12-2015, tome 112, 3, 2015, p. 517-542 - Tous aux abris ! Les occupations du Paléolithique final et du Mésolithique dans les cavités naturelles du Massif armoricain Grégor Marchand et Nicolas Naudinot

Cliquez sur la photo pour zoomer

Fichier(s) attaché(s)

 

12-2015, tome 112, 3, 2015, p. 517-542 - Tous aux abris ! Les occupations du Paléolithique final et du Mésolithique dans les cavités naturelles du Massif armoricain Grégor Marchand et Nicolas Naudinot

Télécharger le document

 

Si elles sont communes et même emblématiques dans le domaine de la Préhistoire, les fouilles dans les grottes ou abris-sous-roche sont très rares sur le Massif armoricain pour le Paléolithique final et le Mésolithique. Les travaux de terrain menés depuis 2007 dans certaines cavités du Finistère (Pont-Glas à Plounéour-Ménez, Kerbizien à Huelgoat et le Rocher de l'Impératrice à Plougastel-Daoulas) offrent un nouvel éclairage sur les manières d'habiter des espaces restreints. Une nouvelle étude est également proposée du mobilier de la seule grotte continentale de Bretagne, Roc'h Toul à Guiclan (Finistère), fouillée de manière pour le moins expéditive en 1868. Une première typologie de ces cavités et de leur remplissage pu être ensuite établie pour certains substrats géologiques seulement (granites et quartzites). A l'exception notable de la grotte marine effondrée de Ménez Drégan à Plouhinec (Finistère) occupée au Paléolithique ancien, la modicité des remplissages sédimentaires des cavernes armoricaines ne donne pas prise à des modélisations chronologiques de grande ampleur. L'intérêt qu'on leur porte est davantage dans la complémentarité fonctionnelle que leurs contenus archéologiques nous offrent en regard des sites de plein air. Dans le mobilier lithique, les phases d'usage et de réfection de l'outillage sont clairement surreprésentées, au détriment des phases de production. Cela plaide pour des usages logistiques de courte durée, dans des systèmes de mobilité différents, plus amples au Paléolithique final qu'au Mésolithique. Malgré leurs limites que cet article s'emploie à bien définir, ces premières fouilles ont bel et bien livré des signatures fonctionnelles très originales dans l'outillage ou dans les manifestations symboliques, qui engagent à réfléchir au rôle de ces petites cavernes dans les réseaux d'habitats du Paléolithique final et du Mésolithique.

 

While common and even iconic in the field of Prehistory, excavations in caves or rock-shelters are very rare in the Armorican Massif for the final Palaeolithic and Mesolithic. Why are these sites largely ignored by research in this part of France, when they form the basis of archaeological work in other parts of Europe? Are rock-shelters missing in Brittany? Of course not' Rock-shelters are actually numerous in Brittany. These sites differ however from the large caves of southern France. They have been neglected because of their small size and the shallow depth of their sedimentary deposits, and also because they do not correspond to the romantic idea people have of caves and rock-shelters??? This particular situation is essentially due to the geological context of Brittany, where the crystalline massif contains no limestone. Rock shelters are thus found in various specific contexts: naturally supported granite boulders, the foot of sandstone or schist cliffs, or marine erosion faults in granite, sandstone or schist contexts.

A new collective research programme, focused on the Late glacial and the Mesolithic, is now aiming to integrate these little known sites with a view to gaining a better understanding of the development the prehistoric palaeo-economic system. While much information has been obtained about Late glacial and Mesolithic societies in this part of Europe over the last twenty years, it essentially comes from large open-air sites that do not reflect the broad spectrum of site types making up the prehistoric palaeo-economic system. Whatever the dominant mobility system of a hunter-gatherer group, small sites are numerous and are essential to understanding the complexity of the system. These small sites are generally invisible in survey or are not examined by preventive archaeology. Investigating rock-shelters appears to be an excellent and easy way to study small sites since they have very small surface areas and are particularly easy to spot in the landscape. Since the socio-economic systems vary throughout Prehistory, our diachronic approach, which considers different environments and climates, enables us to study the development of cultural systems through time and examine how rock-shelter use changed over this long period. Although this research programme has just started, fieldwork has already been conducted in Finistère by the two authors of this paper (Pont-Glas in Plounéour-Menez, Kerbizien Huelgoat and the Rocher de l'Impératrice in Plougastel-Daoulas).

These sites offer a new insight into ways of living in small spaces. The first site is a fallen block shelter, very common in the granitic massif of Brittany. The site was first excavated in 2007 and 2008. Two main Mesolithic occupations have been identified, as well as a later occupation dated to the Iron Age (around 300 BC). The spatial analysis of the site, together with the study of the lithic assemblage, suggest that Pont-Glas is a logistic site where hunters or warriors sheltered over a short period. Kerbizien is another rock-shelter in a granitic context. The excavation carried out in July 2011 identified Tardiglacial (Azilian) levels at the back of the cave in a two metres wide band along the wall. The Early Azilian lithic industry is made up of curved back points, very utilized large blades and burin spalls. Its functional signature is original and seems quite compatible with the hypothesis of a logistic station (i.e. hunting camp) for very mobile human groups. Their knowledge of the local geological resources is a significant aspect to be kept in mind, showing that hunts were regularly made in this area.

Excavations at the Rocher de l'Impératrice are still in progress. This rock-shelter, at the foot of a high sandstone cliff overlooking the Brest roadstead, was mainly occupied during the Early Azilian. The evidence suggests this site was used by small groups of people for short periods. Activities seem to have focused essentially on game acquisition and processing. Yet contrary to Pont-Glas, there is some evidence for other activities in addition to hunting. The discovery of several dozen engraved (and sometimes coloured) schist tablets seems to point in this direction, indicating manufacture or/and use of these "symbolic" elements during the occupation of the shelter by the Early Azilians. Despite the limitations of their archaeological context, these initial excavations produced original functional signatures in tool-kits and symbolic activities, highlighting the role of small caves in settlement networks during the final Palaeolithic and the Mesolithic. These results contribute to a first attempt at modelling the use of rock-shelters in Brittany and the evolution of paleo-economic systems between the Late Glacial and the Late Mesolithic.