This paper is presented exactly as it appeared in the 1987 publication with the exception that the graphics havebeen improved. The page numbers of the original publication are listed at the top of the page.
Department of Anthropology
Field Museum of Natural History
(Currently: Department of Anthropology, UCLA)
Introduction
The need for combined ethnohistorical and archaeological strategies in the construction and testing of prehispanic anthropological models has been a persistent theme in Andean studies (Murra and Morris 1976). The Central Andes is one area in the New World where rich archival materials provide details of early Colonial culture. Utilizing legal, administrative, ecclesiastical and other documentary sources, Andean ethnohistorians have reconstructed many aspects of prehispanic society, economy and ideology.
In spite of its wide acceptance in the literature however, the methodological ideal of combining archaeological and ethnohistorical research has been considerably more difficult to achieve in practice. One of the principal reasons for the absence of such interdisciplinary research is that many ethnohistoric documents either lack definable archaeological correlates or have been analyzed in terms not testable by traditional archaeological methods. On occasion, the early chroniclers describe institutions or behaviors of direct interest to prehistorians working on regional problems. In general, however, such observations are confined to specific historical events or, at the other extreme, refer to phenomena so broad as to be virtually untestable with archaeological data. The use of this class of documents by archaeologists has therefore been unsystematic and eclectic, rather than as a source of models for testing in prehispanic contexts.
Murra and Morris (1976) comments on Cieza de Leon, a major figure in Spanish Colonial historiography, underscore the fact that the classic chroniclers were concerned with general questions of Inca and Spanish Colonial society and history to the exclusion of more specific social, political and economic relationships of interest to contemporary archaeologists:
Cieza de Leon was the best ethnographer of the early European observers. But his main aim was a faithful general picture with only occasional local detail, and most of the questions about structure and organization which are of interest to modern students of the Inca could not have occurred to him. More important, he began his travels in Peru fifteen years after the arrival of Pizarro. Although this was early enough to observe many native institutions still functioning, the upper levels of the Inca state had already been dismantled or subverted to European ends (Murra and Morris 1976:272)
Although the classic accounts provide fundamental insights into the broadest aspects of Inca and non-Inca society, they rarely can be used for the more specific problems of interest to contemporary archaeologists. The fact that surveys of Inca institutions such as Rowe (1944) and Murra (1956) still remain the most authoritative works (Murra's doctoral thesis has just recently been reprinted (Murra 1981) is a reflection of the empirical limits inherent in these sources, limits that these two scholars reached a generation ago. To date, archaeological research problems have not been based on systematic data found in the chronicles. With a few notable exceptions, the chronicles are generally used as corroborating materials in an ex-post factor manner. While this is without question a valid methodological use of ethnohistoric data, it falls short of the ideal set up in the literature.
The implicit ideal of combining ethnohistory with archaeology includes the use of highly detailed information found in documentary sources as the basis of archaeological research designs. In spite of the problems with many of the classic early chronicles discussed above, certain types of ethnohistoric documents do lend themselves to archaeological applications. In particular, official censuses and records of legal proceedings are extremely useful in reconstructing patterns of political and economic organization (Murra 1956, 1964, 1968, 1972; Murra and Morris 1976:274; Rostworowski 1977). These documents differ in fundamental ways from the better known sources generally used by Inca students. Ironically, it is the institutional and bureaucratic nature of these documents that give them a systematic and comprehensive structure not found in the more commonly used sources.
The Garci Diez de San Miguel Visita of 1567
The official records of a royal inspection made by the Spanish crown official Garci Diez de San Miguel fall into the category of documents that are of greatest potential use to archaeologists. Arriving in the Titicaca Basin in 1566-1567, Garci Diez sought to document the social and economic status of the Lupaqa kingdom, one of the principal indigenous states of the region at the same time of Spanish conquest.
The area of Lupaqa influence was restricted to the southwest side of the lake in what is today the department Puno in Peru (Hyslop 1976). At the time of the Spanish Conquest, the kingdom was firmly incorporated into the Inca Empire. In spite of this
subordinate position the Lupaqa held in the imperial system, it remained the dominant indigenous power in the Titicaca Basin. The Lupaqa "capital" was located at the town of Chuquito, from which the region also derives its name. Six other cabeceras, or principal settlements, were spread more or less equidistantly along the lake shore (Figure 1).
The Lupaqa were one of only three indigenous ethnic groups in the entire Andes not granted to individual Spaniards in encomienda (Murra 1964). This rich and powerful indigenous polity was maintained as a crown holding directly under royal control and protection. Unlike other native populations held in encomienda, the Lupaqa paid tax directly to the state. This economic fact underlies the purpose of the Visita: Garci Diez was sent to record the populations of able-bodied tributaries and determine earnings from herding, farming and other economic activities in order to assess their capacity to pay taxes (Diez de San Miguel 1964: x, 5,10).
The Visita is composed of a series of questions and answers from extensive interviews with Lupaqa elite as well as a number of Spanish notables. Given that Garci Diez was primarily concerned with assessing the tax base of the region, the elicited answers focus on the more mundane aspects of Lupaqa life. The result is a comprehensive and systematic document containing information of profound social, economic and political value. Furthermore, Garci Diez was careful to repeat questions to different individuals providing a series of cross-checks on critical issues. Although any document, including the Visita, is replete with subjective biases, Garci Diez provides extensive censuses that serve as a de facto database. As such, the Visita of 1567 constitutes a far superior ethnological document of everyday indigenous life than do the classic chroniclers.
The pertinent information in the Visita includes declarations of all subject towns to Martín Curi and Martín Cusi, the caciques of Hanansaya and Hurinsaya respectively. Other types of socio-economic information include a list of all ayllus in the subject populations, the number of Catholic priests in each town, payments to the church, the nature of tribute roles during the Inca occupation, the size of camelid holdings, earnings from various economic activities and the resources controlled by various elite.
Zonal Complementarity in the South Central Andes
In 1964, John V. Murra published the first statement of the model of vertical control or zonal complementarity as an appended commentary to the Garci Diez Visita. This initial article was followed by several important publications, particularly those of 1968 and 1972. Combined, these served to offer a clear and explicit model of Andean regional political economy based upon the verticality principle. The fundamental premise of verticality is that Andean communities seek economic self-sufficiency through the control of as many different ecological zones as necessary to provide a complement to the natural products existing in the local territory. The mechanism of control is through direct colonization of the different ecozones or resource areas by resident populations. Such as discontinuous pattern of land use results in a distinctive "archipelago" distribution of territory.
The nature of Andean colonies in the zonal complementarity models differs radically from western concepts of colony. In non-Andean contexts, "colony" generally refers to a structural relationship of inferiority vis á vis the home country or territory. In most
cases, such colonial lands are composed of populations ethnically distinct from the home territory. In other cases involving territorial expansion, the colonists are ethnically similar, but almost inevitably are politically subordinate to the colonizing country.
At the folk village level, an Andean colony is characterized by a unique structural relationship in which colonists are culturally identical to the home community. Traditional economic and political relationships of reciprocity and redistribution serve to integrate the colonists within a single, but geographically discontiguous territory. A major implication of Murra’s model of direct control is regional and/or ethnic heterogeneity over the cultural landscape. In his example of the Chillon valley, for instance, three distinct polities shared lands in a single valley. A series of overlapping archipelago polities would, in fact, give rise to marked ethnic/settlement heterogeneity on a regional or even valley-wide level. As Murra points out, it is likely that this site heterogeneity has been misinterpreted by archaeologists as separate cultural phases instead of contemporary, ethnically distinct settlements.
Verticality models have undergone considerable change in the past two decades. Presently, the term "zonal complementarity" is preferred over "verticality". Such a distinction serves to emphasize that a greater range of political and economic mechanisms exist, and have existed, in the Andes to integrate discontiguous ecological zones (Masuda, Shimada and Morris 1985). A number of authors have made the conceptual distinction between "direct" and indirect control (Dillehay 1977; Forman 1978; Mujica 1985; Salomon 1978). Direct control refers to colonization of various ecological zones in the sense originally defined by Murra. Indirect control, on the other hand, is a generic term referring to exchange, alliance or conquest which serves to secure access to complementary ecological zones.
One of the major databases for the original zonal complementarity model as offered by Murra, is the Garci Diez Visita. The Visita specifically mentions a number of regions outside of the Titicaca Basin in which the Lupaqa elite controlled, or had direct access to productive agricultural lands. These areas included the Sama, Moquegua, and Locumba valleys in the southern Peruvian drainages as well as the Larecaja region of eastern Titicaca. A typical reference in the Visita to such lands is as follows:
Y este testigo ha visto tienen [the Lupaqa elite] muchas ch·caras en la costa en los pueblos de Moquegua y Sama con sus subjetos donde cogen mucha cantidad de maÌz y trigo en los cuales pueblos todas las cabeceras de esta provincia tienen chacaras sobradas / y que asimismo ha oido decir que tienen ch·caras en Larecaja de maÌz en Capinota de coca y esto sabe de esta pregunta. (Diez de San Miguel 1964:50)
This pattern of Lupaqa colonization of the maize producing zones outside of the Titicaca Basin during the Spanish Colonial period has been used as a model for the immediate pre-Inca periods when Lupaqa political influence was at its peak. In his review of Titicaca Basin prehistory for instance, Luis Lumbreras accepts this model: "The Lupaqa provided themselves with maize by maintaining colonies in the coastal valleys" (1974b: 201-202). Likewise, in a study concerned primarily with the Tiwanaku occupations of the western coastal valleys, Mujica, Rivera and Lynch (1983:94) hypothesize a sequence of Tiwanaku outliers followed by altiplano colonies in the Tambo, Moquegua and Sama Valleys.
This model of direct Lupaqa control of maize growing lands in the Osmore (Moquegua) drainage was tested with archaeological survey and excavation in the northernmost tributary of the drainage named Otora (Figure 2). The Osmore drainage is mentioned more than 20 times in the Visita as having some form of colonial relationship to the Lupaqa polity. While the total number of Lupaqa mitimae is not great (25 households in Torata, for instance¼), the consistent and repetitive listing of these lands leaves little doubt that such colonies existed and were considered a principal component of the Lupaqa political economy in the Spanish Colonial Period. Moquegua was one of the most frequently listed and important of the lands controlled by the Lupaqa state outside of the circum-Titicaca Basin. It therefore constitutes an ideal area to test this ethnohistoric model of Andean political economy with archaeological data.
An Archaeological Test of pre-Inca Lupaqa Colonization
The Osmore drainage
The Osmore drainage is located on the western slopes of the Andes in extreme southern Peru at 17† latitude x 71† longitude. The western slopes are characterized by
one of the most arid deserts in the world (see Rice this volume). In such an ecological context, the river valleys such as Moquegua provide the only areas of agricultural potential and have therefore been the locus of intense cultural development throughout prehistory.
From the coastal mouth to approximately 1000 meters above sea level (m.a.s.l.) the Osmore drainage is deeply entrenched. This topographical feature precludes irrigation
systems except for the first 10 km up from the beach. In this latter area, the potential for irrigated fields plus the marine/littoral resources provides for an ecologically rich environment. The coast has traditionally been an area of precocious and complex cultural developments throughout the Andes (Moseley 1975). The Peruvian south is no exception.
The second major area of agricultural potential in the drainage is located between 1000 and 2000 m.a.s.l. at the foothills of the Andes. In this mid-valley zone, maximum hydrological potential is combined with a topography conducive to irrigation systems. The most extensive irrigation systems are presently found in this area. The rapid elevation rise and corresponding rugged topography, from 2000 to approximately 3500 m.a.s.l, forces numerous river channel bifurcations resulting in a dendritic stream pattern. This has the effect of isolating small pockets of agricultural land between vast expanses of mountainous and agronomically unproductive desert (see Figure 2).
Prior archaeological research in the Osmore drainage indicated three major settlement systems centered in geographically distinct areas. These included: 1) a massive Tiwanaku occupation in the mid-valley and coastal region (Disselhoff 1968b), 2) a post-Tiwanaku Chiribaya settlement system located in the coastal valley of Ilo (Belan 1981; Ghersi 1956; Lumbreras 1974a) and 3) a post-Tiwanaku system, known as Estuquiña, centered in the upper sierras with several smaller sites in the mid-valley (Moseley et al. n.d.).
The two possibilities for Lupaqa colonies were the Chiribaya and Estuquiña sites. The former did not conform to expectations of Lupaqa colonies due to the absence of sites in the mid or upper valley, the area suggested by the Visita as the location of altiplano colonies. Furthermore, an analysis of Chiribaya ceramics indicated virtually no demonstrable relationship to published post-Tiwanaku Titicaca Basin styles. Finally, the sheer size of Chiribaya sites and the close stylistic relationship of associated ceramics to the Moquegua valley Tiwanaku assemblages suggested that this polity was an autochthonous development out of the earlier Tiwanaku occupation. Subsequent research confirms this chronological sequence and in fact, suggests that the Chiribaya occupation actually antedates the height of Lupaqa political and economic development when colonization would have been most likely (Stanish 1985).
The Estuquiña sites, on the other hand, were the most likely settlements that could be linked with the Lupaqa region. A research program was initiated to test these sites for colonial relationships to the Lupaqa kingdom. A hypothetical cultural sequence was formulated, one that conformed with the previous ethnohistorical and archaeological research (Figure 3). Lupaqa colonies were hypothesized to originate after the collapse of Tiwanaku influence in the region as an attempt to reestablish economic access to these complementary zones. This model was then tested with the Otora data.
The Otora Valley
Preliminary reconnaissance indicated that each of the four main tributaries of the upper Moquegua drainage had complex settlements systems characterized by fortified hilltop sites next to massive abandoned agricultural terracing. The Otora Valley is one of these small, upper sierra tributaries located between 2300 and 3000 m.a.s.l. Given its manageable size and the existence of representative Estuquiña period sites, the Otora Valley was chosen for intensive testing of the Lupaqa colony model.
The Otora Valley is located in a prime maize-growing zone. All agricultural is dependent entirely upon irrigated terrace systems. More than 200 ha of abandoned terraces and approximately 150 ha of modern, in-use terraces, line the sides of the Otora valley. The valley is on a minor camino or trail that today leads into the Moquegua puna providing easy access to the Titicaca Basin. Given its geographical location, agricultural potential and the existence of Estuquiña sites, the Otora Valley served as an ideal area to test this model of Lupaqa colonization.
The archaeological indications of direct colonization by the Lupaqa state include settlement heterogeneity due to the multi-ethnic character of colonial exploitation, distinctive Lupaqa artifacts in both domestic and non-domestic archaeological contexts, intrusive settlements as opposed to autochthonous cultural sequences and architectural similarities to the Lupaqa sites.
Settlement heterogeneity is anticipated either on a valley-wide basis such as the Chillon valley example documented by Murra (1972) or the Codpa Valley in northern Chile reported by Mujica et al. (1983), or on a regional basis as suggested by Lumbreras (1974b). That is, multi-ethnic colonization of the Osmore drainage may be recognized
archaeologically as distinct settlements within a single valley, or as site variability throughout the region with different ethnic groups controlling single, small tributaries.
Previous research by Tschopik (1946), Julien (1978), Lumbreras (1974b) and Hyslop (1976) provides a series of cultural material correlates of the Lupaqa polity. Hyslop's survey of the entire Lupaqa zone in the Titicaca Basin indicates that domestic structure architecture is almost exclusively round, with diameters ranging from three to eight. Sites are generally surrounded by massive defensive walls along the base of hills. Tschopik's work indicates that two types of ceramics are directly associated with the Lupaqa region in the post-Tiwanaku periods: Allita Amaya and Chuquito polychrome. Allita Amaya is associated stylistically with Churajon (near Arequipa) and Chiribaya and appears to be solidly pre-Inca in origin, with possible continuations into the Late Horizon. Chuquito polychromes, on the other hand, are most likely Late Horizon styles produced on Lupaqa sites in the Titicaca Basin after Inca geo-political control had been established (Lumbreras 1974b; Lumbreras and Amat 1968).
An intensive surface reconnaissance located 17 sites in the Otora Valley ranging from small sherd scatters near the valley bottom canals to the larger (120 room) Estuquiña site of Porobaya (Figure 4). The working hypothesis, based upon the previous observations and literature review was that the three large Estuquiña sites (P1, P2 and P3) were Lupaqa colonies while the smaller, hillside sites on the associated canal systems were contemporary, specialized agricultural settlements. The lower valley bottom sites were hypothesized to be either earlier or later than the Estuquiña settlements.
The surface collections tended to confirm this model in that most of
the smaller, peripheral sites had plainwares that fit within
Estuquiña assemblages. Initially, it appeared that the entire
valley was dominated by a huge Lupaqa settlement system characterized
by a site hierarchy with specialized agricultural and
non-agricultural sites. Surface recovery was inadequate to concretely
establish these dates however, and a small excavation program was
initiated. Eight major sites were selected for sampling by excavation
of complete domestic structures.
The first site tested (P5) was a small outlier associated with 15 hectares of abandoned terracing on the same canal as the large Estuquiña period site of (P1) (see Bermann et al., this volume). It originally appeared to be an ideal example of an ancillary site associated with the larger Lupaqa colony and limited text excavations were begun on a single room. In the first structure excavated, we discovered huge quantities of Tiwanaku-like finewares in unequivocal domestic context. It was painfully obvious that the 17 Otora Valley sites represented a cultural sequence of considerably greater complexity and time depth than had been originally anticipated. Additional excavations in other major sites in the valley indicated that there were at least five prehispanic phases represented in Otora. Each phase was characterized by a different system of inter-regional economic and political integration.
Based upon ceramic chronologies already developed in the Titicaca Basin and northern Chile (e.g. see Foccaci 1969; Foccaci and Erices 1971; Julien 1978; Lumbreras 1974b; Lumbreras and Amat 1968; Tschopik 1946), a five-phase prehispanic sequence was developed (Figure 5). This chronology was constructed with a data base of 37 complete structure excavations, a number of tomb excavations, and intensive surface reconnaissance. Architecture and funerary styles were also chronologically sensitive and were
utilized in constructing this sequence as well (for a complete
discussion of the method and results of the Otora investigations, see
Stanish 1985).
-----page 313-----
Tumilaca Period
This is the earliest occupation in the Otora Valley. It is represented by a single hamlet of three to four domestic units (10-15 structures). Stylistically, the ceramics are firmly associated with the post-expansive Tiwanaku settlements in the Moquegua Valley (Goldstein 1985). This period is outside of the interest of this paper. Additional information is available in other publications (Stanish 1985, 1987a; Bermann et al. this volume; Bawden this volume).
Otora Period
The newly proposed Otora Period is represented by six sites. It is defined as the time period including all of the sites that fall between the collapse of the Tumilaca Period occupation and the founding of the Estuquiña period settlements. The six Otora Period sites are therefore not necessarily contemporary, but ceramic and architectural criteria, which were utilized in constructing the chronology, were not sensitive enough to subdivide this period.
One Otora period site, known as Cuesta Alta or P7, appears to be a local development out of the earlier Tumilaca Phase hamlet of P5. The site is on the same canal system, has a few scattered Tiwanaku-derived potsherds and is architecturally similar to P5. Three other specialized sites (P15, P9 and P12), which include a corral and a quarry, are probably contemporary with P7.
A second major site, known as Cuajone or P8, is interpreted as an altiplano colony linked to the northern side of the Lake Titicaca Basin (specifically the Colla region). The evidence comes in the form of 1) room architectural similarities to published Titicaca Basin types, 2) a peculiar type of funerary architecture, known as the collar tomb, which has affinities to M. Tschopik's type 3 chulpa found only in the northwest side of the lake basin, and 3) decorated ceramic styles found in domestic contexts suggestive of published Colla ceramics (Stanish 1985; Tschopik 1946).
The third and final major Otora period site (P4 or Porobaya Chica) is also interpreted as a colony, but one linked ceramically and architecturally to the coastal and lower Ilo Valley Chiribaya polity. Judging from the ceramics, both P4 and P8 may be later than P7 and its related sites. On the other hand, given the uncertainty of the ceramic chronologies, these style variations may reflect ethnic differences in contemporary sites, as suggested by Murra.
Regardless of the exact chronological position of these six sites, the Otora Period in the Otora Valley is, in fact, a classic example of Murra's direct verticality model in which a productive valley is colonized by different ethnic groups (Murra 1968; 1972). His example from the Chillon Valley is a particularly appropriate analogy for the Otora Period (Murra 1968; 1972). It is important to note, however, that there is no link whatsoever to the Lupaqa region. All altiplano "influence" seen in the post-Tiwanaku Otora Valley would appear to derive from the northern side of the lake associated with the Colla polity. That is, the collar tombs, which are architecturally similar to Tschopik's type 3 (although much smaller) are found only on early sites in the northern side of the Lake. Likewise, the Colla or Colla-like ceramics suggest northern Titicaca Basin linkages as does the domestic architecture on P8.
Estuquiña and Estuquiña-Inca Periods
The next two periods, Estuquiña and Estuquiña-Inca, are represented by three, possibly four, architecturally similar sites in the valley. The Estuquiña-Inca Period is defined by the presence of Inca polychromes on culturally pre-Inca settlements. The site of Porobaya or P1 (Figure 6), was the principal Estuquiña Period site excavated in the valley. Fifteen structures were tested. As these were hypothesized to be the Lupaqa colonies, it is necessary to examine the excavation and surface data in some detail.
One of the most sensitive archaeological indicators of ethnic affiliation or exchange is decorated ceramics in domestic contexts (Stanish 1985). The dominant exotic ceramic in the Otora Valley Estuquiña Period sites was a pre- and post-Inca Sillustani variant (Figure 7). Characterized by black-on-red burnished bowl forms, the distribution of this family of ceramics extends throughout a wide area in the South Central Andes often having distinctive local names. The Sillustani tradition is unequivocally associated with the northern side of the Lake Titicaca Basin where it is found in greatest abundance (Hyslop 1976; Julien 1978; Lumbreras and Amat 1968; Lumbreras 1974b; Tschopik 1946).
Two other classes of exotic ceramics as well as local decorated type (Porobaya Tricolor) were found in the Estuquiña Period sites: Inca polychromes and a number of Gentilar fragments. The Inca styles are generally those known as Cuzco A and B (Rowe 1944) and probably were imports from the Titicaca Basin which may have been under Inca control by this period. Gentilar is a ceramic tradition closely linked to the Azapa/Tacna areas on the extreme southern Peruvian and northern Chilean coast. It is unquestionably a coastal and low sierra tradition.
The two most diagnostic and common plainware types, bootpots and Estuquiña bowls (Figures 8 and 9), have almost never been reported from the Titicaca region, but do occur in great numbers in the Ilo and Arequipa areas associated with Chiribaya and Churajon sites respectively. In sum, the ceramic data indicate that the Estuquiña plainware diagnostics are not found in the Titicaca Basin and only one out of four main decorated types are altiplano in origin. The one altiplano ceramic --Sillustani-- is not associated with the Lupaqa region.
A second major ethnically sensitive archaeological indicator in the Otora Valley is the architectural plan of domestic structures. John Hyslop had reported that the dominant house shape in the Lupaqa area is round. In contrast, rectangular or squarish forms tend to dominate in the later prehistoric sites of the northern side of the lake. The architectural plan at all Estuquiña sites in the Osmore drainage is rectangular, without exception. The only round structures in Moquegua are found on very early Pukara-affiliated sites (see Feldman and Goldstein 1986; Feldman this volume) and are not associated with Tiwanaku or post-Tiwanaku settlements.
Based upon my analysis of the excavation and survey data, I have proposed that the site of Porobaya is representative of a widespread, politically independent Estuquiña settlement system. The Estuquiña sites developed directly out of the earlier Otora Period multiethnic valley. There are a number of strong material continuities between the Estuquiña sites and the earlier Otora Period ones. These include funerary architecture (collar tomb to chulpa), room construction, agricultural land use and ceramic remains in both domestic and non-domestic contexts. In general, room plan and tomb styles are northern Titicaca Basin influenced (perhaps introduced by the colonists at P8) while the ceramic assemblages fall quite clearly within lower sierra and coastal traditions.
In contrast to the Otora Period, where sites were very different from one another, the Estuquiña period is characterized by an extreme site homogeneity across the upper sierra of the Osmore and southern Tambo drainages. What we see in these two periods --Otora to Estuquiña-- is the cultural transformation from heterogeneous, multiethnic, low population density settlements to a culturally distinct, homogeneous, nucleated and politically independent settlement system. All evidence suggests that this cultural transformation occurred in the same biological population there is no evidence of major population movements or colonizations after the end of the Otora Period.
The Inca Period
The Inca sites in the Moquegua drainage provide what I believe is the answer to the paradox presented by the data found in the Garci Diez area while the archaeological evidence presented above indicates no such influence prior to the Inca Period. Unlike the Estuquiña sites, Late Horizon local and Inca administrative settlements have both Inca ceramics as well as Chuquito polychromes. Chuquito polychromes are currently linked with the Lupaqa Kingdom located on the southwest side of Lake Titicaca (Hyslop 1976; Lumbreras and Amat 1968; Tschopik 1946). The first archaeological evidence of Lupaqa influence in the Moquegua drainage therefore correlates to the establishment of Inca administrative settlements in the lower valleys.
Figure 7.
-----page 317-----
The date from Moquegua indicate that Sillustani or Sillustani-like ceramics, (again, linked stylistically with the Colla polity on the northwestern side of Lake Titicaca) antedate the appearance of Chuquito polychromes in the region. Ethnohistoric data indicate that the Colla and Lupaqa were bitter enemies and that the Lake Basin in the post-Tiwanaku Altiplano Period was characterized by internecine warfare (Lumbreras 1974a; Hyslop 1976). These data also suggest that the Inca Empire formed a political alliance with the Lupaqa against the Colla (Lumbreras 1974a). Archaeological data from the Titicaca Basin support this historical scenario. Chuquito polychromes are found throughout the Colla region and cover principal Colla sites (e.g. Hatuncolla ñ Julien 1978) but the Sillustani ceramic styles are rarely found on the Lupaqa sites (e.g. see Hyslop 1976).
The implications for the Moquegua region become obvious. The initial altiplano colonization of the Otora Valley during the Otora Period is linked with the Colla region from the northwestern Titicaca Basin. These newly founded and small populations share an agriculturally productive valley with coastal groups who were also engaged in an active colonization program during this post-Tiwanaku, pre-Inca period. In the following Estuquiña Period, the multiethnic composition of the valley gives way to an autochthonous settlement system with extensive exchange relationships between both the coast and Titicaca Basin (as indicated by Gentilar and Sillustani ceramics respectively). The system of direct control through colonies transformed to an indirect control system based upon exchange.
The destruction of the Colla polity by the Inca/Lupaqa alliance is reflected in the ceramic distribution of Otora Valley and the Osmore drainage sites. Pre-Inca sites have Gentilar, Sillustani and Inca vessels, as well as the indigenous polychrome known as Porobaya Tricolor. The Inca sherds are generally finely-made imperial styles suggesting that they were part of an economic exchange system between the Inca-dominated Titicaca Basin, conquered possibly one generation before the Osmore area, and the sierra itself. Likewise, the few Inca sherds may represent a regional exchange system involving Inca, Colla and the Osmore sierra populations before actual Inca geo-political control of the Titicaca Basin and Inca-Lupaqa alliance. These models have yet to be tested and await future research in the region.
During the actual domination of the Moquegua drainage by the Inca state, Colla-affiliated ceramics were replaced by Chuquito polychromes suggested that the political and economic exchange relationships controlled by the pre-Inca Colla were co-opted by the Lupaqa elite in conjunction with Inca imperial policy. During the Spanish Colonial period when Garci Diez de San Miguel completed his census, the Lupaqa elite justified their claim to the Moquegua lands by reference to the status quo imposed during the Inca administration. The lists of Lupaqa colonies compiled by Garci Diez may therefore be the result of Inca political strategies employed in the subjugation of the South Central Andes. In short, there is no evidence of Lupaqa colonization of Moquegua before the Late Horizon.
Conclusion
A model of the pre-Inca Lupaqa colonization in the Moquegua drainage is not supported by the Otora Valley data. The most direct Titicaca Basin influence during the
pre-Inca/post-Tiwanaku periods comes from the north side of the lake where the Colla polity was located. The inter-regional political and economic relationships in the areas impacted by Titicaca Basin societies were complex and dynamic. The nature of these relationships fluctuated between two poles: actual direct colonization and indirect exchange systems between politically independent groups. No one model of interzonal exchange or colonization is adequate to understand the 2000+ years of agriculturally based settlements in the South Central Andes.
-----page 319-----
The first Lupaqa presence is correlated to initial Inca geo-political control of the Osmore drainage. The ethnohistorically documented alliance between the Lupaqa and Inca against the Colla is reflected in the Moquegua archaeological remains. In contrast, the immediate pre-Inca regional political economy was characterized by Colla dominance of sierra-Titicaca Basin exchange. This system was replaced by a more formalized Inca administration in which their allies, the Lupaqa polity, exercised a marked influence.
The Inca period policy of using Lupaqa elite in their administrative hierarchy profoundly affected the cultural landscape of the Osmore drainage and Titicaca Basin. Lupaqa political and economic influence continued into the Spanish Period, as the Garci Diez Visita clearly demonstrates. This document, however, reflects the realities of Inca policy in the drainage; the pre-Inca periods must be studied with the data which can be provided only by the archaeological record.
-----page 320-----
Notes
1 ". . . y los subjetos a él le dan sesenta indios de servicio que se ocupan en las cosas siguientes los diez indios en guardarle el ganado de la tierra que tiene y veinte y cinco indios que tiene en Moquegua que le labran sus chácaras de maíz que allí tiene" (Garci Diez de San Miguel 1964:21)
2 One passage in the Garci Diez Visita actually suggests Aymara speaking Colla groups were living in the Sama Valley town of Torata [sic?, probably "Tarata"]:
"En el pueblo de Torata . . . el señor Garci Diez de San Miguel visitador de la provincia de Chuquito y su partido estando presentes con don Martín Ninaca y don Juan Curí otros indios de este repatimiento les preguntá· se ellos y los dem·s indios cristianos que hay en él se confiesan la Cuaresma y cuando se quieren morir dijeron que cuando algunos indios entienden la lengua del Cuzco y la entiende también el padre de la doctrina se confiesan algunos de ellos y lo mismo los enfermos y cuando no la entienden no se confiesan porque ninguno de los frailes que viene a esta provincia saben la lengua colla que es la de esta provincia [emphasis added] (Garci Diez de San Miguel 1964:194).
Copyright 1989 Charles Stanish (stanish@anthro.ucla.edu)
See Chip's web page for further publications and other information about his work.