- retour

- sommaire

- résumé

- zusammenfassung

- réf.bilbio

dAf 78

JULIEN Michèle et RIEU Jean-Luc

Occupations du Paléolithique supérieur dans le sud-est du Bassin parisien

Three Upper Palaeolithic sites have been discovered during the construction of the A5 motorway between Melun and Troyes. Marolles-sur-Seine, in the Seine-et-Marne area, is related to the other Magdalenian sites discovered in the Paris basin. The excellent preservation of the finds (silex, animal bones, heated stones) has made possible a palaeoethnological study of this horse hunting site, a type of research rarely possible in a rescue excavation. Technological analysis of the lithic from Lailly (Aurignacian), in the Yonne area, and Fontvannes (Perigordian ?), in the Aube area, has revealed the production techniques for three kinds of artefacts: blades, flakes and bladelets. These two sites increase the importance of the Vanne valley in the understanding of occupation of the south - east of the Paris basin. Following the publication about the Middle Palaeolithic sites of the Sensarea (dAf number 47), . this volume renews the study of early Prehistory in this region.

Abstract

Abstract

Part 1 The Magdalenian occupations at Marolles-sur-Seine/Le Grand Canton (Seine-et-Marne)

1 Study framework

The deposits at Marolles-sur-Seine/Le Grand Canton lie in the interfluvial region of the Seine and the Yonne, between the Brie and the Gâtinais plateaux. The archaeological context of the region of the Yonne and the Loing is rich, with some fifteen known Magdalenian sites. These include Marsangy-Le Pré des Forges, upstream on the Yonne river, La Grande Paroisse/Pincevent and Ville-Saint-Jacques/Le Tilloy downstream on the Seine, as well as those of the Loing valley around Nemours.

2 The discovery of the site and its excavation

The deposits were discovered during a detailed site evaluation and excavated in several phases over an area of some 1,000 m². Of the three sectors exposed, only the sector 2 seems to be a habitat : the features comprise many hearths, associated with lithic material, abundant animal remains and vast dumping areas.

3 The chronological, stratigraphical and palaeoenvironmental approaches

Geological, stratigraphical and micromorphological analyses indicate that the Magdalenians were settled on the brown-yellow sands and silts which cover the coarser alluvial deposits already shaped into hummocks and depressions. This all sits on injections of the chalky subsoil, brought to the surface by intense soil movements caused by frost action, below some 40 cm of early Holocene silts. Intrusions in the pollen profiles, and the poor concentration of fossil mollusc remains, hinder accurate dating of the occupation of the site. The ¹⁴C dates are dispersed and give a calibrated date between 15,590 and 13,740 BP.

4 The land and its resources

The interest of the site lies in its diversity, which is a consequence of its position between several geographical zones. The bulk of the flint and stone resources (sandstone and other rocks) comes from the sediments carried by the Yonne river. Several nautilus shells discovered may come from nearby Cretaceous deposits, or were carried in by the Seine. The animal species most frequently encountered during the Magdalenian period in the centre of the Paris basin are horse and reindeer. These are sometimes found in association with smaller species (wolf, fox...). The discussion concerns the factors which could have determined the predominance of horse over reindeer at Marolles.

5 Hunting and the processing of the animal remains

With 5,500 identified remains, the fauna from sector 2 of Marolles/Le Grand Canton forms the most important series in the region for the Magdalenian period. Horse predominates (95% nrd), followed by reindeer (4,4%). Three wolf remains, one bovidae piece and a mammoth’s molar complete the range of animals present. No seasonal indices can be drawn from the age of the animals : the few reindeer are 7-9 year-old adults and 2 year-old immature individuals. The majority of the horses are 5-8 year-old adults. Preservation conditions apart, the low proportion of young animals correlates with that of sub-adults (2-5 years). The general profile of the slaughtered animals could be the result of successive episodes of close hunting of small isolated groups. The study of the skeletal remains has shown that the horses were killed close to the site and brought in for butchery. The small number of ribs and vertebrae suggest either differential conservation, or removal to other sites.

6 The lithic industry

A technological and typological study was carried out on the lithic material from sectors 1 and 2. This represents around 25,000 knapping products, of which 1,045 tools, with a total weight of 360 kg. The raw material is local, with a small number of imported stones. The knapping objectives seem to be the production of blades and bladelets and, occasionally, flakes. Using methods typical for the region, the blades were produced from regularly shaped elongated blocks. The bladelet production is more diversified and results from the fashioning of small, medium quality blocks or plaques, the  knapping of waste pieces, of frost broken fragments or of blade nuclei at the end of the production process. The toolkit is represented in 83% of the cases by burins, scrapers, becs, awls or micro-awls, truncated pieces, composite pieces and, in 17% of the cases, by mounted points. Some were lateral (backed blades with two units or modules : one small with direct retouching, the other being larger with direct or crossed retouching), whilst the others, rarer, are axial with curved or angular backs.

7 The combustion areas

The overall surface area exposed yielded some 1,300 kg of heated stones of which 60% were found in rubbish deposits. Cinders and charcoal remains were rarely present, the 12 hearths being recognised because of the circular organisation of more or less dense concentrations of stones. Hearth number 1 consisted of 74 kg of stones arranged in two adjacent depressions. It was sufficiently structured to allow a functional analysis. Several methods were used : mineralogical study,  analysis of the stones with a scanning electron microscope, study of their thermic alterations (distribution of oxidation, fragmentation), restitution of the evolution of the structure by refitting analyses, study of the organic sediments, etc. In its final phase of use, the structure seems not to have served as a hearth, rather hot stones may have been carried in to dry out meat, unless the whole structure served as a food reserve.

8 Analysis of the spatial distribution

Soil dilation and the even distribution of the remains over a large part of the site suggest a certain degree of disturbance and secondary deposition. Nevertheless, the lithic and bone remains in the sector 2 still seem to be organised around the hearths. A detailed spatial analysis, based on the density of each technological category, was carried out on three related hearths in an area of 70 m2. The knapping stations and the differential distribution of the tools and the bone remains were put into relation with the hearths, as were a refuse disposal zone and a common rubbish area. These observations incite us to conclude that the elements centred on the hearths function together.

Conclusion

The site shows the repeated return of small groups of hunters to the interfluvial zone, to ambush horses or the occasional isolated reindeer. The game was then carried to the camp for butchery and processing. A portion of it was probably taken away to other sites, after having been dried out. The density and the extent of the deposits at Marolles/Le Grand Canton may be explained by its formation processes. In a zone where sediment accumulation is very low, the successive installations could hardly be superposed so, over time, they spread out over a vast area.

Part 2 The Upper Palaeolithic of the Vanne valley

1 Study framework

The Vanne valley separates the Senonais plateau from the Othe forest and rejoins the Yonne at Sens. The deposits of Fontvannes/Le Bas du Fort Miroir are situated on the chalky plateau of Lailly/Le Domaine de Beauregard, near the source of the river and on the silty slope of one of its tributaries, the Alain. All the phases of the Upper Palaeolithic are represented in the region, between Arcy-sur-Cure and Marolles-sur-Seine. However, several series have few characteristic tools and only detailed technological study allows us to go beyond a broad attribution to the Upper Palaeolithic.

2 The deposits of Lailly/Le Domaine de Beauregard (Yonne)

This Upper Palaeolithic level was discovered during the excavation of a Mousterian assemblage. It is situated at the summit of a series of yellow silts from the lower or middle Pleniglacial period, 10 cm below a brown Holocene soil and contained three hearths associated with lithic industries of local flint (660 fragments). Despite the paucity of tools (20), the presence of several elements is the best indicator of the Aurignacian period. These include a Dufour type bladelet, a burin similar to the Vachon types and several streamlined scrapers which are considered to be nuclei. Flint working is organised around the hearths which are composed of blocks of sandstone without any obvious depression. Apart from the debitage on site and the immediate use of the products (scrapers, burins and unretouched blades), absences in the refitting of the nuclei indicate the carrying away of blades and, above all, bladelets for ulterior use. The weak density of these deposits around each hearth seems more characteristic of small halts to replenish a part of the toolkit rather than true campsites.

3 The deposits of Fontvannes/Le Bas du Fort Miroir (Aube)

Discovered on the surface, then evaluated by mechanical test trenches, this deposit lies in the red-brown clay silts at a depth of 50 cm. The excavation brought to light occasional pieces from the Middle Palaeolithic, two protohistoric structures and, above all, a lithic industry from the Upper Palaeolithic. Of the five zones encountered, the zone 1 was the only one to have conserved a legible structure whilst the others showed signs of disturbance. A technological study was carried out on all the local flint material (2,870 fragments). The nuclei revealed three types of production : blades, bladelets and flakes. The number of tools is low (20). Apart from five backed bladelets, the workmanship of the burins, scrapers and retouched blades is rather crude and shows little choice in the selection of the material. The only particularity of this industry is  the very heavy abrasion of the ridges of fifteen crested blades and blade products. Despite these differences, certain technological analogies could mean that this industry is related to the Gravettian or the Belloisian.

haut de page