Textiles and Ethnicity: Tiwanaku in San Pedro de Atacama, North Chile

by Amy Oakland Rodman
Textiles and Ethnicity: Tiwanaku in San Pedro de Atacama, North Chile
Amy Oakland Rodman
Latin American Antiquity
Start Page: 
End Page: 
Select license: 
Select License


Amy Oakland Rodman

Examining textiles and other usually perishable artifacts, this paper focuses on textile style as an indicator of ethnicity in archaeological textiles excavated in the cemetery of Coyo Oriental, Sun Pedro de Atacarna, Chile. The Coyo Oriental cemetery was occupied durlng a period ofstrong Tiwanaku influence in Sun Pedro de Atacama iA.D. 500-10001 recognized in art$acts decorated with Tiwanaku irnages. The analysis presented here identifies at least two dlstinct textlle styles recognized in tunic striping, embroidered selvage treatments, and headgear, a

fact that possibly indicates a multiethnic use of the cemetery and oasis of Coyo Oriental. Associated textiles and artifacts suggest that one group represents a local style and ethnic group and the other is a dlferent ethnic group closely related to Ti\tanaku. Instead of only mlnor Ti\tanaku influence, I suggest that the oasis was home to a foreign altiplano population who maintained for centuries an ethnic identity visible in a distinct textile style.

Este trabajo examina el estilo de textiles corno indicador de etnicidad en 10s textiles arqueoldgicos del cementerio de Coyo Oriental, Sun Pedro de Atacama, regidn II, del norte de Chile durante el periodo de influencia de Tiwanaku, o aproximadamente desde 500-1000 D.C. La estructura textil y el diserio son analizados en textiles individuales y con textiles de cada bultofunerario completo y tambien con textiles de bultosfunerarios de un solo cementerio. Las textiles de Coyo Orienteson comparados con textiles de una variedadde cernenterios prehispcinicos en Sun Pedro de Atacarna donde la preservacidn de artefactos es casi total corno la de Coyo Oriental. Los resultados indentlfican por lo menos dos estilos textiles distintos en el rayado de /as tcnicas, el bordado de las orillas, y en el tocado, posiblemente indicando un uso multiPtnico del cementerio y del oasis de Coyo Oriental. Identicas ttinicas y tocados a /as de ambos grupos han sido identificados en orros cernenterios de Sun Pedro de Atacarna. El ancilisls de 10s textiles, la cesteria, la cercimica y 10s arttIfactos de madera demuestra que uno de 10s grupos estaba ma's ampliamente distribuido en Sun Pedro de Atacama y que continud, con algunas mod$caclones de estllo, en el periodo Intermedio Tardio. El otro grupo esth asociado mds con el oasis de Coyo, tiene textiles y artefactos tinicos e iconografia Tiwanaku. Estos resultados sugieren que un grupo representa el estilo y grupo dtnico local, y el otro grupo es de un grupo etnico diferente asociado estrechamente con Tiwanaku. El estudio tiene importancia para la lnterpretacidn material a preguntas de etnicidad y a preguntas mcis amplias de inter- carnbio y colonizacidn en el hrea centro-sur andino. Muchos investlgadores han indlcados que la influencia de Tlttanaku era dkbil en Sun Pedro de Atacama. En cambio, el ancilisis que aquisepresenta sugiere que la influencia y 10s artefactos de Tiwanaku han llegado a Sun Pedro de Atacama con un grupo de gente altiplhnica. Este grupo permanecid en el oasis de Atacama por siglos 1.' mantuvo su identidad etnica visible en su estilo textil.

The nature of interaction between the altiplano site of Tiwanaku and the peripheral centers within Tiwanaku's sphere of influence in the south-central Andes has been a concern of many investigations during the past two decades (Berenguer and Dauelsberg 1989; Berenguer et al. 1980; Browman 1978, 1980. 1986; Goldstein 1989; Lynch 1983, 1989; Mujica 1985; Nfifiez and Dillehay 1978; Stanish 1989: Tarrago 1977). Prehistoric burials excavated in San Pedro de Atacama. an oasis community in the highland desert of northern Chile (Figure l), have yielded a plethora of wood and bone tablets and tubes carved with recognizable Tiwanaku images associated with the drug cult or snuff complex (Llagostera et al. 1988; Nuiiez 1963; Torres 1984, 1987). Coupled with the oc- casional Tiwanaku ceramic and Tiwanaku textile import (Oakland 1986a, 1986b) and the local production of Tiwanaku ceramic copies and copies of Tiwanaku textiles (Oakland 1986b:Figure 20). it is certain that San Pedro de Atacama operated within the larger interaction sphere with Tiwanaku and the Lake Titicaca Basin. But there is no agreement as to the particular mechanisms that united the local polities and the Tiwanaku center. Whether Tiwanaku artifacts, especially

Amy Oakland Rodman, Department of Art, California State C7niversity, Hayward, CA 94542

Latin American Antiquity, 3(4), 1992, pp. 3 16-340.
Copyright C 1992 by the Society for American Archaeologq

Figure 1. Map showing the the oasis of San Pedro de Atacama, located between the Vilama and San Pedro rivers in the Atacama Desert of northern Chile near the Bolivian border.

portable artifacts, arrived at distant sites through exchange (Berenguer and Dauelsberg 1989; Brow- man 1980; Nufiez and Dillehay 1978; Pollard 1984) or colonization (Kolata 1982, 1983; Ponce Sangines 1972; Rivera 1991) is currently a contested topic.

Berenguer has long argued for a consideration of the differential distribution of material goods between northern Chilean sites, which he interpreted as evidence for Tiwanaku colonies in Arica but as exchange relationships within the San Pedro de Atacama region (Berenguer et al. 1980). Berenguer's more recent assessment is based on Tarrago's (1968, 1976) ceramic analysis from the Argentine puna and San Pedro de Atacama and the Thomas et al. (1984) ceramic and tablet analysis in San Pedro de Atacama. With Dauelsberg he proposed that during the Coyo phase (A.D. 700- 1000) in San Pedro de Atacama at least four groups might be recognized through ceramic and decorated artifact association in tomb lots: Tiwanaku elites, foreigners associated with Tiwanaku, locals, and locals associated with Tiwanaku (Berenguer and Dauelsberg 1989.: 16 1). Their analysis, while open to suggestions of a multiethnic community, continues to reject the possibility of Tiwanaku colonization in San Pedro de Atacama. Because most burials contain locally manufactured ceramic types, they favor instead a view of local elite management of local resources with Tiwanaku affiliation through status objects and ritual paraphernalia. Their assessment does allow for the possibility of at least one representative Tiwanaku group directly associated with the Tiwanaku heartland ("foraneo representante de la autoridad de Tiwanaku") (Berenguer and Dauelsberg 1989: 161). That local ceramics are equivalent to local populations (Berenguer and Dauelsberg 1989:160) needs to be argued, not assumed. Nufiez et al. (1 975) observed that colonies did penetrate San Pedro de Atacama. Lynch (1989:7) suggests that the "high level of stylistic variation and 'foreign' traits back at least to the Tiwanaku horizon" in San Pedro de Atacama describes an area where an investigation into the question of multiethnicity might be profitable. And Llagostera (1976) has identified a long-term Lake Titicaca Basin and San Pedro de Atacama interaction sphere that probably predates Tiwanaku influence in northern Chile.

The analysis presented in this paper does not pretend to resolve the colony vs. exchange debate nor does it assume any simplistic either/or scenario for San Pedro de Atacama archaeology where both relationships may have united the peripheral zones with the altiplano center. Additional mechanisms of diffusion, ritual and religious associations, and economic, social, and political in- teraction could all have played some part. What it does try to accomplish is to provide a different set of data-data rarely used in archaeological reconstructions of the past-that have the capability of identifying individuals and their relationships to larger inclusive groups.

The parameters necessary to identify Tiwanaku colonies within the south-central Andes were defined by Mujica et al. (1 983) and more recently by Moseley et al. (199 1). In addition to cultural criteria, Moseley et al. (1991) have argued for measures of ethnicity in genetic and other biological factors in determining the presence of nonlocal populations within the provinces. While genetic or biological markers might be recoverable if one were fortunate to discern archaeologically the first wave of foreign immigrants within San Pedro de Atacama, many have suggested that the area was part of a long series of altiplano migrations and host to a long-term interaction sphere that predated Tiwanaku influence in the oasis (Bittmann et al. 1979; Llagostera and Costa 1984; Nuiiez and Dillehay 1978). In addition, ethnic distinctions may not necessarily include genetic or biological criteria. Current literature abounds with discussions of ethnicity as self-referential, unconcerned with genetic process (Royce 1982). Stylistic behavior serves as an essential reference point for ethnic groups identifying one as different from another. Textile style is one of the most widely used visible markers of ethnic behavior, often standing for the group as a whole and reflecting shared ideological concerns through construction techniques, patterning, and use.

In this paper, I argue that clothing and textile style, more than any other artifactual category,

identify distinct ethnic groups in prehistoric contexts. The analysis, directed to materials from the

cemetery of Coyo Oriental in San Pedro de Atacama, is not capable of presenting an historical

overview of the entire oasis, for the Coyo Oriental cemetery was apparently in use for a relatively

short period (A.D. 500-900) within San Pedro de Atacama chronology. But it is just this period

when Tiwanaku influence was felt most strongly, and inferences regarding Coyo's place within the

larger oasis and within a longer Atacama occupation will be presented.

The present research is based on the understanding that, in the Andes, textiles have always served

to identify the individual and should continue to do so where preservation allows in archaeological

contexts. The wearing of a garment was not an arbitrary choice in life or death. A garment woven

by the individual, a family member, or close associate and displaying a pattern specific to that

individual within the larger social group is an incomparably useful artifact in ethnic identification.

Within a single burial the evidence provided through the analysis of all garments in a funeral bundle,

the associated artifacts, and a comparison to all funeral bundles and artifacts in a single cemetery

should identify original inhabitants and shared identity representing group ethnicity in archaeological

contexts. The identification rests on close stylistic analysis of all textiles.


Definition of style and its applicability to archaeological reconstruction has animated anthro-

pological literature within the past decade (Binford 1984, 1989; Conkey and Hastorf 1990; Hodder

1982, 1986; Sackett 1982, 1986; Wiessner 1983, 1984, 1985, 1989). Style is recognized here as "a

highly specific and characteristic way of doing something which by its very nature is peculiar to a

specific time and place" (Sackett 1982:63).

As a close correlate, a definition of ethnicity that considers material culture is also a current

anthropological topic (Pyszczyk 1989; Royce 1982; Stanish 1989). Ethnicity as discussed here

conforms to the view that different cultures can live side by side and do the same things, interacting

yet maintaining their distinctiveness, or conversely members of one culture can live far apart and maintain their similarity. This is ethnicity. It is a view of culture where the variability related to the material-culture record is a consequence of that distinctiveness in views and behavior within and between cultural groups (Pyszczyk 1989:245). Using "entire archaeological assemblages to examine ethnicity," Pyszczyk (1 989:2 16) "proposed that the degree ofsimilarity of behavior of the same ethnic group will be relative1.v greater than the degree of similarity of behavior between units of dzflerent ethnic groups" (emphasis in original).

Style and ethnicity merge in the area of self-identification. If style is a particular way of doing something and ethnicity is the way in which persons identify themselves or they are identified by others, then style will reflect the particular ways in which people identify themselves or are identified by others. Sackett (1985, 1990), Wobst (1977), and others (e.g., Pancake 1991) have argued that style resides in and is perceived through a gradient of levels reflecting "ethnic resolution, ranging from individual kin groups within individual settlements . . . to great culture-historical complexes that occupy considerable blocks of space and time" (Sackett 1990:33).

Several authors have recently noted how textile style is commonly used as a basis for expressions of social identity. Weissner (1 984) has stressed the individual and "social identification via com- parison" through beaded headbands in the Kalahari San. The distinct uses of textile style in Yu- goslavian folk culture as a definition of group affiliation have informed the work of Wobst (1977). Hodder (1982) has recently discussed three African tribes where at marriage women and entire families might change costume to join a different ethnic group. The apparent mix does not weaken the ethnic dress style that clearly identifies an individual and his or her tribal relationship: "The individual herself and an outside observer would always come to the same conclusion about the tribe with which a particular person was identifying" (Hodder 1982:21-23).

For the purposes of the present paper it is the understanding of the textile tradition and its applicability in defining the social self that animates discussion. In many areas of the southern Andes, textiles continue as the single most appreciated art form among indigenous populations. Complex patterning representing traditional skills has been passed through generations of weavers and wearers. A discussion of the textile tradition within modem Andean communities will serve to illuminate aspects in the tradition that could be reflected in the material culture of the archae- ological past. Despite the impact of almost 500 years of European colonization, many Andean communities continue to value indigenous clothing style as a specific reference to their cultural and ethnic cohesion.


No Andean community has maintained its textile tradition unchanged. Continuity is not the only aspect which is to be emphasized here, but the dynamic response to change and the strtngth of the basic tradition is of interest to the present investigation as well. Rather than marginalize modem indigenous communities as impassive relics, current ethnographers have chosen to focus upon those aspects that are interpreted as innovative and dynamic. Style is not static: "style is really process masquerading as thing" (Hodder 1990:50).

Recent Andean community studies reveal textiles as significant material markers of cultural identity. Franquemont's (1986:33 1) description of Chinchero, in the Peruvian highlands, begins: "Chinchero women wear their ethnic identity in their many small hair braids, in their red monteras (hats) and jackets, and, especially, in their llijllas (shawls) with broad dark plain areas and narrow design bands." Her sensitive and penetrating description of linguistic terms identifies woven patterns, which in turn contain conceptual and physical associations that unite the Chinchero ethnic group, but she remarks: "The ethnic code is not static or determinate; its correspondences are never one for one. Unlike natural phenomena, designs are made by people and are intentionally structured in keeping with the basic patterns of Chinchero thought" (Franquemont 1986:333). Like Franquemont, Zorn (1 990:248) concludes in her analysis of Sakaka, Bolivia, textiles that: "Sakaka clothing pro- claims a separate ethnic 'Sakaka' identity, but [quite unlike the conservative dress of the Chinchero] the influences on the newest style of Sakaka dress cross ethnic and class lines." Although Zorn (1990: 248-249) suggests that "much of the aesthetic language of the Sakaka textiles draws on an ancient Andean vocabulary, which each generation and weaver reinterprets," she emphasizes that new-style Sakaka clothing also incorporates factory-made and -dyed yams, copies of another ethnic group's garments, new striping techniques borrowed from factory-produced textiles, and "garish neon colors" and "images of motorcycles and Diablada dancers," which she interprets as young weavers' inten- tional subversion of the trade in Andean textiles. The importance of this study is to suggest that economic concerns and stylistic borrowing, which even cross ethnic lines, do not negate the visible ethnic clothing style. The process becomes one of incorporation of the new, a generational change where young weavers prefer to adapt rather than totally reject a definable old textile style.

For the investigator of archaeological textile collections, each of the above ethnographic studies identifies problems that could be transferred to the archaeological record. Is the cultural variability an indicator of ethnic distinction in somewhat static form with perceivable style changes over time reflecting shared ideology? Did style change rapidly in response to outside influence? Are contem- poraneous styles local or multiethnic?

A third ethnographic study of the Calcha region in the extreme southwestern comer of Bolivia is instructive (Medlin 1983, 1986, 1991). As in many communities, the youth of Calcha are not interested in continuing the complex, tradition-laden weaving styles. They are adopting European clothing with few handwoven additions. Community ritual life requires the use of traditional gar- ments in local festivals, and Calchanos preserve fiesta dress when representing the community. In spite of discrimination against handwoven dress and the cash economy that allows access to factory- produced textiles, Medlin (1 99 1:263) notes that "handwoven cloth is a cultural necessity for proper social interaction at mamages, hamlet plantings of corn, and ethnic group celebrations." Medlin (1 99 1:263) asserts "Calcha cloth, because of its design and use of color, is distinctive from that of any other Andean ethnic group." Modern Calcha festival dress includes handwoven garments, items made from treadle-loomed cloth and "still other items crafted by people of nearby ethnic groups or in market towns" (Medlin 1991:265). The handwoven textiles of a Calcha male might include up to four distinctive types of ponchos in addition to highly patterned bags and narrow bands. An investigator of Calcha clothing style would at first be overwhelmed with the variety. But with a close examination of all the individual components of each textile including the yam types, dyed colors, yarn diameters, spin and twist directions, stripe sequences, textile structures, embroidered closures, and a comparison of each of these to all of the textiles of a single weaver's product and to the combined products of the community, Calcha style becomes apparent. Calcha ponchos are distinctive and the variability of styles is confined to a specific universe of choices.

In archaeological collections from the south-central Andes, there is evidence for cultural identity expressed in clothing style like that understood for indigenous Andean communities today. The remarkably good preservation in the cemetery of Coyo Oriental in San Pedro de Atacama (Figure 2) provided an excellent research site to investigate the relation between textile style and ethnicity. The following discussion details those aspects of funeral bundles in the cemetery most useful in ethnic differentiation. Of the original 399 individuals in 216 tombs only 88 bundles contained identifiable textiles, and 56 of these could be separated within one of two groups: A (Figure 3) or B (Figure 4). A detailed analysis of garments defining ethnic groups and a discussion of possible original group affiliation will follow.


The oasis of San Pedro de Atacama, located on the northeastern edge of the Salar de Atacama, Department of Loa, Region I1 in north Chile, is situated along the Vilama and San Pedro rivers running parallel to the Andean Cordillera visible directly to the east. The cemetery of Coyo Oriental, located east of the allyu (community) of Coyo, was first excavated in 1971 and again in 1975, uncovering evidence of 216 tombs representing both single and multiple interments (Le Paige 1973, 1975). Investigators have repeatedly noted that the Coyo Oriental tombs contained a large quantity of Tiwanaku decorated wooden snuff tablets and tubes associated with a locally manufactured style known as casi pulida, a style that follows, but is less carefully formed and not as highly polished

Figure 2. Map situating the modern allyus or oasis communities in San Pedro de Atacama. The Coyo Oriental cemetery is located east of the Coyo oasis.

as, the earlier local style known as negra pullda (Le Paige 1973; NGfiez 1963; Thomas et al. 1984; Torres 1984, 1987). Recently the phase designation Coyo was proposed for the period A.D. 700-

1000 when casipulida ceramics and snuff paraphernalia with Tiwanaku iconography were placed in burials throughout San Pedro de Atacama cemeteries (Berenguer et al. 1986, 1988). Berenguer et al. (1988:345) even proposed that "there is an apparent episode of cultural discontinuity between the development before A.D. 700 and after 1000 that might be interpreted as the end of the San Pedro complex and the beginning of a new cultural complex." Tiwanaku influence in San Pedro, whatever the actual mechanism of interconnection, should be considered to have disrupted the previous local cultural development.

Textiles are rarely addressed in discussions of San Pedro culture or chronology. The Coyo cem- eteries, however, have yielded the largest quantity of Tiwanaku tapestry textiles yet described for San Pedro, four tunics and two mantles (Oakland 1986a, 1986b). Differential preservation of organics

TYPE I (1 red stripe) 3909 4118 5289 5290

TYPE 11 (3 stripes) 4091 4035 41 50 4118

TYPE 111 (4 stripes) 41 35 5290

TYPE IV (5 stripes) 4064 4112 41 86

TYPE V (5 stripes, warp patterned) 3928 3978 3979 4124

Figure 3. Schematic drawings of Group-A tunic styles (Types I-V) and lists of Coyo Oriental tombs associated with specific tunic types.

has created this skewed view representing hundreds of wooden tablets and tubes with Tiwanaku images and few Tiwanaku textiles. Recent excavations in the Solcor-3 cemeteries of San Pedro de Atacama have determined that elaborate narrow bands and bags were originally part of the snuff paraphernalia that wrapped the tablets and associated items, but although the tablets remain, the textiles have not been preserved (Bravo and Llagostera 1986; Llagostera et al. 1988; Oakland 1986a). My examination of these textiles associated with the snuff complex in archaeological collections from the Bolivian highlands, intermontane valleys, and Chilean coast and desert highlands has established a wide distribution of related textile types in Tiwanaku style, which I associate with direct Tiwanaku influence in those areas (Oakland 1986a).

Chronology for preceramic and ceramic periods in San Pedro was first established through the excavations of more than 3,000 tombs in 47 cemeteries by the Jesuit priest Gustavo Le Paige (1964: 5 1). The chronology currently in use is based primarily on eight ceramic phases (I-VIII) established by Tarrago (1 968, 1976) and further refined to five phases with thermoluminescence determinations and given phase names: Toconao (300 B.C.-A.D. loo), Sequitor (A.D. 100-400), Quitor (A.D. 400-


3934 3996


3937 4008


3935 4107 4105 5333


4008 4105 4090 5348

Figure 4. Schematic drawings of Group-B tunic styles (Types I-VII) and mantles and lists of Coyo Oriental tombs associated with specific tunic and mantle types.

700), Coyo (A.D. 700-1000), Solor (A.D. 1000-1470), and Catarpe (A.D. 1470-1535) (Berenguer et al. 1986, 1988). The ceramic associations and radiocarbon determinations made during the present research establish that the Coyo Oriental cemetery was actually used during both the Quitor and Coyo phases or between A.D. 500 and 900 (Table 1).

The predominant tomb form in the cemetery of Coyo Oriental is an unlined, packed-earth cist that in 135 cases accommodated a single individual, and in others was apparently reopened to add individuals in multiple burials. Le Paige recorded 68 infants in the Coyo Oriental cemetery associated with adult burials of both sexes. The tightly prepared and tied funeral bundle was accompanied with a variety ofweapons and tools, as well as ceramics, baskets and bags of food, and bags containing stones, shells. and carved tablets and tubes of the snuff complex (Bravo and Llagostera 1986; Le Paige 1964, 1965, 1973; Llagostera et al. 1988).

Le Paige (1973:170-17 1) did not record any noticeable differences in burial patterns within the Coyo Oriental cemetery except to note the presence of three artifacts, all new to his knowledge of San Pedro de Atacama archaeology: a furry hat with a chin strap (Figure 5), a distinctive basket

location and successive numbering. Although the ceramic type is usually considered to be charac- teristic of the Coyo phase, the calibrated radiocarbon age for tomb 5383 of A.D. 639 i20 (Table 1) is consistent with an earlier Tiwanaku presence (Berenguer et al. 1988:344). Two tombs 5340 and 5341 in the collective tomb 5334-41 also contained furry hats with chin straps, and tomb 5341, with a calibrated radiocarbon age of A.D. 672 i60, contained similar casi pulida ceramics, hafted hammers with handles, snuff paraphernalia, small silver plaques, and a carved Tiwanaku-style tablet.

Tiwanaku tapestry mantles in collective tombs 40 12- 13 and 4084-86 were associated with fewer grave goods: a freshwater shell, four casi pulida ceramics, a bag of copper stones, and one food bag in tomb 40 12- 13; and a broken bow, two casi pulida ceramics, a hafted hammer with handle, two undecorated snuff tablets, and a bone tube in tomb 4084-86. Both burials were located south of the fence. Chronology in these tombs may also relate to a Tiwanaku presence during the Quitor phase as the calibrated radiocarbon age of A.D. 677 i 50 for tomb 4012-13 suggests. The structure, patterning, and embroidery features in the warp-faced tunics in 40 12- 13 and embroidery structure and design preserved in 4084-86 place these tombs in the B Group described below. The latest Group-B date of A.D. 9 10 & 90 (calibrated radiocarbon years) is associated with the only Tiwanaku four-cornered hat yet described for San Pedro de Atacama in tomb 4026.

The diagnostic Group-A tomb 4064 with a calibrated radiocarbon age of A.D. 677 i80 strongly suggests that the two groups discussed in this paper are contemporary. Unfortunately tomb 4064 contained no ceramics, but Group-A tombs 4035,4 1 12, 4 1 17,4 135, and 4 150 were associated with negra pulida or casi pulida ceramics. Few burial goods are described for Group-A tombs in Coyo Oriental, but a wooden spoon, a hafted hammer, and two snuff tablets uncovered in the Group-A burials mentioned above suggest contemporaneous tombs. Where Le Paige noted furry headdresses with chin straps as new to San Pedro and the B Group described for Coyo Oriental is confined to this cemetery and to Solcor-3, it is probable that Group A is best represented in the larger San Pedro oases where identical tunics and headdresses are common to Solor, Quitor- 1, Quitor-5, Quitor- 6, Catarpe, Solcor, and Solcor-3.

Hafted hammers were found in Coyo burials in the San Pedro archaeological inventory, associated in eight of 40 examples in Group-B burials and in one Group-A tomb. This supports Le Paige's (1973) suggestion that the Coyo oasis was home to copper miners but not metallurgists (mineros per0 no herreros) (Bird 1975; Oakland 1990). Aniaza (1990) has noted severe arthritic deformities in both males and females in the Coyo Oriental population that may be occupationally related to the physical stress associated with extracting and transporting minerals.


Of the 88 tombs in which textiles were preserved in the Coyo Oriental cemetery, sample size and recognizable features allowed for style designation in 56 burials with 20 in Group A and 36 in Group B. Border-striping sequences, embroidery structure, and design and headdress types were the most important features used in distinguishing between Group A and Group B. The A and B groups described in this paper are based entirely on textile style. There is a high probability based in associated ceramics, burial furnishings, and radiocarbon determinations that the two groups were contemporaneous. Tombs in each style category have been selected on the basis of headdress type alone (Group A: 4 172, 4 173; and Group B: 3933, 5340, 534 l), but multicolored tunic striping and embroidered selvages separate the styles most securely. Once a tunic was warped (the warp yams were stretched on the loom), a particular group affiliation was reflected. The embroidery further emphasized a specific selection since there was no ambiguity between one group's embroidered style and the other's. In the Coyo cemetery, no burials contain examples of both A and B styles within the same tomb (Table 2).

In all burials the tunic was the most prevalent garment form. Apparently, males and females wore the tunic, a single rectangular warp-faced textile, folded at the shoulder with the neck slot formed in the weaving process with discontinuous wefts. The sides were sewn closed leaving space for the armholes. Males and females used old, worn, and repeatedly mended tunics as a soft, flexible undershirt. This warp-faced tunic type was always placed closest to the body. Construction and patterning of the tunic of the outer bundle layers and the headdress types provided the most diagnostic features useful for the purposes of this paper to hypothesize original ethnic identity. There is evidence that some females in Group B also wore a long, wide textile as a wrapped skirt or dress (aksu), and additionally the large, rectangular mantles described for other Group-B burials could have originally served this purpose.

There is one category, the single red-striped tunic which, if not further embellished with an embroidery type from one or the other group (Group-A Type I or Group-B Type Vb), or accompanied with a specific headdress definable to one or the other groups, remains ambiguous as to style affiliation. Four red-striped tunics have been placed in each style category, Group A (3903, 41 18, 5289, 5290) and Group B (4008, 4090, 4105, 5348) based in embroidery or in associations with additional tunics in one or the other category. Red-striped tunics from seven burials (4075, 4079, 5292, 5293, 53 16, 53 17, 5347) could not be placed in either Group A or Group B, creating at least one nebulous category in the cemetery where originally, in life, one's personal identity perhaps remained negotiable, or was at least unrecoverable in the present analysis. Most individuals' clothing reflects clear, consistent choices made in life and placed with them on death.

Group-A Textile Style

Twenty Coyo Oriental tombs contained evidence of the tunic and headdress type that identifies

Group-A affiliation. Group-A tunics are patterned in five warp-stripe variations (Types I-V). Stripes

are arranged along the side selvages flanking a solid light-brown central field. Striping variations

include either one red stripe (Type I) or three (Type 11), four (Type 111), or five solid red, white, and

maroon stripes (Type IV). These occur between wider stripes that appear to be a single color but

are actually composed of one-half blue and one-half green yams. In the final variation (Type V) the

five solid red, white, and maroon stripes are formed of complementary-warp patterning, again placed

between wider blue-green stripes (Figure 7).

The outer tunic and two headdresses placed on the funeral bundle of an adult male in tomb 4064

are representative of Group-A style (Figure 8). Following from the selvage the striping sequence of

tunic 4064.3 begins with a greedblue combination stripe and continues with red, greedblue, cream,

greedblue, maroon, greedblue, red, greedblue, maroon, greedblue stripes and repeats in mirror

image opposite the light-brown central panel. Red, maroon, and cream stripes are almost equal

widths (1.3-1.6 cm) and the greedblue stripe is composed of individual green and blue colors (1.5

cm each), which are difficult to separate from a distance, and appear as a wide (3-cm), single-colored


The selvage embroideries are just as distinctive as the striping patterns in determining Group-A

style. Embroideries in Group A are, in general, finely executed, elaborate, and tend toward curved

and rounded forms, serrated spirals and diagonal lines ending in loops. Technically termed satin

stitch, the embroidery style is, in some cases, wrapped around warp yams in a diagonal rather than

only horizontal manner. Group-A side-selvage embroideries often continue around the armhole

opening and contain from three to nine vertical rows of polychrome satin stitches in curvilinear

style. The neck plaque, a rectangle embroidered just below the neck slit, must have been a very

diagnostic feature of dress and social position. It is oriented on the center of the chest when worn.

Group-A neck plaques are narrow and long and embroidered in the same curvilinear style as the

side embroideries. Tabbed embroideries surrounding the neck opening are entirely distinct to Group

A, usually accompany the thin, curved neck-plaque, and clearly associate a tunic to Group-A style.

The neck plaque in 4064.3 (2 x 14 cm) is executed in fine (2-mm) vertical rows in maroon, cream,

blue, green, red yams in a curved "x"-and-dot motif. Side-selvage embroideries repeat a similar

curved motif as an expanded design. The tabbed neck-slot opening is embroidered in fine blue step

motifs on a maroon ground.

The headdress type for Group A (Figure 9) is formed over a twisted circular plant fiber foundation.

Strips of camelid skin and hair have been cut and wrapped around the core. In Group-A textile

style the hat center is a separately formed disk in simple looping or loop and twist, usually patterned

the structured universe, two parts paired. Checkerboard neck plaques were centrally positioned on Tiwanaku tapestry tunics from the Coyo Oriental cemetery and the Pica cemetery (Berenguer and Dauelsberg 1989:Figure 8), and should be examined where evidence of Tiwanaku tapestry has been excavated in other Chilean sites such as Arica, Pisagua, Tarapaci, Pica, Chiu-Chiu, and in Peruvian sites near the far south coast in Omo and Loreto Viejo all within the south-central Andes, an area of direct Tiwanaku influence. The checkerboard plaque has even been borrowed from the Tiwanaku repertoire and embroidered on a miniature Huari tapestry tunic, now in the collection of the Brooklyn museum (Anton 1987:Figure 86), and is used as a Huari four-comered hat design in knotted pile (Frame 1990:Plate 8, left).

The analysis presented here supports much of the Berenguer and Dauelsberg synthesis. Strong Tiwanaku influence undoubtedly did disrupt the San Pedro cultural sequence, but there were most likely many people physically present whose original home was the Bolivian altiplano. Kolata (1 99 1) has pointed out that Tiwanaku's strategy was an adaptive, manipulative one rather than one of direct confrontation. If the analysis presented in this paper is correct, foreign Tiwanaku groups coexisted with local populations and were buried in spatially integrated, though partially segregated bounded cemeteries. The boundedness projected through comparison (Wiessner 1989) as an essential aspect of ethnic identification could hardly be more explicit in an oasis community surrounded by desert.

Elaborately decorated textiles with clear Tiwanaku iconography probably represent the larger social, political, and religious center and perhaps elite representatives in provincial regions. But headdresses and warp-faced plain-weave garments connect these Tiwanaku textiles with a larger group of men, women, and children. In the southern Andes, most people wove and wore warp- faced textiles. The Group-B neck plaque embroidered on tapestry and plain-weave tunics alike places Tiwanaku-style tapestry tunics and elaborate Tiwanaku iconography and warp-faced garments as a coherent unit of interrelated style. While it would be possible that both Group A and Group B represent two forms of local style with Group B adopting Tiwanaku-related clothing and decorated paraphernalia, the B Group would have to have changed their entire textile assemblage to become what is here recognizably distinctive to Group B. Some part of the local population may have adopted the Group-B style, perhaps through mamage alliances, but such a distinction between styles without a resident altiplano population seems improbable. Ethnographic data suggest that borrowing and style associations would be visible as additive features of a local ethnic style. The new tools, headdress types, mantles, and textile-style patterning argue instead for the introduction of a foreign element into San Pedro de Atacama. It is suggested that Group B might be understood as originally deriving from a distinct altiplano source, a group that has maintained an ethnic identity in textile style within San Pedro de Atacama for centuries. The specific patterning and structure common to garment styles make textiles one ofthe most useful indicators of cultural identity in the archaeological record.

Acknowledgments. This research was made possible through support of the National Endowment for the Humanities (RO-2 1963-89) under the auspices of the Instituto de Investigations Arqueologicas and the Museo

R.P. Gustavo Le Paige S.J. I owe a debt of gratitude to my colleagues Agustin Llagostera, Maria Antonieta Costa, Francisco Tellez, and Lautaro Nufiez who graciously provided the facilities and permission for textile analysis and generously allowed access to their own analyses and interpretations. I sincerely thank Vicki Cassman who designed and, with the assistance of Vivian Recine and Cecelia Ramirez, implemented textile conservation and storage, and Bernardo Arriaza, M. A. Costa, and Jose Cosilovo who camed out human skeletal analysis. Discussions with colleagues in La Paz, Arica, Austin, and Berkeley have helped clarify various aspects of this article, and I would like to thank Paul Goldstein, Gray Graffam, Charles Stanish, Lilliana Ulloa, Bill and Barbara Conklin, Johan Reinhard, Lynn Meisch, Elayne Zorn, Mary Ann Medlin, and Wolfgang Schuler. I especially appreciate the critical comments provided by my husband Mike Rodman, by Vicki Cassman, Bernardo Amaza, Alan Kolata, Terence Grieder, David Browman, Prudence Rice, and two anonymous reviewers whose advice on an earlier version of this manuscript helped focus the interpretations presented here.


Anton, F. 1987 Ancient Peruvian Testiles. Thames and Hudson, London.

Amaza, B.

1990 Paleo-salud de 10s habitantes de Coqo Oriental: Sun Pedro de Atacarna. Final report submitted to the National Endowment for the Humanities, Washington, D.C., Interpretive Research Project RO-21963-89. Copies available from the National Endowment for the Humanities, Washington, D.C.

Berenguer, J., and P. Dauelsberg

1989 El norte grande en la orbita de Tiwanaku. In Culturas de Chile, Prehistoria. Desde Sus Origenes Hasta Los Albores de la Conquista, edited by J. Hidalgo L., V. Schiappacasse F., H. Niemeyer F., C. Aldunate del S.. and I. Solimano R., pp. 129-180. Editorial Andres Bello, Santiago.

Berenguer, J., V. Castro, and 0. Silva 1980 Reflexinnes acerca de la presencia de Tiwanaku en el norte de Chile. Estudios Arquaolbgicos 5:8 1-93. Universidad de Chile, Antofagasta. Berenguer, J., A. Deza, A. Roman, and A. Llagostera 1986 La secuencia de Myriam Tarrago para San Pedro de Atacama: Un test por termoluminiscencia. Recista Chilena de Antropologia 5: 17-54. Santiago. 1988 Testing a Cultural Sequence for the Atacama Desert. Current Anthropology 29:341-346. Binford, L. 1984 An Avawara Day: Flour, Spinifex Gum, and Shifting Perspectives. Journal ofAnthroL?ological Research 40:157-18.2. 1989 Styles of Style. Journal of Anthropological Archaeology 8:5 1-67 Bird, J.

1975 The "Copper Man": A Prehistoric Miner and His Tools from Northern Chile. In Durnbarton Oaks Conference on Pre-Colurnbian .tfetallurg?. of South America. Dumbarton Oaks Research Libary and Col- lection, Washington, D.C.

Bittmann, B., G. Le Paige, and L. Nuriez 1979 Cultura atacarnefia. Departamento de Extension Cultural del Ministerio de Education, Santiago. Bravo. L., and A. Llagostera 1986 Solcor 3: Un aporte a1 conocimiento de la cultura San Pedro. Periodo 500-900 d.C. Revista Chungarci 16-17:323-332. Universidad de Tarapaca, Arica. Browman, D. 1978 Toward the Development of the Tiwanaku (Tiahuanaco) State. In Advances in Andean Archaeology, edited by D. Browman, pp. 327-349. Mouton, The Hague. 1980 Tiwanaku Expansion and Altiplano Economic Pattern. Estudios Arquuolbgicos 5: 107-1 12. Universidad de Chile, Antofagasta. 1986 Prehispanic Aymara Expansion, the Southern Altiplano and San Pedro de Atacama. Estudios Ataca- rnefios 7:236-252. San Pedro de Atacama. Conkey, M., and C. Hastorf (editors) 1990 The Cies of Stj,le ln Archaeology. Cambridge University Press, Cambridge. Focacci, G. 1983 El Tiwanaku clasico en el valle de Azapa. In Documento de Trabajo 3:94-113. Universidad de Tarapaca, Arica. Frame, M. 1990 Andean Four-Cornered Hats. Metropolitan Museum of Art, New York. Franquemont, C. 1986 Chinchero Pallays: An Ethnic Code. In The Junius B. Bird Conference on Andean Textiles, edited by

A. Rowe, pp. 331-338. Textile Museum, Washington, D.C. Fuentes, J. 1965 Tejidos prehispcinicos de Chile. Editorial Andres Bello, Santiago. Goldstein, P. 1989 Omo, A Tibvanaku Provincial Center in Atfoquegua, Peru. Unpublished Ph.D. dissertation, Department of Anthropology, University of Chicago, Chicago.

Hodder, I. 1982 Sj*rnbols in Action. Cambridge University Press, Cambridge. 1986 Reading the Past: Current Approaches to Interpretation in Archaeology. Cambridge University Press,

Cambridge. 1990 Style as Historical Quality. In The CSes of Style in Archaeology, edited by M. Conkey and C. Hastorf, pp. 44-5 1. Cambridge University Press, Cambridge.

Kolata, A. 1982 Tiwanaku: Portrait of an Andean Civilization. Field Museurn of Natural History Bulletin 53(8): 13-28. 1983 The South Andes. In Ancient South Americans, edited by J. Jennings, pp. 241-285. W. H. Freeman,

San Francisco. 199 1 The Technology and Organization of Agricultural Production in the Tiwanaku State. Latin American Antiquit!) 2:99-125. Le Paige, G. 1963 La antigiiedad de una tumba cornprobada por C-14 y el ambiente que la rodea. Revista Universitaria

26. Universidad Catolica de Chile, Santiago.

1964 El precerbmico en la cordillera atacarnefia 4.10s cementerios del period0 agroalfarero de Sun Pedro de Atacarna. Anales de la Universidad del Norte 3. Universidad del Norte, Antofagasta. 1965 Sun Pedro de Atacama 1)su zona, 14 ternas. Anales de la Universidad del Norte 4. Universidad del Norte, Antofagasta. 1973 Tres cementerios indigenas en San Pedro de Atacama y Toconao. Actas del I'I Congreso de Arqueologia Chilena, pp. 163-187. Universidad de Chile, Santiago. 1975 Coyo Oriental: Tomb Descriptions, Tomb Lists and Artifact Sketches. Ms. on file, Museo R.P. Gustavo Le Paige S.J., San Pedro de Atacama. Llagostera, A. 1976 Hipotesis sobre la expansion incaica en la vertiente occidental de 10s Andes meridionales. In Homenaje a1 R.P. Gustavo Le Paige, pp. 203-2 18. Universidad del Norte, Antofagasta. Llagostera, A,, and M. A. Costa 1984 Los Atacarnefios. Museo Arqueolbgico R.P. Gustavo Le Paige, Sun Pedro de Atacama. Ministerio de Educacibn, Santiago. Llagostera, A,, M. C. Torres, and M. A. Costa 1988 El complejo psicotropico en Solcor-3 (San Pedro de Atacama). Estudios AtacameAos 9:61-98. San Pedro de Atacama. Lynch, T. 1983 Camelid Pastoralism and the Emergence of Tiwanaku Civilization in the South-Central Andes. World Archaeology 15: 1-14.

1989 Regional Interaction, Transhumance, and Verticality: Archaeological Use of Zonal Complementarity in Peru and Northern Chile. In AWichrganDiscussions in Anthropology 8: 1-1 1. University of Michigan, Ann Arbor.

Medlin, M. A. 1983 Surnaj Awayqa Calchapi: U'eaving, Social Organization, and Identity in Calcha, Bolivra. Unpublished Ph.D. dissertation, Department of Anthropology, University of North Carolina, Chapel Hill. 1986 Learning to Weave in Calcha, Bolivia. In The Junius B. Bird Conference on Andean Textiles, edited by A. Rowe, pp. 275-288. Textile Museum, Washington, D.C. 199 1 Ethnic Dress and Calcha Festivals, Bolivia. In Textile Tradrtions of .Vlesoamerrca and the Andes, edited by M. Schevill, J. Berlo, and E. Dwyer. pp. 26 1-279. Garland, New York. Moseley, M.. R. Feldman, P. Goldstein, and L. Watanabe 199 1 Colonies and Conquest: Tiwanaku and Wan in Moquegua. In Huari Admrnistratr~~e

Structure, edited by W. Isbell and G. McEwan, pp. 12 1-140. Dumbarton Oaks Research Library and Collection. Washington,

Mujica, E.

1985 Altiplano-Coast Relationships in the South-Central Andes: From Indirect to Direct Complementarity. In Andean Ecoloavand Civilization, edited by S. Masuda, I. Shimada, and C. Morris, pp. 103-1 40. University of ~okyo Press, Tokyo.

Mujica, E., M. Rivera, and T. Lynch 1983 Proyecto de estudio sobre la complementariedad economica Tiwanaku en 10s valles occidentales del centro-sur andino. Revista Chungarci 1 1:85-109. Universidad de TarapacL, Arica. Nuiiez, L. 1963 Problemas en torno a las tabletas de rape. In Anales de la Universidad de1,Vorte 2: 149-1 67. Universidad del Norte, Antofagasta. Nufiez. L., and T. Dillehay 1978 .Vlocilrdad giratoria, armonfa social y desarrollo en 10s Andes merrdronales: Patrones de trcifico y in- teraccidn econdmica. Universidad del Norte. Antofagasta. Nuiiez, L., V. Zlatar, and P. NGfiez 1975 Un circuit0 transhumLntico entre la costa de Pisagua y el borde occidental de la Pampa del Tamarugal. In Estudios AtacarneAos 3:49-52. San Pedro de Atacama. Oakland, A. 1984 Moquegua Textile Analysis from the 1983-1984 excavations by Paul Goldstein in Omo, Moquegua (Osmore) Valley, Peru. Ms. on file, Department of Art, California State University, Hayward. 1986a Tiwanaku Textile Style from the South Central Andes, Bolivia and h'orth Chile. Unpublished Ph.D. dissertation. Department of Art History, University of Texas, Austin.

1986b Tiwanaku Tapestry Tunics and Mantles from San Pedro de Atacama, Chile. In The Junius B. Bird Conference on Andean Textiles, edited by A. Rowe, pp. 101-122. Textile Museum, Washington, D.C. 1989 Puqui Textile Analysis from the Oruro Archaeological Textile Collection in the Museo de Etnografia

y Folklore. La Paz, Bolivia. Ms. on file, Department of Art, University of California, Hayward. 1990 Tiuanaku Textiles, a View Frorn the ProlSinces. Paper presented at the 30th Annual Meeting of the Institute of Andean Studies, Berkeley. Orellana, M. 1984 Influencias altiplanicas en San Pedro de Atacama. Estudios Atacarnefios 7:197-208. Universidad del Norte, Antofagasta.

Pancake, C. 199 1 Communicative Imagery in Guatemalan Indian Dress. In Textile Traditions of Mesoamerica and the Andes: An Anthology, edited by M. Schevill, J. Berlo, and E. Dwyer, pp.45-62. Garland, New York. Pollard, G.

1984 Interregional Relations in the Southern Andes: Evidence and Expectations for Understanding the Late Prehistory of Northwest Argentina and North Chile. In Social and Economic Organization in the Prehispanic Andes, edited by D. L. Browman, R. L. Berger, and M. A. Rivera, pp. 205-247. BAR International Series

194. British Archaeological Reports, Oxford. Ponce Sangines, C. 1972 Tiwanaku: Espacio, tiernpo ), cultura: Ensayo de sintesis arqueol6gica. Publication No. 30. Academia Nacional de Ciencias de Bolivia, La Paz. Pyszczyk, H. 1989 Consumption and Ethnicity: An Example from the Fur Trade in Western Canada. Journal ofAnthro- pological Archaeology 8.2 1 3-249. Rivera, M. 1991 The Prehistory of Northern Chile: A Synthesis. Journal of World Prehistory 5:l-47.

Rolandi de Perrot, D. 1973 Los textiles tastileiios. In Tastil. una ciudadpre-Incaica Argentina, edited by E. M. Cigliano, pp. 229-

399. Ediciones Cabargon, Buenos Aires. Royce, A. 1982 Ethnic Identity. Indiana University Press, Bloomington.

Sackett, J. 1982 Approaches to Style in Lithic Archaeology. Journal ofAnthropological Archaeology 1:59-112. 1985 Style and Ethnicity in the Kalahari: A Reply to Wiessner. American Antiquit)) 50:151-159. 1986 Isochrestism and Style: A Clarification. Journal ofAnthropologica1 Archaeology 5:266-277. 1990 Style and Ethnicity in Archaeology: The Case for Isochrestism, In The Cses of Style in Archaeology,

edited by M. Conkey and C. Hastorf, pp. 33-50. Cambridge University Press, Cambridge. Stanish, C. 1989 Household Archeology: Testing Models of Zonal Complementarity in the South Central Andes. American Anthropologist 9 1 :7-24. Stuiver, M., and B. Becker 1986 High-Precision Decadal Calibration ofthe Radiocarbon Time Scale, A.D. 1950-2500 B.C. Radiocarbon 28:863-9 10. Tarrago, M. 1968 Secuencias culturales de la etapa agro-alfarera de San Pedro de Atacama, Chile. Actas y .Wernorias del XXXVII Congreso Internacional de Arnericanistas 11: 119-144. Buenos Aires. 1976 Alfareria tipica de San Pedro de Atacama, Norte de Chile. Estud~os Atacamefios 4:5-63. San Pedro de Atacama. 1977 Relaciones prehispanicas entre San Pedro de Atacama y regiones aledanos: La Quebrada de Huma- huaca. Estudios Atacamefios 5:5-63. San Pedro de Atacama. Thomas, C., C. Massone, and M. A. Benavente 1984 Sistematizacion de la alfareria del area de San Pedro de Atacama. Revrsta Chilena de Antropologia 4: 49-1 19. Santiago. Torres, M. C. 1984 Tabletas para alucinogenos de San Pedro de Atacama: Estilo e iconografia. In Tesoros de Sun Pedro de Atacama, pp. 23-36. Museo Chileno de Arte Precolombino, Santiago. 1987 The Iconography of South American Snuff' Traj7s and Related Paraphernalia. Etnologiska Sudier 37. Goteborgs Etnografiska Museum, Goteborg. Ulloa, L. 1982 Estilos decorativos y formas textiles de poblaciones agromaritimas en el extremo norte de Chile. Revista Chungarb 8: 109-1 36. Universidad de Tarapaca, Arica. Wiessner, P. 1983 Style and Social Information in Kalahari San Projectile Points. American Antiqurt)) 48:253-276. 1984 Reconsidering the Behavorial Basis for Style: A Case Study Among the Kalahari San. Journal of Anthropolog~cal Archaeology 3: 190-234. 1985 Style or Isochrestic Variation? A Reply to Sackett. American Antiquit)' 50:l-224. 1989 Style and Changing Relations Between the Individual and Society. In The Meanings of Things, edited by I. Hodder, pp.56-63. Unwin Hyman, London. Wobst, M. 1977 Stylistic Behavior and Information Exchange. In For the Director: Research Essays in Honor ofJames

B. Gr&, edited by C. E. Cleland, pp. 241-252. Anthropological Papers No. 6 1. Museum of Anthropology, University of Michigan, Ann Arbor.

Zorn, E. 1990 Modem Traditions: The Im~act of the Trade in Traditional Textiles on the Sakaka of Northern Potosi. Bolivia. In Proceedings of the ~Atile Society qfArnerica Biennial Symposium, pp. 241-252. Washington,


Received Februar~, 8, 1991; accepted ,May 27, 1992

  • Recommend Us