Skip to Content

Rescue Excavations at the Site of Tumilaca la Chimba, 2006

Patrick Ryan Williams, The Field Museum, Chicago; Michael E. Moseley, University of Florida; Maria C. Lozada, University of Chicago; Nicola Sharratt, University of Illinois at Chicago, Project Grant 2006–2007


Fig. 1: The Moquegua Valley.
Fig. 1: The Moquegua Valley.

The collapse of the Tiwanaku state is a subject that has commanded significant interest in recent years (Bermann et al 1989; Goldstein 2005; Kolata 1993; Owen 2005; Sims 2006; Williams 2002). Goldstein and Owen have both argued that Tiwanaku expansion represents a folk dispersion of altiplano peoples across the south-central Andes (Goldstein 2005; Owen 2005). Owen (2005) argues that the collapse of the highland state led to a second, composed of Tiwanaku refugees fleeing from the centers of state control. The Moquegua Valley, located in southern Peru 300 km from the site of Tiwanaku, was one of the most important centers of Tiwanaku culture (Fig. 1). In Moquegua, the late manifestation of Tiwanaku (post 1000 AD) has been termed Tumilaca, after the type site of Tumilaca la Chimba (Bawden 1989). Despite considerable research on settlement patterns, architecture and material culture from the Tumilaca phase (Bawden 1989; Goldstein 1989; Owen 1993, 1996; Sims 2006; Stanish 1985), questions still remain about the cultural genesis of Tumilaca and the biological relatedness of Tumilaca populations to antecedent Tiwanaku populations. In this paper, we suggest that mortuary data is central to ongoing understandings of who the inhabitants of Tumilaca sites were. Following a brief summary of current interpretations of the period, we present results from excavations undertaken in the cemetery sector of Tumilaca la Chimba.

Tumilaca Phase in Moquegua

The end of Tiwanaku influence across the south central Andes was accompanied by the destruction of monumental architecture and the cessation of particular icons of the state, most notably the staff god image (Janusek 2004). Changes in material culture and shifts in settlement patterns in both the altiplano and more distant colonies followed (Berman et al 1989; Goldstein 2005; Janusek 2005). In Moquegua, Tumilaca ceramics show greater variation from altiplano types in both form and decoration than earlier Tiwanaku vessels. Iconography is more abstract than during the earlier phases, with greater use of geometric designs (Goldstein 1985; 1989; 2005). Recent analysis of surface material from Tumilaca sites links this material to Late Tiwanaku V material from the altiplano (Janusek 2003).

Settlements with Tumilaca style ceramics are also more variable than their classic Tiwanaku counterparts and are found in more dispersed ecological zones than earlier Tiwanaku settlements. Goldstein (1985) ascertained Tumilaca presence in the area of previous Tiwanaku occupation of the middle valley (1000–1500 masl), specifically at the site Omo M11, but Owen (1993; 1996) also documents several late Tiwanaku settlements in the coastal Moquegua valleys as well as in the upper valley reaches (2000–2500 meters above sea level), and Stanish (1985) identified a pioneering Tumilaca settlement in the Otora valley. Settlements are typically located in less accessible locations than the classic period Tiwanaku sites. Architectural changes between the phases are apparent. Evidence from Tumilaca la Chimba indicates that inhabitants of Tumilaca phase sites maintained classic Tiwanaku residential practices, continuing to construct residential units with internal spatial differentiation and behavioral separation (Bawden 1989). However, while the classic period Omo complex is known for its temple (Goldstein 1989), Tumilaca sites are notable for their absence of non-domestic architecture (Bawden 1989).

Mortuary data for Tumilaca

Despite these advances in understandings of Tumilaca settlements and material culture, there is a paucity of excavated mortuary data. To date, excavated burials for the phase include some twenty to thirty excavated by Owen (1993) in the coastal Osmore drainage, as well as fourteen excavated at Tumilaca la Chimba by Pari (1980). Skeletal remains from the latter excavations are unable for study. The relative absence of Tumilaca mortuary remains contrasts with the extensive mortuary collections recovered for the period of classic Tiwanaku in Moquegua which count over 4000 tomb contexts and over 400 intact interments (Owen 1997; Vargas 1988). The skeletal and cultural material excavated from Chen Chen provides evidence that populations at Chen Chen were immigrants from the alitplano, who were both biologically related to altiplano populations and who expressed cultural affiliation with the Tiwanaku capital (Blom 1999; Blom et al 1998; Knudson 2004). Mortuary data for the Tumilaca phase will prove equally important in explaining who these later populations were. In particular, comparisons with the biological and cultural data from the Chen Chen cemeteries will illuminate whether these populations continued to identify themselves as Tiwanaku in the later years of Tiwanaku influence.

We propose several hypotheses regarding the origins of these late Tiwanaku peoples:

  1. Tumilaca peoples are actually the biological descendants of the pre-Tiwanaku inhabitants of the coastal valleys, and that these peoples adopted Tiwanaku pottery and some other material culture, but were rooted in the original Formative populations to occupy these valleys.
  2. Tumilaca populations are the direct descendents of altiplano peoples, as were there Chen Chen predecessors. A subsidiary hypothesis is that Tumilaca were actually descended from the populations interred at Chen Chen. This hypothesis corresponds most closely with Owen's suggestion of a second Tiwanaku diaspora.
  3. The Tumilaca are of mixed descent that may include Tiwanaku altiplano, local Wari groups, and coastal predecessors.

In effort to evaluate these hypotheses, excavations are underway in the Tumilaca phase cemetery at the site of Tumilaca la Chimba.

The Tumilaca la Chimba Site

Fig. 2: The site of Tumilaca la Chimba.
Fig. 2: The site of Tumilaca la Chimba.

The Tumilaca la Chimba site lies about 15km up-valley from the modern city of Moquegua. In the 1980s, Programa Contisuyo researchers defined a large cemetery associated with a defended town on a ridge above the Tumilaca River and identified late Tiwanaku style material (Bawden 1989) (Fig. 2). Overlying part of the Tumilaca town are later Estuquiña (AD 1200–1500) constructions. The cemetery extends down both the east and west slopes of the ridge, covering an area of almost 3300 square meters. A conservative estimate, based on the excavated sample, suggests that there are as many as three hundred burials in the western slope cemetery alone. Tombs are arranged on artificial terraces, likely a mechanism for coping with the extreme gradient of the slope. Mortuary contexts at the site have been badly disturbed by modern looting; the surface is littered with ceramic fragments and, to a lesser degree, human bone.

Based on visible tomb architecture, the cemetery has been separated into four sectors; three on the eastern slope of the ridge, and one on the western slope. Future excavations will determine whether the three sectors on the eastern slope are in fact separate cemeteries. The excavations discussed in this paper were undertaken on the western slope, where the cemetery covers almost 1130 square meters. During the 2006 season, we excavated 120 square meters, essentially cutting a transect from the upper to lower areas of the ridge, and encountered approximately one tomb per four square meters. A total of twenty-seven tombs were excavated, and despite the looting, the excavations furnished interesting data. Looting had largely been restricted to cultural materials. Six (twenty-two percent) excavated tombs were determined to be intact, and even those which had been damaged revealed information on grave architecture and body position. Laboratory analysis of both cultural and biological materials is ongoing. However, observations in the field allow us to comment upon several aspects of mortuary behavior at Tumilaca la Chimba. In this paper we deal specifically with grave architecture, interment treatment, and grave goods, and compare the evidence with that from earlier and later cemeteries in Moquegua.


Grave Architecture

Considerable variation in grave architecture is evident. All excavated tombs were roughly cylindrical in shape. All were subterranean, reaching an average depth of one meter, although many had above ground markers. Tombs were classified into three categories, based on wall construction; eleven were stone lined (tumba) (Fig. 3), ten were partially stone lined (hoyo con alguna construcion de piedras) (Fig. 4), and six were unlined (hoyo sin revestimiento de hoyos) (Fig. 5). There appears to be no spatial variation in tomb types, as variation in wall construction was evident throughout the western slope, and tombs of differing wall construction are found in association with one another.

Fig. 3: Stone-lined tomb.
Fig. 3: Stone-lined tomb.
Fig. 4: Partially stone-lined tomb.
Fig. 4: Partially stone-lined tomb.
Fig. 5: Unlined tomb.
Fig. 5: Unlined tomb.

Fig. 6. Outer rings surrounding tombs
Fig. 6. Outer rings surrounding tombs

Fifteen tombs were further distinguished by an outer ring (Fig. 6). These encircle the subterranean structure and reach a height of approximately 0.5m. Outer rings are not exclusively associated with one type of wall construction, as examples of outer rings were found with all three tomb types. However, there does seem to be a spatial component to the distribution of outer rings, as outer rings were far more common in the middle or the bottom of the slope. The only example found in the upper slope was unusual in that it encircled two separate juvenile burials, whereas in all other cases the ring enclosed a single tomb.

The capstones of many tombs had been removed. Where present, they were often composed of several stones, rather than just one large slap. Floors were untreated in all but two cases. In these two cases, both infant burials, the individual had been placed in a seated position on a flat rock.

Variation in tomb structure not only cross-cuts space, but also age at death. Adults, juveniles and infants were all found in the three different tomb categories, and outer rings were associated with both adults and children. As skeletal analysis is pending, it has not yet been determined if grave architecture corresponds with other variables, such as biological sex or cranial modification.

Grave architecture from Tumilaca la Chimba makes for an interesting comparison with that encountered at both the Chen Chen site and Late Intermediate sites in the Moquegua valley. At Chen Chen, Owen (1997) reports cists – cylindrical tombs lined with stones, a few of which had a stone placed on the floor, and hoyos – also cylindrical tombs but with no stone lining. Both of these tomb types (in addition to partially lined tombs) were identified at Tumilaca la Chimba. Owen does not report outer rings. However, outer rings are mentioned by Williams et al. (1989) for the Late Intermediate Period site of Estuquiña. Both those discussed by Williams et al and those at Tumilaca la Chimba differ from the collared tombs reported by Stanish (1985).

Three types of tomb are evident at Tumilaca la Chimba, although all three forms have the same rough cylindrical shape, and all are subterranean. There is no difference in the spatial distribution of tomb types, but there does appear to be a clear spatial distribution of outer rings. The question of spatial distribution will be further explored when excavations are extended into the eastern slope of the cemetery. Examination of construction phases at Tumilaca la Chimba may elucidate whether the different tomb types represent a temporal shift, rather than act as indicators of individual status or identity.Elements of classic period Tiwanaku tomb construction clearly continued into the Tumilaca phase, but there is also evidence for the introduction of new forms, which arguably continued into the Late Intermediate Period in Moquegua.

Interment Treatment

Excellent preservation of original body position was encountered, even in most of the looted tombs. All burials were of single individuals. The 2006 excavations revealed no multiple burials. In all cases in which body position could be determined, individuals were in a flexed position. Nearly all individuals were seated, although two were lying on their right sides (CB06-46-0075 and CB06-47-0081). Given the shape of the tomb and evidence for decomposition, we are confident that at least one of these individuals was interred lying on their side. All individuals, without exception, were facing towards the east or, in two examples, south east. This pattern of interment positioning illustrates strong continuities with the classic Tiwanaku cemeteries, where individuals are also interred in a flexed seated position facing east (Goldstein, personal communication).

Although conditions in the upper valley mean that preservation is not as good as at middle valley sites, such as Chen Chen, there is evidence that individuals were buried with woolen textiles and likely wrapped in fiber rope. Fragments of brown woven woolen cloth were found in four tombs. Braided fiber rope was in two contexts. In one of these (CB06-47-0081), the rope appeared to have been wrapped around the left arm, in the other (CB06-47-0078) the rope was by the top of the vertebral column, as well as by the arms and feet. Given that traces of rope have been found around the arms and shoulders of skeletons, we suggest that the body was held in a flexed position by the rope. This would mirror practices at Chen Chen where better preservation has revealed seated, flexed individuals wrapped in woven cloth and held in position by braided fiber rope (Owen 1997). Interment treatment practices are consistent across the western cemetery areas, again investigation into the eastern areas will illuminate whether they were standard for the entire cemetery at Tumilaca la Chimba. Further, interment treatment practices appear to have been maintained from the classic Tiwanaku period into the Tumilaca phase.

Grave Goods

Looting at the Tumilaca la Chimba cemetery has affected cultural materials. Most ceramics retrieved during excavations were recovered through surface collection. Forms included keros, tazons, and jars. A fragment of an elaborately decorated painted zoomorphic incensario was found lying on the surface. Four complete vessels were recovered from intact contexts (two keros, a tazon, and a jar). Ceramics include both plain and decorated shards. Decorated ware was red-slipped with black, white and orange paints used in decoration. Geometric designs predominated, but one kero (CB06-47-0026) interred in an intact infant burial was decorated with a stylized trophy head. The material recovered here is comparable to that found by Pari (1980) in mortuary contexts on the eastern slope of the ridge. It is evident that the practice of including ceramic vessels in graves cross-cut the cemetery, and ceramics were associated with adults and juveniles, as well as with all three types of tomb. Offerings were not ubiquitous, however. Four intact contexts included no ceramic offerings at all.

At Chen Chen, individuals were typically interred with one or two ceramic vessels, sometimes a wooden spoon, and occasionally other types of material (Owen 1997). At Tumilaca la Chimba, there was evidence for wooden fragments in six of the tombs excavated in 2006, and it is reasonable to suggest that these were also spoons. Preservation issues make it likely that they were included in other graves but did not survive, particularly given the disturbed nature of many contexts. As at Chen Chen, there is evidence that spoons were placed inside ceramic vessels.

Summary and Discussion

The large cemetery at Tumilaca la Chimba has great potential for informing investigations into late Tiwanaku populations in Moquegua. Although considerable evidence has been gathered on the settlement patterns and material culture of the Tumilaca phase, still lacking is a clear conception of who these people were biologically, and to what extent they continued to express an affiliation with earlier Tiwanaku populations. Mortuary data has proved instrumental to understandings of classic period Tiwanaku in Moquegua (Blom et al 1998). Biological and cultural data from Tumilaca la Chimba can similarly help answer questions about both the biological ancestry of late Tiwanaku populations and the social identity of those populations. Such investigations are central to ongoing understandings of the implications of state collapse on regions that had been subsumed within the Tiwanaku polity.

Analysis of the skeletal materials recovered during 2006 is pending, and we are not yet in a position to comment upon the biological relatedness of those interred at Tumilaca la Chimba with earlier Tiwanaku populations. Further, thorough systematic analysis of cultural materials has yet to be done. However, we do suggest that mortuary behavior at Tumilaca la Chimba was a combination of both tradition and innovation. Significant continuity from the classic period, as indicated by comparison with Chen Chen, is evident in interment treatment of individuals. As in the classic phase, individuals were buried in a seated, flexed position facing east, arguably wrapped in similar textiles and bound by braided fiber rope. Notions about appropriate grave goods were maintained – at least some individuals were accompanied by one or two ceramic vessels, and possibly wooden artifacts. However, mortuary behavior at Tumilaca la Chimba was not simply a seamless continuation of earlier practices. The ringed tombs suggest changes in, at very least, grave architecture. These are especially interesting, given their apparent similarity to those found at the Late Intermediate Period cemetery at Estuquiña (Williams et al 1989). Ongoing investigations into the relative chronology of different tomb forms, particularly the outer rings, will contribute to a clearer understanding of grave differentiation at the site.

As the largest late Tiwanaku cemetery in the Moquegua valley, Tumilaca la Chimba has the potential to answer questions about the biological heritage and cultural identity of Tumilaca populations, and by so doing to shed light on the effects of the end of Tiwanaku influence in the south-central Andes. The data discussed in this paper will be supplemented by further investigations throughout the cemetery.


We thank Donna Nash, Maria Cecelia Lozada, Karl La Favre, Sabrina Scholtz, and Sofia Chacaltana for assistance in the field. Work was permitted by the Instituto Nacional de Cultura del Peru by RDN 1208/INC dated 26 July, 2006, under direction of Patrick Ryan Williams and Maria Elena Rojas Chavez. All errors or omissions are the responsibility of the authors.

References Cited

  • Bawden, Garth
    1989 The Tumilaca Site and Post-Tihuanaco Occupational Stratigraphy. In Ecology, Settlement and History in the Osmore Drainage, Peru. Don S. Rice, Charles Stanish & Phillip R. Scarr (eds). BAR International Series 545(i) pp 287–302.
  • Bermann, Marc, Paul Goldstein, Charles Stanish & Luis Watanabe M.
    1989 The Collapse of the Tiwanaku State: A View from the Osmore Drainage. In Ecology, Settlement and History in the Osmore Drainage, Peru. Don S. Rice, Charles Stanish & Phillip R. Scarr (eds). BAR International Series 545(i) pp 269–286.
  • Blom, Deborah
    1999 Tiwanaku Regional Interaction and Social Identity. Ph.D. Dissertation, Department of Anthropology, University of Chicago.
  • Blom, Deborah, Benedikt Hallgrimsson, Linda Keng, Maria C. Lozada & Jane E. Buisktra.
    1998 Tiwanaku 'Colonization': Bioarchaeological Implications for Migration in the Moquegua Valley, Peru. World Archaeology 30(2): 238–261.
  • Goldstein, Paul
    1985 Tiwanaku Ceramics of the Moquegua Valley, Peru. MA Thesis, Univeristy of Chicago, Chicago.
  • Goldstein, Paul
    1989 Omo: A Tiwanaku Provincial Center in Moquegua, Peru. PhD Thesis, University of Chicago, Chicago.
  • Goldstein, Paul
    2005 Andean disapora: The Tiwanaku Colonies and the origins of South American Empire. University of Florida Press, Gainesville.
  • Janusek, John W.
    2003 Vessels, Time and Society: Toward a Ceramic Chronology in the Tiwanaku Heartland. In Tiwanaku and its Hinterland: Archaeology and Paleoecology of an Andean Civilization, vol 2: Urban and Rural Archaeology. A. L. Kolata (ed). Washington DC: Smithsonian Institution Press. Pp 30–91.
  • Janusek, John W.
    2004 Identity and Power in the Ancient Andes. London, Routledge.
  • Janusek, John W.
    2005 Collapse as Cultural Revolution: Power and Identity in the Tiwanaku to Pacajes Transition. In Foundations of Power in the Prehispanic Andes. K. J. Vaughn, D. Ogburn & C. A. Conlee (eds). American Anthropological Association: 175–210.
  • Kolata, Alan
    1993 Tiwanaku: portrait of an Andean civilization. Cambridge, Blackwell.
  • Owen, Bruce
    1993 A Model of Multiethnicity: State Collapse, Competition, and Social Complexity from Tiwanaku to Chiribaya in the Osmore Valley, Perú. PhD Thesis, University of California at Los Angelese.
  • Owen, Bruce
    1997 Inventario Arqueológico del Drenaje Superior del Río Osmore: Informe del Campo y Informe Final. Report to the Instituto Nacional de Cultura.
  • Owen, Bruce
    2005 Distant Colonies and Explosive Collapse: The Two Stages of the Tiwanaku Diaspora in the Osmore Drainage. Latin American Antiquity 16(1): 45–80.
  • Pari Flores, Romulo Elias
    1980 Excavacion Arqueologica En La Necropolis de Tumilaka – Moquegua. BA Thesis, Universidad Catolica Santa Maria, Arequipa, Peru.
  • Sims, Kenny
    2006 After State Collapse: How Tumilaca Communities Developed in the Upper Moquegua Valley, Peru. In After Collapse. Glenn M. Schwartz & John J. Nichols (eds). Tucson, University of Arizona Press. Pp 114–136.
  • Stanish, Charles
    1985 Post-Tiwanaku Regional Economics in the Otora Valley, Southern Peru. Unpublished Ph.D. dissertation, Department of Anthropology, University of Chicago.
  • Vargas, V. B.
    1988 Informe final del proyecto: 'Rescate arqueològico del cemeterio de Chen Chen.' Report to the Instituto Nacional de Cultura.
  • Williams, Patrick Ryan
    2002 Rethinking Disaster-Induced Collapse in the Demise of the Andean Highland States: Wari and Tiwanaku. World Archaeology 33 (3): 361–374.
  • Williams, Sloan A., Jane E. Buikstra, Niki R. Clark, Maria Cecelia Lozada Cerna & Elva Torres Pino
    1989 Mortuary Site Excavations and Skeletal Biology in the Osmore Project. In Ecology, Settlement and History in the Osmore Drainage, Peru. D. Rice, C. Stanish & P. Scarr (eds). BAR International Series 545(i): 329–346.