The early Upper Palaeolithic in Haute-Normandie, and more generally in northern France, is poorly known. The industries from Saint-Martin-Osmonville and Epouville are particularly interesting because the first analyses conducted during the 1970s and 1980s led to the conclusion of a possible Châtelperronian for the first site and a mixture of Aurignacian and Mousterian levels for the second one. The stratigraphic position of the industries could be an issue. In Saint-Martin-Osmonville, stratigraphic studies in the 1980s set the industry in the Nagelbeek level (22000 BP), usually sterile in this area. Recent data have shown the existence of an important hiatus covering the previous period of time, circa 70000 - 30000 BP. At Epouville, the archaeological level is systematically associated with a thin layer of "grumeleux" [gritty] alluvium which does not always rest on the same underlying deposits. More generally, this layer tops the Mesnil-Esnard stratum, which can be changeable, dilated, divided and sometimes non-existent. The industry, included in the immediately posterior alluvium, might be dated back to 45000- 40000 BP. The "série 3 " from Saint-Martin-Osmonville consists of 1134 knapped flints. Even though the patinas may vary, the assemblage remains technologically homogeneous. There are few laminar products as opposed to the number of points. The use of a softer hammer has been attested on some pieces. The study of operational schemata shows a conceptual continuity with the production of Levallois flakes and laminar production. The knapping method using unipolar sequences dominates the others. The knapping evolves progressively from a facial to a semi-rotating technique. The management of flaking surface convexities can vary greatly. There are few retouched elements, thus it is difficult to propose cultural interpretations. Unlike A. Michel's conclusions, there are no Upper Palaeolithic type tools, there are no unfinished Châtel-perronian points. This assemblage may present some Châtelperronian features including laminar products, a more rotating knapping technique, the use of a softer hammer... But some elements are missing, partly understood operational schemata (the beginning of core exploitation is missing), rare tools and few specific ones... Nevertheless, it seems that the assemblage belongs to the Middle Palaeolithic. In Epouville, the technological analysis has shown the presence of two different archaeological levels, in which two major assemblages can be found. The most important assemblage, also know as "série grise " [grey series], consists of 1034 artefacts. Its main features are a grey patina as well as variable gloss; it has been interpreted as Mousterian. The second one, also called " série vert-marron " [green-brown series], consists of 655 artefacts. Its main features are a brownish-green patina and an important gloss. This assemblage has been extremely disturbed by frost, and has been interpreted as Aurignacian. This second assemblage was the matter under discussion. The "série vert-marron " is characterized by varied production — on the one hand, we have long, regular flakes (first generation striking), and on the other we have shorter, less regular elements (second generation). These two types of production belong to the same evolutional operational schema (a few cores show successive traces of the two sequences). The first laminar knapping is done according to the semi-rotating unipolar method. The operating table is shaped by either lateral crests or a frontal crest. The use of " neo-crested " blades is frequent. The softer hammer is the sole hammer used for this operational schema. As for second generation elements, hard hammer percussion is obvious. Their knapping corresponds to a more opportunistic management of the core. The toolmaker always uses the same method (unipolar), but in a less sophisticated way and with irregular striking. There are few retouched elements, with no sign of the Aurignacian. The " série vert-marron " seems to belong to the early Upper Palaeolithic but it is difficult to be precise. The lack of technological and typological references for these times in Northern France makes it difficult for us to propose a precise interpretation of this series.
Read full abstract