Ancient Cuzco

by Brian S. Bauer

  CHAPTER 7

  The Wari Period (AD 600–1000) in the Cuzco Region

  THE WARI PERIOD (also generally known as the Middle Horizon) encompasses a broad span of time during which much of the Andean highlands came under the influence of two empires: Wari and Tiwanaku. Current research suggests that the Wari began to expand from their traditional homeland in the Ayacucho region of Peru sometime after AD 550 and that the expansion continued through at least AD 900, after which the empire appears to have suddenly collapsed. Though less is known concerning the development of Tiwanaku, it seems that by AD 300 the city of Tiwanaku, near the southwestern shore of Lake Titicaca in Bolivia, was of considerable importance. Expansion of Tiwanaku may have begun around AD 500 and waned, like Wari, near the end of the first millennium.

  Researchers have long noted the presence of Wari materials in the Cuzco region, and it is recognized that Ayacucho influenced the thriving polities of the region for several centuries. Thus, the Wari Period is a fascinating phase for the Cuzco region. It is a time of both foreign occupation as well as indigenous development. Fortunately, it is also one of the better-studied periods of the region’s history. In this chapter I outline the course of Wari Period research in the Cuzco region and examine the processes of culture change that occurred there as a result of its incorporation into the Wari Empire.

  Indicators of Wari Influence

  Wari influence in a region is commonly inferred from architectural remains as well as from the presence of Wari ceramics or other portable artifacts. The Wari built many of their buildings in a distinct architectural style that featured large high-walled rectangular enclosures. Made of fieldstones and mud mortar, these enclosures generally contained a central patio and a series of distinctly long and narrow galleries. Archaeologists frequently use the presence of these enclosures, along with their patios and galleries, to identify installations built by the expanding Wari Empire.

  An even more common means of documenting Wari influence in a region is through ceramics. Among the various ceramic styles used in this book to identify Wari influence are a series of styles that were actually produced in the Ayacucho area and then imported into the Cuzco region. These include the Ayacucho styles of Chakipampa, Okros, Viñaque, Huamanga, and Robles Moco (Knobloch 1991; Glowacki 1996, 2002).1 Various examples of possible imported Wari ceramics have been found in excavations at Pikillacta. Neutron activation analysis by Glowacki and her associates of selected pieces indicate that they were in fact produced in the Wari heartland and then imported into the Cuzco region (Montoya et al. 2000). During our 1999 excavations in the Cuzco Valley, we recovered various pieces of Wari ceramics (Bauer and Jones 2003). Carbon from a trash pit at the site of Tankarpata, which contained a fragment of Viñaque pottery, provided a date of 1290 ± 50 BP (calibrated 95.4% probability AD 650–880).2 In addition, a midden at the site of Pukacancha containing a piece of Huamanga pottery dated to 1210 ± 45 BP (calibrated 95.4% probability AD 680–960).3 These dates fall well within the time traditionally believed to frame the Wari expansion.

  There are also a number of other ceramic styles in the Cuzco region that appear to be locally produced but that closely imitate ceramics of the Wari heartland. For example, the finer wares recovered at the site of Pikillacta have been shown to imitate the Okros ceramics of the Wari homeland, but they were produced in the Cuzco region (Knobloch 1991: 253–254; Glowacki 1996; Montoya et al. 2000). Another example is the recently defined style of Arahuay (Torres Poblete 1989; Glowacki 1996; Bauer 1999, 2002). Arahuay ceramics are characterized by the use of broad red bands outlined with narrow black lines over a buff slip (Photo 7.1). Glowacki (1996) has demonstrated that Arahuay pottery closely imitates the Huamanga ceramics of the Ayacucho region. Through neutron activation, she and her colleagues have also shown that Arahuay ceramics are made from local clays (Montoya et al. 2000). Surveys have found that Arahuay ceramics are widely spread in the Cuzco region, and we currently use them as a marker for Wari influence in the area. Because the physical composition of Arahuay pottery is similar to that of Qotakalli (Montoya et al. 2000), it appears that many of the ceramic workshops that produced Qotakalli vessels before the arrival of the Wari gradually accepted and began production of the Wari-style pottery, such as Arahuay, after the region fell under the influence of the Ayacucho state. Although Arahuay ceramics represented one of the dominant ceramic styles of the Cuzco region during the Wari Period, its period of production was little understood before our 1999 excavation season (Bauer and Jones 2003).

  PHOTO 7.1. Arahuay pottery from the Cuzco Basin

  Theodore McCown and John Rowe first proposed Wari influence in the Cuzco region in the early 1940s (Rowe 1944: 53). Manuel Chávez Ballón and Rowe later confirmed it in the early 1950s as they studied excavation and surface collections from a number of archaeological sites southeast of the city of Cuzco (Rowe 1956: 142). The most important information came from the site of Batan Orco, a small knoll that juts out into the Vilcanota River Valley near the town of Huaro (Reichlen 1954; Rowe 1956: 142). The discovery by looters of an elite tomb at Batan Orco in 1952 brought the site to the attention of Cuzco officials as well as the general public (El Comercio 1952a–k; Reichlen 1954). Chávez Ballón (personal communication, 1990) dug at the site that same year, but his collections were destroyed before he could finish his analysis. Barreda Murillo (1973) conducted additional excavations at Batan Orco in 1952, and Tom Patterson and Rowe made surface collections there in the 1960s (Patterson 1967). Initially, the finds at Batan Orco were classified as Tiwanaku-related (Reichlen 1954). However, further examination of the materials by Chávez Ballón and Rowe suggested a closer relationship to the Ayacucho state than to Tiwanaku. This conclusion is supported by recent work at Batan Orco by Zapata (1998) and at the nearby town of Huaro (Glowacki and Zapata 1998; Glowacki 2002).

  The Development and Expansion of the Wari Empire

  Developing out of a culture in the Ayacucho region currently known as Huarpa, the city of Wari grew to enormous proportions. At the height of its influence, around AD 700–800, the city covered many square kilometers with densely packed buildings and rectangular compounds. As Katharina Schreiber (1992: 80) describes it, “The site, as known today, is immense. The architectural core of the site, comprising diagnostic rectangular compound-style architecture, measures some 200 hectares. Around this core is an area of dense surface scatter; including this scatter, the site covers perhaps as much as 300 hectares.” There is evidence of craft specialization within the city as well as indications of royal tombs and elite compounds (Benavides C. 1991; Isbell et al. 1991). Elaborate ceramic production was also occurring in villages near it (Pozzi-Escot B. 1991). With the current end to more than a decade of political upheaval in the region, we will soon learn a great deal more about the core region of Wari development.

  PHOTO 7.2. Aerial photograph of Pikillacta (Neg. no. 334819, photo by Shippee Johnson, courtesy the Library, American Museum of Natural History)

  Current scholarship suggests that the Wari rapidly expanded beyond their core area of development sometime in the late sixth or early seventh century AD. As the Wari conquered and incorporated various regions into their empire, they built a series of administrative centers across the central Andes. Two of these are qualitatively larger than the other centers, and, fascinatingly, they are positioned near what appear to be the northern and southern extremes of Wari influence in the highlands.

  In the north, in the area of Huamanchuco, lies the extensive Wari ruin called Viracochapampa. The site covers more than 30 hectares and displays many classic features of Wari architecture, such as a great enclosing wall and large-scale rectangular courts with central patios and multiple long, narrow galleries. It also contains some local features such as niched halls. Research at the site indicates that Viracochapampa was never completed, although when construction began and when it stopped is not well understood (Topic and Topic 1985).

  Near the southern extreme of the Wari Empire is the
site of Pikillacta. As discussed below, this site shares many interesting features with Viracochapampa, not least of which is the fact that it was never completed. Other Wari centers are known in the central highlands, but none come close to the dimensions of Viracochapampa, Pikillacta, and Wari (Schreiber 1992: 96–114). The relative sizes of the three largest Wari sites, and their geographical locations spread across the central Andes, have encouraged scholars to suggest that the capital of the Wari Empire was the city of Wari in the Ayacucho region and that the Wari attempted to build two large installations (Viracochapampa and Pikillacta) to serve as provincial capitals near the northern and southern borders of their empire.4

  The Site of Pikillacta

  The largest single Wari occupation in the Cuzco region is the site of Pikillacta, which is located in the Lucre Basin approximately 30 kilometers southeast of the city of Cuzco. Cieza de León (1976: 261 [1551: Pt. 1, Ch. 97]) visited this well-preserved site soon after the European invasion, and centuries later Squier (1877: 419–422) commented on its large size and apparent antiquity.5 More than a thousand years after its abandonment, the site of Pikillacta is breathtaking in both its horizontal and vertical scales (Photos 7.2 and 7.3).

  Because of its large size and remarkable state of preservation, Pikillacta remains an important feature of the landscape today (Photos 7.4 and 7.5). It has, nevertheless, suffered at the hands of looters. Cieza de León notes that looting began at the site soon after the Spaniards arrived in Cuzco:

  There were great buildings in Mohina, but they are now destroyed and fallen down. When the Governor Don Francisco Pizarro entered Cuzco with the Spaniards, they say they found near these buildings and in them a large amount of silver and gold, and even more of the fine, valuable clothing I have mentioned on other occasions. And I have heard some of the Spaniards say that there was at this place a stone statue of a man with a kind of long robe and beads in his hand, and other figures and statues. (1976: 261 [1551: Pt. 1, Ch. 97D6

  PHOTO 7.3. Aerial photograph of Pikillacta (Courtesy of Servicio Aerofotográfico Nacional, Peru)

  In the early 1920s, looting at Pikillacta produced two caches of turquoise figurines (Cook 1992). The figurines were part of two dedicatory offerings placed within a building in the central sector of the site. Various other offerings, although less elaborate, have been found in later excavations (McEwan 1996). Over the past fifty years many individuals have conducted research at Pikillacta, including Emilio Harth-Terré (1959), Barreda Murillo (1964), William T. Sanders (1973), and Alfredo Valencia Zegarra. Beginning as early as the 1930s, the Peruvian government, most recently under the auspices of the National Institute of Culture (Cuzco), has also directed a series of renovation projects at the site. Finally, McEwan supervised a long-term project at Pikillacta beginning in 1979.

  PIKILLACTA AS A WARI ADMINISTRATIVE CENTER

  Pikillacta is a well-planned and excessively large installation that was built and occupied in a deliberate sequence. The site was constructed on a large mountain shelf—one that had not been occupied by previous cultures—overlooking Lake Lucre. Excavations in Pikillacta and in trash middens outside the massive walls of the site indicate that its occupants used several different ceramic styles, including a few exotic vessels imported from Nazca and Cajamarca (Knobloch 1991: 253; Glowacki 1996). Investigations have also revealed much about the construction sequence of the site. McEwan (1996) divides the central zone of Pikillacta into four sectors, each of which has a different construction and occupational history (Map 7.1). Sector 1, the northernmost sector, is composed of eighty-one large square compounds with some internal galleries and structures. The layout is remarkably rigid, with little variation in compound types. Research in this sector suggests that it was nearly complete at the time of site abandonment, although it was never occupied.

  PHOTO 7.4. A thousand years after its abandonment, Pikillacta remains a very impressive site.

  PHOTO 7.5. Some of the walls at Pikillacta still stand several stories high.

  MAP 7.1. Map of Pikillacta (Courtesy of Gordon McEwan)

  The central area of Pikillacta, Sector 2, is by far the most elaborate. It is composed of many large square and rectangular compounds and a mazelike collection of internal galleries. There is also an expansive central plaza and a smaller, secondary plaza to one side. Excavations indicate that Sector 2 was completed and occupied for a considerable period. Most of the walls and floors of the structures were covered with a thick, white gypsum plaster. McEwan (1996: 182–183) writes, “Excavation illustrates that Sector 2 was the first to be built, and featured fully plastered walls and floors in multi-story buildings with thatched roofs. This sector was occupied long enough for some buildings to have undergone at least two remodeling episodes.” A large burning event in the sector resulted in the superimposition of the floors in the multistoried galleries as their crossbeams failed and the floors collapsed, one on top of the other.

  The construction sequence and layout of Sector 3, the southernmost area of the site, is the least understood, partly because a large-scale remodeling of the area by the Peruvian government may have occurred in the early 1930s. The section seems vacuous in comparison to other parts of the site, and the fact that some of the walls appear unfinished suggests that the construction of this sector was still in progress when Pikillacta was abandoned.

  The western area of the site, Sector 4, is composed of five compounds of densely packed single-room structures. Four of the compounds also contain large rectangular areas. Since the first mapping of the site, it has been suggested that this sector represents a storage area. This idea has been readily accepted, based on features of the sector such as its limited access, the five hundred identical cell-like rooms, and large rectangular areas perhaps used for the drying of products. However, excavations in the rooms have revealed domestic remains, including informal hearths and light trash middens (McEwan 1991).

  THE CONSTRUCTION OF PIKILLACTA

  The construction techniques used in building Pikillacta are not complex. Its walls are made of local fieldstones held together with mud mortar. The gypsum plaster that once covered many of its structures is also found locally. Nevertheless, its great size, formal layout, and impressively high walls mark Pikillacta as an unprecedented construction project in the south-central Andes.

  The building of Pikillacta was an immense investment by the Wari Empire in the Cuzco region, and some unusual architectural features of the complex hint at how the local labor was organized during its construction phases. Traditionally, large public works projects in the Andes are organized around village-level ayllus (kin groups). Rather than organizing themselves into a single massive labor force when undertaking large public labor projects, Andean populations generally work in a multitude of small groups, the coordination of which is left in the hands of overseers. Each ayllu is assigned a distinct aspect of the larger project, and that section or task is to be completed independently of the other groups’ contributions. Evidence of such labor organization has been noted in the production of adobes at the Moche Pyramid of the Sun on the north coast of Peru (Hastings and Moseley 1975), and other facts suggest that similar labor and construction techniques were used to build Pikillacta.7 For example, the great enclosure wall of the site is divided into a large number of uneven sections (Photo 7.6). These seams introduce a structural weakness to the wall, but they make sense if the seams represent the labor assignments of separate kin groups.

  We can also assume, using well-documented Inca analogies, that the construction of Pikillacta occurred through mit’a (turn) labor. Under the Inca system of governance, each village was required to provide a certain number of workers for state projects for a set time period. When they had completed their turn, the workers were replaced by laborers from other settlements, and the first group of workers returned to their homes. Although we may never know which local groups were most heavily involved, there is no doubt that the construction of Pikillacta and the maintenance of the people wh
o occupied it placed a tremendous strain on the local population.

  THE ABANDONMENT OF PIKILLACTA

  Unlike normal villages and towns, whose population may wax and wane over the years, imperial installations frequently experience a boom-and-bust cycle. For example, many of the largest Inca provincial centers, such as Huánuco Pampa (Morris and Thompson 1985), Pumpu (Matos M. 1994), and Machu Picchu, were abandoned soon after the Spanish invasion. The high-level Inca administrators fled, and the local, lower-level providers of the centers returned to their villages as the power of the Inca Empire collapsed. Many of the Inca provincial capitals were artificial cities, in the sense that there were no permanent residents with strong traditions that tied them to the centers. The provincial capitals were quickly built under the strong hand of the empire and even more quickly abandoned as soon as the hegemonic power of Cuzco waned.

  A similar pattern of construction, habitation, and abandonment can be seen in Pikillacta. As a planned Wari center, Pikillacta appears on a mammoth scale in an area of the Lucre Basin that had not attracted previous occupations. The resident population would have been made up of foreign elite, no doubt closely affiliated with those of the Wari heartland, as well as an untold number of locals serving their labor obligations. Further excavations may reveal the presence of lower-level administrators, craft specialists, and some military personnel as well.