There has been a heated debate for years over the existence of chiefdoms in pre-Columbian Amazonia (Carneiro 1995; Drennan 1995; Heckenberger 2003; Lathrap 1968a, 1970a; Meggers 1993-1995; Neves 1999; Roosevelt 1980, 1991b, 1999; Stahl 2002; Viveiros de Castro 1996; Whitehead 1994). If they were indeed chiefdoms, how was power financed and transmitted? Where were the political boundaries among these polities and how did they change over time? What was the role of warfare, long-distance trade and labor exploitation in the constitution of these social formations? Finally, how were hierarchies materialized in groups of objects and monumental architecture? Surveying the academic exchange, it is fair to state that the debate has been based on scant archaeological data, a few sixteenth and seventeenth century reports, and a lot of speculation. This is no one’s fault for the underdevelopment of Amazonian archaeology leaves room for it. New data from several areas where intensive research has been conducted during the last ten years, some of it yet not fully published, provides a basis for the construction of a model that moves away from typological discussion in terms of old evolutionary stages.
This new scenario is based on the premise that late pre-colonial societies of the Amazonian floodplain were cyclical, with alternating periods of political centralization and
Decentralization, the latter inferred from events of settlement abandonment and regional population decline. For instance, in the early sixteenth century AD, there was, in the Santarem area, what seems to have been a highly centralized chiefdom based on longdistance trade (Nimuendaju 2004), intensive agriculture (Woods and McCann 1999), and the production and circulation of sophisticated pottery (Gomes 2002) centered around a quasi-urban settlement (Roosevelt 1999). Something similar seems to have happened in the upper Xingu as well, far away from the main Amazonian floodplain (Heckenberger 2005).
In other areas, such as Marajo Island, the peak of political centralization, represented by mound building, happened earlier, around the eleventh and twelfth centuries AD. By the early sixteenth century AD, the social formation that generated the mounds had already collapsed.
On the other hand, along the Guianese shore evidence of political centralization that might be compatible with chiefly social formations is lacking: settlements were not large, they were occupied for short intervals and there is no sign of settlement hierarchy or monumental architecture (Boomert 2004; Rostain and Versteeg 2004).
A closer look at the archaeology of the central Amazon may help to better understand these cycling processes. Research verifies that the peak of population density and human occupation happened from the fifth to the eleventh centuries AD (Neves 2005; Neves and Petersen 2006). During this time, large and dense sedentary occupations generated the anthropogenic dark soils (ADEs) associated with artificial mounds. After this interval archaeological sites get smaller, with good evidence for short occupation spans. This process of change, also marked by the constructions of defensive structures, such as moats around sites, is probably associated with the beginning of Polychrome Tradition occupations in the area (Neves 2006). Interestingly, some of the abandoned sites are located in very productive settings, such as high bluffs overlooking the fertile Amazonian floodplain, but they were only scantily, if ever, reoccupied after the eleventh century AD (Neves and Petersen 2006).
World History
