INTRODUCTION

The study of the pre-Columbian occupation on Marajo Island dates back to the beginnings of archaeology as a field of inquiry in Brazil during the late nineteenth century. Elaborate funerary vessels, together with other exquisite pottery objects excavated from Marajoara cemetery mounds soon filled museums in Rio de Janeiro and Belem, while short notes and articles published in important journals attracted worldwide attention to the unexpected traits of “civilization” just discovered in the tropics. For decades to come, the origins of the people who built the 10- to 12 m-high earthen mounds and the meanings of the decorative designs on their pottery were a matter of speculation.

During the first half of the twentieth century, scientists, journalists, and nonprofessional archaeologists visited and excavated the mounds located in the seasonally inundated savannas, disturbing them so much that some were turned into piles of broken sherds mixed with the sediment that had been used for mound construction. A site distribution map published by Helen Palmatary in 1950 indicated the existence of some thirty mounds or mound groups dispersed over an area of roughly 20,000 km2, leaving open the possibility that many more were yet to be found. Although funerary practices and mound features were described similarly for all locations and the ceramics were all taken as belonging to the same tradition, the excavators noticed both horizontal and vertical variation in the archaeological record; pottery sub-styles varied across the region and burial practices changed through time (Figure 19.1).

A change in research objectives and methodology took place with the arrival of Betty Meggers and Clifford Evans, who carried out the first regional survey on Marajo Island during the late 1940s, providing a comprehensive account of Marajoara culture and previous occupations. Based on ceramic attributes, Meggers and Evans (1957) defined five different archaeological phases for Marajo Island alone. With the exception of the

Handbook of South American Archaeology, edited by Helaine Silverman and William H. Isbell.

Springer, New York, 2008

Figure 19.1. Distribution of Marajoara phase sites and pottery styles. (Denise Schaan)

Marajoara Phase, which is the fourth one, the others were called “tropical forest phases,” after Steward (1948).

Meggers and Evans demonstrated that for each cemetery mound there were several habitation mounds. Assuming that ecology would limit cultural evolution, Meggers and Evans suggested that Marajoara populations had migrated from western South America, where they found similar traits of social complexity (Meggers 1954; Meggers and Evans 1957: fig. 148). They explained that mound building, necessary for survival in a seasonally flooded environment, was a cultural trait brought with the settlers.

In the 1980s, Anna Roosevelt challenged Meggers and Evans’ interpretations through her investigation of two cemetery mounds, one located in the southeastern savannas, and another at the Anajas River headwaters (see Figure 19.1). She demonstrated that these mounds contained not only burials but also abundant domestic structures, indicating that they were used for habitation, feasting, and funerary rituals. Based on radiocarbon dates from her research and previous investigations (Meggers and Danon 1988; Simoes and Figueiredo 1965), Roosevelt (1991) established that the Marajoara Phase lasted from AD 400 to 1350. Analyses of microfaunal and macrobotanical remains indicated that the diet was based on “annual cropping of seed crops, plant collection, and intensive seasonal fishing,” a diet “supplemented by tree fruits and seeds and occasional game” (Roosevelt 1991: 26, 405). Roosevelt (1999) also stated that Marajoara mounds were administratively

And economically independent, organized in a “heterarchical” system. This explanation of the complex traits exhibited by Marajoara society contrasted with the one offered by Betty Meggers.

Meggers maintained that the Marajoara people were intrusive; given the assumed ecological limitations, autochthonous development of social complexity would have been unlikely, if not impossible. Meggers (2001) explained the long duration of the Marajoara Phase by the ability of those populations to adapt to the local ecology, exploiting the region’s abundant wild resources, especially palm starch. Meggers’ migration hypothesis, however, implies a “cultural decline” through time, which has not been clearly demonstrated. In fact, while there are signs of a collapse in the regional political economy after AD 1100, expansion characterizes the period from AD 700 to 1000 (Schaan 2004). Moreover, there is a clear continuity in material culture between previous occupations by small scale, autonomous villages and the mound-building societies; there is also some evolution in settlement patterns through time, indicating that the advent of Marajoara culture was a result of a long-term process of cultural change.

Roosevelt (1999) argued that only a heterarchical mode of social organization could explain functional similarities between mounds, based on her understanding that all the known mounds were part of a single settlement system. The absence of political centralization, in her view, was due to a generalized subsistence economy, which lacked the necessary surplus to finance social complexity. Nevertheless, I see a problem of scale of analysis in Roosevelt’s view of the lack of a political center, since groups of ceremonial mounds associated with several smaller habitation mounds probably represent distinct settlement systems, some small chiefdoms of perhaps 1000 to 3000 inhabitants each.

Although Meggers and Roosevelt express different ideas on how to interpret the cultural development on Marajo Island, both explanations stem from the same assumption: that only intensive farming could support the emergence of regional, centralized, and hierarchical social systems. In the light of the restricted agricultural potential of the savanna soils and the absence of any evidence for intensive planting of seed crops, archaeological evidence for social stratification and in situ development of social complexity were overlooked by these researchers. However, a different picture emerges when the island’s particular ecology is considered, and the archaeological record is revisited.