Wari and Tiwanaku shared SAIS iconography, and its diffusion from the two centers defines the Middle Horizon. Other material culture was also spread, including striped tunics of interlocking tapestry, woven on long, horizontal looms, and decorated with distinctively stylized religious images. A tall drinking vessel called a “kero,” pottery decorated with painted polychrome designs, burnished black ceramics, four-cornered hats, metal shawl pins known as “tupu,” Anadananthera colubrina symbols, and bronze metallurgy seem also to have spread throughout the Andes as part of the package.
Early archaeologists considered Middle Horizon cultural diffusion to be a single wave originating at Tiahuanaco, and gradually diminishing with distance. As more research accumulated in the second half of the twentieth century it became clear that there were two variants, each quite consistent within its territory. Tiwanaku is characterized by megalithic stone monuments, vernacular architecture of small adobe rooms, grinding stones with a push-pull mano, distinctive llama mandible tools, conical “trompo” stones, small stemmed projectile points, interlocked tapestry tunics of one piece woven entirely of wool, wooden snuff tablets and other drug-snuffing paraphernalia, as well as pottery that is easily differentiated from Wari ceramics by experienced archaeologists. Tiwanaku people carefully placed trash in deep pits, as though it were ritually polluting.
By contrast, Wari is characterized by rough stone buildings and the indiscriminant dumping of trash in abandoned houses. Public architecture emphasized quadrilateral compounds divided into rectangular apartments so rigidly organized that it has been called orthogonal cellular architecture (Isbell 1991). Less common are D-shaped ceremonial structures. Vernacular buildings are also rectangular, but less regular in plan. Huari people used the rocker grinder. Llama pelvis tools and ovate stone points were popular, but snuff tablets were rare or absent. Interlocked tapestry tunics were very similar to Tiwanaku’s in appearance, but different in construction. They were woven in two pieces and seamed up the middle. Usually they had cotton warp and woolen weft. Four-cornered hats from the two cultures were also similar in appearance, but different in manufacture.
Each city spread its own material culture creating a Wari realm in the north and a Tiwanaku sphere in the south (Figure 37.2). Based on distributions of material remains, especially pottery, the boundary between the Wari and Tiwanaku territories ran through southern Peru, across the Department of Puno, some distance north of Lake Titicaca, to the coast of Arequipa and Moquegua (Rowe 1956). In Arequipa Wari ceramics are found in the Ocona-Cotahuasi Valley (Jennings and Yepez 2002), and farther south to at least the Sihuas Valley. Tiwanaku ceramics dominate northern Chile and Peru’s southernmost Department of Tacna (Berenguer and Daulesberg 1989). In between, from the city of Arequipa to the Moquegua coast, Tiwanaku ceramics occur, but there are also Wari sites and enclaves (Cardona 2002).
From the boundary with Wari, Tiwanaku ceramics and other artifacts extend south through Bolivia and Chile, reaching Potosi, Tarija, and northwestern Argentina at their southeastern extreme. In Chile they are spread throughout the dry Atacama region, at least to San Pedro de Atacama (Torres 2002).
Wari art and artifacts extend far to the north, reaching Cajamarca and the Chotano River in the highlands. On the coast they stretch from Moquegua to at least the Lam-bayeque Valley.
Does the Middle Horizon diffusion of pottery styles and other material culture reveal the spread of empires, perhaps through conquest? Some Andeanists answer “Yes,” pointing to archaeological evidence for intrusive colonies and provincial rule. On the other hand, it is clear that in many places, Wari, and in others, Tiwanaku, cultural remains appear only as prestige goods, or as stylistic influences that do not look like imperial invasion or administrative control.
A puzzling feature of Wari and Tiwanaku spread is action in concert, similar to what is implied for the synthesis of SAIS iconography. Early in the Middle Horizon, perhaps by AD 650 to 700, Wari and Tiwanaku colonists invaded the Moquegua Valley, on the far south coast of Peru (Goldstein 2005; Williams and Nash 2005). Wari occupied the upper valley, and Tiwanaku the middle sector. There is little or no evidence of warfare between the two, or between the new arrivals and the locals, although the possibility cannot be ruled out.
In Moquegua Wari established its primary center on a vertical-sided mesa named Cerro Baul, as well as several neighboring hills. Naturally defensive, Cerro Baul was further strengthened with baffles on the ascending trail, and perhaps additional walls. On the flat summit Wari settlers constructed 4 or 5 ha of stone walled buildings, including orthogonal cellular patio groups, D-shaped temples, feasting halls, a brewery, and other facilities appropriate for residents, administrators and guests, although barracks have not been discovered. A long and impressive canal was dug, claiming new land for cultivation, and bringing water to the foot of the mesa-top town (Williams 2003). A Huari heartland authority at Cerro Baul was commanding labor, and negotiating with Tiwanakan neighbors.
Wari settlements on the surrounding slopes include a range of residential facilities varying from large elite homes to small, humble dwellings (Nash and Williams 2002). Some residents appear to have come from the Huari heartland, while others seem “Warified” locals. Indeed, the archaeological record reveals Huari intrusion, as well as the emergence of a new, Wari identity based on innovative material culture combining Wari with local traditions. At the Middle Horizon Beringa site, several valleys north, Owen (in press) documents similar processes in pottery. Apparently, a new international Wari identity was being forged.
Tiwanaku established two sets of communities downstream from the Wari settlements. They were not fortified, and were characterized by virtually pure heartland material culture, in two versions known as Omo and Chen Chen. Material culture as well as the people themselves, seems to have remained faithfully heartland, with no “Tiwanakanized” locals or hybrid, international assemblages or identities (Blom and Buikstra 1999; Goldstein 2005; Sutter 2005).
At the Omo enclave 20 km from Cerro Baul a medium-sized monument was constructed of adobe. Originally including several walls faced with cut stones, the complex was about 45 m wide and 120 m long. It consisted of three in-line enclosures, each elevated above its predecessor. The first two are open assembly areas. The uppermost was divided into a series of rooms around a small, central sunken court. Goldstein (2005) emphasizes similarities between the M10 building and altiplano temples, interpreting M10 as a heartland installation where religious processions took place. However, M10 is remarkably similar to later Andean palaces, at least as they were described for the Incas, so it may represent the seat of political authority for Moquegua’s Tiwanaku colonists. Was it occupied by a Tiahuanaco governor, or a colonist patriarch? Was control of the colonies centralized rule, or more ceremonial and ideological, as Goldstein now infers?
Tiwanaku’s Moquegua colonists were altiplano immigrants, but there is confusion about how the colonies should be understood. When Paul Goldstein (2005) began investigating Tiwanaku intrusion into Moquegua he emphasized the awesome power of the Tihua-naco state, and colonists’ dependence on it. But progressively, and heavily influenced by
Analogy with modem highland squatters in Moquegua, he stresses the agency and choice of the migrants themselves, even suggesting that the colonists may have migrated to escape state control. Or, perhaps they were refugees from heartland convulsions. Be that as it may, what is clear is that the local Moqueguans were not drawn into an expansive and international Tiwanaku political identity. Even if the Moquegua colonies were ruled from Tiahua-naco, through a governor installed at M10, Tiahuanaco’s behavior is puzzling. Empires promote new international identities. Empires exploit provincial people. Empires do not ignore the conquered locals.
In the southern altiplano heartland, Lukurmata, Khonko Wankani, Pajchiri, Oje, Chucaripupata and Pukuro-Uyu are all Middle Horizon (many have earlier occupations as well) sites that have monumental architectural remains reminiscent of the monumental buildings at Tiahuanaco. None is located more that 50 to 75 km from the ancient capital. Few of these sites have been investigated well enough to fully understand their role in the rise of Tiahuanaco, or in Middle Horizon political organization, but the spatial distribution seems to define a core territory spreading no further than 75 km from the capital. It probably represents the early cultural hearth within which Tiahuanaco developed, as well as an area that was subject to direct rule. What was the history and nature of Tiwanaku beyond this core territory?
The Moquegua colonies had a temple/palace at Omo M10, perhaps representing direct rule from the metropole. So the Middle Moquegua Valley was probably an imperial province of Tiahuanaco, with its capital at Omo M10, although lack of interest in local residents seems contradictory, and peacefully (?) sharing the Valley with Wari even more curious.
Significantly, no other provincial Tiwanaku capitals have been identified by the presence of monumental administrative architecture. A Tiwanaku settlement on the northern Lake Titicaca shore, at modern Puno’s Isla Esteves (Figure 37.2) may have had more than domestic architecture and agricultural terraces, but construction of a luxury hotel destroyed most remains. Stanish (2003) argues that Tiahuanaco directly controlled this area, as well as a corridor connecting the two locations, that skirted the entire west side of Lake Titicaca. But with no administrative capital, archaeologists cannot be sure. Perhaps future research, determining whether there were intrusive colonies, plus “Tiwanakanized” locals, will resolve the question.
During the Middle Horizon, Tiwanaku art and artifacts became extremely common in the agriculturally rich Cochabamba Valley. This eastern Andean valley, long inhabited by complex societies, was famous for maize, and chicha beer brewed from maize. The Inca Empire completely reorganized the valley to extract maize - and left a spectacular mountain-top administrative town, Incarracay, overlooking the basin. Some archaeologists argue that earlier Tiwanaku did the same, citing as evidence Cochabamba’s adoption of Tiwanaku-style pottery (Ponce 1981; Stanish 2003). However, settlement patterns changed little (Higueras 1996), and no Tiwanaku-style administrative architecture has been identified, even at the site with the most Tiahuanaco-like ceramics, Pinami (Cespedes 2000). Furthermore, a study of biodistance based on skulls from the Cochabamba region implies demographic continuity, not the intrusion of altiplano people during the Middle Horizon (O’Brien 1999), as in Moquegua. Much more research is required. Perhaps new investigations will change our ideas about relations between Tiahuanaco and Cochabamba, but at present it seems more likely that Tiahuanaco was an influential religious center, trade partner, and international cultural model emulated by Cochabambans, not an imperial administrator. It is apparent from Cochabamba-Tiwanaku style artifacts that a new international Tiwanaku identity had emerged.
Arica, on the Chilean coast, is another debated case of Tiwanaku colonization. Goldstein (2005) argues that highland Tiwanaku colonists were present, but administrative architecture is lacking. Uribe and Aguero (2002) conclude that the impressive Tiwanaku materials can be accounted for by trade with the neighboring Tiwanaku colonies in Moquegua.
Throughout arid northern Chile, south of Moquegua and Arica, Tiwanaku artifacts occur in select burials. Spectacular textiles appear, as do keros and other ceramics, occasional gold objects, and of course, snuff paraphernalia that bear SAIS icons. Some scholars infer Lake Titicaca colonists (Kolata 1993; Rodman 1992), but most archaeologists point out that the number of Tiwanaku-style objects is actually small, and they occur as a minority of artifacts among a great many more local artifacts in elite graves. The more convincing interpretation is that llama caravans conveyed traders and goods across vast deserts, dispersing products of different microenvironments and craft traditions in a long-standing system of trans-cultural interaction (Berenguer 2000; Berenguer and Daulesberg 1989) that Stovel (Chapter 49 in this volume) describes as a “cultural field.” Apparently, Tiahuanaco was not extending political control or sending colonists into Chile, but participating in an ancient sphere of relations that involved poorly understood movements of people and goods. What was probably new was the promotion of a multinational Tiwanaku identity associated with practices involving Tiwanaku-style material culture.
Wari stands in sharp contrast to Tiwanaku. Administrative architecture similar to that constructed at Moquegua’s Cerro Baul appeared throughout the Wari sphere (Isbell and McEwan 1991, Schreiber 1992) early in its development. The most impressive provincial administrative capital is Pikillacta (McEwan 2005) (Figure 37.8), in the Cuzco Valley about 300 km southeast of the Huari metropole. A rectangular complex encloses 1 sq km
Figure 37.8. Pikillacta, the Wari administrative center in Cuzco. (Map redrawn from McEwan 2005: fig. 2.7; photo by William H. Isbell)
Of orthogonal cellular architecture. Within are walled streets; patio groups containing large plazas surrounded by multi-storied lateral rooms and halls; rows of small, oval rooms that might have been storehouses, and ceremonial courts. There also are roads and irrigation canals, an aqueduct, great walls, and other features that reshaped the entire valley during 2 to 4 centuries of Wari occupation - about AD 650-700 until 900-1000 (McEwan 2005). Twenty km south of Pikillacta is Huaro, another intrusive Wari community, larger in area than Pikillacta, but more residential in nature (Glowacki 2002). Elite tombs, perhaps for Huari governors, were found on Huaro’s Batan Urqu hill (Zapata 1997). Wari ceramic styles significantly altered the pottery traditions of Cuzco, and the labor invested in Wari public architecture ran into millions of man days. There is little doubt that Cuzco became a province administered more or less directly from Huari, and that exploited locals were participating in an empire that promoted a new international identity.
Almost 1000 km north of Cuzco, in the Huamachuco Valley, is Viracochapampa, a Wari orthogonal cellular enclosure similar to, but somewhat smaller than, Pikillacta (Topic 1991; Topic and Topic 2001). Huamachuco had an impressive architectural tradition of its own, which probably contributed as much to Wari innovations as it received. Furthermore, Viracochapampa was not occupied for long, and much of its great enclosure remained vacant of interior compounds. So Wari presence in this northern region is unlikely to have been simple conquest and control as implied for Cuzco. More likely, it represents accommodation to indirect rule, or perhaps an even more mutual process, but surely involving some degree of control as well as the international Wari identity and material culture, at least among Huamachuco’s elites. Perhaps a similar process explains Honcopampa, a town in the next valley to the south, the large Callejon de Huaylas. It seems to have been a center for indirect Huari rule, consisting of several palaces and mortuary monuments that combined local and foreign architectural canons, as well as ceramics.
The most intensive investigation of an intrusive Wari occupation outside the heartland is in the Sondondo Valley, about 130 km south of Ayacucho (Figure 37.2; Schreiber 1992, 2005). A small orthogonal cellular administrative compound, 125 by 255 m was constructed. Katharina Schreiber shows that Wari built roads through the area, relocated all the settlements, constructed a shrine, and terraced and irrigated the valley sides. This is an extremely convincing example of imperial conquest, colonization, and then administrative reorganization and control for imperial benefits. Huari exploited Sondondo.
Schreiber (2005) argues from her Sondondo study that similar behavior was more or less characteristic of Wari, and that consequently, it must be classified in the unilinear stage “empire.” While I agree that Wari is best understood as an empire, classification often results in the imposition of the ideal characteristics on the past, creating unverified knowledge that goes untested because it conforms to our expectations. I suspect that variation in Huari’s provincial policies was extremely great, as were the forms of its control and other influences. In many cases neither conquest, colonization, nor provincial rule seem indicated by the archaeological record. Conversely, even when these classic diagnostics of empire are lacking, as in Peru’s north coastal valleys, Wari cannot be discarded as irrelevant. The presence of Wari ceramics in elite tombs, as at the Jequetepeque Valley site of San Jose de Moro (Castillo 2001), reveals at least some degree of participation in Wari’s new international identity.
Schreiber’s case study in Sondondo, fascinating as it is, may not represent typical Wari provincial rule, as she infers. Wari also occupied neighboring Chicha Valley, but the settlements and the landscape were modified very little (Meddens 1985). Why such difference in policy? A hint may be preserved in Sondondo - megalithic tombs. Not only was Sondondo radically reorganized, but one of the sites has megalithic stone chambers that were probably sepulchers for nobles. Since comparable tombs are limited to the Huari capital itself, I suspect that the Sondondo Valley had a Wari royal residence, a country palace or royal retreat. Inca rulers built palaces some distance from Cuzco, where they could retreat from the public. The most famous example is Machu Picchu, but there are numerous others. Like Sondondo, the surroundings were intensively terraced, and royal mummies were kept there some of the time.
The Wari provincial administrative capitals of Peru’s highlands, and other less known examples, reveal a profound difference between Wari and Tiwanaku. Wari behavior, albeit still far from adequately understood, is much more consistent with expectations of an empire than Tiwanaku, which contradicts fundamental expectations, even in its best case, Omo in the Middle Moquegua Valley. Huari did establish and administer provinces for the benefit of its homeland elite. It “Warified” provincials, and everywhere seems to have promoted a new international Wari identity with distinctive material culture. Tiahuanaco has only one known case of what might have been provincial administrative architecture, but the residents of the site seem to have rigidly maintained heartland Tiwanaku identity, without interfering with locals, who show no evidence of participating in a unifying social process. Elsewhere, there is no trace of administration, but much more evidence for material culture employed to create a common new Tiwanaku identity.
Of course, definitive conclusions are still premature. Along the coast Wari’s orthogonal cellular compounds are very rare, consistently smaller than in the highlands, and they often have such sparse occupational debris that it is difficult to confirm their cultural affiliation. In the far south coastal Camana Valley, excavations by Malpass (2002) at a small orthogonal cellular complex produced barely a handful of Wari-related sherds. Significantly, the south coastal valleys of Nazca, Ica, and Pisco experienced radical ceramic changes during the Middle Horizon, with Wari-related pottery styles replacing millennia-long local traditions. This certainly suggests Wari colonization, followed by the formation of a new identity. To date, however, only one tiny orthogonal cellular compound has been identified, Pataraya, high in the Nazca Valley (Schreiber 2001).
Peru’s central coast has little evidence for specialized Wari administrative architecture either, in spite of great cultural transformations during the Middle Horizon. Wari-style pottery and textiles appear in burials at Pachacamac (Kaulicke 2001), Huaca Malena, Chimu Capac, and other sites. SAIS icons became ascendant, innovative ceramic styles appeared, and new, but still poorly defined, architectural forms materialized at centers such as Cajamarquilla. Settlement patterns experienced change, but the small Socos site in the Chillon Valley (Isla and Guerrero 1987) seems to be the only example of orthogonal cellular construction.
On the north coast Wari-style pottery and ovate obsidian projectile points appeared in burials at San Jose de Moro (Castillo 2001). Wari ceramics and icons were found at Huaca del Sol/Huaca de la Luna, and other centers. Wari tapestries were buried at Huaca Cao Viejo, and other cemeteries with important local elites. A new polychrome ceramic style called “Huari Norteno” (Larco 1948), inspired by Wari but with strong local components, gained widespread popularity. In later north coast religious art several SAIS-inspired icons accompany older Moche supernaturals.
In the mid-twentieth century archaeologists agreed that north coast cultures were so completely transformed during the Middle Horizon that they must have been conquered and reorganized by Wari. However, recent research (Bawden 1996; Chapdelaine 2002) identifies cultural patterns thought to have been introduced - including great rectangular administrative enclosures, planned urban centers, and probably even state government - in pre-Middle Horizon Moche culture. Emphasizing continuity in north coast culture as well as political capitals such as Pampa Grande that reveal no Wari presence (Shimada 1994), the convulsions of north coast civilization during the Middle Horizon and Late Intermediate Period are now attributed to internal factors, not Wari warriors. However, this reversal may represent extreme swings of the intellectual pendulum, from inferring Wari conquest and absolute control to the inference that Moche culture remained totally independent of Wari throughout the Middle Horizon. Wari may have been resisted by the Moche, at least from about the Viru Valley north (Topic and Topic 1987). But of course, even if these scholars are correct, the north coast still reveals enough Wari influence to indicate significant participation in the newly emerging international Wari identity.
Definitive interpretations of the spread of Wari and Tiwanaku culture during the Middle Horizon are still elusive. There seems little doubt that some kind of imperialistic organization and expansionism was involved. But the Andean archaeological record does not confirm the imperial organization or goals typical of “empire” as an evolutionary stage. Tiwanaku is more divergent than Wari, for at least in some parts of the highlands, Huari behaved much like a typical imperial power. However, extremely sparse occupation in some orthogonal cellular compounds and even seemingly never-completed enclosures, demands more investigation if Wari provincial organization and political economy are to be understood. If the transformed and highly administered Sondondo Valley characterizes Wari provincial activity, imperialism was much more like the evolutionary ideal than implied by the archaeological remains from the central coast, for example. But the Sondondo Valley may not have been a province at all, but a royal estate. Many coastal valleys, some revealing remarkable cultural transformation during the Middle Horizon, have little or no evidence for direct Huari administration. Perhaps indirect rule was employed, but this remains to be documented. On the other hand, the adoption of SAIS icons, the popularity of new, innovative styles in pottery, the prominence of clothing of interlocking tapestry - often depicting SAIS themes - and other material culture do attest to the emergence of a shared new Wari religion and identity.
Tiwanaku is more confusing. It seems to have created mechanisms for provincial rule in Moquegua, only to ignore the exploitative goals of empire. In other areas the nature and processes of Tiwanaku incursion are far from understood, but seem to be based more on the promotion of new religion and identity through consumption of distinctive material culture. Was this an immature step toward imperialism that was never realized, or some kind of cultural formation that is not adequately understood by archaeologists?
Among many practices that promoted new Wari and Tiwanaku identities, one of the most prominent was drinking maize beer from a distinctive chalice-like kero. Was a single new system of etiquette shared by Wari and Tiwanaku? Did both cultures subscribe to a higher-level identity, perhaps like Christianity, with competing centers, like Rome and Constantinople?
World History
