Cave and rockshelter sites

The classic cave and rock-shelter sites have largely dominated research on the Middle Palaeolithic of southwestern France throughout this century. It is only during the past two decades that a systematic attempt has been made to extend this research to open-air sites in the region (Rigaud 1969,1982; Le Tensorer 1981; Turq 1977a,b, 1978,1988a, 1989a, 1992b; Geneste 1985). How far this may have distorted our understanding of Middle Palaeolithic settlement and activity patterns will be discussed in the following section. Nevertheless, the various river valleys of the Perigord and immediately adjacent areas have produced a remarkable concentration of cave and rock-shelter sites (Fig. 8.1), many of which contain exceptionally detailed records of human activity spanning long periods of the Middle Palaeolithic succession. As a reflection of highly intensive patterns of occupation of cave and rock-shelter locations, this concentration of sites is unique within Europe.

It is surprising that no systematic study has ever been devoted to the distribution, location and topographic aspects of all Middle Palaeolithic cave and rock-shelter sites in this region - though several partial studies focused on specific areas have been pub-

O Chez-Pourré

Laussel La Chapelle-aux-Saints

Q Vauírey cLcombe-Capelle

O Regourdou

w La Gane °Combe-Grenal

O Roe de Combe

O Moulin du Miiieu Las Póiónos O

O Chez-Pourré

O Regourdou

i«Merveillós Le Moustier Q^aDBIanchard

Laussel La Chapelle-aux-Saints

Q Vauírey cLcombe-Capelle w La Gane °Combe-Grenal

O Roe de Combe

Mas-Viel O

O Moulin du Miiieu Las Póiónos O

Mas-Viel O

Figure 8.1 Distribution of the principal Middle Palaeolithic cave and rock-shelter sites in the Perigord and adjacent areas of southwestern France.

lished (Rigaud 1982; Duchadeau-Kervazo 1982, 1984, 1986; Le Tensorer 1981; Geneste 1985; Jaubert 1984). Any study of these issues encounters widely recognized problems - the varying intensity of site prospecting in each area, the possibility of selective erosion or destruction of occupation deposits in different geological contexts and the inevitabil ity that previous research will have tended to concentrate on the richer and more deeply stratified sites in preference to the smaller and archaeologically poorer ones (cf. White 1985; Rigaud 1982; Geneste 1985). Even allowing for these potential sources of bias in the recorded distributions, there are several fairly clear and well documented patterns which have emerged from these studies which are worth reviewing here.

1. The first and most obvious point is the strong concentration of sites on the various limestone regions which outcrop in a broad band approximately 40-60 km wide around the western and southern foothills of the Massif Central (see Fig. 8.1). Obviously the dominating factor here is the natural distribution of caves and major rock shelters, which are largely confined to these geological outcrops. However, the pattern is not quite as simple as it might at first appear. As shown in Fig. 8.1, sporadic cave or rock-shelter occupations can be traced in certain areas some way to the east of the main limestone outcrops, as for example in some sandstone regions of the eastern Perigord (e.g. Chez-Pourrez in the Corrèze) and extending into the granitic and crystalline outcrops of the Massif Centrale (e.g. Baume Vallée and Rond du Barry in the Haute-Loire). To the west of this zone, only a single well documented site is known - the cave of Pair-non-Pair, located on an isolated outcrop of limestone in the Gironde department, close to the Atlantic coast.

As Turq and others have pointed out, however, some of the critical factors in these distribution patterns may have been related not only to the inherent distribution of cave and rock-shelter sites but also to the availability of lithic raw material supplies. There seems to be a tendency for both cave/rock-shelter and open-air sites to be concentrated mainly on outcrops which provided the most abundant, accessible and high quality flint, such as the extensive Senonian outcrops in the southern Perigord and some of the outcrops of Maestrichtian and related deposits further to the north-west. Other ecological and environmental factors which may have influenced these distributions have been discussed in Chapter 2 - notably, the potential importance of sheltered, valley habitats as refuges for several species of game during the colder, full-glacial episodes of the Pleistocene and the probable role of these valleys as primary migration routes for species such as reindeer and horse between the higher elevations of the Central Massif and more low-lying areas of the Atlantic Plain (see also Mellars 1985). For many reasons these sheltered, limestone valley regions of southwestern France are likely to have provided an almost ideal combination of environmental and ecological resources of critical importance to human groups throughout the whole of the last glacial succession.

2. When the distribution of cave and rock-shelter sites is examined in relation to the main river valley drainage, some other clear patterns emerge. As seen in Fig. 8.1, the highest densities occur in three major river catchments - those of the Dordogne and its immediate tributaries (c. 15 sites), the Vézère and related tributary valleys (c. 20 sites) and the various valley systems of the Dronne and Charente in the northern Dordogne region and southern Charente (c. 15 sites). Beyond this core zone, sites are distributed more sporadically, for example, in the catchment zone of the Lot river immediately to the south of the Dordogne and in some more northern valleys in the adjacent Departments of the Vienne, Corrèze, Maine-et-Loire and Deux-Sèvres (Fig. 8.1). As discussed further in Chapter 12, these distributions reveal some subtle but interesting contrasts when compared with the distribution of Upper Palaeolithic sites in the same regions (Fig. 8.2), which may indicate significantly different patterns of mobility and economic strategies during the two periods.

Neanderthal Sites
Figure 8.2 Distribution of Upper Palaeolithic sites in the Perigord and adjacent areas - including both cave/ rock-shelter and open-air sites. After Demars 1982.

3. If the site distribution is examined in relation to more localized topographic features, other interesting patterns are seen. Several workers have noted that many of the most intensively occupied sites (e.g. Combe Gre-nal, Pech de l'Aze, Laussel, Combe Capelle) tend to occur not within the main course of the major river valleys (of the Dordogne,

Vezere, Lot etc.) but in some of the minor tributary valleys which feed into them. Local concentrations of sites in these locations have been recorded for example in the small valley of the Ceou to the south of the Dordogne (Rigaud 1982, 1988), in the valley of the Beune on the north bank of the Vezere, and in the valleys of the Boulou and Rebieres, to the

Was this article helpful?

0 0

Post a comment