A Pre-Hispanic Chiefdom in Barinas, Venezuela: Excavations at Gaván-Complex Sites

Document Type


Publication Date



Between 1983 and 1992, the authors conducted an archaeological project that involved five years of survey and excavation in a 450 km2 study region that included portions of the high llanos (savanna grasslands) and adjacent Andean piedmont in the state of Barinas, Venezuela. Fieldwork (in 1983–1988) was followed by four years of laboratory analysis in the Departamento de Antropología at the Instituto Venezolano de Investigaciones Científicas (IVIC) in Altos de Pipe, state of Miranda. Our project was designed to investigate whether during pre-Hispanic times the study region had witnessed the development of a chiefdom, which we defined as a regional (multivillage) polity led by a paramount chief who ruled from a regional center and presided over a chiefly administration that was centralized but not internally specialized. Before beginning the fieldwork, we conducted a review of the historical literature concerning the indigenous societies encountered by the first Europeans to explore this part of Venezuela. These early accounts described various independent regional-scale societies comprising numerous villages, politically unified under the leadership of a paramount chief. This led us to wonder about the antiquity of chiefly societies and their territorial extent in our study region, so we developed a research design to explore whether chiefdoms had emerged there in pre-Hispanic times. Drawing upon the abundant theoretical and empirical literature on chiefdoms, we set forth a series of expected archaeological manifestations of chiefdom organization, grouped as follows: regional hierarchy and integration, size of the regional polity territory or territory administered, regional political center, social differentiation, political economy and village economy, ceremonialism, exchange, and warfare. We were also interested in evaluating explanatory models of chiefdom formation that emphasize a variety of causal factors, including local resource control, agricultural intensification, population growth, warfare, long-distance exchange, and ritual-ideological legitimization. Our fieldwork comprised three seasons of regional survey, during the summer months of 1983–1985, followed by two dry seasons (January–May) ofexcavation in 1986 and 1988. On survey we recorded a total of 103 archaeological sites in our study region that was centered on the Canaguá River valley, extending across the high llanos (savanna grasslands) and adjacent Andean piedmont. Site occupations pertained to two chronological periods: an early period dating to a.d. 300–1000 and a later period dating to a.d. 1000–1850, taking our coverage into the early historic period. We called the earlier of these occupations on the high llanos the Gaván complex, divided into the Early Gaván phase (a.d. 300–550) and the Late Gaván phase (a.d. 550–1000), the latter of which exhibited many of the characteristics consistent with the expected archaeological manifestations of a chiefly society. There was convincing evidence of a regional hierarchy. We recorded 34 habitation sites and two drained-field agricultural sites dating to the Late Gaván phase. Taking both site size and mounded architecture into account, we were able to define a three-tier settlement hierarchy in the region, with the 33 ha site of B12 as the sole first-tier center with the region’s largest earthen mounds, five other sites (B17, B21, B25, B30, and B97) as second-tier centers with smaller earthen mounds and ranging in size from 4.6 to 9.4 ha, and the remaining 28 habitation sites as third-tier villages that lacked mounds and varied in size from 0.5 to 5 ha. Regional integration during the Late Gaván phase was evidenced by an elaborate network of calzadas (earthen causeways or raised roads) that linked the first-tier center of B12 to four (and possibly five) of the second-tier sites and many of the third-tier sites. Our excavation strategy was designed to recover samples from all three tiers of the settlement hierarchy. We carried out excavations at five Gaván-complex habitation sites (B12, B97, B17, B21, B26) as well as at the drained-field site of B27. The size of the regional polity during Late Gaván times is consistent with the expected territorial extent of a chiefdom, which for theoretical reasons is not likely to exceed a radius of about one half-day of travel from the regional center; the reason is that such a territory could be managed by the regional leadership at the center without having to establish specialized subsidiary centers of regional administration. The distance between the first-tier center of B12 and the farthest village within its political region was about 17–18 km, and the total territory would have been roughly 290 km2. No part of the regional polity’s domain would have lain beyond a half-day of travel by foot from the first-tier center, as we would expect for a chiefdom. The site of B12 is consistent with the expectations for a regional chiefly center: larger in size and with a more formalized and imposing community layout than the other sites in the polity. During the Late Gaván phase, B12 covered more than three times the area of the next-largest site. The most impressive earthen mounds in the region were found at B12, which featured a 500 m long avenue lined by house mounds, with the two largest mounds, probablyceremonial in nature, facing each other from opposite ends of the avenue. B12 was also the only site circumscribed by an earthwork (as can be seen on the cover of this monograph) that probably had defensive functions. We judge the amount and diversity of public architecture at B12 to be consistent with the expected range for a chiefdom. Our excavations at several Gaván-complex sites found evidence of pervasive social inequality or differentiation during the Late Gaván phase, manifested by the differences between individual burials, between households, and between residential sectors within sites. Such institutionalized social differences would have helped to legitimize and reinforce the centralized but not internally specialized administrative organization of the Late Gaván chiefdom. Our Late Gaván phase data are also consistent with the expectations for the political economy and the village economy of a chiefdom. Evidence of agricultural intensification was found in the form of drained fields capable of doubling the crop yields of ordinary fields. Since there were no signs of local or regional population pressure, we conclude that the main purpose of these drained fields was to produce a surplus, which was sent to the regional leadership at B12 by way of the calzada system that linked the drained fields to the first-tier center. At the same time, local villages pursued a variety of productive activities consistent with basic economic self-sufficiency. The centralized political economy of the regional chiefdom was grafted onto a locally self-sufficient village economy that was capable of generating a surplus as part of its contribution to the regional polity. The largest and most elaborate ceremonial facilities of the Late Gaván phase were found at the regional center of B12, a pattern consistent with the expected archaeological manifestations of a chiefdom, where ritual can play an important role in legitimizing regional authority. Also, the distribution of imported luxury goods at Late Gaván sites was consistent with the expectations of a model of chiefly prestige-good exchange; exotic items obtained by the regional elite through long-distance exchange may be sent to secondary elites within the region, as a form of payment for the latter’s allegiance and assistance in mobilizing surplus on the local level. Probably as a consequence of such a two-way flow of surplus and prestige goods, certain exotic items (polished stone ornaments from the Venezuelan Andes and beyond) were found in relative abundance in our excavations at a second-tier center. It was at the regional center of B12 where we recovered the most evidence of warfare, consistent with models that view elite-directed warfare as an effective strategy for fomenting political cohesion in a regional chiefdom. B12 was circumscribed by an impressive oval earthwork that our excavations showed was topped by a palisade, similar to the defensive constructions described for 16th-century Venezuelan chiefdoms. B12 apparently suffered repeated attacksduring the Late Gaván phase, as evidenced by the recurring layers of charcoal and burned earth in the profile of the site’s largest mound. The widespread distribution of burned daub fragments in our test pits, especially in the uppermost levels, indicated that the regional center was completely abandoned after a final, catastrophic attack, which may have been launched against the Late Gaván chiefdom by a rival polity based in the Acequia–Anaro River drainage, the next major valley to the southwest. Since our Late Gaván phase data are consistent with the proposed archaeological expectations of a chiefly society, we also assess a series of explanatory models of chiefdom formation that have been proposed in recent years by anthropologists and archaeologists. Our method for testing these models involves an analytical comparison between the Early Gaván phase, when our study region shows no evidence of chiefdom organization (but was occupied by three small villages, one of which, B12, was larger than the others), and the Late Gaván phase, when multiple lines of evidence indicate that our study region was occupied by a chiefdom. To test each model of chiefdom formation, we determine whether the proposed causal variable was differentially associated with the B12 site during both the Early Gaván and Late Gaván phases, that is, over the time frame in which B12 was transformed from the largest of three small villages into the first-tier center of a regional chiefdom. Using data on changes over time in occupation area, architecture, and artifact distributions, we assess models that place causal importance on local resource control, population growth, warfare, long-distance exchange, and ritual-ideological legitimization in the formation of chiefdoms. Our analytical results are not consistent with models of chiefdom formation that attribute causal importance to ritual and long-distance exchange. Although these factors figured significantly in the dynamics of the developed Late Gaván chiefdom, they do not appear to have been instrumental in this chiefdom’s emergence. By contrast, our results do provide support for models that highlight local resource control, population growth, and warfare in the formation of chiefdoms. Not only were these factors important in the operation of the Late Gaván chiefdom, but, according to our analyses, they also played key causal roles in the initial appearance of chiefdom organization around a.d. 550.