"Eco-logistical" Practice over the Longue-Durée: Intermediate Elites and Hybrid Community Structures in the Colca Valley, Peru
Steve Wernke, University of Wisconsin-Madison
Paper Presented at the Annual Meeting of the Society for American Archaeology, March 20-24, 2002, Denver, Colorado, during the syposium "Intermediate Elite Agency in Precolumbian States and Empires," organized by Christina M. Elson and R. Alan Covey.

This paper complements previous studies of Andean "verticality" relationships by focusing on community-scale changes in "eco-logistics"-practices recursively structuring and structured by economy and ecology-in the context of the Colca valley. First, settlement pattern data from the period of autonomous Collagua rule during the Late Intermediate Period are compared to the Inka and early Spanish colonial occupations. Secondly, analysis of land tenure data from 16th and early 17th century Spanish censuses tracks the eco-logistical practices of community elites. The two data sets combined provide a view of the hybridization of community power structures under Inka and Spanish rule.

By the 15th century, the expanding Inka empire had developed varied and flexible imperial strategies for accommodating and appropriating a diverse range of ethnic polities, and the equally diverse ecologies that they inhabited. These strategies were so successful in part because they manipulated key cultural principles and practices familiar to local polities and their leaders. Central among these, as hypothesized by John Murra thirty years ago, were the principles and practices of "verticality", whereby the marked altitudinal ecological zonation of the Andes was matched by "vertical archipelagoes" of settlements linked through exchange in order to maintain a "maximum control of ecological tiers" (Murra 1968; Murra 1972). These systems were both imagined and real in their status as ideals and as built features arrayed over local and regional landscapes. Understanding the process of the imperfect reproduction of such ideals in the context of Inka and Spanish imperial occupation remains a critical research frontier.

Here I discuss the relationship between such ideal social structures and their physical expression on the landscape in the case of the Collagua ethnic polity of highland southern Peru during Inka and early colonial times. I will talk in most detail about an aspect of the documentary research with which I am currently most engaged--the ways in which structuring principles of local communities were manifested on the landscape through analysis of early colonial land tenure patterns, using a series of detailed colonial censuses, or visitas. I will show how data recorded in these censuses can be used to reconstruct what I call the eco-logistical practices of Collagua intermediate elites during Inka times. I use the term eco-logistics to refer simultaneously to economic decision making regarding the deployment of labor and its products, and ecological decision making regarding the energetic and material requirements of agro-pastoral goods within varied Andean microclimates.
I build on recent critiques of the verticality model which point out that the construction of stable agro-pastoral production systems in the Andes is predicated on a simplification by humans of more complex ecological gradients. In other words, as pointed out by Enrique Mayer (1985) and others, the "ecological tiers" in Murra's model are as much economic as ecological. This means that the asymmetrical systems of redistribution common to late prehispanic Andean economic organization must not be epiphenomenal to an adaptive process of securing ecologically-diverse resources, as implicit in Murra's paradigmatic framework (see Stanish 1989a, 1989b, 1992; Van Buren 1993, 1996) . So acknowledging the unmistakable anthropogenic nature of Andean ecological tiers begs questions about not only how the particular ecological strictures and possibilities of Andean ecology influence "suprahousehold" economic organization (Guillet 1978, 1981), but also how different community structures affect patterns of production, distribution, consumption, and, in turn, the Andean landscape.
By emphasizing human agency in the environment in this way, I argue that the general structure-agency framework of Bourdieu (1977) and Giddens (1979) is applicable to the relationship between economy and ecology. Within my formulation, eco-logistics are recursively related to the built environment, such that they at once structure and are structured by it. I argue that Collagua native lords, or kurakas, were pivotal in this recursive process because of their intermediate elite status as redistributive brokers and extractive agents for the state.

This general approach forms the point of departure for my dissertation project, which also integrates archaeological data I collected during a full-coverage survey of a 90 square kilometer area in the same part of the Colca River Valley in which the Spanish censuses were recorded (Wernke and Guerra Santander 2001). Although my survey covers a large portion of the central territory of the Collaguas, archaeological evidence for Collagua influence or presence, mostly in the form of ceramic assemblages in the Collagua style, have been identified in lower-lying valleys to the south (Wernke 2001). As we'll see, these distributions coincide with documentary evidence for significant outlying Collagua "island" settlements

My settlement pattern data provide strong evidence for decentralized, but coordinated Inka rule in the valley. Along with Peruvian archaeologists Willy Yépez and Erika Simborth, I recorded 162 archaeological sites with 300 occupational components using a mixed strategy of regional and site-specific survey. No site dominates the settlement pattern in terms of size, centrality, or elaborate architecture. Instead, the Inka occupation of the valley is marked by overall locational and organizational continuity from the village-based pattern of the preceding period of autonomous Collagua rule. All but two of the 14 Inka period settlements recorded also have occupations from the preceding Collagua Period. Archaeologically, Inka rule clearly appears to have been mediated by local elites, resulting in hybrid Collagua-Inka settlement configurations. For example, rustic forms of Inka architecture associated with state-sponsored public ritual appear in association with high status Collagua domestic architecture at sites such as Uyu Uyu, where a great hall, or kallanka borders the central plaza.

Such hybridization of site layout continues into early colonial times. The opposite side of this plaza is dominated by a structure I have hypothesized to be an early Spanish chapel. A similar building, also with clear colonial architectural features, such as fired roof tiles, is found at the largest settlement in the survey, San Antonio. Aside from the style and placement of these buildings, ecclesiastical documents indicate that Franciscan missionaries established a series of early doctrinas or missionary settlements in the valley by the mid to late 1540's. These data thus indicate that, in a manner similar to the way in which the Inka had capitalized on earlier established centers of Collagua elites, Franciscans capitalized upon already established local centers of Inka power and ritual by locating doctrinas at sites such as Uyu Uyu and San Antonio.

However, this dispersed village pattern was abruptly truncated in the early 1570s with the forced resettlement of Andean populations into nucleated reducción villages by the Viceroy Francisco de Toledo. Although this political imposition radically altered the settlement pattern, it did not change the valley's elaborate terracing and irrigation infrastructure. This set up a dialectic between the two, in which eco-logistical practices had to accommodate or reconfigure both prehispanic and colonial features of the built environment.

Descriptions of the organization of the Collagua province under Inka rule in early colonial documents provide independent evidence for how such physical manifestations accorded with ideal sociopolitical models. An account from 1586 in the Relaciones Geograficas de Indias by provincial magistrate Juan Ulloa Mogollón relates how a nested hierarchy of kin-based corporate groups, called ayllus in Quechua, were organized according to the logic of high, middle, and low status designations, called Collana, Payan, and Cayao, within overarching upper and lower-ranked moieties (Ulloa Mogollón [1586] 1965). Tom Zuidema and Brian Bauer have both noted parallels between this tripartite hierarchical structure and the ceque system of Cuzco, which was also organized by these categories (Bauer 1998; Zuidema 1964)
(3). Also, the pataca designation of the minimal ayllus, meaning 100 in Aymara, refers to a unit of 100 tributary households, suggesting that these ayllus were treated as bureaucratic equivalents for levying labor tribute within Inka decimal administration (4).

These same structural principles were applied to the overall organization of the Collagua province, which was divided between Yanquecollaguas, Laricollaguas, and Cabanas. In terms of sociopolitical organization, this means that the leader of the highest ranked ayllu of the upper half of Yanquecollaguas was also lord of the entire province, and the leader of the highest ranked ayllu of the upper half of Laricollaguas was second in charge of the entire province (Benavides 1987; Benavides 1988; Cock 1977-78; Galdos 1984; Pease 1981). As was common in many other provinces in the Inka empire, direct intervention by state personnel was almost certainly limited to the top one or two tiers of this hierarchy. This would have limited imperial administrative burdens by minimizing points of contact with intermediate elites, while also providing a means for gauging productivity and engendering competition between equivalent tribute units.

The detailed data recorded in local colonial censuses provide a means for analyzing how these ideal structures were mapped out onto local and regional agro-pastoral landscapes. Yanque and Coporaque, the two villages within my survey area, are the best-documented villages in the censuses. Each household declared all landholdings by predominant crop and size, using the Andean measure of the topo, which for our purposes today can be equated to about a third of a hectare. Land tenure patterns in the visitas can be cartographically reconstructed because each field was also located by toponym, or place name. Because of the historical durability of toponyms in this and other areas of the Andes, the location of fields claimed in the visitas can be mapped out by matching them up with their modern toponym areas. Using this method, I currently have 82 toponyms mapped in ArcGIS. In ArcGIS, each of the toponym polygons is linked to a relational database with the colonial census data, allowing for sorting and querying of the database, and viewing their results on the map. At this point, I have entered the visita data for 450 households, with a total of 993 inhabitants, from the lower moiety of one village, Coporaque. My toponym mosaic maps can presently account for 44%, or 223 out of 507 topos of agricultural fields declared in the 1604 visita. However, I can map out 68% of the fields declared by ayllu leaders, the subset of interest today for their status as intermediate elites.

In the visita, Coporaque households are grouped in five ayllus: Collana Ayllu, Payan Collana Pataca, Payan Taypi Pataca, Payan Cayao Pataca, and Cayao Pataca. Here I will focus primarily on the tributary population of these ayllus, that is, the 161 families with male heads of household between 18 and 50 years of age. Amongst these tributary households, the largest group, 51%, claim between 1 and 2 topos. By contrast, the average total area of landholdings claimed by kurakas and mandones is 5.6 topos, thus occupying the top 6 percent of households in terms of total field area. As is common in high altitude Andean settings where drought and frost are constant threats, the general risk-reduction strategy used by households was to disperse many small holdings over different vertically-distributed production zones, rather than maximizing field size in any single location. This is reflected in the small mean size of fields claimed, just one third of a topo, or about 1200 square meters. The crop mosaic for all non-elite households was about 60% maize and 39% quinoa, with 1% divided among potatoes and Cañiwa. Within the local area, the fields claimed by ayllu leaders share nearly the same proportion of maize to quinua as the rest of the tributary population, that is, 58% maize, and 42% quinua. However, when kuraka landholdings from outside the valley are calculated, their relative percentage of maize increases to 69%.

These additional maize fields are located in lower-lying valleys to the south, here shown with the distributions of Collagua ceramic assemblages I mentioned before, in the villages of Huanca and Lluta, and also in 13 toponyms within the valley of Arequipa. While ayllu leaders did not hold exclusive access to these lands, they did hold a disproportionate amount of them. That is, while they represent only 1% of total households, they held 25% of the 30 topos among these distant maize lands. This example points toward significant islands in a Collagua vertical archipelago specializing in the production of highly valued maize in valleys to the south.

There is also strong evidence regarding how the diverse production zones within the Colca valley were managed prehispanically, and how that system was mapped onto the Spanish-imposed village settlement system. The ayllus based in Coporaque also had satellite members in eight other reducción villages within the valley and in the surrounding high altitude grasslands, or puna, shown here. I reconstructed this by tracking the name of the leader listed for each ayllu in each of the villages in the census. For example, a resident of Coporaque named Martin Chuquianco is recorded as the Kuraka of the same ayllu in six of these villages in the upper part of the valley. This leader was responsible for collecting colonial taxes from tributaries in each of these villages, as I suspect his structural equivalent would have been during Inka times. This is significant first of all because it shows that these villages remained linked under a regional indigenous sociopolitical structure. But also because all of these other villages, with the exception of Yanque, are herding villages in the upper reaches of the valley.

In the terms that I discussed earlier, this example shows how Collagua intermediate elites mediated local and regional scale eco-logistics. Tracking the geographical distribution of the authority of ayllu leaders in this way gives a top-down view of how Collagua community structures were arrayed over the landscape, but equally important are the ways in which the domestic economy of individual households articulated with this supra-local political economy from the bottom-up.

A critical issue here is whether ayllu members directly accessed production zones through trade with fellow ayllu members in outlying areas, or indirectly through redistribution by ayllu leaders. As we have seen, ayllu leaders held a disproportionate amount of maize fields in the valleys to the south, and a small subset of the tributary population also had direct access to them. But how was access to pastoralist products from the high altitude grasslands mediated?

I address this question by analyzing the landholdings of commoner households within Coporaque and these herder villages. Interestingly, amongst the small percentage of households that declared agricultural fields in these herding communities, maize was the most common crop claimed. While I haven't yet analyzed this data systematically, these fields must have been located in the lower reaches of the valley, since maize agriculture is impossible in the altitudes of these pastoralist villages
(5). In fact, these maize fields were often located with toponyms that I have mapped in the area around Coporaque. This suggests that their access to agricultural goods from around Coporaque was not mediated by ayllu leaders, but their access to the lands to produce them probably was. That is, they either had rights to these maize fields as members of their ayllu, or were provided with them by their ayllu leader.
In summary, these examples illustrate how I am reconstructing what I call the eco-logistical practices of Collagua intermediate elites, and how those practices recursively structure and were structured by the built environment. On the one hand, the archaeological data I briefly summarized show how the Inka minimized administrative outlays by establishing public architecture at already established high status sites from the previous period of autonomous Collagua rule. On the other hand, early colonial documentation provided a complementary view of the hybridization of Collagua communities by showing how local corporate groups were reworked into a nested hierarchy according to Inka structuring principles. Finally, my colonial period land tenure pattern analysis illustrates how these ideal, or imagined communities, were mapped out onto the landscape, and the critical role of Collagua elites as brokers between local and regional-scale production systems.
Bauer, B. S.
1998 The sacred landscape of the Inca: the Cuzco ceque system. University of Texas Press, Austin.
Benavides, M.
1987 Apuntes históricos y etnográficos del valle del río Colca (Arequipa, Perú) 1575-1980. Boletín de Lima 50:7-20.

1988 La división social y geográfica hanansaya/urinsaya en el valle del Colca y la provincia de Caylloma (Arequipa, Perú). Boletín de Lima 60:49-53.
Bourdieu, P.
1977 Outline of a theory of practice. Translated by R. Nice. Cambridge University Press, Cambridge.
Cock, G.
1977-78 Los kurakas de los Collaguas: poder político y poder económico. Historia y cultura (10):95-118.
Galdos Rodriguez
1984 Expansión de los Collaguas hacia el valle de Arequipa. Derecho 296:81-152.
Giddens, A.
1979 Central problems of social theory. University of California Press, Berkeley.
Guillet, D.W.
1978 The Supra-Household Sphere of Production in the Andean Peasant Economy. Actes du XLLe Congrés International des Américanistes, 4:89-105. Paris.

1981 Land Tenure, Ecological Zone, and Agricultural Regime in the Central Andes. American Ethnologist 8(1):139-156.
Murra, J. V.
1968 An Aymara kingdom in 1567. Ethnohistory 15(2):115-151.
1972 El control vertical de un máximo de pisos ecológicos en la economia de las sociedades andinas. In Visita de la Provincia de León de Huánuco II, edited by J.V. Murra, pp. 427-476. Universidad Hermilio Valdizán, Huánuco, Peru.
Pease, F.
1981 Ayllu y parcialidad, reflexiones sobre el caso de Collaguas. In Etnohistoria y antropología andina, segunda jornada del Museo Nacional de Historia, edited by M. K. d. Paredes and M. M. d. Pease, pp. 19-34. Museo Nacional de Historia y la Comisión para Intercambio Educativo entre los Estados Unidos y el Perú, Lima.
Stanish, C.
1989a An Archaeological Evaluation of an Ethnohistorical Model in Moquegua. In Ecology, Settlement and History in the Osmore Drainage, Peru, edited by D.S. Rice, C. Stanish and P.R. Scarr, pp. 303-320. B.A.R. International Series 545 (ii):303-319.

1989b Household Archaeology: Testing Models of Zonal Complementarity in the South Central Andes. American Anthropologist 91:7-24.

1992 Ancient Andean Political Economy. University of Texas Press, Austin.
Tomka, S. A.
1987 Resource ownership and utilization patterns among the Yanque-Collaguas as manifested in the visita de Yanque-Collaguas, 1591. Andean Perspective Newsletter, Institute of Latin American Studies, University of Texas at Austin 5:15-24.
Ulloa Mogollón, J. d.
(1586) 1965 Relación de la provincia de los Collaguas para la discrepción de las Indias que su magestad manda hacer. In Relaciones Geográficas de Indias, edited by M. Jimenez de la Espada, pp. 326-333. vol. I., 3 vols. Ediciones Atlas, Madrid.
Van Buren, M.
1993 Community and Empire in Southern Peru: The Site of Torata Alta under Spanish Rule. Unpublished Ph.D dissertation, Anthropology Department, University of Arizona. University Microfilms, Ann Arbor, Michigan.
1996 Rethinking the Vertical Archipelago; Ethnicity, Exchange, and History in the South Central Andes. American Anthropologist 98(2)338-351.
Wernke, S. A.
2001 A reassessment of Collagua and provincial Inka ceramic styles of Arequipa, Peru. Paper presented at the 66th Annual Meeting for the Society for American Archaeology, New Orleans, LA.
Wernke, S. and E. Guerra Santander
2001 El Proyecto Prospección Regional Valle del Colca. Report to the Instituto Nacional de Cultura del Peru, Lima, Peru.
Zuidema, R. T.
1964 The ceque system of Cuzco: The social organization of the capital of the Inca. E.J. Brill, Leiden.
(1) This is a slightly expanded version from the paper I presented, but is also a working draft; bibliography is minimal. Please contact me if you would like to cite this paper, so that I may give the most up-to-date information.

(2) I am sensitive to Stanish's cogent critiques of over-dependence on portable artifacts as indicators of verticality systems (Stanish 1989a, 1989b, 1992). In this case, however, Collagua ceramics appear to make up the majority of the LIP-LH assemblages in the areas mentioned (i.e. they are much more ubiquitous than exotics in grave lots, for example), and preliminary findings from recent full-coverage survey by Erika Simborth and Clorinda Orbegoso in Huanca and Lluta suggest that architecture in these areas shares some distinctively Collagua attributes (Simborth and Orbegoso pers. comm. 2002).

(3) Zuidema noted that the organization of the nine minimal ayllus of each moiety of the Collaguas directly parallels the organization of the ceques of three of the four suyus of Cuzco: Chinchaysuyu, Antisuyu, and Collasuyu. In each of these suyus, there are nine ceques organized by a repeating pattern of three groups of three ceques, named Collana, Payan, and Cayao.

(4) Two additional lines of evidence within the visita indicate penetrating Inka presence in local ayllus: 1) the presence of three classes of official Inka imperial craft specialists embedded within the ayllu structure: weavers of sumptuous cumbi cloth (cumbicamayoc), silversmiths (listed as plateros), and potters (listed as olleros oficiales), and 2) the presence of two individuals exempt from tributary obligations for their status as descendents of the Inka Huayna Capac, suggesting marriage alliances between local elites and the panaca of Huayna Capac (Capac Ayllu).
(5) Cf Tomka 1987. Tomka reports a predominance of maize amongst households that declare agricultural fields in the upper reaches of the valley in the 1591 visita to Yanquecollaguas Urinsaya.
The field research for this project was funded by a Dissertation Grant from the Wenner-Gren Foundation for Anthropological Research. Write-up was supported by a Junior Fellowship in Precolumbian Studies at Dumbarton Oaks, and a Wenner-Gren Lita Osmundson Fellowship. I would like to especially thank Maria Benavides for providing photocopies and transcriptions of the 1604 visita. Special thanks also to Nicholas Tripcevich for all of his detailed advice in designing the GIS.