dAf 70
VACHER Stéphane, JÉRÉMIE Sylvie, BRIAND Jérôme
Amérindiens de Sinnamary (Guyane).
Archéologie en forêt équatoriale
This work introduces the initial results of research at sites occupied by Native Americans in the Sinnamary basin (1660-220 BP), carried out at the time of the construction of the Petit Saut dam by Electricité de France in French Guyana.
A methodological approach pertaining to the Tropical sylvan and riparian milieu is followed by a study of the open-air sites, and a commentary focusing on the phenomenon of terra preta and palaeolithic fires. For the first time a typological framework and a vocabulary are proposed for reference to lithic series and polishing-tool sites. This work also includes a study of ceramics using two complementary approaches : determination of place of manufacture based on the chemical make-up of the pastes employed, and codification of the forms and motifs of the 22,000 objects registered.
Following up on the publication devoted to the colonial period in this region (dAf 60), this volume makes a significant contribution to the renewal of historiographical work on the Guyana plateau.
Abstract
Abstract
The first programmed archaeological rescue project in French Guyana was carried out between 1990 and 1995, before the construction of a hydroelectric plant at Petit Saut on the middle Sinnamary river. The future reservoir, covering some 310 km2 of tropical rain forest, was subjected to a research project which marked a new stage in the development of american indian archaeology on the Guyana Plateau.
1. The methodological approach
During the stage of prospecting for sites, two methods were found to be particularly efficient, despite the problems caused by a dense forest environment. The first of these, river prospecting, consists of the systematic visiting of the rocky outcrops during periods of low water. The second method is a variant of field-walking aimed at the examination of uprooted trees which allowed a great number of subsoil observations. 273 sites (polishing stones and open-air sites) were discovered in this way. Underwater remote sensing and aerial photography gave no significant results. The sites were mapped at a scale of 1/25,000 which was accurate enough to allow test pitting of 52 open-air sites and the excavation of 5 of these after extensive topsoil stripping.
2. The open-air sites
The implantation of the sites depends on many and very variable environmental factors. They are sometimes situated on the river banks but the preference seems to be for upland areas. The inter-fluvial zones were not studied and their archaeological potential remains to be determined.
Test pits on these sites, cut by mechanical diggers, brought to light a stratigraphy which shows the systematic presence of two horizons of terra preta under the humus. The creation and evolution of this lever is uncertain but seems to be tied to human presence. Following the isolation of the intrusions due to animal and plant activities. no early soils or structures were observed. After the abandonment of these sites, soil movements and leaching wiped out the mineral components of the terra preta horizons and displaced the artefacts. This uniformisation of the soil colours masked the presence of any eventual structures (post-holes or rubbish pits). These were only visible when they had been cut into the subsoil. No spatial organisation of these features was observed, despite their density on the excavated sites.
The stratigraphy contained several layers of wood charcoal in contact with the terra preta. This showed the exista!1ce of early forest fires, brought on by periods of warmer climatic conditions. Their presence casts doubt on the validity of the 131 charcoal samples taken from the anthropic layers and used for the 14C dating.
Despite this possible contamination, different chronological markers did appear: from 10,000 to 8,000 b.p. and from 6,000 to 4,000 b.p. for the early forest fires. The american indian presence is noted between 1,660 and 220 b.p.
3. The organic material
Our understanding of the environmental conditions for the early forest fire periods is based on the results of the anthrocological analyses and give an idea of the vegetation around the sites during their transformation by human activity.
The osteological studies brought to light carbonised human and animal remains. These came from structures containing deposits of pottery, dispersed amongst the occupation levels of the open air sites.
The wooden objects were uncommon and came from the riverbed. They are treated in a descriptive catalogue (spatulae, bow staves, paddles or oars and the shafts of spindle whorls).
4. The lithic material
The study of this corpus of material was directed towards the artefacts found in situ or on the riverbed. The rock types are varied: quartz, quartzite, magmatic and lateritic rocks. It is quite clear that the functions of the products were determined by the nature of the stone used.
The study of the stone tools (knapped or roughed-out and polished) is part and parcel of the creation of a descriptive catalogue. Its aims are the normalisation and unification of the rather too . disparate terminology and the creation of a well illustrated typology.
The products of knapping were all found in situ, but away from the known workshops. They form the first known assemblage in French Guyana. The . nature of the stone determined the production techniques. The products derived from quartz are particularly frequent. They show the use of an anvil, a technique which caused a high incidence of failures and is thus accompanied by many breakages.
Despite the large amount of data represented by these finds, it is too early to propose a seriation of the sites based on the lithic material. Too many factors remain unknown, such as the specialisation of each site or the relative cultural value attached to a given type of tool or production technique.
The fixed and portable polishing stones formed the second line of research for the lithic material. A study of the workshops, which ranged from 1 to over 400 pieces, shows a regular distribution of the sites, depending on that of the rocky outcrops by the water courses. The distances between them don't exceed 9,000 m. The typology of the forms present was established using the measurements of over 1,450 stones. It allows us to distinguish, albeit in a qualitative manner, a chain of production specifically for the polishing process and, perhaps, even for particular objects.
5. The pottery
This represents the greater part of the archaeological material recovered. After a critical analysis of the means used up until now in the Amazon basin, this study proposes two different but complementary approaches used to treat the 180,000 sherds from the 74 open air sites. The first of these concerns the description and the classification of the fabrics using a binocular microscope to determine their colour and the nature and the frequence of the inclusions. This is completed by physical and chemical analyses, X-ray spectroscopy and thin sections. This first line of research allowed the confirmation of the hypothesis that each site had its own production. It also allowed the localisation of the sources of the raw materials exploited as well as the nature of the treatments used in their processing (purification, addition of tempering material, etc). The uniformity of the firing techniques became evident, as well as the identical natures of the slips. The second line of research is much more classic in its approach, even if it represents an innovation for the French Guyana. It consists in establishing a typology of forms and decorations based on a series of codes which allowed a computer-based analysis of the 21,851 recorded objects. In common with the study of the lithic material, the result was the creation of an illustrated catalogue.
6. A synthetic and critical approach
It was difficult to put the results of the Petit Saut project into relation with earlier studies. Although the themes aborded and the methods and analyses used are different, six groups of sites appear. Only the first three of these have any typological or chronolgical value. They concern the" Arauquinoid- Koriabo" assemblage which belongs to the already known traditions and complexes, as well as the «fine red ware» and "incised body" assemblages which were previously unknown. The last three groups include the sites. having no decorated elements, those having yielded insufficient material and those where the abundance of material does not allow a real comparison with other known sites in the Amazonian basin. These results put the existing typologies and chronologies into question without being able to provide new ones to replace them. A better characterisation of the phases of early forest fires remains to be done, along with the multiplication of extensive excavations. These should allow us to obtain significant assemblages of artefacts, structures and stratigraphical sequences for each of the different "cultures" observed.
Keywords
Amerindian. anthrocology. Arauquinoid. rain forest. Koriabo. osteology. early forest fires. Guyana plateau. polishing stone sites. occupation and burial sites, terra preta, typology. chronology.
