Huamachuco

----->
REGIONAL AND HISTORICAL CONTEXT

The qollqa discussed above can be better understood if they are placed in a wider regional and historical context. Huamachuco is located at a junction of two parts of the Inka road: the main north/ south route in the highlands, and a secondary east/west route leading to the coast. Therefore, it is likely that the Huamachuco qollqa were not the only storage facilities in the province.

During the latter half of the eighteenth century, Fray Juan de Santa Gertrudis (1970: 225) traveled through the area and mentions that he saw qollqa between Cajabamba and Huamachuco. A modern caserío (hamlet) located 13.5 km north of Huamachuco alongside the Inka road is called Colcabamba. Although there are no preserved qollqa known from the immediate area, the toponym is suggestive. Hyslop (1984: 61,65) reports no tampu between Huamachuco and Cajabamba, so qollqa in this vicinity may have been part of a production storage complex. At an elevation of 3,050 m, the region is well suited to both maize and tuber agriculture.

There is documentary evidence6 of another set of qollqa at Chuquibamba in the Condebamba Valley, approximately 25 km northwest, of Huamachuco. The document is primarily concerned with a lawsuit concerning the succession in a curacazgo (polity controlled by a local level lord) of a waranqa of mitmaqizuna in the province. In the course of the testimony, it is mentioned that the mitmaqkuna were moved to Chuquibamba to serve at some depósitos (storehouses) that the Inka had for his soldiers. Chuquibamba survives as the name of an hacienda on the left bank of the Condebamba River about 22 km northwest of Huamachuco. On the opposite bank of the river, 3 to 4 km from Chuquibamba are located the Hacienda Colcas and a caserío called Colcas. Again, the toponyms are suggestive. These modern settlements are located between 2,200 and 2,400 m elevation; the lnka may have produced and temporarily stored ají and tropical fruits grown at these lower elevations.

Finally, about 55 km west of Huamachuco there is a site, Llagaday Saddle, which Janet and Mark Mackenzie surveyed in 1979 (Mackenzie 1980: 304-10). Llagaday Saddle is located in the general area of a tampu called Cucho on the route from Huamachuco to the coast (Tópic and Topic n.d.). The site itself is only 20 m from a walled footpath and includes the foundations of two rows of conjoined rooms that curve along the contours of a gentle slope. There are at least 21 rooms in the upper row and 9 in the lower row; average room measurements are about 3 x 5 m. Below the two rows of rooms are several poorly preserved, or perhaps unfinished, terraces, each about 1.5 m wide. There were sherds of large storage jars in the local Late Intermediate (pre-Inka) Period style. L.lagaday Saddle does not appear to be located directly on the lnka road and thus is unlikely to be a tampu. It was probably a storage facility associated with agricultural production. At an elevation of 3,750 m, the emphasis was probably on tubers, which continue to be important in the area today.

Although information about these three sites is still limited, it does suggest that there were a number of other storage complexes located throughout the province of Huamachuco. These storage facilities occur in zones at widely varying altitudes, suggesting provision was made to store a variety of crops in the zones in which they were produced.

Returning to Huamachuco, we should emphasize that although the excavated qollqa are generally similar to examples elsewhere in the Empire, they are different in detail. Specifically, only rectangular storerooms occur, although these include at least two distinctive types: One type, with an elevated floor, has not been reported in Inka contexts south of Huamachuco; the type with subfloor canals resembles examples from Huánuco Pampa (Morris 1981) but may have functioned quite differently. Ceramics are exceedingly scarce in the Huamachuco storerooms, and those that do occur are predominantly in the local style. In contrast, at both Huánuco Pampa (Morris Chapters 5 and 8) and Hatun Xauxa (D'Altroy and Hastorf 1984 and Chapter 9) diagnostic Inka sherds were commonly associated with grain storage. These differences certainly reflect distinctive variations in storage technology.

Because storage was so important to Andean society, it is useful to trace, to the extent possible, the history of the development of this distinctive technology. Although we as yet only have hypotheses about the history of qollqa with subfloor canals, we have more concrete evidence of the history of qollqa with elevated floors. This type of structure was first identified at Cerro Amaru, located a few kilometers from Huamachuco (Topic and Topic 1984: 45-51). Two storerooms dating to the Early Intermediate Period or early Middle Horizon were excavated (Figure 7-6). These structures are round with floors raised above ground level on a masonry support. The support is a stone-faced bench or ledge around the inside of the circular wall; this is similar in concept to the third face inside the C shaped walls found in some of the Cerro Santa Barbara qollqa. In the Cerro Amaru structures, a low central wall bisects the building and supports the middle of the floor joists, much as the central pier does at Cerro Santa Barbara.

One of us (Topic in press; Topic and Topic 1984) has suggested that the Cerro Amaru storerooms are similar to possible storerooms at Jargampata and Azángaro. These are slightly later in time, dating to Middle Horizon 2, and are associated with the Wari culture of the Central Highlands. At Jargampata (Isbell 1977: 37), there are two conjoined rooms with a row of corbels for a second-story floor located only about a meter above a packed clay first-story floor; very narrow doors (65 x 66 cm) open into the first story. Isbell (1977: 38 ,49) interprets these buildings as possible storage structures. We concur and suggest that, as at Cerro Amaru, the actual storage space was on the "second" story, the "doors" are ventilators rather than entrances, and the "first" story functioned only to raise the storage area off the ground. Although Anders (1986: 213-14) considers the 340 rooms arranged in 40 rows in the central sector of Azángaro to be temporary residences related to calendric rituals, they are quite similar to the two rooms at Jargampata. The rooms within a row are conjoined, there are stone corbels located only 1 meter above the ground, and very small "doors" (60 x 60 cm) opening into the space below the corbels (Anders 1986: 213 and Figure 10.8).

Thus, Inka qollqa on Cerro Santa Barbara are conceptually similar to Early Intermediate Period or Middle Horizon storerooms at Cerro Amaru, which in turn are similar to Wari storerooms in the Central Highlands. However, the historical connections between these storerooms from different time periods are still difficult to unravel. As yet, we know of no Late Intermediate Period examples from Huamachuco that could serve as an historical link between the Cerro Amaru and Cerro Santa Barbara structures. There are also no reported examples from the Central Highlands either earlier or later than the Wari ones. On the other hand, Wari and Wari-influenced ceramics occur at

Cerro Amaru, together with other exotic ceramic styles and trade goods. This evidence of wide-ranging contact a mechanism for the diffusion, in either direction, of this type of qollqa during the Middle Horizon.

We should mention one other possible case of qollqa with elevated floors. Schjellerup (1984: 172-76) provides very clear descriptions and drawings of Inka qollqa in Chachapoyas province (see Figure 1-1). These occur in two of 8 and 15 buildings at quite different elevations. Qollqa in the upper row have a 20 x 25 cm rectangular opening in the front wall al ground level, a door 1 meter above the ground, and six 10 x 10 cm square openings in both the front and back walls just below the level of the door. If the 10 cm square openings are interpreted as beam sockets, these qollqa would have elevated floors similar to the Wari examples discussed above. However, Schjellerup (1984: 172-76, personal communication 1990) disagrees with this reconstruction and interprets these openings as ventilators. We admit that in the illustrated example the corresponding openings on the front and back walls are not as well aligned as one would except of beam sockets. The lower row of qollqa was less well preserved, so that any features located in the upper parts of the walls are probably lost. There were 40 x 14 cm rectangular openings at ground level in each of four walls and partially preserved masonry bench along the back walls. The bench and ground level ventilators she describes suggest a buildings similar to those al Cerro Santa Barbara and Cerro Amaru.

In the excavated examples of qollqa with elevated floors, sherds of storage jars are scarce or absten. This statement applies not only to the Cerro Santa Barbara qollqa, where is it clear that maize was stored but also to the structures at Cerro Amaru , Jargampata, Azángaro, and Chachapoyas (see Figure 1-1). The lack of storage jars in all these complexes contrast with the storage practices documented at Huánuco Pampa and Hatun Xauxa (Morris Chapter 8; D'Altroy and Hastorf Chapter 9) where grain was commonly stored in jars. Perhaps in qollqa with elevated floors, grains were stored in bags. Another possibility, suggested by the Agustinian reference to findings bodies in a pile of maize (Primeros Agustinos 1918: 36), is that grains were stored loose and in bulk. A third possibility is that there were internal bins as Garcilaso describes; the straw-impressed plaster al Cerro Santa

Barbara and the cane-marked plaster from Cerro Mamorco (from a qollqa with evidence of maize but no elevated floor) may be evidence of these bins.

Only the excavations at Cerro Santa Barbara and in Building C at Cerro Cacañan produced significant quantities of sherds. As we pointed out earlier, most of the sherds from Cerro Santa Barbara are from an Early Horizon occupation, but sherds in the local Late Intermediate Period style, called Huamachuco-on-White, were also present. The diagnostic sherds from Building C were also in the Huamachuco-on-White style. From both sites the sample of Huamachuco-on-White sherds is very small: The sherds from Cerro Santa Barbara may represent only one jar, whereas those from Building C represent at least one jar and one bowl. Other excavations show that Huamachuco-on-White continues into the Late Horizon and that jars in this style sometimes have rope nubbins, which are an indication of Inka influence. In fact, the one diagnostic lnka sherd from the entire Huamachuco storage complex is a rope nubbin that may be from the Huamachuco-on-White jar found in Building C.

Although the sample of sherds is small, it is surprising that the Huamachuco-on-White style occurs, whereas the Provincial Inka style is lacking. Provincial Inka ceramics, and even a wooden kero (drinking cup), have been found in Huamachuco itself (Topic and Topic n.d.; McCown 1945: 310), but we know of no Huamachuco-on-White pottery from around the modern town. Because of the small number of vessels from both Huamachuco and the qollqa excavations, we cannot say more about the lack of Inca pottery in the olla except that it contrasts with the situation in Guanaco Pampas and Haxtun Xauxa.

Retornar Avanzar