Page 107


Art of the Wari Empire and Its Contemporaries

by Patricia J. Knobloch

Research Associate, Institute of Andean Studies

Berkeley, California

Submitted for publication in ANCIENT ART OF THE ANDEAN WORLD, edited by Shozo

Masuda and Izumi Shimada. Iwanami Shoten, Publishers, Tokyo.

© Spring, 1989

English Translation (© Knobloch, 1988) with pagination from the Japanese volume added.

Continuing research on this topic can be found at: Who Was Who in the Middle Horizon Andean Prehistory?

      The site of Wari is located 10 km north of Ayacucho in the south-central highlands of Peru (Fig. 1) at an altitude of 2800 m. Situated defensively on a mesa with precipitous valleys on three sides, the durable remnants of buildings, pottery, lithics and ground stone cover 5 km2. The site is situated between a lower montane savannah and alluvial valley bottom, which supplied agricultural acreage to support a large occupation. During the Middle Horizon, social interaction among numerous small polities on the central and south coasts and in the highlands of Peru and Bolivia produced evidence of state formation with a centralized authority located at the ancient urban site of Wari.

      Style change in Middle Horizon art is an analytical tool for discovering and explaining the sequence of prehistoric events during the development, expansion and demise of the Wari state. As different Middle Horizon populations interacted and exchanged information, the artisans borrowed and created new designs in clothing, ornaments, and pottery and even built walled enclosures that represented their new social environment. Their art also combined many older styles (Bennett 1954: Fig. 30; J. Rowe 1967; Rowe and Brandel 1969/70). Middle Horizon iconography is notable for elaborate, full-bodied anthropomorphic figures with supernatural attributes. Of the more prominent figures, the most important is the Front-faced Deity icon, also known as the Male Deity (Menzel 1964: 19), Wari Deity (Menzel 1977: 33), or central staffed deity (Cook 1987: 50-51) (Fig. 2A, B) Other important figures depicted in profile are known as attendants or angels
Page 108
(Menzel 1977: Figs. 66, 67). These figures can be found on artifacts depicted either individually or combined. When combined, these figures represent an artistic model of hierarchy known as the Central Deity Theme (Cook 1987) that appears to have been used by Wari leaders as an ideological model to unify diverse polities into a state. Another key attribute of Wari unification is the installation of large rectangular compounds with central courts and narrow side rooms, and connecting corridors. The size and spatial distribution of these unique installations indicate a complex Wari government that exhibited extensive economic planning and political control.

      In this chapter, I investigate Middle Horizon archaeological data to discover how artisans used art to document their social environment and to explain their art as an integral medium of political, economic and social communication. I assume that art is not a purely cultural event, but can be interpreted as recorded information of intellectual expressions.


      To organize the archaeological data I began with a general chronology of temporal units, such as 'Middle Horizon Epoch IA', and stylistic units, such as 'less fancy Chakipampa A', established by Dorothy Menzel's
page 109
(1964, 1977) studies of Middle Horizon archaeology. Each general unit was further defined on the basis of stratified occupational refuse excavated at the sites of Wari and Conchopata, a site located 10 km south of Wari along the edge of a steep ravine (Knobloch 1983; Isbell 1987). The outcome was a more detailed chronology of events with briefer temporal units and more precise stylistic units for further analysis. Such a chronology can aid the archaeologist in defining ancient processes of interaction and in differentiating social changes within a population's local history from changes caused by foreign intrusion. The deposition of pottery in the stratified refuses also provided evidence for revising Menzel's chronology of pottery manufacture for the Wari area and empire (Knobloch 1983). Results of the evaluation indicated that the less fancy Chakipampa style could be separated into her earlier A and later B sub-styles, but that the fancier iconography and designs of the Chakipampa A style occurred after the Chakipampa B style. Also, Conchopata style sherds, assigned to Middle Horizon IA by Menzel (1964), were excavated from layers containing Chakipampa B pottery at the site of Conchopata. Subsequently, pottery making events of Epoch I were sequenced differently. This sequence reveals most importantly that the Nasca 9 style was not developed by coastal populations but by highland artisans from innovations in the Chakipampa B style. By definition, the Middle Horizon began when Nasca 9 style pottery arrived in the Ica Valley. Therefore, in the new sequence, I suggest that Middle Horizon I is a brief transitional period dating to A.D. 750-800, later than Wari's initial expansion into such areas as the Lucre Basin at the site of Pikillacta (McEwan 1984) near Cuzco and the south coast Moquegua Valley at the mesa top site of Cerro Baul.

      In this paper I continue to refer to the names of the Wari styles as A (A.D. 550-650) or B (A.D. 650-750/800) in reference to Menzel's (1964) previous chronological placement in Epochs IA and IB of the Middle Horizon, but these cross date to the Ica Valley pottery style of the Early Intermediate Period Phase 8, which I suggest is later (Table 1). Units of time based on detailed attribute analysis of style change are necessary because C-14 dating produces ranges of time that are too broad for the sequencing of numerous unique events, such as making a pot or weaving a tapestry. A C-14 date also estimates the time in which an artifact may have been produced.
Page 110
Most C-14 dates for late Early Intermediate Period and Middle Horizon archaeological finds have ranges of 50 to 100 years. At a 95% chance of accuracy in dating these finds, the C-14 dates overlap such that all of the associated prehistoric events could have occurred from 650 to 700 A.D. This unit of time suggests a very rapid social change over great distances, which is not supported by present archaeological analysis. However, C-14 dates remain useful for distinguishing large chronological divisions and for estimating ranges of absolute dates. Thus my discussion of social changes within large chronological divisions will be based on style change, and I will suggest absolute dates for these divisions based on C-14 dates.


1. A.D. 400-500

      The chronology of events for Wari state development begins with the Early Intermediate Period. At A.D. 400 the Ayacucho population, known as the Huarpa culture, can be described as a valley-wide, multi-community society that occupied several large sections of the Wari mesa and areas near Ayacucho. They lived in small communities with agglutinated rectangular-roomed buildings (Lumbreras 1974). By A.D. 500 the society in the Ayacucho Valley was interacting with the coastal Nasca Valley population some 200 km to the southwest. This interaction was variable in intensity yet persisted for at least four centuries. Evidence of interaction occurs in the ceramic data. Artists embellished the geometric black-on-white Huarpa pottery designs with the addition of circles, filler dots and orange and red colors. Nasca art contains numerous examples of curvilinear designs, such as elaborate anthropomorphic icons; yet, Huarpa artisans still chose Nasca's more geometric designs. The Nasca potters borrowed little more than the checkerboard, alternating pyramid and chevron band designs. (Fig. 3)

2. A.D. 550-650

      Due to early interaction with coastal Nascans, Wari potters used more colors and curvilinear band designs on their pots. They retained the Huarpa style's matte-white backgrounds, fine, black outlining and structured or paneled design fields. At this time the predominate Wari style is known as the less fancy Chakipampa A style (Menzel 1964). Coastal Nasca art contains few if any less fancy Chakipampa A style designs. This period when the social interaction of the Wari/Nasca populations appears to have waned cross dates to a period of environmental stress including a devastating 32-year drought, A.D. 562-594, detected in the deep ice core samples from Quelccaya north of Lake Titicaca (Shimada in this volume; Shimada, 1991). Data from the Nasca 7 site of La Tinguińa in the upper Ica Valley provide additional support (ibid; Menzel 1971). The art of Chakipampa B potters indicates a renewed interaction with the coastal cultures, including the Ica Valley population.

3. A.D. 650-750/800

      Wari's Chakipampa B potters changed the vessel shapes to include straight-sided open bowls and more face-necked jars.
Page 111
They continued to make the Huarpa style face-necked jars that were made with only slight modeling of the straight-walled necks to represent chins, noses, ears and eyes (Knobloch 1983). Early Chakipampa B potters painted more colorful designs, such as multi-colored, nested zigzag bands and vertical-banded rectangles. They also began painting animal icons on their pottery; though rarely found in early Chakipampa B refuse deposits, these icons occurred more frequently in the middle and late Chakipampa B pottery samples indicating a gradual development. Many of the designs and icons were clearly derived from Nasca art, such as the 3-fillet band design, multi- colored, circled dot designs, an icon of an animal viewed from above with a triangular tail (derived from Nasca 7 and 8, Lawrence Dawson, personal communication 1983), and another animal icon with a jagged band of triangles that formed a rectangle as the body and tail. Design antecedents for the last icon came from the killer whale or fish fin design in Nasca art (see Seler 1961:Abb.333, 334, 334a) which Nasca 5 and 6 artists drew as equilateral or isosceles triangles but which Nasca 7 artists drew more curved or as right-angled or obtuse-angled triangles on a curved back due to inferred Moche interaction (Harcourt 1924: Pl. 57, center) (Lawrence Dawson, personal communication, 1988).

      This design of a jagged band of triangles was also found on sherds from the surface of a site in the Pisco Valley, from excavations at Conchopata in association with early Chakipampa B pottery (Benavides n.d.: Lam X, Fig.1) and from excavations at Wari (Benavides 1984: Lam. X, Fig. g; Lam. XI, Fig. e, f) in association with middle Chakipampa B pottery. The Pisco Valley design contains highland innovations in the use of a red-orange background delimited into panels by chevron bands with designs painted in red, grey and white colors. Because the icons retain Nasca 7 design attributes, I believe that this icon and the triangular-tailed animal icon are the earliest evidence of an exchange of mythical art between coastal Nasca and highland Wari populations dating to A.D.700-750 (Fig. 4) .

      The hump-backed animal icon is a depiction that went through many changes from its Nasca (Tello 1942: Lam. 17; Schlesier 1959: Abb. 202) origins. Hump-backed, feline or monkey depictions have the essential design similarities of a profile-viewed animal with snout, toothed-mouth, curved back, front and back legs and curled tail and may also have ties to Moche examples. Nasca 7 potters made the icon appear monkey like with long hand-like front paws, human-looking face and tightly curled ear and tail (Muelle and Bias 1938: Lam. 4a). Nasca 8 potters painted highly abstract variations of the head only (Gayton and Kroeber 1927: Fig. 10a; Strong 1957: Fig. 17e). Highland Wari potters painted the icon as though the icon represented a dog, feline, or, possibly, small rodents. A variation occurs on middle Chakipampa B pottery in which the head is divided and painted in two colors. Again there appears to be no evidence that Wari potters borrowed or modified the Nasca Epoch 8 bodiless variation; thus the Chakipampa B potters were more likely influenced by earlier Nasca 7 and Huarpa styles which were passed on from one generation to the next and culminated in stylistic changes with their own mythological interpretations.

      Other animal icons include the Ayacucho serpent with a body of three joined concentric circles (Menzel 1964: PI. 6, Fig. 17), and a full-bodied animal drawn in profile with a zigzag band across its back
Page 112
(Menzel 1968: Pi.28, Figs.1-6). These designs may also have had Moche antecedents. Animal icons with bodies of joined circles occurred in Moche art (Sawyer 1975: Fig. 29; Kutscher 1954: Pls. 54c, 75). The second animal icon is distinctive for its fangs, two horn-like projections, ear flopped back and unusual zigzag back appendage. Its Moche design antecedents may be from the Strombus galeatus (conch shell from Ecuadorian tropical waters) monster (Donnan 1976:Fig. 92) which Wari artists developed into one of the minor attendant animal icons in the Central Deity Theme (Conklin 1970:Fig.1, lower left corner). These Chakipampa B icons and designs, including chevron bands and cross-banded designs on vessel rims, were assimilated gradually by Wari artisans (Knobloch 1983).

      Late Chakipampa B artisan contributions to Wari iconography can be found on sherds excavated from the Moraduchayoq sector at Wari (Brewster-Wray 1983). Several sub floor cists with large, disk-like capstones were discovered. In ancient times, looters opened the cists by digging trenches between them rather than by removing the heavy capstones. Excavated pottery was extremely scattered among the cists (Cook 1985:67) and mixed with later Epoch 2 pottery when fill from the surrounding matrix eroded into the cists. Most of these Chakipampa B sherds indicate a vast improvement in design execution compared to the ceramic samples from early and middle Chakipampa B deposits. Designs on the cist pottery also increase the number of representational depictions of human and animal icons. The cist pottery was mostly high quality ware found associated with a ritualized event whereas pottery from the stratified deposit previously mentioned was occupational refuse and contained mostly low quality ware.

      This gradual development of animal icons supports a change in the chronological sequence. The fancy Chakipampa A pottery that Menzel had placed in Epoch IA is most similar to the late Chakipampa B cist pottery and coastal Nasca 9 style pottery. I suggest that the fancy Chakipampa style of icons was fully developed by late Chakipampa B artisans at which time the Wari interaction with coastal Ica and Nasca valley populations was renewed and strengthened. When the late Chakipampa B /Nasca 9 style was borrowed by coastal potters they used the designs on their own vessel shapes, such as round-based, short-necked jars, and rarely followed the Wari practice of placing the animal icons within paneled areas on the vessel's body (Fig. 5). Because Wari vessel shapes remained somewhat different from Nasca vessel shapes, the designs were probably transferred linguistically or by means of artistic representations, such as textiles. I suggest that the late Chakipampa B style defines a brief Middle Horizon Epoch 1. Thus the beginning of the Middle Horizon began much later in Wari history.

4. Iconographic Interpretation of Chakipampa Icons

      The animal icons occur frequently in the late Chakipampa style of Wari art. Though some design origins appear to be from coastal Nasca and Moche cultures, the Wari population created more abstract versions of their own mythological realm. This mythology does not, however, appear to be of political importance to the initial expansion of the Wari state.
Page 113
Few, if any, ceramic examples of the animal icons occurred at this time at Wari sites, such as Pikillacta and Cerro Baul, that occur outside the south coast area. At these sites most pottery is Wari secular ware with geometric designs, such as a rectangular design of multi-colored vertical bands - known as the "pendent rectangle" design - usually drawn from lip to base on the interior of straight-sided, open bowls. This design was also painted on the cheek areas of face-necked jars that represented Wari individuals. In general, early Wari art depicts an assimilation of older styles that involved the depiction of animal icons. These icons, however, do not clearly relate to natural animals as other depictions of jaguars, pumas, llamas and monkeys tend to do. Chakipampa artists depicted animal icons from various viewpoints (e.g., from above, from the side, upside-down) as though the positions of the animal were unrestricted and only on two-dimensional media, such as pottery and textiles. The animal designs on the Nasca desert plains created by the Nasca artists are a similar example of such unrestricted, two-dimensional representation (see Shimada in this volume). Thus, I suggest that the highland artisans, presumably including weavers whose products did not survive, were documenting a mythical system that existed on a two-dimensional or flat, visual surface and that lacked directional restrictions, such as a top and bottom or left and right sides to which icons could be aligned. Thus, this style change from Huarpa's geometric style to Wari style animal icons represents a newly learned ability to illustrate abstract concepts, albeit mythical ones.

      For several reasons, I suggest that Wari artists were documenting their relationships with a supernatural world of mythical animals that represented their belief and explanation of the two-dimensional field of stars in the night sky. It has already been suggested that the circled dots may represent stars and Menzel (1977: Figs. 63 upper right, 124) labeled several animal icons as 'star animals'. Pantheons of mythical beings, of course, typify European astronomical iconography, but are also integral to astronomical observations in many complex religions. In Inca religion some stars and constellations were represented by animal spirits (Menzel 1977: Fig. 34). I greatly doubt that Wari populations ignored stellar phenomena. The following reasons should also be considered. The night sky contains numerous combinations of white points and dark areas. Rather than three or four icons, the extensive number of animal icons depicted on pottery indicates that potters were working in a realm of numerous iconic descriptions. Even late Nasca weavers who were contributing an abundance of innovative creature designs (A. Rowe 1979:117) may have had constellations as models, though the angular, fillet-banded executions do not include most Chakipampa B icons. Furthermore, there are star combinations that are observably dominant such as the three stars in a row in Orion, the zigzag pattern in Cassiopeia, and circlets of stars, especially in Scorpio. These have iconic equivalents, such as the Ayacucho serpent, zigzag-backed animal, and the many versions of the hump-backed animal icons. Moreover, the features coinciding with star patterns are consistently depicted by the body structure of the icon; only at the level of small design elements - such as eyes, appendages, and teeth which would have lacked stellar equivalents to follow - did the artists show variation. Artisans also may have depicted the half moon phase as the divided eye on some of these nocturnal icons as it passed into a constellation.

      The rare occurrence of these icons other than at Wari and its south coast sites suggests that Wari leaders contained this mythology and allowed little if any participation by colonists or foreigners. I suggest that the brief histories of late Chakipampa art and early Wari colonial occupation may represent the failure of Wari leaders to use mythical symbols of power in managing their political domain. Epoch 2 art provides evidence that perhaps Wari leaders eventually attempted such a strategy in state government that allowed for more egalitarian participation and political dependence on a Wari belief system.

5. A.D. 750/800: Transition to Epoch 2

      Returning to the Wari chronological sequence based on stratified occupational refuse I will now discuss the evidence for the second mythological movement that eventually dominated Epoch 2 art.

      In 1942, Julio C. Tello discovered and excavated numerous, large urns that were ritually smashed and buried at the site of Conchopata (Menzel 1964: 6).
Page 114
The second iconographic change in Wari art is the depiction of anthropomorphic mythical beings as shown on these urns, known as the Conchopata style and assigned to Epoch IA (Menzel 1964; Cook 1987; Isbell 1987). In 1977, a second smashed and buried offering of large, face-necked jars was discovered at Conchopata and identified by Menzel as Conchopata IB style pottery and assigned to Epoch IB (Cook 1979; 1985). The ceramic offerings at Conchopata, however, appear alien to the site due to a lack of evidence for iconographic antecedents.

      The question here is whether Wari artisans were responsible for the replacement of the iconography of mythical icons having animal attributes by mythical icons having human attributes. A few sherds from Conchopata excavations provide evidence for a possible interpretation of events. At Conchopata stratified refuse layers in one long excavation trench (Isbell 1987: Fig. 9) contained many Chakipampa B ceramics and a few sherds from Conchopata style vessels. The depth of occupational refuse associated with Chakipampa pottery averaged one meter at Conchopata and 2.5 meters at Wari. Strata into which the elaborate offering of face-necked jars was deposited (Cook 1987) contained pottery examples of large urns with poorly executed chevron bands at the rim, simple face-necked jars with the pendent rectangle design on the cheek area and other Chakipampa B pottery with a black background. This evidence indicates that Wari occupants settled the area before the introduction of this offering. Above these strata, the next stratum, which may have existed prior to the face-necked jar offering, contained pottery examples of fancy Chakipampa B style icons of the ventrally-extended, triangular-tailed creature and Ayacucho serpent, designs of vertical and horizontal chevron bands, symmetrical rays, recurved-rays, s-shaped rays, and straight-sided open bowls with vertical wall angles, and, again, simple face-necked jars. This stratum's sample included one sherd from the face-necked jar offering, but this sherd came from an excavation unit 6 meters south and perhaps occurred due to ancient surface disturbance.

      Of great importance is one sherd from this stratum that shows part of a Conchopata style angel icon. The partial design has a tripartite plume on a short, horizontal, 3-fillet stem connected to a mouth with interlocking canines, a nose as a circled dot, and a small rectangular section of the headdress above the nose. This example is not duplicated in the 1942 Tello excavated sample.

      The next stratum above was an occupation surface of hard-packed sandy soil. This stratum contained more Chakipampa B ceramics. Cut into these last two strata just north of the offering was a trench, presumably for a wall, which was filled with Chakipampa B sherds and two Conchopata style sherds. One of the latter shows a section of the bodiless angel head identical to those found on pottery from the Tello excavation (Spielvogel 1955: Pl. 55; Cook 1987: Fig. 38). The second contains a section of another variation of the angel icons in which the nose was drawn as a curl and the tripartite plume caps a single-fillet band that curves up from the mouth in front of the nose. These examples suggest that other offering deposits may exist, and that the offering of smashed urns excavated by Tello occurred after the offering of the face-necked jars and long after the beginning of Wari occupation.

6. Sources of the Epoch 2 Iconography

      Given this chronology of events and the intrusive nature of the offering pottery, I suggest that this second iconographic development came from a population foreign to Wari artisans. As stated above, the Central Deity Theme represents more ancient iconography (Bennett 1954: Fig.30). J. Rowe (1977) elaborated upon the origins of the deity figure with plumes radiating from its head and the numerous creature configurations and design elements that came from Chavín (J. Rowe 1967), an Early Horizon site in the north highlands, and Pucara (Rowe and Brandel 1971), an Early Intermediate Period site located 50 km northwest of Lake Titicaca. Who maintained and reintroduced this ancient iconography?

      In the Wari excavation of stratified occupational refuse, Chakipampa B pottery fragments of face-necked jars provide evidence that potters were depicting individuals with long black hair, black side-burns,
Page 115
head bands with a chevron band design, and cheek designs or tattoos of rectangles divided into vertical bands of different colors. Similar individuals were depicted on the large face-necked jars in the Conchopata B style offering deposit yet with more elaborate facial designs, mustache and chin whisker attributes, or nothing at all (Cook 1987). On the bodies of the jars, artists painted the Central Deity Theme. These renditions were transcribed in a highly abstract manner in which many details are correctly included yet incorrectly drawn when compared to the examples excavated by Tello. I suggest that the Wari artisans were at this time not in direct contact with the source of this iconography that the iconography was transferred linguistically by these elite individuals with chevron band headdresses and was then reinterpreted by the artisans.

7. Pachacamac

      The homeland for these people wearing chevron headbands may have been at Wari. However, such individuals were also depicted by Nasca (Spielvogel 1955: Pl. 119, top) and Pachacamac potters (Fig. 6) (Spielvogel 1955: Pl. 56, top). What is of great importance is the fact that Pachacamac potters depicted these individuals and many others. Such a statement implies that the Pachacamac style art is not exclusively an Epoch 2 phenomenon as presently hypothesized (Menzel 1977), because the Conchopata B examples are contemporary with late Chakipampa B style pottery and date to Epoch 1. Based on a preliminary analysis of the Pachacamac style using the pottery collections of Berkeley's Lowie Museum and Berlin's Volkerkunde Museum and the Uhle collection at the University Museum in Philadelphia, I found that specific vessel shapes and icons can be seriated to form an Epoch I style group.

      This early Pachacamac, here labeled Pachacamac IB (see Table 1), style is defined by a creature icon known as the Pachacamac Griffin, the Front-faced Deity icon and human figures. The Griffin icon represents a mythical animal with bird head, wings, and tail with mammalian fore- and hind-legs. The Griffin and Deity icons occur most often on cups. The cups are straight-sided with flat bases and often display the Chakipampa dark orange background. There are about 50 examples in the Berlin collection with the Griffin and Deity icons that indicates that the activity involving these cups, perhaps ritual drinking was practiced frequently or by many individuals.

      Though these cups were numerous in the Berlin collection, they are rare or non-existent in burial associations with Epoch 2 Pachacamac materials, for example, in the published examples of grave lots cited in Menzel's (1964: 53-55) description of the Pachacamac style. In the Epoch 2 Pachacamac material excavated by Uhle only one cup (Uhle 1903: Fig. 19) with a griffin icon was found in one of the oldest graves and, therefore, was probably an heirloom. According to Uhle's field notes, the only cup with a staff-bearing, Front-faced Deity icon (Uhle 1903: Fig. 16) was 'presented' to him and did not come from his excavations.

      Nineteen Pachacamac-style face-necked jars are in the Berlin collection (Schmidt 1929:Abbs.267, upper right, 283; Tafel III, left). The face on the neck of the vessels was created with slight modeling and icons were painted in paneled areas just as on Wari examples (Fig. 5). Again, Uhle found no similar examples in the Epoch 2 Pachacamac excavations. At Ancón a similar face-necked jar was found yet its modeling and designs indicate non-Pachacamac manufacture (Strong 1925:Pl.46f). These face-necked jars depict several different individuals wearing similarly styled headbands and tunics. Some had earplugs, one had a nose plug, eight had gray painted noses, nine had pendent rectangles painted on the cheek areas; various geometric designs, bird and feline animals were depicted in the paneled areas on each vessel's body. Further analysis of Pachacamac material will, I expect, seriate other vessels and designs into Epoch -1 and 2 temporal units. I suggest that Pachacamac, which was known historically as a place of oracles, earned the reputation of a pilgrimage site during Middle Horizon I with priestly individuals who proselytized the new mythology.
Page 116

8. The Central Deity Theme

      The dominant icon in this new Middle Horizon mythology was the Front-faced Deity with a headdress of radiating plumes. Menzel (1977:61) suggested that the Wari and Inca shared the concept of sky deities, such as the sun, moon and thunder. Perhaps this icon represents the sun. As the sun deity, this icon could represent great power or control by dominating the daytime sky (Fig.7). The Pachacamac Griffin icon is also significant because it complements the earlier Wari mythology of animal icons, as well as more ancient examples of the Chavín and Pucara icons that combined four limbs and wings, thereby endowing the mythical animals with the supernatural attribute of transition between running and flying. Artists gradually modified this icon into a well-defined combination of designs that occurred on Epoch 2 Pachacamac (Schmidt 1929:Abbs.284, 281, lower left) and Tiwanaku pottery (Posnansky 1958:Pls.10a, b) (Menzel 1977:31). I suggest that the animals and the Front-faced Deity icons formed a complete mythological explanation for the sun and stars in which supernatural animals could run across the earth, fly into the night sky and return to earth, thereby accompanying and attending to the sun deity. Epoch I Pachacamac art may provide the necessary icons of Front-faced Deity and Griffin, but the Central Deity Theme is less apparent. This theme requires a hierarchical relationship of animal icons as angel attendant on a central figure to complete the highland version represented on the Tello offering vessels. It appears that the Pachacamac potters were not afforded much space to combine secondary figures with a central figure on these small cups, unlike the large Conchopata urns that were manufactured to provide large design areas. Fortunately, one cup in the Berlin Museum's Pachacamac collection was decorated with the Griffin as a secondary figure - moreover, as a standing attendant - with profile head and facing a central front-faced figure holding two staffs ((Fig. 8A, 8B). This mapping of icons onto a conceptualized canvas of constellations
Page 117
and sun provides a possible model as to how artisans and religious leaders developed the Central Deity Theme.

9. A.D. 800-900: Epoch 2

      What happened next has often been described as a political upheaval or hiatus, in that Epoch 2 Wari occupation and art changed rapidly and spread to new areas rather abruptly (e.g., Huamachuco (Topic and Topic 1984), Azangaro (Anders 1986), Jincamocco (Schreiber 1987), Chimú Capac (Menzel 1977), Jargampata (Isbell 1977)). I suggest that the character of the change in Wari art is evidence of a new political strategy that did not disrupt Wari leadership but empowered its leaders and enhanced its appeal to foreign polities.

      The Central Deity Theme represents the development of a religious mythology with a hierarchical structure. However, Middle Horizon artifacts indicate that the theme was rarely depicted; rather, icons were displayed individually on stone sculpture, pottery, textiles (Fig. 9) and ornaments. Dissemination of this information evolved from the limited, one-time events of large vessel offerings at Conchopata and Pacheco, a site in the Nasca Valley, to numerous, reusable small cups. Some cups display only the face of an icon with the supernatural attributes of the divided eyes and interlocking canines. Other cups, some formed into the well-known 'lyre' shape, display a bodiless, square head with a plumed headdress (Fig. 10) .
Page 118
The face often lacks supernatural attributes; rather, the block-headed icon shows a Wari leader wearing the symbols of supernatural power. The broad spatial distribution of this icon on cups as well as on other vessels (Fig. 11) indicates a more egalitarian access to Wari state identity. By wearing tunics displaying the attendant angel icons (Fig. 12) and headdresses of gold plume ornaments (Eisleb and Strelow 1980:Abbs.305, 306), Wari leaders could embody mythical powers of a central deity. By allowing religious leaders to carry the rituals into various polities, Wari leadership could be recognized as a centralized power. Thus, the Central Deity Theme was transformed into a political strategy as the Central Leader Theme.

      As in Epoch 1, Epoch 2 potters also depicted numerous individuals on face-necked jars. Although the ethnicity and social status of these Wari individuals demands further analysis, several patterns are pronounced. Those wearing four-cornered hats (Fig. 13) seem to be high status individuals in Wari society. They are often depicted in ritual events that involve cups. The drinking cups do not necessarily show mythical iconography (Fig. 14; Eisleb and Strelow 1980:Abb.260). One individual, in particular, whose face was painted in a geometric, step-fret pattern appears on many artifacts and may be representative of an ethnic group identity.
Page 119
On one textile from Ica, this individual is depicted three times and is central to a scene where the weaver documented the bipartisan relationship of numerous individuals or ethnic groups ( See Hearst 4-4556 for Fig. 15). On the textile, the individual faces a full-bodied depiction of another individual; both are held at the top of their heads by standing, full- bodied Griffins. Perhaps the weaver was documenting a political event in which the supernatural power of the mythical Griffin animal was necessary to unify opposing social groups.

10. Tiwanaku

      Epoch 2 artisans translated the mythical world into icon representations on artifacts that were easily transported throughout the Wari realm and into the Lake Titicaca area where another "empire", Tiwanaku, is assumed to have existed. The site of Tiwanaku is located 16 kilometers from the southeast edge of Lake Titicaca at an altitude of 3,660 meters. Archeological remains include a central core of monolithic-walled buildings, sunken courtyards, pyramids and sculptures, and a surrounding area of artifacts measuring about 3 km2. Between the Kisma-Chatta mountain range to the south and the Achuta mountain range to the north, the site is strategically located as a gateway between the lake and the altiplano. From many years of excavations at the site, archaeologists have argued that several building phases were involved. Historical looting of the site created a great deal of damage including fallen walls, smashed statues and a large, gaping hole in the middle of the large, pyramidal structure of Akapana. Though the ceramic sequences appear to complement each other (Bennett 1934, Ponce Sanginés 1972, Lumbreras 1974), the chronologies being used today are not well supported by stratigraphic data or radiocarbon dates. Bennett (1934) excavated Tiwanaku to gain a "stratigraphic pottery series" to support the stylistic sequence of pottery. This sequence had three divisions: Early, Classic and Decadent. Unlike his later report on Wari (Bennett 1953), Bennett's Tiwanaku report is almost devoid of illustrations of his excavated material. Even though the strongest evidence for his sequence came from Pits V and VIII, particularly level 2 of Pit VIII that contained a solid mass of some 1,878 sherds, none were illustrated. Only whole vessels were illustrated and these examples came from unprovenienced museum collections
Page 120
or from excavated deposits with Classic and Decadent style pottery found together and, therefore, contemporary.

      Data in Bennett's tables do not support his sequence either. The "Early" phase is defined by museum pieces (Fig. 16). Only plain ware shapes or pottery with simple linear or wavy lines were found in the levels he designated as Early. The Early bowl with modeled animal heads, 'Shape Bf", was found in both Early and Classic deposits. In reading his report, it is literally impossible to discover examples of the Early Tiwanaku angular design, long-necked decanter shape, and spittoon vessel shape (Bennett 1934:Fig. 13c', 14b) in any of his stratigraphic tables. Thus, the Early style pottery is chronologically unassociated in the sequence. In fact, eight thermo luminescent tests made on Early style pottery date between 830 and 940 A.D. (Eisleb and Strelow 1980: 171).
Page 121
The Early style pottery may simply represent contemporary, local manufacture or post-date the other types. Bennett's sequence of Classic (Fig. 17) and Decadent (Fig. 18) style pottery are not well supported by the stratigraphic findings either. His findings showed that the two types were either isolated in different pits indicating a spatial and functional distinction of this pottery or mixed within the same pit indicating a contemporaneity in their use. No pit showed distinct stratification to support a temporal distinction in the use of these types. For example, the wide open, flaring rim bowl, Shape C, has typical Classic designs and not Decadent designs, yet was found in greatest frequency in deposits that Bennett labeled as Decadent in date, including Pit IV, which is suppose to date entirely to the Decadent period. Bennett (1934: 403, 455) admitted that further analysis is necessary and that his stratigraphy is not well supportive of the pottery sequence. His style of excavating by arbitrary levels contributed to the confusion of temporal distinctions, also. Thus, Bennett's report only reiterated the pottery sequence based entirely on a logically deduced analysis of development from Classic pottery of richly painted and varied vessels to Decadent pottery of drab colors and restricted designs. Artistic content went logically from full-bodied depictions to abstract elements. In contrast, if the pottery types were viewed as contemporary, then the evidence might suggest that the Tiawanaku site was inhabited and visited by many diverse social groups who participated in a pan-Andean religious, political and economic system.

      Carlos Ponce Sanginés did not excavate Tiwanaku to clarify Bennett's sequence. His work was aimed at determining the construction phases of the site. He accepted and renamed the pottery phases. He added two phases, Phases I and II, that pre-date Bennett's Early phase and represent a formative period. Bennett's Early and Classic phases are Ponce Sanginés' (1978) Phases III and IV, respectively, that are designated as Tiwanaku's urbanization and regional development period dating to approximately 30-824 A.D. The Decadent phase is Ponce Sanginés's Phase V during which a Tiwanaku empire expanded and lasted until approximately 1200 A.D. The relevant radiocarbon dates are somewhat difficult to interpret. Of the eight samples obtained by A. Kidder II and W. R. Coe in 1955 (near Bennett's test pits) and published by Ralph in 1959 (Ponce Sanginés 1981: Table 2), five samples date the Early phase as 30-690 A.D. and three samples date the Classic phase as 145-824 A.D. Again, this evidence is more supportive of contemporaneity of the pottery types found in the excavations. Thus, Ponce Sangines' work contributed little to the pottery sequence problem, but does contribute to our understanding of the building phases of the site. He suggested that construction of the temples in the central area of the site, such as Kalasasaya and the Semi-subterranean Temple, dated to Phase III. Modification of these temples, construction of peripheral buildings and the addition of sculpture that display icons of the Central Deity theme date to Phase IV.

      Tiwanaku iconography as found on the stone sculpture of Phase IV is most similar in artistic execution and content to the urns excavated by Tello in 1942 at Conchopata, and the textile, metal and wood carved artifacts found at Middle Horizon 2 coastal sites, as well as on coastal Atarco style pottery dating to Epoch 2. Tiwanaku iconography is also limited in what the artisans were attempting to document. They depicted only the mythical animal and Central Deity icons of the second mythological movement, a few human participants, but little else for which an urban environment would provide models. In contrast, the Nasca, Moche and Wari artisans depicted their local flora and fauna, daily activities with both human and animal participants, as well as mythical icons.
Page 122
I suggest that Tiwanaku art was used for very focused, ritual activities. Since the site's iconography is restricted in composition and appears to have been added after construction of the principal temples, I further suggest that Wari political leaders introduced the mythical symbols and rituals of centralized power to the local Tiwanaku leaders and their people. Either by negotiation or, perhaps, by force, Wari may then have transformed Tiwanaku into another center of pilgrimage, whereby Andean populations could participate peacefully in a vital, mythical belief system that interfered minimally with local, everyday activities yet refaced the ritual functions of major, ancestral temples.


      Obviously, there can be numerous factors involved in the demise of a prehistoric state and abandonment of its urban capital and associated sites. For example, social revolt, environmental catastrophe, and/or urban health problems may have occurred at Wari and disrupted the state's leadership. Based on iconographic documentation, I would like to discuss another possible factor. At the Temple of Pachacamac, Uhle excavated numerous Middle Horizon graves. One of the oldest graves, H7c, underlay the last Middle Horizon reconstruction phase of the temple of Pachacamac and contained many vessels and a badly damaged large textile with an elaborate scene (Fig.19-21) (Uhle 1903: Fig.18; Pl.4, Figs.1-3; Pl.5, Figs.1,9,10,11). The artist documented an event in which a central figure held a tumi knife and trophy head and was flanked on either side by attendants. Two breaks in the cloth and additional fragments suggest that there may have been other attendant figures. Along the right edge four other individuals were painted, two full-bodied and two bodiless. This textile is a remarkable document in several ways. First, the Central Deity Theme continued into Epoch 2 as the Central Leader Theme; all figures are clearly human and hierarchically arranged. Secondly, several designs such as the T-shaped plume on the central figure's headdress and the animal-headed rays are similar to Moche art (Donnan 1976: Fig. 6). Of the four side figures, the headdress of the top and bottom figures and the curved body of a middle figure are antecedent to carved figures on the wooden Idol of Pachacamac (Jímenez 1985). This wooden idol displays animal designs found on Early Sicán pottery (Shimada 1990) and on blackware from the Middle Horizon site of Chimú Capac, located in the Supe Valley, (Menzel 1977: Figs. 60, 61).
Page 123
Thus, the artist documented the gradual abandonment of Wari mythology and assimilation of North Coast Moche culture at Pachacamac. By A.D. 900, concurrent with the Sicán on the northern North Coast, the Pachacamac artisans had blended Moche and Wari iconography of central figures and animals into what has come to be called Chimú art. Over the next several generations, the growing political unity of Chimú societies may have drawn heavily from the Wari political strategy. For example, it appears that the authority and prestige of the individuals who wore the Wari four-cornered hats persisted, as is documented by potters in Sicán blackware, particularly that of Middle Sicán, A.D. 900-1100. Perhaps these individuals, by wearing symbols of mythical power, established and sustained an elite authority over this new populace as Wari political power waned.


      Ancient artisans provided numerous artifacts that aid archaeological analysis of complex societies. I emphasize that such artifacts are more than cultural products for describing the functions of day-to-day prehistoric behavior. More importantly, I interpret the iconography of art as a prehistoric record by which artisans documented their observations of social events. In the case of the Wari empire the archaeological data provides evidence of two mythological belief systems. In the earlier one artisans depicted icons that represented mythical star animals. In the later one, Wari artisans borrowed Pachacamac religious iconography of anthropomorphic deities that translated the mythical symbols of the sun and stars into political symbols of centralized power. Epoch 2 Wari leaders wore these symbols to identify themselves as central authorities. As the icons became more accessible artisans depicted many individuals of high-ranking status who wore these symbols of power. Consequently, the demise of the Wari state may have been partially attributed to the increasing secularization of mythical power and dissipation of a centralized source of religiously sanctified authority.

Figures in Original Publication - Figures provided for this web version

Fig. 1 Map of Peru Same

Fig. 2A Wooden lime container shows anthropomorphic Central Deity with plumed headdress and the supernatural attribute of interlocking canines. Provenience: Pachacamac. Museum für Völkerkunde, Berlin. VA40419.

Fig. 2B Reverse of 2A Same with additional side view as Fig. 2C.

Fig. 3 Tall, Nasca vase shows borrowed Huarpa designs of alternating triangles and checkerboard pattern. Provenience: Majoro Grande, Nasca. Robert H. Lowie Museum of Anthropology, University of California at Berkeley. 4-8559.

Fig. 4 Tall, ceramic goblet shows jagged band of triangles that represent the final stylistic abstraction of the fish fin design. Provenience: possibly Nasca or Ica. Denver Art Museum. 1984.637. Single spout, tube-handled bottle with "fish fin" design. Museum für Völkerkunde, Berlin VA51171

Fig. 5 Face-necked Nasca jar shows the borrowed Chakipampa B version of the ventrally-extended creature icon with a triangular tail. Provenience: Nasca. From the Collections of the Putnam Museum, Davenport, Iowa. AR7346. Tall goblet/kero with the "ventrally extended, triangular tailed" animal. Provenience: Nasca, Ht. 20.1 cm. D. 15.4 cm. Museum für Völkerkunde, Berlin VA51130 (Eisleb 1977:Fig.243)

Fig. 6 Face-necked jar represents a Pachacamac elite wearing a chevron band headdress and tunic of mythical icons. Provenience: Pachacamac. Museum für Völkerkunde, Berlin. VA49531. Same with additional view. The 1989 publication is most similar to image on right. Also, the photo was printed reversed so chevron band is incorrectly pointing to the right in the 1989 publication.

Fig.7 Tall cup shows Pachacamac I style Central Deity icon with supernatural attributes of the divided eyes and interlocking canines. Provenience. Pachacamac. Museum für Völkerkunde, Berlin. VA 49212. Same with extra views

Fig. 8 One side of a cup showing a standing Griffin icon and the two staffs held by a central figure icon on the other side. Provenience: Pachacamac. Ht.13.8 cm. Museum für Völkerkunde, Berlin. VA49237.

Fig. 8B Reverse of 8A. Same though the museum staff kindly restored the cup completely for the 1989 publication photos.

Fig. 9 Wari textile with Griffin head and feline tail icon woven into mirrored symmetry that creates bicolor-headed icons. Provenience: Huarmey, 160 x 62 cm. Fundación Museo Amano. R.0191. Stylized feline attendant icon with split face and curled tail designs on Wari textile. Provenience: between Ica and Pisco. 210 x 51.5 cm. Museum für Völkerkunde, Berlin. VA16297

Fig. 10 Lyre-shaped short cup shows the block-headed icon of a Wari leader. Provenience: Huarmey, Ht. 9.2 cm., D. 9.7 cm. Fundación Museo Amano. R.1509. Lyre-shaped cup shows the block-headed icon of a Wari leader. Provenience: Nasca. Ht.9.8 cm. Museum für Völkerkunde, Berlin. VA51110

Fig. 11 Double-spout, strap-handled jar shows the block-headed icon of a Wari leader. Provenience: Huayuri. Nasca. Ht.12.5 cm. Fundación Museo Amano. R.0971. Double-spout,strap-handled jar shows the block-headed icon of a Wari leader. Provenience: Pachacamac. Width 17.3 cm. VA49489

Fig. 12 Tapestry-woven textile shows attendant angels. Provenience: Palpa. 222 x 113 cm. Fundación Museo Amano. R.0206. Textile of attendant angels with a stylized icon of the hallucinogenic plant, Anadenanthera colubrina, below chin band on angel's chest. Provenience: Pachacamac. 110 x 68 cm. Museum für Völkerkunde, Berlin VA30963 (Eisleb and Strelow 1980:Fig.328a,b)

Fig. 13 Four-cornered hat with Griffin icons. Provenience: unknown. Ht. 14 cm. D. 55 cm. Fundación Museo Amano. R.0206. Four-cornered hat with stylized image of the hallucinogenic plant, Anadenanthera colubrina. Ht. 10 cm. Width 43.7 cm. Museum für Völkerkunde, Berlin VA63996 (Eisleb and Strelow 1980:Fig.343)

Fig. 14 Cup shows the white outlined black bands and nested, white triangles that occur on Epoch 2 Wari pottery that includes a face-necked jar of an individual wearing a four-cornered hat. Provenience: south coast. Anthropology Collections, Stanford Museum, Stanford University. 60-1009. A similarly designed cup in the hands of an elite with a four-cornered hat (i.e., the Epoch 2 Wari example mentioned above), possibly indicating a ritual toasting to the gods. Ht. 15.6 cm. Museum für Völkerkunde, Berlin VA8057 (Eisleb and Strelow 1980:Fig. 260).

Fig. 15 Textile tapestry fragment shows elaborate arrangement of Middle Horizon individuals or ethnic group icons. Provenience: Site E, Ocucaje, Ica Valley. 92.5 x 52 cm. Robert H. Lowie Museum of Anthropology, University of California at Berkeley. (see Hearst 4-4556). Same, museum is now the "Phoebe Apperson Hearst Museum of Anthropology".

Fig. 16 Wide rimmed bowl of Bennett's Early style or Ponce Sanginés's Tiwanaku III style pottery. Provenience: Tiwanaku. Ht.17.2 cm. Museum für Völkerkunde, Berlin. VA 30836. (Eisleb and Strelow 1980:Fig.12a,b). Wide rimmed bowl of Bennett's Early style or Ponce Sanginés's Tiwanaku III style pottery. Provenience: Tiahuanaco. Ht.14.1 cm. D. 22.8 cm. Museum für Völkerkunde, Berlin. VA 16708. (Eisleb and Strelow 1980:Fig.13).

Fig. 17 Bowl with modelled puma head of Bennett's Classic style or Ponce Sanginés's Tiwanaku IV style pottery. Provenience: Tiwanaku. Museum für Völkerkunde, Berlin. VA 64591. (see Eisleb and Strelow 1980:Fig.119a,b)

Fig. 18 Kero of Bennett's Decadent style or Ponce Sanginés's Tiwanaku IV style pottery. Provenience: Tiwanaku. Ht.21.9 Dm.16 cm. Museum für Völkerkunde, Berlin. VA 10475. (Eisleb and Strelow 1980:Fig.84) Same with extra views

Figs. 19-21. Painted textile of the "Central Deity/Leader" theme. Provenience: Grave H7c, Pachacamac. University Museum of Archaeology/Anthropology, University of Pennsylvania. 26721 A, C, D. (See Uhle 1903:Pl.4, Figs.1a,b,c [Reprinted by Shimada 1991).


Anders, Martha B.
1986 Dual organization and calendars inferred from the planned site of Azangaro: Wari administrative strategies, v.1-3. Unpublished Ph. D. dissertation, Department of Anthropology, Cornell University.

Benavides, Mario C.
1963 Estudio de la ceramica decorada de Qonchopata. Tesis de licenciatura, Universidad Nacional de San Cristóbal de Huamanga. Ayacucho.
1984 Carácter del Estado, Wari. Universidad Nacional de San Cristóbal de Huamanga. Ayacucho.

Bennett, Wendell C.
1934 Excavations at Tiahuanaco. American Museum of Natural History Anthropological Papers, No. 35, pt. 3. New York.
1953 Excavations at Wari, Ayacucho, Peru. Yale University Publications in Anthropology, No. 49. New Haven.
1954 Ancient Arts of the Andes, with an introduction by René d'Harcourt. The Museum of Modern Art, New York (distributed by Simon and Schuster, New York).

Brewster-Wray, Christina
1983 "Spatial patterning and the function of a Huari architectural compound." Investigations of the Andean Past, edited by Daniel H. Sandweiss, pp. 125-135. Cornell Latin American Studies Program, Cornell University. Ithaca, New York.

Conklin, William J
1970 "Peruvian textile fragment from the beginning of the Middle Horizon." Textile Museum Journal 3(1):15-24. Washington, D.C.

Cook, Anita G.
1979 The iconography of empire: symbolic communication in 7th Century Peru. Unpublished Master's thesis, Department of Anthropology, State University of New York at Binghamton.
1985 Art and time, in the evolution of Andean state expansion. Unpublished Ph. D. dissertation, Department of Anthropology, State University of New York at Binghamton.
1987 "The Middle Horizon ceramic offerings from Conchopata." Ńawpa Pacha, 22-23(1984-1985):49-90.

Dawson, Lawrence
n.d. Personal communication, 1983.
n.d. Personal communication, 1988.

Donnan, Christopher B.
1976 Moche Art and Iconography. UCLA Latin American Center Publications 33. University of California, Los Angeles.

Eisleb, Dieter
1977 Altperuanische Kulturen II: Nazca . Museum für Völkerkunde. Berlin.

Eisleb, Dieter and Renate Strelow
1980 Altperuanische Kulturen III: Tiahuanaco. Museum für Völkerkunde. Berlin.

Gayton, Anna and Alfred Kroeber
1927 The Uhle pottery collections from Nazca. University of California Publications in American Archaeology and Ethnology 24(1). University of California Press, Berkeley.

Harcourt, Raoul and Marie d'Harcourt
1924 La Céramique Ancienne du Pérou. Paris.

Isbell, William H.
1977 The Rural Foundation for Urbanism: Economic and Stylistic Interaction between Rural and Urban Communities in Eighth-Century Peru. Illinois Studies in Anthropology, No.10. University of Illinois Press, Urbana.
1987 "Conchopata, ideological innovator in Middle Horizon IA." Ńawpa Pacha 22-23(1984-1985):91-126.

Jiménez Borja, Arturo
1985 "Pachacamac". Boletín de Lima. 7(38):40-54.

Knobloch, Patricia J.
1983 A study of the Andean Huari ceramics from the Early Intermediate Period to the Middle Horizon Epoch 1. Unpublished Ph. D. dissertation, Department of Anthropology, State University of New York at Binghamton.

Kutscher, Gerdt
1954 Nordperuanische Keramik: Figürlich Verzierte Gefässe der Früh-Chimu. Verlag Gebr. Mann. Berlin.

Lumbreras, Luis Guillermo
1974 The Peoples and Cultures of Ancient Peru. Smithsonian Institution, Washington, D. C.

McEwan, Gordon
1984 The Middle Horizon in the Valley of Cuzco, Peru: the impact of the Wari occupation of Pikillacta in the Lucre Basin. Unpublished Ph. D. dissertation, Department of Anthropology, University of Texas at Austin.

Menzel, Dorothy
1964 "Style and time in the Middle Horizon". Ńawpa Pacha 2:1-106.
1968 "New data on the Huari Empire in Middle Horizon Epoch 2A". Ńawpa Pacha 6:47-114. Berkeley.
1977 The Archaeology of Ancient Peru and the Work of Max Uhle. R. H. Lowie Museum of Anthropology, University of California, Berkeley.

Muelle, Jorge C. and Camilo Blas
1938 "Muestrario de arte peruano precolombino." Revista del Museo Nacional 7(2):163-280. Lima.

Ponce Sanginés, Carlos
1972 Tiwanaku: Espacio, Tiempo y Cultura: Ensayo de Síntesis Arqueológica. Primera edición.. Academia Nacional de Ciencias de Bolivia, Publicación 30. La Paz.
1978 Panorama de la Arqueología Boliviana. Instituto Nacional de Arqueologia, Publicación 27. La Paz.
1981 Descripción Sumaria del Templete Semisub-terráneo de Tiwanaku. 5a ed. Librería y Editorial "Junventud", La Paz.

Posnansky, Arthur
1958 Tihuanacu: la Cuna del Hombre Americano. Tihuanacu: the Cradle of American Man, Vols. III and IV. Ministerio de Educación, La Paz.

Rowe, Ann P.
1979 "A late Nasca derived textile with tapestry medallions." Bulletin of the Detroit Institute of Arts 57(3):115-123.

Rowe, John Howland
1967 "Form and meaning in Chavín art." In: Peruvian Archaeology: Selected Readings, edited by John H. Rowe and Dorothy Menzel, pp. 70-103. Peek Publication, Palo Alto.
1977 "El arte religioso del Cuzco en el Horizonte Temprano." Ńawpa Pacha 14(1976):1-20, pls. XI-XVI.

Rowe, John Howland and Catherine Terry Brandel
1971 "Pucara style pottery designs." Ńawpa Pacha 7/8(1969-70):1-16, pls. I-XVIII.

Sawyer, Alan R.
1975 Ancient Andean Arts in the Collections of the Krannert Art Museum. University of Illinois, Champaign-Urbana.

Schlesier, K. H.
1959 "Stilgeschichtliche Einordnung der Nazca-Vasenmalereien: Beiträge zur Geschichte der Hochkulturen des vorkolumbischen Peru. Annali Laternanensi 23:9-236. Citta del Vaticano.

Schmidt, Max
1929 Kunst und Kultur von Peru. Propylaen-Verlag, Berlin.

Schreiber, Katharina J.
1987 "From State to Empire: the expansion of Wari outside the Ayacucho Basin." The Origins and Development of the Andean State, ed. by Jonathan Haas, Sheila Pozorski and Thomas Pozorski, pp. 91-96. Cambridge University Press, Cambridge.

Seler, Eduard
1961 Gesammelte Abhandlungen zur Amerikanischen Sprach- und Altertumskunde. Akademische Drück - U. Verlagsanstalt, Graz, Austria.

Shimada, Izumi
1990 "Cultural Continuities and Discontinuities on the Northern North Coast of Peru, Middle-Late Horizons." The Northern Dynasties: Kingship and Statecraft in Chimor, edited by Michael E. Moseley and Alana Cordy-Collins. pp. 297-392. Dumbarton Oaks, Washington, D.C.

Shimada, I., Schaaf, C. B., Thompson, L., and E. Mosley-Tbompson
1991 "Cultural impacts of severe droughts in the prehispanic Andes: application of a 1,500-year ice core precipitation record." World Archaeology 22:247-270.

Spielvogel, Rosalind Brueck
1955 Wari: a study of Tiahuanaco style. Unpublished Ph. D. dissertation, Department of Anthropology, Yale University, New Haven.

Strong, William D.
1925 The Uhle pottery collections from Ancón. University of California Publications in American Archaeology and Ethnology 21(4):i-ii, 135-190. Berkeley.
1957 Paracas, Nazca and Tiahuanacoid Cultural Relationships in South Coastal Peru. Memoirs of the Society for American Archaeology 13.

Tello, Julio C.
1942 "Origen y desarrollo de las civilizaciones prehistóricas andinas". Actas Cientificos del 27th Congreso de Americanistas de 1939. Reprint. Librería e Imprenta Gil, S.A. Lima.

Topic, Theresa and John Topic
1984 The Huamachuco Archaeological Project: preliminary report on the third field season, June-August 1983. Trent University Occasional Papers in Anthropology, no. 1. Peterborough.

Uhle, Max
1903 Pachacamac: Report of the William Pepper, M. D., LL. D., Peruvian Expedition of 1896. Department of Archaeology of the University of Pennsylvania, Philadelphia. [Reprinted 1991 University Museum of Archaeology and Anthropology, University of Pennsylvania, Philadephia. Introduction by Izumi Shimada.]

ACKNOWLEDGEMENTS: I am grateful for valuable editorial and substantive comments provided by Patricia J. Lyon and lzumi Shimada. And I sincerely thank Lawrence Dawson who was so generous in sharing his insights on Moche-Nasca stylistic interaction.
*Dr.Patricia J. Knobloch, 9229 Dillon Drive, La Mesa, CA 91941 (619)460-9443