Excavations Excavations Introduction Survey of the Taraco Peninsula Excavations at Chiripa Excavations at Cala Cala

Excavations at Cala Cala (Alto Pukara)

Robin A. Beck, Jr.

2001 Formative Communities and Sacred Space: Excavations at Alto Pukara, Lake Titicaca, Bolivia. Northwestern University. April 5, 2001.

Please do not cite without the author’s permission.

In Bolivia’s Lake Titicaca Basin, Middle Formative period villagers at the archaeological site of Chiripa built a large earthen platform that visually dominated their community. Atop this platform, they erected a series of ritual facilities, the best known of which is probably the Upper House complex discovered by Wendell Bennett in 1934. In the past 15 years, the construction of this platform at Chiripa has been identified as a major development in social complexity. Now, following the completion of Matt Bandy’s full-coverage survey of the Taraco Peninsula, archaeologists have data on similar platforms at several nearby sites. This paper offers a preliminary summary of excavations I conducted at one of these recently discovered sites--Alto Pukara--during the 2000 field season. This project contributes a set of comparative data on the nature of early ritual architecture in the Lake Titicaca Basin, as well as the broader Andean region.
Alto Pukara, measuring 3.25 ha and located 4 km east of Chiripa, provides an important opportunity to study transformations in the scale of Middle Formative political and ritual institutions (following convention, I date the Middle Formative from 800-200 BC). The Taraco Archaeological Project (TAP), directed by Christine Hastorf, has conducted fieldwork at Chiripa over four field seasons--1992, 1996, 1998, and 1999. TAP’s primary research goal has been to address the development of complex political institutions in the southern Titicaca Basin, and as a result of this long-term research, Chiripa is now the best understood Formative settlement within this region. Until recently, however, archaeological investigations along the Taraco Peninsula had concentrated almost exclusively on the site of Chiripa. With the recent excavations at Alto Pukara, we may now begin to view Chiripa and neighboring communities within a broader, regional context.
The excavations conducted at Alto Pukara from July to November of last year exposed 60 square meters of the earthen platform. The Alto Pukara platform is much smaller than the Chiripa mound, measuring only 33 meters east-west by 30 meters north-south, and rising to a maximum height of about 4 meters above the surrounding terraces. Based on both the sunken appearance of the mound surface (slide 5) and a topographic map of the site (slide 6), I expected to find a sunken court in the center of the platform, similar to the sunken courts discovered at Chiripa and at Ch’isi on the Copacabana Peninsula. Likewise, I suspected that the visible ridges surrounding the platform’s sunken center were the remains of structures like those excavated at Chiripa.
Surprisingly, we discovered no sunken temple at Alto Pukara, and the ridges turned out not to be the ruins of structures, but rather, a thick layer of cobbles that, based on preliminary ceramic data, dates to post-Tiwanaku times. The ridge layer probably formed as Pacajes and later farmers cleared cobbles from their fields and piled them on top of the mound, much the same as modern farmers at the site.
What the excavations did reveal was a modest ritual complex consisting of two small buildings located on opposite sides of the platform along the same north-south axis. An associated plaza, consisting of at least six superimposed, red clay floors, was located between the two structures. This platform complex is relatively undisturbed by later cultural deposits; we recovered no Tiwanaku or Pacajes ceramics below the plowzone and the cobble zone just described. All ceramics associated with the structures and the plaza, and with the fill episode immediately above the architecture, are classic Chiripa, fiber-tempered wares. The structures associated with this complex are very close to the ground surface, as we often located the tops of the structure walls less than 20 cm deep. However, due to the lack of later disturbance, the structure walls are actually quite well preserved.
Structure 1, on the south side of the platform, measured 5.3 meters north-south by 4.2 meters east-west, and was oriented 10 degrees off of true north, the same orientation as the terraces which surround the platform. The walls of this structure were single course, and lacked the double-wall, “storage bin” construction for which Chiripa’s Upper House complex is so well known. The walls were faced with cobbles on both the interior and exterior sides, with the space between facings filled with mud, cobbles, and gravel; the walls were also quite thick, averaging 75 cm, and were preserved to a height of about 85 cm. A doorway measuring 90 cm wide was located in the center of the east wall, and unlike Chiripa’s Upper Houses, it lacked evidence of either a sliding door slot or other means of closing the entrance. Though lacking inner wall bins, Structure 1 did have two small niches, one in each corner of its west wall, opposite the doorway; at least one, and probably both, niches extended down to the floor of the structure. This slide illustrates how Structure 1 may have originally appeared, reconstructed to its preserved wall height. (add figure here)
Structure 2 was located on the north side of the platform, and measured 5.6 meters north-south by 4.5 meters east-west. Though we defined all four walls of Structure 1, we only exposed two walls and three corners of Structure 2--this was sufficient, however, to estimate the total size of the building. The walls of Structure 2 were oriented precisely to true north, but otherwise followed an architectural style identical to that of Structure 1: single-wall construction with niches (but lacking inter-wall bins), cobble-faced interior and exterior sides, and an average thickness of 75 cm; the walls were preserved to a height of 95 cm. (add figure here)
Though the style of Structure 2 was identical to that of Structure 1, there are some major differences in the organization of its architectural space. Unlike Structure 1, with its doorway on the east, Structure 2 had a doorway in the center of its west wall (also measuring 90 cm in width). As such, the two structures are inverted, with Structure 1 opening to the east and Structure 2 opening to the west. Structure 2 also had two adjoining niches located on the same side of the structure as the entranceway: again, Structure 1 had single niches in the wall corners opposite the doorway. Structures 1 and 2 thus seem to be contemporaneous chambers built in the same style and according to the same construction techniques, but with variability in both niche location and doorway placement.
Excavations inside and immediately outside Structure 1 revealed a sequence of superimposed, prepared clay floors, or floor resurfacings. These floors were quite thin, and were composed of well-prepared brown clay. Outside the doorway of Structure 1, the floors inclined sharply to the east, and there was evidence here of both a graveled footpath and preserved patches of a thin yellow wash alongside this graveled approach to the doorway. Each floor surface was covered by what appear to be organic, use-related deposits. Inside the structure, these deposits were very thin; outside the doorway, though, they became thicker, and grew more so as one moved downslope away from the doorway. This pattern of deposition suggests that the interior floors were regularly swept clean, and that the sweepings accumulated on the exterior surfaces outside of the doorway. It is also possible, however, that these deposits outside the doorway are alluvial in nature; micromorphology analyses will help to resolve this problem. Few artifacts were recovered from either the interior surfaces or from related deposits outside the doorway; this suggests that a limited range of activities took place inside the structure, and that these activities did not produce typical domestic debris.
Excavations in the interior northeast corner of Structure 1 revealed a large, oval fire pit measuring 1.3 meters north south by 80 cm east west. While we lacked sufficient time to completely excavate this feature prior to the close of the 2000 season, there were few artifacts--either of cooking vessel fragments or of burned bone--associated with its upper ash lenses, suggesting that this feature did not perform the functions of a typical domestic hearth. Ash deposits were not restricted to the fire pit, but were also encountered on the floor surfaces surrounding the pit. If the fire pit was used to consume burnt offerings, then it is possible that the surrounding floor surface was used for similar kinds of ritual activities. Elizabeth Bonnier suggests such an interpretation for ash deposits located on the floors of Mito-style buildings in the Central Highlands of Peru.
Stratigraphic evidence suggests that the ritual use of Structure 1 ended with a sequence of intentional closing events. First, the fire pit was deliberately capped with a pile of cobbles. Next, a layer of soil was spread across the final interior floor surface, probably to protect the floor, after which the structure was filled with a thick layer of cobbles. Much of this cobble fill derived from the upper walls being intentionally pushed inward; the resulting rubble was then spread across the protective layer of soil. Significantly, there is no evidence of either roof fall or burned roofing material in this sequence of closing events, suggesting that the roof was intentionally dismantled prior to the capping of the floor. Finally, a thin layer of soil was deposited on top of the cobble fill. I suggest, in sum, that this sequence of events represents a ritual entombment of Structure 1’s final floor surface.
The final surface outside Structure 1 appears to have been capped with a thick layer of clayey soil at this same time. Prior to the deposition of this fill, a grooved and polished bone tube was placed on the floor immediately outside the doorway, between the faces of the door; stone labrets and a copper band are among the special artifacts deposited within the subsequent fill. Although we do not yet have sufficient data to describe the full sequence of closing events associated with Structure 2, this building was also filled with cobbles. Further, the fill deposit immediately outside the doorway of Structure 2 also included stone labrets, as well as two sheet gold necklaces and fragments of human bone.
To move on now to the broader, regional level of the Taraco Peninsula, Structure 1 and Structure 2 at Alto Pukara appear to have been built to the same architectural conventions as the two Lower Houses--House Sub-1 and House Sub-2--defined by Alfred Kidder and William Coe at Chiripa in 1955. These structures were situated immediately below the famous Upper Houses in the mound stratigraphy, and carbon samples taken from strata above and below the original floor surface of House Sub-2 bracket the time of its construction and use to between 800 and 400 BC. Karen Mohr Chavez identifies these two Chiripa structures as antecedent to the Yaya-Mama tradition, and provides a summary of Coe’s field notes. Like the Alto Pukara structures, House Sub-1 and House Sub-2 were single-walled, without storage bins, and were separated from one another, unlike the interconnected Upper House complex. Both structures were built entirely of cobbles, without adobe; their walls were thick and cobble-faced on the interior and exterior sides, with the space between faces filled with mud, cobbles, and gravel. In these construction techniques, Coe’s Lower Houses are nearly identical to Structures 1 and 2 at Alto Pukara, but represent a very different style than the Upper House complex.
More striking, Coe discovered a narrow niche in the northwest corner of House Sub-2, the base of which coincided with the floor of the structure. This niche measured 53 cm long by 35 cm wide; the corner niches in Structure 1 at Alto Pukara were this same size, averaging 50 cm in length and 30 cm in width, and also extended down to the floor surface. Additional niches may have existed inside this Lower House, but Coe only excavated one quarter of the structure. No data on the presence or absence of niches are available for House Sub-1; by the time of Coe’s excavations, this structure had been largely destroyed--save for one wall--by the construction of a tennis court for the hacienda. Finally, a red clay surface extended at least 4 meters to the east of Coe’s structures. There is no evidence to suggest that this floor was part of a sunken court. Rather, it was probably a simple open plaza like that connecting Structures 1 and 2 at Alto Pukara.
I suggest that the Lower Houses Coe excavated at Chiripa are of the same architectural tradition as Alto Pukara Structures 1 and 2. Given both the early radiocarbon dates from the Chiripa structures, and the position of these structures under the Upper Houses, it is clear that this architecture predates the widespread Yaya-Mama tradition, with which the Upper House complex is usually associated; as noted, Karen Mohr Chavez considered Coe’s Lower Houses to be architectural antecedents to Yaya-Mama. With the excavations at Alto Pukara, we have evidence of similar structures at a nearby site, suggesting that this is the earliest regional style of public ritual architecture yet reported in the Lake Titicaca area, though its distribution may have been restricted to Bolivia’s Taraco Peninsula. Radiocarbon dates from Alto Pukara will help to clarify the age of this architectural style, and perhaps more important, will address the issue of whether communities such as Alto Pukara ritually closed their own platform facilities when the Upper House complex was built at Chiripa.
If so, this suggests that an important social transformation took place during the Middle Formative period. Prior to the construction of the Upper House complex, fieldstone chambers built atop a centrally located earthen platform--such as Structures 1 and 2 at Alto Pukara, and Coe’s Lower Houses at Chiripa--probably served as the foci of ritual life in each village along the Taraco Peninsula; each village, that is, was ritually and politically autonomous. After 400 BC, however, I suggest that leaders at Chiripa formalized the ritual and political relationships among communities along this part of the peninsula. The orderly and careful entombment of Structure 1 at Alto Pukara suggests that events associated with this social transformation were persuasive, rather than coersive. In sum, Middle Formative villagers seem to have abandoned their communal, chamber-oriented ritualism in favor of a regional ceremonialism that focused upon the monumental, Upper House complex at Chiripa.
Before concluding, I would like to thank the families of Cala Cala, each of which participated in the fieldwork, the community of Chiripa, my DINAAR representative, Victor Plaza, and the other members of TAP for making this project possible.