The Aurignacian station of Villeneuve-sur-Lot "Brignol" (Lot-et-Garonne, France), between taphonomy and palethnology
La station aurignacienne de Villeneuve-sur-Lot "Brignol" (Lot-et-Garonne, France), entre taphonomie et palethnologie
Résumé
The open-air site of Villeneuve-sur-Lot “Brignol”, attributed to the Caminade endscraper facies of the Late Aurignacian, poses some initial difficulties when we attempt to study it from a palethnological perspective. This small site, a single occupation level on the northern banks of the Lot River, presented no evident structuration of archaeological materials during excavation and is organised as a weakly concentrated nappe of various categories of material. Clearly, “Brignol” is not a “Pompeii”, making it a poor candidate for palethnological study. We propose, however, that an overly binary distinction between “well” and “poorly” preserved Early Upper Palaeolithic sites creates a barrier to research by imposing a false dichotomy. If the overwhelming majority of existing sites cannot provide any palethnological information, then palethnological research projects for these early periods would seem to be destined for failure before they even begin. By moving past such a duality, we are forced to consult our taphonomic and archaeological data to determine the degree of conservation of a site. Such a nested approach is applied here with “Brignol” serving as a case study. Firstly, the bottom up analysis will be presented, concentrating on the different natural aspects (geomorphological, diagenetic processes, etc.) that could have influenced the organisation of the occupation level. The top down study will then use quantitative and qualitative methods of spatial analysis while making reference to certain known experimental and ethnoarchaeological models. The article finishes with a comparison of these two analyses and the presentation of a holistic vision of the site. The site is located 200 m north of the current right bank of the Lot River, in the lower portion of a sandy-silt deposit that covers the Fw2 terrace, tentatively attributed to the Middle Pleistocene (fig. 1). The Fy-z alluvial aquifer, assumed to date from the Late Pleistocene to the Holocene, is located 12 m below the occupation level. This suggests the site was likely relatively well protected from flooding at the time of occupation, but not from the rare extreme cases of flooding known historically to have reached 15 m. The single-occupation level (fig. 2), consisting of lithic materials (mostly flint and quartz) and some charcoal (faunal remains are not preserved), was excavated continuously over 71 m2 and water-screened with 2 mm mesh sieves. The nappe has an oblong shape that seems to be stretched in the direction of the slope (2° ; fig. 3). Granulometric data of the lithic fine fraction on a sample of sub-squares indicates the possible first phases of residualisation, with a general lack of 2-4 mm elements, yet this effect is absent when considering the fraction > 5 mm (fig. 4). A preferential orientation in the axis of the slope exists in the flint refits, which could be the result of water run-off, while the preferential axis is perpendicular to the slope for quartz refits (fig. 5). Finally, while flint pieces show a white patina and gloss of chemical origin, none of these items are polarised, and there is no clear evidence of post-site abandonment freeze/thaw on the pieces. Activities on site can be separated by general raw material category (fig. 7) or by type of production (fig. 10), but regardless of how one looks at the data it is clustered, rather than uniform or random (fig. 6), and the clusters are dominated by different raw materials and/or production objectives (figs. 7, 8, 10, and 11). This spatial organisation of production provides a competing/complementary explanation, both qualitatively and quantitatively, of varying lithic granulometries across the site and in the axis of the slope (fig. 8). Density contour analysis of differing categories of material allowed us to propose, using densities of burned flints, the approximate placement of a hypothetical indeterminate structure associated with a combustion event (in situ hearth ? dump ? etc.) (fig. 7e) and a relationship with the centrally dispersed heated quartz stone cobbles (fig. 7f ; fig. 11), possibly used partially or in their entirety for stone-boiling. The superposition of lithic refit data, and their quantitative and qualitative analysis, allowed us to define areas of production relative to their objectives (Caminade endscrapers, simplified blade, and raw material testing/sorting ; fig. 10 ; fig. 11), briefly discuss movement between these and other areas of the site (fig. 11), and propose a minimum number of individuals (n = 3) having frequented the Caminade endscraper production area via the seemingly synchronous and repetitive use of space (fig. 10a). A tentative palethnological reading of the site (fig. 11) provides estimations for differing and sometimes superimposed functional areas over the small surface, which fits with expectations for a short-term occupation with few people. Crossing the spatial data with data on differing degrees of know-how (two on site) allows us to provide a hypothetical minimum estimate of four individuals having used the surface area studied. A comparison of the bottom up and top down analysis allows us to confidently state that while some run-off clearly occurred its effect was minimal, and other high-energy formation processes can likely be excluded. These results in turn allow us to propose a spatial organization of the site, with varying degrees of confidence, and an initial palethnological interpretation of the use of space. While the site was still not Pompei, the nested approach that was adopted made it possible to evaluate the degree to which both natural and anthropic processes contributed to the organisation of the site, and the story told is ultimately one of relatively high resolution. The latent structuration of space at “Brignol”, and the palethnological reconstruction, serves as yet another example of the necessity to treat open-air Early Upper Palaeolithic sites with such a nested approach.
La station de plein air de Villeneuve-sur-Lot « Brignol », attribuée à l’Aurignacien récent à grattoirs Caminade, présente certaines difficultés si on tente de l’aborder selon une perspective palethnologique. Se présentant comme une nappe lâche de vestiges, sans structuration évidente lors de la fouille, le site n’a en effet rien d’une « Pompéi » préhistorique. Pourtant nous pensons qu’en prenant en compte l’ensemble des données taphonomiques et archéologiques, il est possible de donner un sens palethnologique à cet épandage. De plus, il nous apparaît que la division entre sites de plein air « bien » ou « mal » conservés freine nos interprétations archéologiques en introduisant une fausse dichotomie. Afin de dépasser cette dualité, il est nécessaire de confronter étroitement et systématiquement les données taphonomiques et archéologiques afin de mieux définir le degré de conservation des sites et, dans un second temps, d’interpréter l’organisation spatiale des vestiges. Dans un premier temps, cet article présente une analyse bottom up, en se concentrant sur les différents processus naturels qui ont pu influencer l’organisation de la nappe de vestiges. L’étude top down applique ensuite des méthodes d’analyse spatiale quantitatives et qualitatives en faisant référence à certains modèles expérimentaux et ethnoarchéologiques. Enfin, une confrontation des résultats de ces deux approches complémentaires permet une vision plus holistique du site. La structuration latente de l’organisation des vestiges à « Brignol » ainsi reconstituée et les informations palethnologiques qui en découlent montrent une nouvelle fois la nécessité d’appréhender les sites de plein air du Paléolithique supérieur ancien selon une telle approche croisée.
Origine : Fichiers éditeurs autorisés sur une archive ouverte