

Urteaga Portocarrero, N. P. S.1960 Bosquejo parcial sucinto sobre la cultura

caxamarca-marañón. In Actas y trabajos del IICongreso Nacional de Historia del Perú (época pre-hispánica), pp. 107-113. Volume 1. Lima: Centrode Estudios Históricos-Militares del Perú.

A Colonial Human Burial Excavated in 1965between Portals 5 and 6 at Huánuco Pampa

Monica Barnes (Andean Past and AmericanMuseum of Natural History, e-mail: monica@andeanpast.org), Catherine Gaither (Metro-politan College of Denver, e-mail: gaither@mscd.edu), Robert A. Benfer, Jr. (Universityof Missouri, e-mail: benferr@missouri.edu), andDaniel Shea (sheade@beloit.edu) report on ahuman skeleton excavated under the overalldirection of John V. Murra at the Inca site ofHuánuco Pampa in Huánuco Region.

From 20 July until 23 November 1965,funded by Peru’s Paternato Nacional deArqueología, the government entity then re-sponsible for overseeing archaeological work,Murra undertook the “cleaning” (limpieza) andwall consolidation of the monumental portion ofHuánuco Pampa (Morris et al. 2011:9-10; Murraand Hadden 1966). This was, in effect, a sub-project of Murra’s larger, National ScienceFoundation funded research, “A Study of Pro-

ANDEAN PAST 10 (2012) - 268

vincal Inca Life” (1963-1966). Under Murra’ssupervision was a small team of graduate stu-dents, young professional archaeologists, andPeace Corps volunteers including Mahlon [Lon]Barash, the late Luis Barreda Murillo, the lateGordon Hadden, the late Peter S. Jenson, thelate E. Craig Morris, Daniel Shea, and JamesStanton, the latter a civil engineer. Manuallabor was performed by a team of approximatelythirty local farmers. During the course of thiswork a human skeleton was uncovered. This isone of the few human remains ever found atHuánuco Pampa (c.f. Morris et al. 2011:31). It1

was excavated on 4-5 September, 1965 by Sheaand Barreda assisted by Barash and witnessed2

by Morris and Hadden.

The information available on this burialconsists of a series of eight black and whitemedium format photographs by Barash, fifteenblack and white 35 millimeter photographs byBarreda, and one medium format black andwhite photo by Hadden taken during the exca-vation of the human remains, and seven photo-graphs of the skull by Barreda, in addition toShea’s field notes. These are on file as part ofthe John Victor Murra Archive in the JuniusBird Laboratory of South American Archaeol-ogy, Division of Anthropology, American Mu-seum of Natural History in the City of NewYork. The photos are reproduced courtesy of theAMNH. The roll and photo numbers in thecaptions of this report refer to Murra’s originalphoto inventory. In addition, color photographs

of the excavation remain in the possession ofShea. The bones and associated artifacts havebeen destroyed.3

Context

To make the site of Huánuco Pampa morecomprehensible, in his view, to visitors, JohnMurra removed large amounts of what he con-sidered to be overburden from the monumentalor central portions of Huánuco Pampa, includ-ing the ushnu, the buildings incorporating theseries of aligned portals in the eastern portion ofthe site (the palace), the North and South ka-llankas bordering the eastern edge of the site,the bath, and the unfinished temple (templete).Although this was, in effect, an excavation,there is no indication found so far that the soilsremoved were screened for artifacts, that sec-tions were created or drawn, or that environ-mental samples were taken. In addition, JohnMurra’s team cleared vegetation from much ofthe fine masonry at Huánuco Pampa, resetstones that had fallen from walls, and removedboulders thought to interfere with foot traffic(Barnes 2010).

In the course of such work, Murra’s teamencountered the grave of a young person. Theskeleton lay within the building of fine Cusco-Inca style ashlar masonry that encompasses thegateways known as Portals 5 and 6, the eastern-most of the sets of aligned doorways leading tothe Huánuco Pampa ushnu plaza (Figure 1; seeMorris and Thompson 1985: figures 11, 13 forthe location of this building). The individualwas in a simple grave cutting a canal or drainthat ran at or below the floor level of the build-ing. The channel was lined with flat stonesrunning along its floor and sides, and capping itstop. The channel parallels the interior of the

In May, 1964 John Murra and Peace Corps volunteer1

Peter S. Jenson encountered a looted stone burialchamber with a circular plan still containing human longbones in the Quellaycancha portion of Huánuco Pampa(Murra Archive, AMNH, Roll 15, photos 7, 8), anothertomb (Roll 15, photo 9) and an apparently unlooted cairntomb nearby (Roll 15, photo 10).

Field notes by Daniel Shea, John V. Murra Archive,2

Junius Bird Laboratory of South American Archaeology,Anthropology Division, American Museum of NaturalHistory, p. 7.

Letter from Craig Morris to Terence N. D’Altroy,3

November 17, 1976, Terence N. D’Altroy file, Junius BirdLaboratory, American Museum of Natural History.

269 - Research Reports

south wall of the building, then makes a right-angle corner near Portal 5. From the informa-tion preserved and discovered so far, it is notpossible to follow the entire course of this fea-ture.

However, given its position at HuánucoPampa, where it intersects a probable proces-sional line from the eastern quadrant of the siteto the ushnu, Huánuco Pampa’s sacred center,as well as the importance of liquid offerings toushnus (Meddens et al. 2010:186), it is temptingto suppose that this canal has some connectionto the site’s ushnu. A similar arrangement can beobserved at Inca Pachacamac on the centralPeruvian coast (Barnes, vidi, September 2011).

Because the creation of the grave wouldhave interfered with the functioning of thecanal, the grave must postdate the canal’s use.An iron pin was found near the skeleton’s neck(Shea, field notes, p. 8). This was not illustratedby Shea and does not appear in the photo-graphs. Whether the pin was a shroud pin, or atupu, an indigenous woman’s shawl fastener, itsmaterial is a certain indicator that places theburial in colonial times, probably between 15August 1539 when the Spanish settlement atHuánuco Pampa was officially founded and 2February 1541 when it was abandoned. Sheaalso noted a few animal bones and Inca sherdsassociated with the burial but pointed out thatthese could have been elements of the grave’sfill and not contemporaneous with the burialitself (Shea, field notes, p. 7). Some of thephotographs show three or four small, light-colored conical objects and one small, light-colored round object in the center of the chest.These are not mentioned by Shea in his fieldnotes. It is tempting to think of these objects asthree bone arrow heads and one oxidized leadball, and the proximate cause of death. If that isso, they represent exceptionally precise shoot-ing by two assailants, one using a bow and the

other a long gun or pistol. Because the possible4

arrowheads do not resemble those used bySpanish bowmen, at least one of the shooterswould have been an Indian. However, thephotographs are an insufficient basis for theidentification of these artifacts. Nevertheless,the stratigraphy and the position of the body,coupled with the iron pin, indicate that theburial is post-Inca.

No statistical treatment can be given to asingle skeleton found in isolation. However, theburial context is important, although it cannotbe fully interpreted. Publication of individualburials may, gradually and collectively, lead toinsights that cannot be anticipated at present.Without publication, the information inherentin these burials remains unavailable to scholars for analysis and synthesis. For other examples ofpublished single burials see Gaither et al. 2009;and Quilter et al., this volume, p. 106, figure 12.An important caveat must, however, be stated.The authors had no access to the bones them-selves, and worked only with photographs of theskeleton in situ and with views of each of theaspects of the skull. Furthermore, distortionsmay have been introduced by camera angle andlens. We do not know what photographic equip-ment was employed, except that it included a 35millimeter camera, probably a rangefinder, anda medium format camera, probably a twin lensreflex. This situation poses obvious limits toanalysis.

Description and analysis of the osteological anddental remains

The skeleton is extended, lying on its back,facing east, with the left forearm crossing thethird lumbar vertebra and then crossing theright arm with the left hand resting on the right

For an unambiguous example of an Andean Indian killed4

by a firearm in the immediate post-conquest period seeMurphy et al. 2010:641-642, figure 4.

ANDEAN PAST 10 (2012) - 270

Figure 1: Skeleton discovered in building encompassingPortals 5 and 6, Huánuco Pampa, viewed from the feet.Portal 5 is behind the skeleton and other portals leadingto the ushnu can be discerned in the distance. Roll 26,photo 5.

Figure 2: Skeleton discovered in building encompassingPortals 5 and 6, Huánuco Pampa, viewed from the trueleft side. Roll 26, photo 4.

elbow. The right arm is extended with the elbowdescribing a 90 degree angle. The left humeruslies parallel to the vertebral column. The rightleg is slightly flexed laterally at the knee (Figures1, 2).

The deceased was possibly placed in thegrave in a state of rigor mortis, which would putdeath anytime between three hours and threedays prior to burial.

Sex. This cannot be established for a certaintyfrom the pelvis. The subpubic angle of the os coxae appears to be narrow, indicating a male,with the caveat that views of the pelvis are notgood, and there are some feminine indications.However, there is no ventral arc, further sug-gesting that the individual is male. The greatersciatic notch is indeterminate to probably male.The ischiopubic ramus is wide, and possiblyblunt, also suggesting a male. However, a sub-pubic concavity, a female trait, is possibly pres-ent. The pubic symphysis appears to be beveledmesially, but this is uncertain from the photo-graph (Figure 2).

Cranial features including the mastoidprocess, glabella, nuchal crest, and the temporalline are indeterminate. There are large supra-orbital arches concentrated near the mid-line.

Age. All limb bone epiphyses visible appear to befused. There are no visible osteophytes orenthesiophytes, suggesting an age at death ofless than 40 years. Cranial sutures are all open, although the metopic, basilar, and incisivesutures are fused (Figure 3), suggesting anindividual older than 20. The anterior medianpalatine suture is open, indicating a young adult.There is incomplete fusion of the zygomaticbone with the maxillary bone, also suggesting ayoung adult, someone who died between the ageof 20-35. The pubic tubercle is clearly visible inthe right pubic symphysis. Because this feature

271 - Research Reports

disappears by about 25 this could be taken as anupper boundary of age.

Post-cranial pathology. Little can be said aboutthe post-cranial bones on the basis of the photo-graphs alone. However, Figure 2 shows a possi-bly arthritic surface for the proximal left radius.Alternatively, the photograph may be showingporosity common in the young. The pitting inthe lateral superior right clavicle may be due tothe stronger muscle attachment of a right-hand-ed person, or it may represent the result ofhealing after torn ligaments. There is no obviousevidence of the right arm having been madeunusable. This probably fits with the obviouslateral deviation to the left of the nuchal mus-cles (see cranium, below), but there is no evi-dence of tumpline use. There are no obviousparry fractures on the left forearm, but thesewould probably not be visible in these photo-graphs. The right humerus appears to be quitestraight, typical of agriculturalists. There ispossible osteoarthritis on the right distal hu-merus on the lateral margin of the capitulum.

Cranium

Posterior. The posterior view of the cranium(Figure 3) shows pronounced asymmetry. Theneurocranium is slightly flattened at lambda,with greater deflection to the right. There is agreater expression on the attachment of nuchalmuscles to the left; the sagittal suture deviatestowards the right parietal bone. The parietalforamina are moderate to large in size; it isunusual that the right is not closer to lambda.

There is occipital parietal flattening, possiblyindicating deliberate cranial modification. Thereis slight, healed porotic hyperostosis in thesquamous portion of the occipital bone. Thissuggests the individual may have suffered fromanemia at some point in life.

Figure 3: Posterior view of cranium of individual de

picted in Figures 1 and 2. Roll 26, photo 15.

Figure 4: Right lateral view of cranium of individualdepicted in Figures 1 and 2. Roll 26, photo 14.

Right lateral cranium. (Figure 4) The canal isnearly circular. There is a possible puncturewound at the superior articular border of thetemporal bone. There is an accessory bone atpterion. The right temporal line is heavy for theindividual’s age. Alveolar prognathism is notpronounced. The ratio of the right mastoidlength to the height of the skull at mastoid, asmeasured in the photographs, is 0.23; the ratio

ANDEAN PAST 10 (2012) - 272

of the left is 0.20. The right zygomatic arch alsoappears more developed than the left, indicatinga larger masseter muscle for chewing. Theseobservations are confirmed by the lack of molarson the left side.

Figure 5: Left lateral view of cranium of individualdepicted in Figures 1 and 2. Roll 26, photo 17.

Left lateral cranium. There are no auditory osteo-mas on either side (Figure 5). The left auditorymeatus is more flattened than the right, whichsuggests the loss of molars may have been duringthe early teen years (see dental pathology be-low) to have affected the development of themeatus. The left mastoid muscle attachmentmay differ from the right, with possibly moresurface area for the sterno-cleido-mastoid mus-cle to attach. This may explain a greater expres-sion at its insertion in the right clavice. How-ever, the left mastoid appearance may be due tothe angle of the photograph. Damage to thesuperior temporal bone appears postmortem.Canals on the greater wing of the sphenoid arewithin normal range. The specimen exhibits anaccessory bone at the asterion and in the occipi-tal suture, reflecting sutural complexity. Adepression on the left posterior parietal could bea healed wound, but it would require inspectionof the inner table to confirm.

Figure 6: Frontal view of cranium of individual depictedin Figures 1 and 2. Roll 26, photo 14.

Frontal view. The frontal view of the skull(Figure 6) shows expansion of the supraorbital ridges medially. The nose is broad and the lowermargin is guttered. The orbits are rectangular;the zygomatic arches are broad. There is sub-stantial resorption in the sub-orbital regions. The temporalis muscle has marked the rightfrontal bone more than the left, indicating morechewing on that side, the one that retainedmolars. There is a possible cut superior to theright orbit. The supraorbital canal on the rightis enclosed, the left not. Remnants of the meto-pic suture are visible above the nasion.

Inferior view. The right occipital condyle isdivided and with a much smaller articulatoryarea than the left (Figure 7). The left occipitalcondyle shows temporal mandibular joint dis-ease with the mandibular condyle movingforward to articulate with the sphenoid. TheTMJ differences may represent the bad bite thatdeveloped from the loss of the upper left molars

273 - Research Reports

and the need to exert more force upward andlaterally from the mandibular right molars.

Figure 7: Inferior view of cranium of individual depictedin Figures 1 and 2. Roll 26, photo 16.

Dental pathology. The maxillary first and secondleft molars and first, second, and third rightmolars are missing antemortem. Tooth wear isevident, most notably on the anterior dentition.There is evidence of reactive bone, indicatingactive infection at the time of death. There is acarious lesion on the distal interproximalcementoenamel juncture (CEJ) of the rightmaxillary first premolar. The left maxillary firstpremolar demonstrates a distal interproximalcarious lesion on CEJ that could indicate para-masticatory behavior, although it could becoincidental. There is tartar accumulation onthe right maxillary first and second molars. Thesecond premolar is not in occlusion. A possiblygranulating crypt for the right third molar isobserved (Figure 4). This trait establishes alower age boundary in the late teens. All threemolars had been present and erupted indicatinga minimum age of approximately 15 to 18 years.

Teeth show significant resorption of thealveoli, with pronounced porosity, which proba-bly indicates chronic infection of the tissue. Thecentral incisors are winged, deviating mesiallyand posteriorly. Enamel hypoplasia is not evi-dent on the crowns (only extreme lines wouldbe visible in the photograph), but there is some

effect on the roots at the coronas of the incisors.The suborbital (infraorbital) canal (artery) onthe right side shows signs of osteitis. The pala-tine suture is irregular as expressed in the frontalview and also visible in Figure 7, an inferiorview.

Figure 7 shows the lack of molars on the leftmaxilla, the alveoli filled or filling. The level ofgranulation in the upper jaw suggests that theleft molars were lost in the last year or two oflife. The right side shows what appears to be two molars above the occlusal plane but in fact is asecond molar behind it, as can be seen in Figure 6. Its crypt was being granulated, but withoutseeing the floor, one cannot be sure whether itwas present in life. There does not appear to beinterstitial wear on the distal surface of M2,which argues against its having been presentand in occlusion for any length of time. Theremay be a carious lesion on the occlusal surface of M2 and possibly M1. Two of the lower leftmolars were lost in life and resorbed, but theextreme recession of the alveolus in one soyoung suggests that infection could have playeda role.

Conclusions

The last days of Huánuco Pampa wereviolent ones. By the late 1530s the site had beenlargely abandoned by its indigenous inhabitants.As Craig Morris, director of excavations atHuánuco Pampa after John Murra, pointed out:

The seven years immediately following theconquest are of special archaeological impor-tance because it was during that time thatmost of the artifact associations in the Incaicsectors of the site were formed as native usepatterns were discontinued under varying de-grees of disruptions from both the conqueringEuropeans and the indigenous rebellions

against new rule (Morris 1979:210).

ANDEAN PAST 10 (2012) - 274

The person discussed in this report is mostlikely part of the “associations” formed duringHuánuco Pampa’s final occupation. The graveseems to have blocked the free flow of water orother liquids in an important part of the site,suggesting that maintaining this flow was nolonger of importance. The prone position facingeast, crossed arms, and iron pin all indicate apost-Inca burial. However, the cadaver was notplaced in a Christian cemetery. It is not clearwhen such a cemetery was established atHuánuco Pampa. Documentary evidence onlysuggests that a hacienda chapel already existedin 1786 when it appears on a map of HuánucoPampa (Sobreviela and Sierra 1786). In any case,the presence of a single human burial in thispart of the site is intriguing.

ACKNOWLEDGEMENTS

We thank Jack H. Prost and Ellen FitzSimmons Steinbergfor suggestions incorporated into this paper. Lon Barashshared his recollections of excavating the skeleton dis-cussed in this report. Sumru Aricanli facilitated access tothe photographs and field notes.

REFERENCES CITED:

Barnes, Monica2010 John Victor Murra’s Provincial Inca Life Project

and American Anthropology. Antiquity Volume084, issue 326, on-line only, Project Gallery.http://www.antiquity.ac.uk/projgall/barnes326/(consulted6 February 2012).

Gaither, Catherine, Jonathan Bethard, Jonathan Kent,Víctor Vásquez Sánchez, Teresa Rosales Tham, andRichard Busch2009 Strange Harvest: A Discussion of Sacrifice and

Missing Body Parts on the North Coast of Peru.Andean Past 9:177-194.

Meddens, Frank M., Colin McEwan, and Cirilo VivancoPomacochan2010 Inca “Stone Ancestors” in Context at a High-

Altitude Usnu Platform. Latin American Antiquity21(2):173-194.

Morris, Craig1979 The Spanish Occupation of an Inca Administra-

tive City. Actes du XLII Congrès International dese

Américanistes: Congrès du Centenaire, Paris, 2-9Septembre 1976. IX-B:209-219, Paris.

Morris, Craig, R. Alan Covey, and Pat Stein2011 The Huánuco Pampa Archaeological Project:

Volume 1: The Plaza and Palace Complex. Ameri-can Museum of Natural History AnthropologicalPapers 96.

Morris, Craig and Donald Thompson1985 Huánuco Pampa: An Inca City and its Hinterland.

London: Thames and Hudson.Murphy, Melissa S., Catherine Gaither, Elena Goycochea,John W. Verano, and Guillermo Cock2010 Violence and Weapon-related Trauma at Puru-

chuco-Huaquerones, Peru. American Journal ofPhysical Anthropology142:646-649.

Murra, John V. and G[ordon] J. Hadden1966 Apéndice: Informe presentado al Patronato

Nacional de Arqueología sobre la labor de lim-pieza y consolidación de Huánuco Viejo (20 dejulio a 23 de noviembre 1965). Cuadernos deInvestigacion, Antropología 1:129-144 plus 12unnumbered plates.

Quilter, Jeffrey, Régulo Franco J., César Gálvez M.,William Doonan, Catherine Gaither, Jaime Jiménez S.,Hal Starratt, and Michele L. Koons2011 The Well and the Huaca: Ceremony, Chro-

nology, and Culture Change at Huaca CaoViejo, Chicama Valley, Peru. Andean Past 10:99-129.

Sobreviela, Padre Manuel and Lorenzo de la Sierra1786 Plan del palacio destinado para baño de los yncas en

el partido de Huamalies con el nombre de Huanucoel viexo. British Library, Add. MS. 17671.

Mitomarca: A Possible Fortification in theUpper Huallaga Basin

Yuichi Matsumoto (Dumbarton Oaks, e-mail:ymatsumoto1976@gmail.com), Jason S. Nes-bitt (Yale University, e-mail: jason.s.nesbitt@gmail.com), and Denesy Palacios J. (Universi-dad Nacional Hermilio Valdizán, Huánuco,Perú, e-mail: denesy@terra.com).

The Upper Huallaga Basin is on the easternslopes of the Andes and is a mountainous regioncut by the Huallaga River (Figure 1). It is animportant ecological interface and a corridorbetween the central highlands and Amazonia.

During the 1960s, the University of Tokyocarried out intensive excavations at the sites ofKotosh and Shillacoto. Although this pioneer-

275 - Research Reports

ing research established a fine-grained relativechronology of the region’s Initial Period andEarly Horizon, little archaeological research hasoccurred since then because of widespreadviolence during the 1970s and 1980s. As aresult, there are considerable gaps in our under-standing of the culture history of the region,especially for the long span of time between thefirst century A.D. and the years immediatelyprior to the incorporation of the region into theInca empire in the late fifteenth century A.D. Therefore, for the purpose of amplifying archae-ological understanding of the prehistory of theUpper Huallaga Basin, a settlement patternstudy was conducted in 2001 by Kinya Inokuchi(University of Saitama, Japan), Yoshio Onuki(University of Tokyo), Eisei Tsurumi (Univer-sity of Tokyo), Yuichi Matsumoto, and AlvaroRuiz.

In 2005 and 2006, Yuichi Matsumoto andJason Nesbitt conducted a follow-up study tothe 2001 survey to understand the region’sprehispanic water management. In 2005, it wasobserved that at least some of the Initial Periodand Early Horizon sites in the upper HuallagaBasin were near possibly ancient canals. In2006, we examined this possibility throughsurvey of canal systems around the Early Hori-zon site of Sajara-patac. While elements of thecanal near Sajara-patac are clearly modern (e.g.the use of concrete), other parts suggest greaterantiquity. For instance, sections of the canal,especially on nearby slopes, are shallow gravity-fed channels cut into bedrock. Based on itsproximity to Sajara-patac, we hypothesize thatthe canal may have been associated with theEarly Horizon occupation. Therefore, in 2006,we followed this canal with the intent of deter-mining if it was associated with other archaeo-logical sites. We found that the canal passed byanother, larger site enclosed by a massive perim-eter wall. This site is locally called Mitomarca.We hypothesize that Mitomarca may be a

fortification dating primarily to the Late Inter-mediate Period.

Site Description

Mitomarca is in the district of Churubamba,close to the eastern end of the Upper HuallagaBasin. The site stands on a natural hill along thenorthern bank of the Huallaga River at anelevation of 2300 m.a.s.l. The Early Horizonceremonial center of Sajara-patac (Inokuchi etal. 2003; Matsumoto and Tsurumi in press) isapproximately 600 meters to the southeast, onthe lower part of the same hill (Figure 1).

Figure 1: Location of Mitomarca and distribution ofarchaeological sites in the upper Huallaga Basin.

Figure 2: Southern perimeter wall of Mitomarca.

ANDEAN PAST 10 (2012) - 276

Figure 3: Mitomarca from the north, showing perimeterwall. The arrow indicates the location of the plaza whilethe dotted line shows the perimeter wall.

Figure 4: A platform construction in the northern area of

Mitomarca.

Mitomarca extends over an area of at least0.8 hectares, and many stone constructions arevisible on the surface. A massive perimeter wallmore than 1 meter wide and 1.5 meters highsurrounds the hilltop (Figures 2, 3). Because thebase of the wall is not exposed, the wall is prob-ably much taller. In some sections the wall isdouble-faced and could be part of long narrowroom, or hallway-like buildings. This massivewall curves in accordance with the naturallandscape and forms a large ovoid enclosure. Itsstone masonry consists of stacked, quarried, flatstones set with mud mortar. These materials

were locally available and the stones wereprobably quarried from the natural outcropsaround the site. The flat sides of the stones wereused for the wall faces. As noted above, a canalthat may have been used from the prehispanicperiod runs from north to south along theexterior of the perimeter wall.

The area inside the perimeter wall containsseveral rooms and possible platform construc-tions (Figure 4). Although the central part ofthe area is one to two meters higher than otherareas, it was difficult to determine whether aplatform-like construction is buried below, orthe rise is simply part of the natural landscape.Many lines of stone walls can be recognized inthis area and most seem to be double-faced.Room constructions can be seen in the northernpart of the enclosure. Although most of themare square, the size of the rooms varies, and thelargest one measures approximately three byfour meters in plan. To the north of the clusterof square rooms, three rectangular platformsprotrude from the slope to the exterior.

In the southwestern area of the site, near itsedge, there is a square plaza measuring twenty-three meters by twenty-three meters (Figures 3,5). While the northern and eastern plaza wallsare unifacial and the area behind them back-filled, the south and west ones are double-faced.

Figure 5: Plaza from the north.

277 - Research Reports

Chronological Position of Mitomarca

The relative chronology of the Initial Periodto the Early Intermediate Period is well estab-lished in this region (e.g. Onuki 1972). Ceramicstyle is the fundamental tool for establishing thechronological position of the Mitomarca site.Although potsherds were scarce on the surface,we recognize late Early Horizon (Kotosh Sajara-patac Period) and Early Intermediate Period(Kotosh Higueras Period) sherds. However,most of the potsherds do not fit within thetypology established at Kotosh, and thereforeprobably postdate the Kotosh Higueras Period.Nevertheless, stylistic similarities with potteryfrom Kotosh and Sajara-patac suggest that thelatter site was first occupied in the late EarlyHorizon or early part of the Early IntermediatePeriod. A radiocarbon date from Sajara-patac(Pta-9659) yields an uncalibrated date of2140±70 B.P. for this period (Matsumoto andTsurumi in press).

However, the architecture exposed on thesurface at Mitomarca seems to contradict theceramic data somewhat. The architectural styleis clearly distinct from that of the late EarlyHorizon and Early Intermediate Period of theregion which share similarities. For example,while the Sajara-patac site is in close proximityto Mitomarca, the late Early Horizon architec-ture of Sajara-patac is much smaller in scale andthe masonry is not as well constructed as that ofMitomarca (ibid.). For instance, the masonry ofSajara-patac is much coarser in style than Mito-marca and does not use quarried stone (ibid.).Moreover, sites dating to the Early Horizon orthe Early Intermediate Period (Kotosh HiguerasPeriod) did not use massive perimeter walls. These data accord with other lines of archaeo-logical evidence, which suggest that warfare wasnot common in this part of the Huallaga Basinduring the Early Horizon.

Instead, high walls made of stacked flatstones and massive defensive walls enclosing ahilltop show a great deal of similarity with theLate Intermediate Period and Late Horizonarchitectural styles of the central highlands (e.g.Bonnier 1997; Grosboll 1988; Mantha 2009;Morales 1984; Salcedo et al. 2000; Thompson1967). Although the available data are scarce,the architectural style of Mitomarca seems toshow affiliation to architectural styles distrib-uted from the quechua to puna along the Mara-ñón and Huallaga drainages and documented atsites such as Piruru (Girault 1981), Garu (Sán-chez Murrugarra and Palacios 1988; Salcedo etal. 2000), and sites in the Rapayán Valley (Man-tha 2009) and the Province of Dos de Mayo inHuánuco (Morales 1984). The existence ofKotosh Higueras Period Pottery at Mitomarcamay suggest, as Grosboll indicated (1988:354-358), the considerable continuity in the potterystyle from the beginning of the Early Intermedi-ate Period until late prehistoric times. Isbell(1974) refers to a pottery style similar to KotoshHigueras as the “CB series” shared in the centraland south-central highlands. According toOnuki, this style existed until the Late Interme-diate Period in the Upper Huallaga Region(Onuki 1982:219-220).

Mitomarca as a Fortification

Based on the hilltop location and presenceof a massive perimeter wall, we hypothesize thatMitomarca probably functioned as a fortifica-tion. The data from the north coast and high-lands indicate that the late Early Horizon was atime of political instability and social tensionamong different groups following the collapse ofthe Chavín sphere of interaction (Burger 1992).In these regions, radical changes in settlementpatterns are recognized. Many fortified sitesappeared and the population was clusteredaround them. Mitomarca contains a late EarlyHorizon component, and, therefore, it may seemappropriate to consider it as a fortified site that

ANDEAN PAST 10 (2012) - 278

appeared as a response to the political instabili-ties of this period. However, the general settle-ment pattern data clearly contradict this view.During the Kotosh Sajara-patac Period, thenumber of sites radically increased and new siteswere distributed throughout the yunga ecozonein the Upper Huallaga Basin. Although newsites were constructed at higher elevations nearthe border of the yunga and quechua ecozones,bottomland sites also increased in frequency, butsome of them were first occupied in the middle-Early Horizon and continued to be occupiedduring the Kotosh Sajara-patac Period. Further-more, most sites dating to this time did notutilize defensive architectural features. There-fore, in the Upper Huallaga Basin, it may bedangerous to assume the existence of increasedconflict with neighboring regions during the lateEarly Horizon, and, thus, it is difficult to inter-pret Mitomarca as a fortified site of the lateEarly Horizon or beginning of the Early Interme-diate Period.

If we accept that the surface architecture ofMitomarca was constructed some time after theEarly Intermediate Period, probably during theLate Intermediate Period, it seems more proba-ble that Mitomarca was a fort, because of theknown tensions among different ethnic groupsof the Huallaga region at that time. Althoughthe cultural developments after the Kotosh Hi-gueras Period, that is, from the Middle Horizonto the Late Horizon, are not well known in theUpper Huallaga Basin, the survey data of SueGrosboll (1988) provide useful insights perti-nent to this issue.

Grosboll surveyed the area where theChupachu and Quero people lived, and shelocated many of the Late Horizon sites describedin the visita document of 1549 and 1562 writtenby Iñigo Ortiz de Zuñiga (Ortiz de Zuñiga (1967,1972 [1562]). According to Grosboll, the sitesthat belong to the Middle Period (500-1300A.D.) in her terminology, tend to be found in

between 2400 and 2900 m.a.s.l., while the LateHorizon sites were frequently in the elevationsbetween 3000-4000 m.a.s.l. (Grosboll 1988:75).As Daniel Morales (1984) points out regardingthe Late Intermediate Period and Late Horizonsites in the Upper Huallaga Basin, it is veryprobable that the sites that Grosboll registeredas Late Horizon have a Late Intermediate Periodcomponent as well. Therefore, the shift ofsettlements to higher elevations may haveoccurred during the Late Intermediate Period,rather than during the Late Horizon. If thischange in settlement pattern reflected thepolitical disintegration of ethnic groups, it mightbe possible to interpret Mitomarca as a fort.Alexis Mantha (2009) identified multiple sitecategories based on survey of the RapayánValley approximately 80 kilometers to the northof Mitomarca. He found that some of the highelevation sites of the Late Intermediate Perioddisplay both defensive and ceremonial charac-teristics, and named sites of this type as “Defen-sive, Ceremonial, and Communication settle-ment (DCC)” (ibid.:168-169). If, in Mitomarca,the square plaza had a ceremonial function, andthe perimeter wall functioned as a defensivewall, this might be an appropriate term to repre-sent Mitomarca.

Conclusion

Our preliminary observation of Mitomarcagenerated more questions than we had at thebeginning of our field-work, rather than solvingissues of the later prehistory of the UpperHuallaga Basin. Only further study of the sitewill make it possible to evaluate the hypothesespresented in this report. While the regionalcultural context from the Preceramic Period tothe Early Horizon is well studied, the periodafter the Early Intermediate Period is onlypoorly known. In particular, the Middle Horizonand the Late Intermediate Period of the UpperHuallaga lack basic archaeological data. Mito-marca is a multi-component site that probably

279 - Research Reports

contains occupation during these periods. Thus,the site will be an appropriate research site to fillthe gap in archaeological data and extend theregional chronology to the Late Horizon.

REFERENCES CITED

Bonnier, Elizabeth1997 Morfología del espacio aldeano y su expresión

cultural en los Andes centrales. In Archaeológicaperuana 2: Arquitectura y civilización en los Andesprehispánicos, edited by Elizabeth Bonnier andHenning Bischof. Reiss-Museum, Mannheim,Germany: Sociedad Arqueológica Peruano-Alemana.

Burger, Richard L. 1992 Chavin and the Origins of Andean Civilization.

New York:Thames and Hudson.Girault, Louis1981 Fouilles sur le site de Piruru en 1968 et 1970.

Bulletín de l’Institut Français d’Études Andines10(1-2):101-112.

Grosboll, Sue1988 An Archaeological Approach to Demography of

Prehispanic Andean Communities. Ph.D. disserta-tion, University of Wisconsin, Madison.

Inokuchi, Kinya, Yosio Ounki, Eisei Turumi, Yuichi Mat-sumoto, and Nelly Martell Castillo. 2003 Excavations at the Site of Sajara-patac in Peru.

América Antigua 6:35–52 (written in Japaneseand published by the Sociedad Japonesa de Estu-dios sobre la América Antigua).

Isbell, William H.1974 Ecología de la expansión de los quechua-hablan-

tes. Revista del Museo Nacional 40:139-155.Mantha, Alexis2009 Territoriality, Social Boundaries and Ancestor

Veneration in the Central Andes of Peru. Journalof Anthropological Archaeology 28(2):158-176.

Matsumoto, Yuichi and Eisei Tsurumiin press Archaeological Investigations at Sajara-patac in

the Upper Huallaga Basin. Ñawpa Pacha.Morales, Daniel1984 Algunos sitios arqueológicos del Reino de Gua-

núco. Boletín de Lima 6(33):83-95. Onuki, Yoshio1972 Pottery and Clay Artifacts. In Andes 4: Excava-

tions at Kotosh, Peru, 1963 and 1966, edited bySeiichi Izumi and Kazuo Terada, pp. 177–248.Tokyo: University of Tokyo Press.

1982 Una perspectiva prehistórica de la utilizaciónambiental en la sierra nor-central de los Andescentrales. In El hombre y su ambiente en los Andescentrales. Senri Ethnological Studies 10, edited by

Luis Millones and Hiroyasu Tomoeda, pp. 211-228. Osaka, Japan: National Museum of Ethnol-ogy.

Ortiz de Zúñiga, Iñigo1967 Visita de la provincia de León de Huánuco en 1562,

volume 1, Documentos para la Historia y Etno-logía de Huánuco y la Selva Central, Universi-dad Nacional Hermilio Valdizán, Huánuco,Perú.

1972 Visita de la Provincia de Leon de Huánuco en 1562,Tomo 2, Documentos para la Historia y Etnolo-gía de Huánuco y la Selva Central, UniversidadNacional Hermilio Valdizán.

Salcedo Camacho, Luis E., Enrique Molina, Luis E.Cáceres, Patricia Habeteler, Víctor Curay, and Juan Rofes2000 Trabajos de puesta en valor en el complejo

arqueológico Garu. Revista Desafíos 1:60-74.Sánchez Murrugarra, Ricardo and Denesy Palacios J.1988 Garu: Un complejo urbano preincaico. In

Arquitectura e arqueología: Pasado y futuro de laconstrucción en el Perú, edited by Víctor RangelFlores, pp.153-162. Chiclayo, Perú: Universidadde Chiclayo, Museo Brüning, and CONCYTEC.

Thompson, Donald E. 1967 Investigaciones arqueológicas en las aldeas chu-

pachu de Ichu y Auquimarca. In Visita de laprovincia de León de Huánuco en 1562. Tomo I:visita de las cuatro waranqa de los chupachu, editedby John Victor Murra; Ortiz de Zúñiga, VolumeI., pp. 357-367. Universidad Nacional HermilioValdizán, Facultad de Letras y Educación,Huánuco.

An Initial Period Domestic Occupation atHuaca Cortada, Caballo Muerto Complex

Jason S. Nesbitt (Yale University, e-mail: jason.s.nesbitt@gmail.com) reports on his recentexcavations at the Caballo Muerto site in thelower Moche Valley.

Caballo Muerto is well known in the archae-ological literature as a large multi-mound Cupis-nique center in the lower Moche Valley. Pio-neering research at Caballo Muerto in the late1960’s and 1970’s established that each of themounds dated to the Initial Period or EarlyHorizon (Chauchat et al. 2006; Pozorski 1983;Watanabe 1976). Between 2007 and 2008, Iinitiated new excavations at Caballo Muerto(Nesbitt et al. 2010). The goal of the excava-

ANDEAN PAST 10 (2012) - 280

tions was to examine Cupisnique social, reli-gious, and economic organization from a dia-chronic perspective. To that end, my researchconsisted of excavations within the monumentalbuildings of Huaca de la Cruz, Huaca Cortada,and Huaca Curaca. (For a map of CaballoMuerto and the relative position of its variousmounds see Nesbitt et al. 2010: figure 2; Po-zorski 1976: 295, figure 2). This research estab-lished a long sequence beginning in the earlyInitial Period and continuing into the middleEarly Horizon.

Another aspect of my excavations atCaballo Muerto was the location and excavationof domestic occupations. Caballo Muerto pres-ents a number of logistical and methodologicalchallenges to the location of small scale, non-monumental architecture. Earlier excavationsaround Huaca Herederos Chica and HuacaHerederos Grande, located 250 meters to thesouth of Huaca Cortada, determined that theInitial Period ground surface was buried byapproximately four meters of alluvium(Chauchat et al. 2006). Therefore, if residencesare outside the areas of public architecture, thesubstantial depths at which these occupationsare likely to be found make it difficult to un-cover them. During my research at CaballoMuerto, I conducted four test excavations in theagricultural fields between Huaca Cortada andHuaca Herederos Grande (Figure 1). Each ofthe four units was positive and showed evidencefor deeply buried domestic occupations andsmall scale ceremonial architecture. Here Ireport on one of these excavations: Unit 2,which yielded household features, as well asartifactual, zooarchaeological, and archaeo-botanical materials that date to what I term theCortijo Phase, tentatively dated to the early andmiddle parts of the Initial Period.

Excavation of Unit 2

Unit 2 was a two by four meter excavation,in the agricultural fields just to the south ofHuaca Cortada (Figure 1), a large terracedplatform mound that measures one hundred byninety meters at its base and twenty-one metersin height. Radiocarbon dating places the earliestoccupation of this mound at approximately 1500cal BC (Nesbitt et al. 2010). As is discussedbelow, the Unit 2 domestic occupation is con-temporary with Huaca Cortada.

Figure 1: Plan of Huaca Cortada and Huaca Curacawith the location of test units (shaded areas).

Figure 2: The southern profile of Unit 2.

I identified twelve layers during the excava-tions of Unit 2 (Figure 2). The first (uppermost)layer was a mixed surface deposit of cobble-sized

281 - Research Reports

rocks and a small quantity of Chimú style pot-tery. Chimú ceramics and stone wall founda-tions were common in other test units aroundHuaca Cortada, indicating the presence of aChimú occupation postdating the use of themonuments at Caballo Muerto.

Directly underlying this first layer were lightbrown (7.5YR6/3) clayey soils with large cobblesand small boulders that contained Initial Periodpottery fragments (Layer 2). Associated withthese materials were small pieces of charcoal,eroded conical adobe fragments, and numerousburned, orange clay fragments with cane andcordage impressions. In some cases these claypieces had traces of white paint on their exteri-ors. However, no occupation surface was associ-ated with these materials. Beginning at a depth of 1.46 meters belowthe surface was a brown (7.5YR4/4) silty soil(Layer 3) that had fewer rocks, but an increas-ing abundance of broken conical adobe frag-ments, some of which can be seen in thesouthern profile. Within this layer were also thinlenses of light-colored sand and a small circularcarbon concentration. There were no diagnosticpottery rims, but the pastes of the body sherdswere consistent with Cortijo Phase ceramics.Interestingly, this layer was not continuous, butinstead represents a cultural layer that post-dated a probable flood of the Río Seco, now adry quebrada that runs just to the west of HuacaCortada.

This flood deposit was a thick series of layersthat sloped to the west (Layers 4-6). Theselayers together range from fifty-six to ninety centimeters in thickness and were composed ofa brownish yellow (10YR6/6) coarse sand con-taining small pebbles and rounded cobbles.There were no artifactual materials recoveredfrom these layers. In terms of color, composi-tion, and thickness, these layers were quitedistinct from the overlying and underlying

layers. I hypothesize that these layers originatedfrom the flooding of a branch of the Río Secoduring an El Niño event that covered the un-derlying Initial Period domestic occupationdescribed below.

Layer 7 consisted of a brown (7.5YR4/4)silty soil with several broken conical adobefragments, some of which also can be seen in thesouthern profile. Mixed within this layer weredeposits of light-colored sand. Just to the east ofthe adobe fragments was a small, circular carbonconcentration. The artifact assemblage con-sisted of a small number of non-diagnostic bodysherds with Initial Period pastes. As will bediscussed below, this layer was not continuous,but instead seemed to be a layer that pre-dateda flood of a branch of the Río Seco.

Intact Initial Period occupation surfaces(Layers 8-11) were uncovered beginning at adepth of approximately 2.20 meters below theground surface. These layers consisted of a seriesof superimposed domestic occupation layers(discussed below). Each layer is described sepa-rately below.

Layer 8 was a surface or floor made up of abrown (7.5YR4/3) compacted earth mixed inplaces with fine-grained yellow (10YR7/6) sand.The floor also had concentrations and flecks ofcharcoal and ash that began in the westernprofile of Unit 2 and continued to the east. Theprofile exhibited a discrete lens of charcoal andred (2.5YR5/6) earth that formed a probablehearth measuring approximately eighty centime-ters in length. The presence of a hearth in thesouthern profile indicates that the occupationalsurface continues to the south.

The Layer 8 floor yielded some refuse,including mammal, fish, and charred shellfishremains. Burned, orange clay fragments contin-ued to be abundant and a single adobe fragmentwas found resting on the surface. Pottery was

ANDEAN PAST 10 (2012) - 282

relatively abundant in this layer and I foundnumerous blackened, neckless olla rim frag-ments resting horizontally on the surface. Inaddition, I found a small, flat, stone circularornament with two holes in the center of theartifact (Figure 3l).

Layer 9, a second occupation surface, wasfound at a depth of 2.35 meters. It is a com-pacted brown (7.5YR4/3) deposit with few rocksin the matrix. Throughout the surface, therewas red (2.5YR5/6) burned earth, as well ascarbon concentrations. Near the center of theunit I found a long-neck bottle rim that corre-sponds to the Cortijo Phase (Figure 3b), and anassociated rough spherical stone (Figure 3m).Other fragments of pottery were found restinghorizontally on the surface of this layer.

Figure 3: A sample of artifacts from Unit 2; a-k showCortijo Phase pottery; a) short-necked jar, b) bottle, c)bowl, d-h) neckless ollas, i-k) decorated sherds; l and mshow the lithic materials described in the text.

As with other layers, there were abundantburned clay quincha (wattle-and-daub) frag-ments, with cane impressions. Shellfish remainswere also present in this layer.

Figure 4: Photograph of the Layer 10 floor with arrowindicating the location of the hearth. The apparentdifference in dimensions between Figures 2 and 4 is dueto parallax error of the lens employed in taking thephotograph.

Layer 10 represented a third occupationsurface found at an average depth of 2.6 metersbelow the present ground surface. The mainfeature from this layer was a circular hearth inthe eastern part of the unit. The hearth mea-sured approximately fifty centimeters in diame-ter. The outermost part of the hearth was linedby fire-reddened clay. Within the central part ofthe hearth were a series of fire-cracked rocks, aswell as abundant charcoal and ash (Figure 4). Itook a sample of the hearth for radiocarbondating and flotation.

Layer 11 was a thin layer of coarse yellow(10YR7/6) sand underlying the floor. In thisdeposit I found sixteen non-diagnostic sherdswith hearth-blackened exteriors and InitialPeriod pastes. In addition, two small fragmentsof human bone were recovered. Unfortunately,the part or parts of the skeleton they representcould not be identified.

283 - Research Reports

Layer 12 was a sterile deposit consisting ofloosely consolidated cobbles within a coarsesand matrix. No artifacts were recovered fromthis layer. Layer 12 was first found at a depth of2.7 meters below the present ground surface andthe excavations were terminated at just overthree meters below the surface.

I interpret Layers 8 through 11 as represent-ing a probable domestic occupation associatedwith Huaca Cortada. The evidence that sup-ports this conclusion includes the presence ofsuperimposed hearths, cooking vessels, andrefuse such as animal remains. The abundanceof adobe fragments and burned clay fragmentswith cane impressions and, more rarely, piecesof mud plaster, throughout most of the layers inUnit 2 suggest that the domestic unit was awattle-and-daub building.

The superposition of hearths and occupationsurfaces characterizing the domestic occupationsin Unit 2 appears to represent the accumulationof households, perhaps over several generations.As Hirth (1993:24) observed, these types ofaccumulated deposits create problems for inves-tigating household composition synchronically.To get around these problems, Hirth uses theterm “household series” to define the remainsthat amass from a succession of householdsoccupying the same physical space (or “home-stead” in Hirth’s [ibid.: 24] terminology). Layers8 to 11 of Unit 2 can best be interpreted as ahousehold series accumulated during domesticoccupations, probably over a relatively shortperiod of time during the Cortijo Phase.

Chronology

No absolute dates currently exist for theoccupation of Unit 2. However, a pilot radiocar-bon dating project focusing on the earliestphases of monumental architecture at HuacaCortada indicates that its first architecturalphase dates to around 1500 cal BC (Nesbitt et

al. 2010). The small assemblage of potteryassociated with this phase demonstrates closeparallels with Cortijo Phase ceramics recoveredfrom Layers 8-11 in Unit 2. Cortijo Phase pot-tery in Unit 2 consists of unslipped wares withoxidized pastes and coarse inclusions. The mostcommon form from the Cortijo Phase is a neck-less olla. Some examples have simple incised orpunctate decorations (Figure 3 j-k). The mostdiagnostic vessel form from this time, however,is a type of long-necked bottle, with small punc-tations ringing the area where the base of theneck meets the bottle chamber. Analysis ofCortijo Phase pottery at Caballo Muerto showsbroad similarities to a variety of early and middleInitial Period ceramic traditions including theMontegrande (Ulbert 1994) and early andmiddle Guañape (Strong and Evans 1952) stylesfrom the nearby Jequetepeque and Virú Valleys.At present, based on cross dating with HuacaCortada, I tentatively date the Cortijo Phase toapproximately 1500-1200 cal BC. A radiocarbondating project is underway to create a moredetailed absolute chronology, which will beavailable in the near future. More relevant tothis report, the ceramic data clearly show thatthe occupation described for Unit 2 is coevalwith the earliest known phases of use of HuacaCortada.

Subsistence economy

Careful recovery techniques, including fine-grained screening and flotation of Layers 8through 11, allowed the collection of floral andfaunal materials that shed some light on thesubsistence economy of the Cortijo Phase occu-pation of Caballo Muerto. Faunal materials fromLayers 8 and 9 include the remains of shellfish,including clams (Donax sp.) and mussels (Choro-mytilus chorus). In addition, the zooarchaeo-logical analysis found one bone from a gull(Larus sp.) as well one drum otolith. A singlethoracic vertebra from a deer (Odocoileus virgini-anus) was found in Layer 9. Flotation of the

ANDEAN PAST 10 (2012) - 284

hearth from Layer 10 yielded abundant fishremains including fifty-one anchovy (Engraulisringens) vertebrae and four drum (Sciaena delici-osa) vertebrae.

Starch grain residue analysis was conductedby Víctor Vásquez (Universidad Nacional deTrujillo) on a small sample (n=4) of cookingvessel sherds. The results of this study revealedstarch grains from manioc, potato, and maize.

In summary, the faunal, botanical, andresidue data point to a mixed economy duringthe Initial Period. This economy was based onagricultural products as well as on significantprotein resources from the Pacific coast, approx-imately sixteen kilometers to the west. Thesedata accord well with research done on subsis-tence in the Moche Valley by Shelia andThomas Pozorski. They hypothesized thatduring the early Initial Period there was a welldeveloped relationship between the coastalfishing community Gramalote and the inlandInitial Period monumental center of CaballoMuerto (Pozorski and Pozorski 1979). This mayhave been typical of coast-inland relations inthe region. Conclusions

Unit 2 represents one of the first knowndomestic occupations at Caballo Muerto andhas given new insights into its economy and thenature of its Initial Period occupation. Never-theless, many questions remain, which I amattempting to answer in my dissertation-in-progress. In particular, I hope to determine ingreater detail whether occupations like thatdescribed for Unit 2 represent permanent occu-pations around large mounds like Huaca Cor-tada, or if they represent temporary structuresused seasonally by peoples living in other partsof the Moche Valley who came to CaballoMuerto periodically for rituals.

REFERENCES CITED

Chauchat, Claude, Jean Guffroy, and Thomas Pozorski2006 Excavations at Huaca Herederos Chica, Moche

Valley, Peru. Journal of Field Archaeology 31:233-250.

Hirth, Kenneth G. 1993 The Household as an Analytical Unit: Problems

in Method and Theory. In Prehispanic DomesticUnits in Western Mesoamerica: Studies of theHousehold, Compound, and Residence, edited byRobert S. Santley and Kenneth G. Hirth, pp. 21-36. Boca Raton, Florida: CRC Press.

Nesbitt, Jason, Belkys Gutiérrez, and Segundo Vásquez2010 Excavaciones en Huaca Cortada, complejo

Caballo Muerto: Un informe preliminar. Boletínde Arqueología PUCP 12 (2008):261-286.

Pozorski, Shelia and Thomas Pozorski1979 An Early Subsistence Exchange System in the

Moche Valley, Peru. Journal of Field Archaeology6:413-432.

Pozorski, Thomas1976 Caballo Muerto: A Complex of Early Ceramic Sites

in the Moche Valley, Peru. Ph.D. dissertation,University of Texas, Austin.

1983 The Caballo Muerto Complex and its Place inthe Andean Chronological Sequence. Annals ofthe Carnegie Museum 52:1-40.

Strong, William D. and Clifford Evans 1952 Cultural Stratigraphy in the Virú Valley, Northern

Peru. New York: Columbia University Press. Ulbert, Cornelius1994 Die Keramik der formativzeitlichen Siedlung Monte-

grande, Jequetepequetal, nord-Peru. Materialien zurAllegemeinen und Vergleichenden Archäologie 39.Mainz, Germany: Verlag Philipp Von Zabern.

Watanabe, Luis1976 Sitios tempranos en el valle de Moche (costa norte

del Peru). Ph.D. dissertation, UniversidadNacional Mayor de San Marcos, Lima.

Marine Exploitation and Paleoenvironment asViewed through Molluscan Resources at theEarly Horizon Center of Huambacho, NepeñaValley, Coastal Ancash

David Chicoine (Louisiana State University, e-mail: dchico@lsu.edu) and Carol Rojas(Proyecto Arqueológico Huaca del Sol y laLuna, e-mail: carolrojas.vega@gmail.com).

285 - Research Reports

In 2003-2004, twenty-three weeks of field-work at Huambacho yielded important data onEarly Horizon societies in the lower NepeñaValley, Coastal Ancash, Peru. Chicoine directedthe field-work as part of the Proyecto Arqueol-ógico Huambacho. Rojas carried out the first-hand identification and classification of themolluscan remains. Analyses of architecture,material culture, and radiometric measurementsindicate that between 800 and 200 cal B.C.E.(2F) Huambacho represented a small elitecenter with strong public components (Chicoine2006). Settlement patterns and more recentfield-work elsewhere in Nepeña suggest that theHuambacho community was related to thelarger complex of Caylán located further inland,on the north bank of the river, some fifteenkilometers from the littoral (Chicoine andIkehara 2010).

Significant numbers of artifacts and ecofactswere recovered during field-work, includingmollusk remains and fish bones. This reportfocuses on the place of the molluscan assem-blage within the faunal remains and implicationsfor understanding past diets, economies, socialorganizations, and aquatic ecologies. In particu-lar, taxonomic analyses reveal patterns of ma-rine exploitation and bioindicators.Paleoenvironmental reconstructions are, ingeneral, consistent with the cold waters typicalof the Humboldt Current and little evidencesupports the occurrence of major climatic dis-turbances, including those related to the ElNiño-Southern Oscillation phenomenon orENSO.

Background and research at Huambacho

The Early Horizon (900-200 B.C.E.) isusually seen as a time of increased interregionalinteractions tied to the Chavín phenomenon insome regions of what is now Peru. In coastalAncash, economies were mainly based on irriga-tion agriculture, animal husbandry, the exploita-

tion of marine resources, and trade. Popularplant domesticates included varieties of tubers,fruits, cucurbits, and legumes, as well as maize,introduced in the region at the beginning of theEarly Horizon. In a continuation of InitialPeriod (1800-900 B.C.E.) socioeconomic pro-cesses, irrigation farming and inland settlementsgradually gained more importance than settle-ments dependent upon coastal adaptations.Nevertheless, studies of marine resources arescarce, and our knowledge of coastal economiesin Ancash is still preliminary (Pozorski andPozorski 2003). The Huambacho molluscandata allow a preliminary assessment of EarlyHorizon marine economies and their associatedtrade networks, as well as local environmentalconditions.

Huambacho is on the south bank of thelower Nepeña Valley, some eight kilometersfrom the littoral. It consists of two distinctcomplexes of stone-walled enclosures, patios,and monumental plazas originally covering anarea of more than twelve hectares. Our excava-tions have yielded a variety of cultural remains,including more than 4,000 diagnostic potterysherds, and faunal and botanical remains, mostlyfrom secondary refuse deposits. Field-workresults suggest that an elite group permanentlyresided onsite, while most community memberslikely lived in peripheral areas.

The analysis of vertebrates allowed theidentification of a minimum number of 1,300bones (NISP), including both wild and domesti-cated animals (Table 1). Fish and birds includemany species of marine and riverine animals.Rodents include guinea pigs (Cavia porcellus)and also rats and mice. Bone fragments of dogs(Canis familiaris) and sea lions (Otaria flavescens)were also identified.

Animal bones are dominated by camelids (c.40 percent). Further analyses will be needed toassert the relative importance of each category

ANDEAN PAST 10 (2012) - 286

of the vertebrates (e.g. meat weight, quantity ofanimal protein, identity of camelids). The highfrequency of camelids contrasts with InitialPeriod sites in coastal Ancash where theseanimals are usually rare or absent. As packanimals, camelids would have been useful in theEarly Horizon context of intensified interre-gional interactions.

At Huambacho, the virtual absence of meatcutting tools, such as obsidian blades, and thelack of intensive fracturing or burning of camelidbones contrasts with typical patterns of meatconsumption. At the same time, members of theproject collected large amounts of shell remains,highlighting the importance of seafood. Theevidence from Huambacho is consistent withcomparative data from Casma, the neighboringvalley to the south, where the local subsistenceeconomy was mainly based on maize agriculture,as well as the exploitation of marine resourcesand, with less intensity, animal husbandry.

TotalCategory NISP percentCamelid 517 39.77Rodent 464 35.69Fish 144 11.08Bird 55 4.23Other/Unknown 120 9.23TOTAL 1,300 100.00

Table 1: Number of individual specimens (NISP) ofanimal bones, counts, and percentages.

Shell remains

Apart from fish remains, members of theproject also recovered large numbers of seaanimals, including crustaceans (i.e. arthropods),echinoderms, and mollusks. Crustaceans arerepresented by crabs (Platyxanthus orbignii;n=26) and echinoderms are represented byseasonally comestible black sea urchins (Tetra-pygus niger; n=4). Most of the corpus consists ofmollusks; almost 20,000 shell fragments wererecovered. More than 75 percent of the sample

came from secondary trash deposits, mainlyconstruction fill.

The analysis of mollusk remains was facili-tated by their good preservation. Rojas was ableto identify genus and species in more thanninety-nine percent of the cases. She usedreference collections and the published litera-ture (Álamo and Valdivieso 1987; Ramírez et al.2003). The quantification procedures favoredMinimum Number of Individuals (MNI) overthe Number of Individual Specimens (NISP)and weight. For bivalves, MNI were evaluatedby separating and counting left and right hinges,and using the higher of the two counts for eachtaxon. For gastropods, MNI were defined by thenumber of univalve apices.

The analysis identified 11,274 MNI and atotal of 53 molluscan taxa pertaining to theclasses Monoplacophora (MNI=1, 0.01 percent),Polyplacophora (MNI=528, 4.7 percent) Gastro-poda (MNI=1,138, 10.1 percent), and Pele-cypoda (bivalves); (MNI=9,551, 84.7 percent).Rojas was unable to identify 55 fragments (0.49percent). The high number of MNI allows for aconfident assessment of the taxonomic composi-tion of the assemblage (see Evans 1972:12). Theassemblage is taxonomically rich, but not verydiverse, with three species, Perumytilus purpura-tus, Semimytilus algosus, and Donax obesulus accounting for eighty percent of the corpus,which hints at a narrow exploitation spectrum.

287 - Research Reports

Temp

.*

Taxon MNI % Sub-

strate

Supra-

littoral

Meso-

littoral

Infra-

littoral

Monoplacophora

Chiton Sp. 528 468 R •

C Polyplachophora

C Neopilina ewingi 1 0.01 R •

Gastropoda

I Fissurella crassa 145 1.29 R •

I Fissurella limbata 182 1.61 R •

I Fissurella maxima 16 0.14 R •

I Fissurella peruviana 11 0.10 R •

I Fissurella sp. 38 0.34 R •

I Scurria viridula 10 0.09 R •

I Prisogaster niger 329 2.92 R • • •

W Hipponix panamensis 1 0.01 S •

W Hipponix pilosus 1 0.01 S •

I Calyptraea trochiformis 1 0.01 R • •

I Polinices uber 2 0.02 S •

CConcholepas concho-

lepas7 0.06 R • •

C Thais chocolata 125 1.11 S • •

I Thais delessertiana 2 0.02 S • •

I Thais haemastoma 49 0.43 S • •

I Thais sp. 1 0.01 S • •

WAncistromesus mexi-

canus44 0.39

I Tegula atra 1 0.01 S • •

I Tegula sp. 1 0.01 S

W Crepidula lessonii 4 0.04 R

I Crepipatella dilatata 11 0.10 R • •

I Crucibulum lignarium 2 0.02 R

I Crucibulum monticulus 4 0.04 R

W Olivella columellaris 1 0.01 S •

W Prunum curtum 15 0.13 S •

W Umbraculum ovale 1 0.01 • •

— Scutalus proteus 128 1.14 L

— Scutalus sp. 6 0.05 L

C Siphonaria lessoni 1 0.01 R • •

Pelecypoda

C Choromytilus chorus 23 0.20 R • •

CPerumytilus purpura-

tus3404 30.19 R • •

C Semimytilus algosus 4000 35.48 R •

I Atrina maura 1 0.01 S • •

I Ostrea megodon 1 0.01 S • •

W Pecten sericeus 1 0.01 S • •

W Argopecten circularis 19 0.17 S •

C Argopecten purpuratus 5 0.04 S •

I Argopecten sp. 19 0.17 S •

I Leptopecten sp. 3 0.03 S • •

I Lyropecten subnodosus 2 0.02 S • •

ITrachycardium pro-

cerum7 0.06 M •

C Mesodesma donacium 386 3.42 S • •

C Mesodesma sp. 1 0.01 S • •

I Donax obesulus 1617 14.34 S • •

I Donax sp. 15 0.13 S • •

W Tagelus peruvianus 6 0.05 M • •

I Semele corrugata 1 0.01 S •

I Semele sp. 17 0.15 S •

W Chione subrugosa 21 0.19 M • •

W Chione peruviana 2 0.02 M • •

Unknown 55 0.49

TOTAL 11274 100.00

*Water temperature: C, cold water; W, warm water; I, indifferent /un-known (i.e., qualifies poorly as a temperature indicator; based on Álamoand Valdivieso 1987; Peña 1970, 1971; Roselló et al. 2001); Substrate: R,rock; S, sand/silt; M, mangrove/muddy; L, land.

Table 2: Molluscan data from the Early Horizon siteof Huambacho. Spondylus sp. not included.

Mollusks have different tolerance ranges toa number of variables and species dwell indifferent environments. Basic distinctions canbe made between (1) soft (e.g. sandy, muddy)and (2) hard (e.g. rocky) substrates, as well asbetween (1) cold and (2) warm water. As seenin Table 2, the Huambacho data indicate theexploitation of various biotopes. Analyses revealthe exploitation of land snails (Sculatus sp.;MNI=134, 1.2 percent), but the bulk of theassemblage is composed of marine taxa. Mostare typical of the cold Humboldt Current andthe Peruvian Province of the Pacific (0 N-40 S).o o

The sample is dominated by two species of smallrock-dwelling mussels: Semimytilus algosus(MNI= 4,000, 35.5 percent) and Perumytiluspurpuratus (MNI=3,404, 30.2 percent). Otherrock-perching species include limpets (Fissurellasp.; MNI= 392, 3.5 percent), Chiton sp. (MNI=528, 4.7 percent), the mussel Choromytiluschorus (MNI= 23, 0.2 percent), as well thegastropod Prisogaster niger (MNI=329, 2.9percent).

In contrast, sand-dwelling species are lesscommon. The most popular is the small tide-zone clam Donax sp. (MNI=1,632, 14.5 per-cent), while the medium-sized sand-dwellingclam Mesodesma sp. (MNI=387, 3.4 percent),the scallop shell Argopecten sp. (MNI=43, 0.4percent) and the gastropod Thais sp.(MNI=177, 1.6 percent) are found in lesserproportions. The collectors that supplied theHuambacho community also exploited, with lessintensity, muddy biotopes as exemplified by thepresence of the mud-flat clam Chione sp.(MNI=23, 0.2 percent), the razor clam Tagelusperuvianus (MNI=6, 0.1 percent), and Trachy-cardium procerum (MNI=7, 0.4 percent). Al-though the Huambacho assemblage is domi-nated by food refuse, data also point towardshells that were worked and transformed intoartifacts, as well as carcasses too small to beedible that were brought to the site as by-prod-ucts (e.g. the small gastropod Olivella collu-

ANDEAN PAST 10 (2012) - 288

mellaris, MNI=1, 0.01 percent). Several exam-ples of perforated Prunum curtum (MNI=15, 0.1percent) have been noted. Also, a total of 115specimens of the thorny oyster Spondylus sp.were collected. Most specimens came fromsurficial and/or disturbed contexts, often inproximity to looted burials associated with post-Early Horizon occupations. Hence, the Spon-dylus data are difficult to interpret and have notbeen included in the taxonomic quantificationprovided in Table 2.

Shellfish exploitation in coastal Nepeña

Taking into account the position of the siteeight kilometers inland and the virtual absenceof fishing implements at the site, it is likely thatthe Huambacho mollusks were acquiredthrough interactions with groups that inhabitedthe littoral. In the Moche Valley, coastal LaLibertad, research suggests that during theInitial Period, the coastal, marine-orientedcommunity of Gramalote was tied to the inlandcomplex of Caballo Muerto (Pozorski 1979:173). Caballo Muerto animal protein was sup-plied by marine and inland resources whileincreasing importance was given to agriculture,including maize cultivation. Gramalote’s subsis-tence economy was also primarily marine-ori-ented and supplemented by inland animalprotein, in particular deer and llama. Furthersouth, Haas and Creamer (2006) have recentlysuggested the existence of similar coast-inlandexchange networks during the PreceramicPeriod in the Norte Chico region.

Based on geographical proximity, the Huam-bacho mollusks were probably collected fromthe southern portion of the Nepeña coastlineand/or the northern portion of the Bahía deTortugas. Exploitation mainly focused on rockycliffs where species of mussels and, to a lesserextent, limpets and gastropods were available.These species are still abundant, especially inthe coastal area near the modern settlement of

Los Chimús, where prehistoric shell middenshave been reported (Thompson 1962:206).Huambacho is only a two hours’ walk from LosChimús. In contrast, the northern portion of theNepeña coastline, as well as the southern por-tion of the Bahía de Samanco, is dominated bysandy beaches of the Nepeña Delta.

From a comparative perspective, the recentrecovery of more than 20,000 MNI of shellremains during the 2009 excavations at theEarly Horizon site of Caylán brings insights intopatterns of marine exploitation in the lowerNepeña Valley (see Chicoine and Ikehara2010). At Caylán, preliminary analyses indicatethe predominance of the sand-dwelling Donaxsp. (c. seventy percent), while rock-dwellingmussels are represented in lesser proportions (c.twenty-five percent). Both the Huambacho andCaylán assemblages are dominated by the samethree species, but the diametrically inversedemphasis of sandy vs. rocky substrates is striking.This observation is consistent with the locationof Caylán, three-and-a-half hours’ walk from theNepeña delta and the Bahía de Samanco, wheresandy environments are ample today.

Available data from Early Horizon sitessuggest that marine resources in Nepeña trav-elled through multiple networks. Indeed, prelim-inary comparisons argue against a single, cen-tralized redistributive or exchange system.Rather, several different, probably independent,coastal providers and satellites have to be hy-pothesized from the Bahía de Tortugas to theBahía de Samanco. Archaeological surveys willbe necessary to find and document these com-munities.

Finally, preliminary comparisons pointtowards multiple exploitation strategies, perhapsboth specialized and unspecialized. As noted byRoselló et al. (2001:77), the sand-dwelling clamDonax sp. is easily collectible by hand in largenumbers. On the basis of their observations of

289 - Research Reports

modern collectors in Huanchaco, Roselló et al.(ibid.) suggest that individual fisherfolks cancollect up to thirty kilograms of clams in threehours of work using rudimentary rakes. Con-versely, the authors suggest that collecting rock-dwelling species is a more perilous enterpriserequiring more specialized practices. From thatstandpoint, Roselló et al. (ibid.) interpret thepredominance of Donax sp. at the Early Interme-diate Period site of Huacas de Moche (c. 80percent of the faunal assemblage) as an indica-tor of the unspecialized exploitation of beachenvironments. In Nepeña, more research will beneeded, but the available data suggest diversestrategies to exploit molluscan resources.

Molluscan bioindicators and paleoenvironment

Marine mollusks are adapted to specificecological parameters (e.g. water temperatures,salinity, oxygen, pH level, availability of ade-quate substrate). As such, they represent poten-tial bioindicators useful for the reconstruction oflocal environmental conditions. In coastal Peru,molluscan data are particularly helpful in dis-cerning climate fluctuations related to theoccurrence of major ENSO events (Díaz andOrtlieb 1993; Rollins et al. 1990; Sandweiss1996). The most salient marine manifestationsof ENSO are related to an increase in watertemperature and changes in salinity, proportionsof dissolved oxygen, and tidal patterns (Díazand Ortlieb 1993:161,164). These changestypically cause indigenous mollusk species tomigrate or disappear. Only a few taxa are foundin greater abundance during ENSO eventsincluding Argopecten purpuratus, and Thaischocolata (ibid.: 165, Table 1).

In addition to the well-known tropicalizationof the normally cold Humboldt environments,ENSO events impact the long-term regenera-tion of certain species. Researchers have sug-gested that Mesodesma donacium, a sand-dwell-ing species of clam, may take up to a decade or

longer to recolonize its habitat after a majorENSO event. In contrast, rock-perching musselslike Perumytilus purpuratus and Semimytilusalgosus are found six months or less following anENSO event (Rollins et al. 1990:471; Sandweisset al. 2001:604).

In order to reconstruct environmentalconditions, archaeomalacologists need to con-sider the impact of cultural preferences, pre-scriptions, and proscriptions on the exploitationof marine taxa. In other words, the archaeologi-cal absence of a species cannot be straightfor-wardly equated with its unavailability. Con-versely, mollusks can be transported greatdistances and their presence at a site does notautomatically represent local conditions (Rollinset al. 1990:470-471). Finally, some speciespreserve better and have a higher probability forarchaeological recovery.

Keeping these considerations in mind, theHuambacho sample is dominated by cold watertaxa typical of the Humboldt Current. With theexception of two species (Ancistromesus mexica-nus, Argopecten circularis), the presence oftropical species is sparse. Indeed, less than onepercent of the total MNI is represented byspecies better adapted to tropical water. Also,some species were clearly imported for theirvalue as adornments and perhaps currencies.Spondylus is a good example. With the excep-tion of a complete shell, Spondylus specimens atHuambacho are all worked into beads andblanks. The same rationale applies to Prunumcurtum specimens found with drilled holes tohang them. Overall, as viewed through themollusk evidence, the Huambacho data doesnot lend weight to the occurrence of a majortropicalization of local oceanic conditions typi-cal of ENSO.

ANDEAN PAST 10 (2012) - 290

Conclusions

Results of the malacological analyses con-firm the importance of marine resources in localEarly Horizon economies and indicate theexploitation of various biotopes. Rocky cliffs andintertidal sandy substrates were particularlyfavored by the Nepeña collectors. Preliminarycomparative data hint at the existence of di-verse strategies of marine exploitation poten-tially tied to independent economic networks.Meanwhile, the predominance of typical Hum-boldt Current taxa lends little weight to theoccurrence of major climatic disruptions duringthe Huambacho occupation.

ACKNOWLEDGEMENTS

Our gratitude goes to the Lima and Chimbote offices ofthe Instituto Nacional de Cultura (now the Ministerio deCultura) for excavation permission and supervision, aswell as to Víctor Pimentel and Jeisen Navarro who co-directed the 2003 and 2004 field seasons, respectively.

REFERENCES CITED

Álamo, Víctor, and Violeta Valdivieso1987 Lista sistemática de moluscos marinos del Perú.

Lima: Instituto del Mar del Perú.Chicoine, David2006 Early Horizon Architecture at Huambacho,

Nepeña Valley, Peru. Journal of Field Archaeology31(2):1-22.

Chicoine, David and Hugo Ikehara2010 Nuevas evidencias sobre el Formativo del valle

de Nepeña: Resultados preliminares de la prime-ra temporada de investigaciones en Caylán.Boletín de Arqueología PUCP 12 (2008):349-370.

Díaz, Amanda and Luc Ortlieb1993 El fenómeno “El Niño” y los moluscos de la costa

peruana. Bulletin de l'Institut Français d'ÉtudesAndines 22(1):159-77.

Evans, John G.1972 Land Snails in Archaeology with Special Reference

to the British Isles. London: Seminar Press.Haas, Jonathan and Winifred Creamer2006 Crucible of Andean Civilization: The Peruvian

Coast from 3000 to 1800 B.C. Current Anthropol-ogy 47(5):745-775.

Pozorski, Shelia1979 Prehistoric Diet and Subsistence of the Moche

Valley, Peru. World Archaeology 2(2):163-184.Pozorski, Shelia and Thomas Pozorski2003 Paleoenvironment at Almejas: Early Exploitation

of Estuarine Fauna on the North Coast of Peru.In El Niño in Peru: Biology and Culture over10,000 Years, edited by Jonathan Haas andMichael O. Dillon, pp. 52-70. Chicago: FieldMuseum of Natural History.

Ramírez, Rina, Carlos Paredes, and José Arenas2003 Moluscos del Perú. Revista de Biología Tropical

(Suppl. 3):225-284.Rollins, H. B., Daniel H. Sandweiss, and Judith C. Rollins1990 Mollusks and Coastal Archaeology: A Review. In

Archaeological Geology of North America, editedby Norman P. Lasca and Jack Donahue, pp. 467-478. Boulder, Colorado: Geological Society ofAmerica.

Roselló, Eufrasia, Víctor Vásquez, Arturo Morales, andTeresa Rosales2001 Marine Resources from an Urban Moche (470-

600 AD) Area in the Huacas del Sol y de laLuna Archaeological Complex (Trujillo, Perú).International Journal of Osteoarchaeology 11(1-2):72-87.

Sandweiss, Daniel H., Kirk A. Maasch, Richard L. Burger, James B. Richardson III, Harold B. Rollins, and AmyClement2001 Variation in Holocene el Niño Frequencies:

Climate Records and Cultural Consequences inAncient Peru. Geology 29(7):603-606.

Thompson, Donald E.1962 Formative Period Architecture in the Casma

Valley, Peru. Proceedings of the 35th InternationalCongress of Americanists 2:205-11.

291 - Research Reports

CHILE

Tarapacá, Norte Grande

Architecture in the Coastal Desert

The Chilean National Fund to SupportScience and Technology (Fondo Nacional deApoyo a la Ciencia y Tecnología [FON-DECYT]) financed projects 1030923 and1080458, directed by Mauricio Uribe (Univer-sidad de Chile, e-mail mur@uchile.cl). Co-investigators Simón Urbina A. (UniversidadAustral de Chile, e-mail: simon.urbina@do-centes.uach.cl), Leonor Adán A. (UniversidadAustral de Chile, e-mail: ladan@uach.cl), andEstefania Vidal M. (Universidad de Chile, e-mail: estefania.vidal.monetro@gmail.com)report on the prehispanic architecture of Tara-pacá.

According to sixteenth and seventeenthcentury chroniclers, the indigenous societies ofTarapacá were equipped with light architecture,or with structures exhibiting the civilizing influ-ence of settlers arriving from the Altiplanoduring the Early Formative Period (900 B.C.-A.D. 200; Núñez 1971; Núñez and Moragas1977). This viewpoint promotes the interpre-tation that the coastal groups were conservativeand materially precarious, and had been pas-sively integrated into the economic and culturalchanges promulgated from interior farming andherding zones (Sanhueza 1985).

In contrast, the study summarized hereemphasizes the social complexity of Tarapacá’scoastal desert (running from approximately19E55' to 21E43' south) by analyzing the archi-tecture of sites located between the mouths ofthe Camiña and Loa Rivers, a north-southstretch of about 208 kilometers. In FONDECYTProject 1030923, “The Pica-Tarapacá Complex:Proposals for an Archaeology of the Societies ofthe South-Central Andes”, and in FONDECYT

Figure 1. Map of northern Chile showing archaeologicalsites mentioned in the text and other Formative Periodsettlements in the Tarapacá region (900 B.C. to A.D.900).

Project 1080458 “The Formative Period inTarapacá: Progress and Social Tragedy in theEvolution and the Early Cultural Complexity ofChile’s Norte Grande, South-Central Andes”,we systematize the variability and change withina coastal architectural tradition, beginning inthe Late Archaic Period (1500-900 B.C.) andmoving forward through time. Finally, we offera current and diachronic appreciation of thesouthern stretch of the western valleys of thesouth-central Andes.

The initiation of the coastal architectural tradition

In the north of Chile, Archaic coastal archi-tecture (9000-900 B.C.) has been defined as“. . . dwellings of light material . . . in whichperishable substances such as vegetable fiber,

ANDEAN PAST 10 (2012) - 292

bones, and marine mammal leather are used”(Muñoz 2002:489).

The earliest known occupations with archi-tecture on the coast of Tarapacá are docu-mented at the Caramucho-3 site, where anextensive shell midden, dating to 4030 B.C.(Morgas 1995) and incorporating fish hooks,cactus remains, and shells, and with collapsedstone structures, has been recorded (Olmos andSanhueza 1984:146). According to Moragas(1995:77), settlements on the interfluvial desertcoast between Iquique and the Loa Valley occurpreferentially in certain sectors near water holesor where the concentration of mists permitshuman life. The time of occupation has beendetermined by radiocarbon and thermo-luminescence dating (Urbina et al. 2011).

Núñez (1965:24) describes an extensiveshell midden at Bajo Molle 1, graves preparedin sand, and a few potsherds, next to dwellingsbuilt of stone (B-1A sector), that could datefrom the final phase of Chinchorro (MomiasRojas, 2850-1850 B.C.; Olmos and Sanhueza1984:144). Probably these stones supportedprops and flexible covers that were easy totransport during annual migrations betweenwater holes and coves.

The Caleta Huelen 42 site, at the mouth ofthe Loa River (Figure 1), constitutes the earliestevidence of more elaborate and stable structureswithin the area of study, with stone houses withmortar incorporating ash and organic remains.The beginning of occupation is dated at 2830B.C. (Zlatar 1983). There are stratified mounddeposits over the houses, similar to those de-scribed south of the Loa River (Schaedel 1957),particularly at the site of Los Bronces 1, south ofTaltal. None of the thirty structures beneath themound exceeds ten square meters of internalspace (Table 1). Stone pillars were used toorganize the interiors and the floors were sealed.

Clusters of structures were organized around central patios (Figure 2).

Site N o.

Enclo-sures

Built-uparea (m )2

Settledarea,

hectares

Density -residences/

hectare

% groundoccupied

Pisagua N 25 432.21 0.090 277.78 48.02

Los Verdes 1 12.25 N/R N/R N/R

Sarmenia 8 66.68 0.294 27.21 2.26

Pabellón de Pica 1 87.55 N/R N/R N/R

Chomache 5 63.07 0.021 238.10 30.03

Punta Blanca 8 78.04 0.125 64.00 6.21

Caleta Huelén Alto 159 1299.82 1.354 117.43 0.95

Caleta Huelén 42 30 70.79 0.328 91.46 4.29

N/R - Not Recorded

Table 1: Main architectural traits of coastal settlements.

Figure 2: Plan of the Caleta Huelén 42 site (after Zlatar1983: Figure 1).

A notable trait of the coastal desert archi-tecture at the end of the Archaic (2010-870B.C.) is the presence of sealed mortar floors ona base of seaweed ash, within which the dead,wearing mud masks, were buried, making “thesehabitations as well as funeral structures” (Llago-stera 1989:70).

The settlements studied at Punta Patacheand Cáñamo show the Archaic-Formativetransition (2010 B.C. to A.D. 325) of theseconstruction elements. At Patache a habitationpattern is defined by shell middens adjacent torock outcrops (for example, at the Patache A,M, N, and P sites). Sometimes there are small,free-standing, circular structures (for example,

293 - Research Reports

at Patache G) with low walls and interior spacesof less than five square meters. At Patache Gthe body of a young adult, covered with a tex-tile, was buried in an extended position, directlybelow the living surface, without a sealed floor.The textile yielded a radiocarbon date of cal.195 B.C. (Urbina et al 2011). The pattern ofcircular structures beneath shell middens, andnear rock outcrops, is represented farther to thenorth, in the southern part of the Sarmenia site.There, two contiguous structures with interiorspaces of 4.2 cubic meters and 4.6 cubic meters, respectively, were recorded.

Consequently, it is established that thecoastal populations of Tarapacá, from the end ofthe Archaic and the beginning of the FormativePeriod, were related to the Chinchorro traditionfound farther north (Arriaza and Standen 2002)as well as to other, more southern, architecturalmanifestations (Schaedel 1957).

Figure 3. Plan of the Punta Blanca site.

Formative Period architecture

Archaeological survey along the Iquique-LoaRiver coastal transect identified a significantcomplex of Formative Period occupations thatpresented elements of both architectural conti-nuity and innovation. Ninety-one Formative

Period settlements were recorded out of a totalnumber of 1,307 sites.

The first architectural form is an expedientarchitecture recorded at the Punta Blancasite(390-620 cal A.D.). There, sub-rectangularstructures of restricted dimensions and simple,adjacent stone walls are irregularly dispersed atrock outcrops or beneath small rock shelters(Figure 3). They lack prepared mortar andpillars with buried foundations and can bedefined as a type of light architecture.

The second architectural form was recordedat the mouth of the Loa River at the boundaryof the second marine terrace and the coastalmountain range. The Caleta Huelén Alto site,characterized by ceramic types dating betweenA.D. 200 and 900 (Uribe 2009), presents dis-persed stone structures of expedient construc-tion on a terraced hillside extending more thana kilometer. There, small parapets built with low stone walls were associated with the site’s nu-merous footpaths. Notable is the absence offood remains and the abundance of lithic debris,copper minerals, and granite hammer stones,occurring in walled work areas, or in openspaces with otherwise clear floors. Included among the habitation units are ovoid structureswith interior spaces of less than five squaremeters, and others with sub-rectangular plansup to fifteen square meters. Caleta Huelén Altocombines construction elements that resemble those at the southern sector of Sarmenia, and atPunta Blanca and the Chomache 1 sites whichexhibit widely-dispersed occupations.

The Chomache 1 site represents a moreagglutinated form of architecture in a laterchronological context (cal. A.D. 650-1290). Anew pattern can be observed. There are fivestructures with rectangular plans, and witharchitectural elements absent in the ArchaicPeriod, such as terracing of the site, and internal

ANDEAN PAST 10 (2012) - 294

division of spaces by means of partitions orsubstructures (Figure 4).

Figure 4. Plan of the Chomache 1 site. Substructure 5Adoes not appear in this plan.

Located next to a large rock outcrop, is astructure (Edificio 5) with slightly curved cor-ners and pillars with stone foundations, all traitspresent in earlier buildings. A more compactresidential unit with a greater investment oflabor is evident.

The architecture of Chomache 1 representsa transformation in the way the coastal area ofTarapacá was occupied at this time. Similar sitesare Caleta Huelén Alto and Punta Blanca, aswell as a more northern site, Pisagua N. Thislatter site was occupied from the Late FormativePeriod until the Late Intermediate Period orfrom A.D. 290 to 1470 (Urbina et al. 2011). Thelatest ground plan consisted of 25 agglutinatedstone structures with square, rectangular, andtrapezoidal plans. These structures were con-structed over time on terraced surfaces (Figure5). Excavations revealed rows of reeds that wereincorporated into the stone walls, to constructfaçades of organic material (quincha).

At Pisagua N, as at Chomache 1, domesticspaces were made sacred by means of foundationofferings such as the burial of a dog’s head next

to a wall in Pisagua N’s Edificio 19, and that ofa marine mammal cranium in the northeasterncorner of Structure 1 at Chomache 1. Both sitesachieve similar levels of agglutination (Table 1),and both are adjacent to a rocky shore lackingnearby fresh water. Because of this it was neces-sary to develop techniques for condensing thecoastal fog (camanchaca).

Figure 5. Plan of the Pisagua N site.

The architecture of Punta Blanca, Cho-mache 1 and Caleta Huelén Alto reveals a formcharacteristic of the desert coast south of Iqui-que and Tarapacá, where domestic environ-ments are now separate from funerary spaces(Llagostera 1989). On the other hand, the re-occupation of Chomache 1 and Pisagua Nindicate that a pattern with an agglutinativetendency was current until the Late Intermedi-ate Period (A.D. 900-1450). This architecturehad greater functional variability than that ofearlier sites, to judge from the sizes and forms ofits structures.

Late coastal dwellings

Among the Late Intermediate Period sitesincluding Chomache 1 and Pisagua N are otherforms of residences. One is represented byStructure 1 at the Pabellón de Pica-1 site (Fig-ure 6), dated between cal A.D.1280 and 1430.This is a large sub-circular area (85 squaremeters) made of curved, double-faced stone andmud walls more than 50 centimeters thick and

295 - Research Reports

laid in courses, features that resemble those ofthe Caserones-1 site. The unusual height of Structure 1’s walls, greater than 1.5 meters, isremarkable, as is the high proportion of mortarin its facing. In accordance with its formalcharacteristics, position, and visibility, it mayhave been a communal and public space nearresidential spaces.

In addition, three settlements built on shellmiddens in an otherwise flat area, with Archaicand Formative Period evidence on their periph-eries, or directly on top of them, were recorded.At Los Verdes-1 (cal A.D. 1160-1280), Struc-ture 1 is a free-standing building with a rectan-gular plan and double stone walls laid in coursesand joined with ashy clay mortar. We recordedthe same sort of building in the Northern Sectorof Sarmenia at Structures 6 and 7 (Figure 7).This differed from the earlier Southern Sectorwhere we recorded two structures separated bytwenty-nine meters, one of which had been en-larged to form a conglomeration with exteriorstructures (Structures 6, 6A, and 6B).

Figure 6. Pabellón de Pica site, Structure 1.

The placement of these disperse structureswith rectangular and quadrangular plans on flatground with shell deposits also occurred at themouth of the Tana/Camiña Quebrada at the site

of Pisagua B (Figure 8) with dates ranging fromA.D. 1496 to 1556 (Urbina et al, 2011).

Figure 7. Sarmenia site, Structure 5 in the northernsector.

Figure 8. Plan of the Pisagua B and Pisagua Viejo sites(after Núñez 1970: figure 1).

At present we can state that during theLate Periods there were settlements with dis-persed structures built according to a rectangu-lar plan, with stone foundations and wattle-and-daub (quincha) walls, as well as others withhigher stone walls. These latter buildings give anidea of individual houses that were absent in theFormative Period. This format had a correlate atthe Spanish Colonial Period indigenous site ofPisagua Viejo (Núnez 1970:53).

ANDEAN PAST 10 (2012) - 296

Conclusions

The sites analyzed in this report indicatesocial transformations made manifest as agreater diversity and complexity in architecturalexpression. These show a change in the occupa-tion of space from the end of the Archaic Perioduntil the Formative Period, in which an increasein size took place, as well as a change to morepermanent houses, while structures with sealedfloors were abandoned. During the FormativePeriod architecture permitted the developmentof productive activities including the obtentionof marine resources, the collection of fertilizersfor agriculture, and the processing of minerals.It facilitated movement along the coast, forexample, to places with more stable waterresources. Some sites began to become morecompact and internally subdivided, while otherindigenous camps continued to function astemporary accommodation with more disperseand expedient constructions. Until the LateHorizon and Colonial Period, sites dating fromthe Formative Period continued to be reoccu-pied, while new ones were founded with struc-tures that incorporated quincha, and individualmodular stone houses of great solidity.

This report reinforces the idea that “. . .architecture, as a technological innovation,appears to be promoted in social contexts rele-vant to the communities” (Adán and Urbina2007:26). Equally, the data contradict the ideagenerally accepted in Northern Chilean archae-ology that “. . . societies with a long tradition ofmaritime adaptation persisted without largevariations from stages earlier than Chinchorro”(Sanhueza 1985:57). Based on the evidence ofthe distinct types of architecture found on theTarapacá coast, at least, the strictly conservativebehavior of the coastal populations, described inan essentialist manner by most Chilean archae-ologists, dissolves. Instead there seems to havebeen processes of transformation and social and

cultural innovation that were the products oftheir own dynamic histories.

Translation from the Spanish by Monica Barnes

with the assistance of Simón Urbina

REFERENCES CITED

Adán, Leonor and Simón Urbina2007 Arquitectura formativa de San Pedro de Ata-

cama. Estudios Atacameños 34:7-30.Arriaza T., Bernardo and Vivian Standen G.

2002 Momias, muerte y ritos ancestrales. Arica, Chile:Ediciones Universidad de Tarapacá.

Llagostera M., Agustín1989 Caza y pesca marítima (9000 a 1000 AC). In

Culturas de Chile: Prehistoria; desde sus orígeneshasta los albores de la conquista. Edited by JorgeHidalgo, Vergilio Schiappacasse, Hans Nie-meyer, Carlos Aldunate, and Iván Solimano, pp.57-80. Santiago de Chile: Editorial Andrés Bello.

Moragas W., Cora1995 Desarrollo de las comunidades prehispánicas del

litoral Iquique-desembocadura río Loa. Actas delXII Congreso de Arqueología Chilena. Hombre yDesierto 9, Volume 1:65-80. Antofagasta, Chile

Muñoz O., Iván2002 En torno a la formación del proceso aldeano en

los valles desérticos del norte de Chile: El casodel valle de Azapa. Doctoral dissertation,Universidad Nacional Autónoma de México,México, Distrito Federal, México.

Núñez A., Lautaro1965 Prospección arqueológica al Norte de Chile.

Estudios Arqueológicos 1:9-36.1970 Descubrimiento de la aldea española: “Pisagua

Viejo” (Provincia de Tarapacá. Ancora 52-58.1971 Secuencia y cambio en los asentamientos hu-

manos de la desembocadura del río Loa en elNorte de Chile. Boletín de la Universidad de Chile112:3-25.

Núñez A., Lautaro and Cora Moragas1977 Una ocupación con cerámica temprana en la

secuencia del distrito de Cáñamo (costa desér-tica del norte de Chile). Estudios Atacameños5:21-49.

Olmos F., Olaff and Julio Sanhueza T.1984 El precerámico en la Costa-Sur de Iquique.

Chungara 13:143-154.

297 - Research Reports

Sanhueza T., Julio1985 Poblaciones tardías en playa Los Verdes, costa

sur de Iquique, I Región, Chile. Chungara 14:45-60.

Schaedel, Richard P.1957 Arqueología Chilena: Contribuciones al estudio

de la región comprendida entre Arica y LaSerena. Centros de Estudios Arqueológicos,Universidad de Chile 2. Santiago de Chile.

Urbina A., Simón, Leonor Adán A., Cora Moragas W.,Sebastián Olmos T., and Rolando Ajata L.2011 Arquitectura y asentamientos formativos en la

costa desértica de Tarapacá, Norte de Chile.Estudios Atacameños 41:21-34.

Uribe R., Mauricio2009 El período Formativo de Tarapacá y su cerámica:

Avances sobre complejidad social en la Costa delNorte Grande de Chile (900 AC-800 DC).Estudios Atacameños 37: 5-27.

Zlatar, Vjerar M.1983 Replanteamiento sobre el problema Caleta

Huelen 42. Chungara 10:21-28.

ARGENTINA

La Bolsa 1 Site, Tafí Valley, Tucumán Prov-ince: Household Mortuary Practices in aSouth Andean Village (First MillenniumA.D.)

Julián Salazar (Centro de Estudios HistóricosSegreti, Concejo Nacional de InvestigacionesCientíficas y Técnicas (CONICET), Uni-versidad Nacional de Córdoba, e-mail: jjsalba@hotmail.com) presents new data about SouthAndean mortuary practices and the construc-tion of kin-based relationships. The data haveemerged from recent field research on EarlyCeramic Period (c. 500 B.C. to A.D. 850)contexts from the Tafí Valley.

Figure 1: Map of Tucumán Province, showing thelocation of the Tafi Valley.

This small valley lies between the orographicsystems of the Sierras de Aconquija and theCumbres Calchaquíes, and is east of the Yoca-

ANDEAN PAST 10 (2012) - 298

huil Valley, at an average attitude of 2500m.a.s.l. (Figure 1). Throughout the first millen-nium A.D., its inhabitants were sedentaryvillagers who practiced agriculture based onmaize, potatoes, and beans and who herdedllamas. Intensive food production permitteddemographic growth, seen in the proliferation ofvillage settlements, and an increase in socialcomplexity, the latter evident in ceremonialmounds and monolithic carved sculpture (Ber-berián and Nielsen 1988; González and Núñez1962). Nevertheless, this population growth andincrease in complexity appear to have beenbased on autonomous households rather thanon political centralization.

My research aims to establish relationshipsbetween daily practices and social reproductionstrategies and household strategies, and toanalyze the active role of household materialassemblages in the construction of kinshiprelations and identity. Consequently, I carriedout field-work in residential compounds at theLa Bolsa 1 site.

Ancestor Worship in Household Compound LB1-U14, La Bolsa 1 site

La Bolsa 1 is a concentrated village site inthe north of the Tafí Valley. La Bolsa 1 wasinhabited between 2100 B.C. and 850 A.D. Thissettlement includes twenty-one householdcompounds and a twenty-five hectare complexsystem of agricultural structures. The site layoutshows spontaneous expansion rather thanformal community planning. Household com-pounds are spatially segregated architecturalunits of about 200 square meters that incorpo-rate both food storage facilities and craft pro-duction workshops and show strong traces of kinidentity. The dwellings include circular struc-tures around open patios, with walls built oflarge fieldstones (Figure 2).

Figure 2: Plan of the central area of the La Bolsa 1 site,showing some of the architectural compounds.

I excavated two household compounds,LB1-U10 and LB1-U14, and diverse outdoorcontexts in the central area of La Bolsa 1. HereI analyze the burial context recovered in theLB1-U14 residential unit.

LB1-U14 is a 200 square meter residentialunit. It includes a circular patio, R1, four smallerenclosures, R2, R3, R4, and R6 linked to itthrough formal doors, and two peripheral struc-tures entered through R4 and R6 (Figure 3).

The compound’s plan is that of a recurrentpattern, seen contemporaneously in othergeographical settings near Tafí, such as theYocavil Valley and the Cajón Valley (Ascheroand Ribotta 2007; Berberián and Nielsen 1988;Scattolin 2007). Centripetal domestic spaces arearranged around a central patio, the only struc-ture with direct access to the exterior. Thiscentral place is a circular unroofed enclosure ofeighty square meters which controls movementwithin the unit and mediates access to theroofed structures of the house. I excavated theresidential compound completely and recordedevidence of all the construction, depositional,and post-depositional events that left materialtraces.

299 - Research Reports

Figure 3. Plan of the LB1-U14 residential compound.

Analysis of architectural features and arti-factual assemblages, soil chemistry studies, andphytolith identification allowed the recognitionof activity areas in various parts of the residen-tial compound. The occupation floors of five ofthe seven structures were dated from cal AD650 to cal AD 850, corresponding to the lastoccupation of the compound before its aban-donment.

The small peripheral enclosures show evi-dence of specific activities, especially foodprocessing and cooking. In Structure R6 a fewcooking vessels were broken around a centralstone hearth. Maize (Zea mays) phytoliths wereidentified from within the soil samples takenfrom this area. Other activities detected are toolstorage and sleeping and resting.

Diverse activities were detected in thecentral and biggest structure, R1. These include maize grinding, food storage within a specialroom, and ritual deposition of ceramic animalrepresentations. Non-locally manufactured pottery fragments were present. In the centralportion of this open patio, an oval feature was found. It was an underground, bell-shaped,stone-walled chamber, with a false dome closureprotruding thirty centimeters above the occupa-tional floor. This structure,“Cista 1”, is a cist, a

typical tomb type in Tafí Valley Early CeramicPeriod sites.

At first glance, the cist could be seen as asingle feature interpretable as simply evidenceof a single household’s veneration of an ancestor because of the power, wealth, or kinship rela-tions he or she accumulated during life.

Indeed, this is the regular interpretationmade of other, similar contexts. Nevertheless,the excavation of this burial showed that it was the result of complex depositional processes(Figure 4).

Figure 4: “Cista 1” section.

Just after the cist closure was removed,about thirty centimeters below the patio occu-pational floor, a small figurine was found. It is ananthropomorphic stone statuette of a woman.

ANDEAN PAST 10 (2012) - 300

Her face has carved lines descending from hereyes which could represent tears, face paint,tattoos, or scarification (Figure 4, top). Thelower portion of this artifact has a flake scar,suggesting that it was intentionally broken or“killed” before deposition.

Figure 5: Schematic plan of Level II burial.

Next to the Level II base (80 centimetersbelow the occupation floor) I found humanbones, in a very bad state of preservation, associ-ated with a gray pottery bowl (Figure 5). Identi-fiable bone fragments were two skull pieces andtwelve teeth. Studies of tooth wear traces androots allowed us to establish that this person wasan adult. Hundreds of unidentified bone frag-ments were recovered from the central area ofthe burial. Within these maize (Zea mays) andcucurbit (Cucurbita sp.) phytoliths were identi-fied. These could be interpreted as the remainsof the deceased’s final meal, or as part of thegrave goods accompanying the dead. The ce-ramic bowl, buried horizontally, is a fine, un-decorated pottery vessel which could betypologically dated to ca. A.D. 500 to A.D. 900.

Below this level, a thin layer of burnt claycovered the entire structure. After its removal Idetected a more compact stratum, Level III,where traces of another burial were present(Figure 6). It was composed of an adult’sosteological remains, in a very bad state ofpreservation, associated with three coarse fabricvessels. The human remains were fragments ofa skull and a jawbone in the western margin ofthe base of the cist, and hundreds of unidenti-fied little pieces of bone. The skeletal remainswere accompanied by a cup (made of ordinaryred clay, with a rough surface finish, a lip handlevertically attached, and a modeled anthropo-morphic face decoration applied on the sideopposite the handle; Figure 6 top), a jar (tech-nologically similar, with a uniform surface finish,no decoration, and a thick soot layer on theouter surface), and numerous fragments ofvessels with similar characteristics. None of thepottery presents complex decoration, nor corre-sponds to what is known as “fine” craftsmanshipfor this period. Charred wood was recoveredfrom this level and dated to 1799±37 BP, cal.AD 130-260, the earliest date for LB1-U14.

The disposition of material indicated anintentional disturbance of the earliest burial,before depositing the final one. Bone remainswere disposed next to the structure’s wall andthe ceramic jar was between the jaw and theskull.

Archaeological evidence recovered from theburial structure of household compound LB1-U14 allows me to propose that this context wascreated by many events of digging, opening,depositing, firing, closing, and burying, forminga material palimpsest. Although palimpsests arecommon archaeological contexts, they aretypically viewed as a handicap.

301 - Research Reports

Figure 6: Level III burial.

On the contrary, I emphasize the informa-tion provided by this particular mixture (Bailey2007; Lucas 2005; Olivier 1999). It demon-strates that ancestral corpses were not sociallydead: they were continuously taken from theirtombs, exhibited, worshiped, fed, and reburied.In addition, this mixture indicates that theywere part of daily activities. The material config-uration of the burial feature generated a perma-nent interaction with the living. It was locatedin the center of the house, in a place of neces-sary passage, the locus where vital practices suchas maize grinding and food storage were carried

out. A major part of quotidian life took placethere.

Gillespie (2001) states that the constructionof social persons derives from acting within acontext, as part of experience and daily prac-tices. This includes relations between people,dead and alive, groups, and material objects(Brown and Walker 2008). The mortuary pa-limpsest analyzed in LB1-U14 is precisely thematerialization of practices that create andstrengthen kinship relations as living people interact with their ancestors’ physical remains.These practices regarding death have a longduration, from the first centuries A.D. to ca.A.D. 850. They formed the relationships be-tween house dwellers and ancestor, which builtthe foundations of the household competitivegroup identity (Dillehay 1995).

The construction of various kin groups withstrong internal identity is a possible explanationfor the formation of a complex society with a lack of political centralization, a process thatwas characteristic of the first millennium in thestudy area.

REFERENCES CITED

Aschero, Carlos and Eduardo Ribotta 2007 Usos del espacio, tiempo y funebria en El Remate

(Los Zazos, Amaicha del Valle, Tucumán). InPaisajes y procesos sociales en Tafí: Una miradainterdisciplinaria desde el Valle (Tucumán, Argen-tina), edited by Patricia Arenas, Bárbara Manasse, and Estela Noli, pp. 79-94. Tucumán,Argentina: Universidad Nacional de Catamarca.

Bailey, Geoff2007 Time Perspectives, Palimpsests and the Archae-

ology of Time. Journal of Anthropological Archae-ology 26(2):198-223.

Berberián, Eduardo E. and Axel Nielsen 1988 Sistemas de asentamiento prehispánicos en el

Valle de Tafí. In Sistemas de asentamientos prehis-pánicos en el valle de Tafí, edited by Eduardo E.Berberián, pp. 20-52. Córdoba, Argentina:Editorial Comechingonia.

ANDEAN PAST 10 (2012) - 302

Brown, Linda A. and William H. Walker 2008 Prologue: Archaeology, Animism and Non-

Human Agents. Journal of Archaeological Methodand Theory 15:297-299.

Dillehay, Tom D.1995 Introduction. In Tombs for The Living: Andean

Mortuary Practices, edited by Tom D. Dillehay,pp. 1-27. Washington, D.C.: Dumbarton Oaks.

Gillespie, Susan D. 2001 Personhood, Agency and Mortuary Ritual: A

Case Study from the Ancient Maya. Journal ofAnthropological Archaeology 20:73-112.

González, Alberto Rex and Víctor Núñez Regueiro1962 Preliminary Report on Archaeological Research

in Tafí del Valle, N.W. Argentina. Akten des 34Internationalen Amerikanisten Kongresses: Wien,18-25. Juli 1960, pp. 485-496. Horn, Austria:F.Berger.

Lucas, Gavin2005 The Archaeology of Time. London and New York:

Routledge.

Olivier, Laurent1999 The Hochdorf ‘Princely’ Grave and the Question

of the Nature of Archaeological Funerary As-semblages. In Time and Archaeology, edited byTim Murray, pp. 109-138. London and NewYork: Routledge.

Scattolin, C.2007 Santa María antes del año mil: Fechas y ma-

teriales para una historia cultural. In Sociedadesprecolombinas surandinas: Temporalidad, inter-acción y dinámica cultural del NOA en el ámbito delos Andes centro-sur; Taller Internacional deArqueología del NOA y Andes Centro Sur, editedby Veronica Williams, Beatriz Ventura, AdrianaCallegari, and Hugo Yacobaccio, pp. 203-220.Buenos Aires: Artes Gráficos Buschi.