Character of openair sites

Any attempt to identify the activities carried out in open-air sites is inevitably beset by the various problems outlined earlier - especially by the lack of organic remains from all except a small handful of sites and the lack of information from controlled excavations on the overall extent, size and internal spatial organization of sites. Some of the clearer and apparently well documented patterns which have emerged from recent work can be summarized as follows:

1. The feature which has been commented on most frequently in the earlier literature is the high proportion of open-air sites in southwestern France which appear to belong to the Mousterian of Acheulian tradition (MTA) variant. This diagnosis of course rests on the presence of typical cordiform or triangular hand axes which are to all appearances iden tical to those found in the classic MTA industries in local cave and rock-shelter sites. Whether this can be taken to imply a strict chronological synchronism with the analogous industries from the rock-shelter sites is of course a separate question which, in the absence of clear dating evidence for most of the sites, remains difficult to assess (cf. Mel-lars 1969: 147-9). Leaving this aside, characteristic cordiform or triangular hand axes do occur, frequently in large numbers, on a high proportion of documented open-air sites within the Perigord region and also in many adjacent areas to the north, south and west. Precise quantitative data are hard to establish but in the northern Perigord region, for example, over two-thirds of the recorded open-air Middle Palaeolithic sites are said to have produced typical hand axes (Ducha-deau-Kervazo 1982, 1986), while similar patterns are suggested by the studies of Rigaud

# open-air sites

□ cave/rock-shelter sites 50 km

Figure 8.11 Distribution of finds of cordiform and related hand-axe forms within the southwestern French region, as documented by Jaubert & Rouzaud (1985). Although all these finds are conventionally attributed to the MTA it should be noted that almost all the finds from open-air sites are at present undated.

# open-air sites

□ cave/rock-shelter sites 50 km

Figure 8.11 Distribution of finds of cordiform and related hand-axe forms within the southwestern French region, as documented by Jaubert & Rouzaud (1985). Although all these finds are conventionally attributed to the MTA it should be noted that almost all the finds from open-air sites are at present undated.

(1982) in the areas between the Vezere and the Dordogne and by Turq (1988a, 1989a, 1992b), Le Tensorer (1981) and others further south between the Dordogne and the Lot

(Fig. 8.12). The map compiled by Jaubert (Fig. 8.11) shows over 300 separate find-spots of apparently Mousterian hand axes from open-air localities, distributed over a broad zone

■ Cave/rock-shelter site o Open-air site

Figure 8.12 Distribution of MTA industries in cave/rock-shelter and open-air sites in the areas between the Dordogne and Lot valleys, as documented by Turq 1992b.

Figure 8.12 Distribution of MTA industries in cave/rock-shelter and open-air sites in the areas between the Dordogne and Lot valleys, as documented by Turq 1992b.

extending from the northern foothills of the Pyrenees through to the Loire valley and the western Atlantic Plain.

While these figures are impressive and leave no doubt as to the widespread distribution of technologically MTA industries within the open-air sites of southwestern France, any interpretation of these statistics should be treated with caution. As noted above, the great majority of recorded open-air sites are represented by finds of either single artefacts or by small groups of associated finds (see Fig. 8.7, and Duchadeau-Kervazo 1982, 1986). Many of the finds of supposedly MTA industries at open sites are therefore likely to represent either tools which were casually lost or discarded in the course of foraging activities or at best scatters from very brief episodes of activity on the sites. Equally significant is the relative visibility of hand axes in the course of surface collecting (due to their size and distinctive appearance) and the obvious attractions which these tools have had for generations of amateur flint collectors over the past century (Rigaud 1982). As Jaubert (1984) has emphasized, the apparent abundance and widespread distribution of hand axes in surface collections have almost certainly been greatly inflated by the combination of these two factors in the course of previous surface collecting on the sites.

Any impression that other industrial variants of the Mousterian are effectively lacking at open-air sites, however, has been clearly contradicted by discoveries over the past 20-30 years. Apparently typical occurrences of Quina-Mousterian assemblages, for example, have been recorded from the sites of Chinchon (Gironde) and Puycelsi (Tarn), as well as from at least five or six sites on the higher plateaux between the valleys of the

■ Cave/rock-shelter site o Open-air site

Figure 8.13 Distribution of Quina Mousterian industries in cave/rock-shelter and open-air sites in the areas between the Dordogne and Lot valleys. After Le Tensorer (1981) and Turq (1992b).

Figure 8.13 Distribution of Quina Mousterian industries in cave/rock-shelter and open-air sites in the areas between the Dordogne and Lot valleys. After Le Tensorer (1981) and Turq (1992b).

Lot and the Dordogne (Fig. 8.13) (Sireix & Bordes 1972; Tavoso 1987a; Le Tensorer 1973, 1981). Similar occurrences of Ferrassie type industries have been reported from open-air sites at Pons (Charente-Maritime) and Rescoundudou (Aveyron) (Lassarade et al. 1969; Jaubert 1983, 1989), while occurrences of seemingly characteristic Denticulate assemblages have been reported from a number of sites in the Euche valley in the northern Dordogne (Geneste 1985: 76-105) and further south from the sites of La Borde in the Department of Lot (Jaubert et al. 1990) and Mauran in the Haute Garonne (Girard et al. 1975; Jaubert & Brugal 1990). Occurrences of taxonomically Typical Mousterian assemblages are more difficult to document from open-air sites (largely owing to the poorly characterized nature of this variant: see Chapter 6), but have been reported provisionally from the sites of Corbiac and Fon-

seigner (Dordogne) and Tour-de-Faure (Lot) (Guichard 1976; Bordes 1984; Geneste 1985; Turq & Dolse 1988). Even if the majority of documented open-air sites appear to relate to the MTA variant, it is now clear that occurrences of most if not all the other recognized industrial variants of the Mousterian are well represented in these locations.

2. Information on total areas of occupation at open-air sites and the intensity or duration of occupation episodes is inevitably difficult to derive from the available field data. Obviously, this information can only be used with any confidence from extensively excavated sites where the overall spatial and strati-graphic extent of the occupation zones have been accurately defined. For surface collections we face not only the inevitable problem of occupational palimpsests but also the likelihood that sites which may have originated as relatively small, discrete concentrations will have been extensively disturbed and redistributed in the course of repeated ploughing or other agricultural activities on the sites.

The tendency in the recent literature has been to make a general distinction between the categories of relatively rich, extensive sites and poorer or more restricted sites (Gen-este 1985; Turq 1988a, 1989a). Some of the better documented examples of the latter category have been reported by Geneste (1985: 76-105) at a number of sites in the Euche valley in the northern Dordogne (Fig. 5.13) and by Turq (1988a, 1989a) for sites in the Lot and Lot-et-Garonne regions. Two general features of these smaller and poorer sites have been reported: first, that they tend to occur predominantly in lower locations, either along the slopes of valleys or close to the floodplain zones of valley bottoms (Turq 1988a: 104; 1989a; Duchadeau-Kervazo 1986); and second that the lithic assemblages recovered from the sites are not only relatively restricted in quantative terms, but often seem to reflect some clear specialization in the particular forms of retouched tools or different categories of flaking debitage represented (Turq 1988a, 1989a). Some of the possible implications of these patterns are discussed further below.

As noted earlier, one clear consensus which has emerged from recent work is that the majority of the much richer and more extensive open-air sites occur predominantly on higher and more exposed locations of the major plateaux of the Perigord (Rigaud 1982; Duchadeau-Kervazo 1986; Turq 1988a). Examples of such sites have been reported by Rigaud (1969, 1982) at a number of localities on the Meyrals plateaux between the valleys of the Vézére and the Dordogne, and by Turq on some of the similar high plateaux areas between the Dordogne and the Lot. To the north of the Perigord similar examples of extremely rich, extensive open-air sites have been reported at Fontmaure in the Vienne

(Pradel 1954, 1963) and at La Groix-Guémard in Deux-Sèvres (Ricard 1980).

The central problem in all these rich and extensive open-air sites is to know whether they genuinely reflect substantial episodes of occupation by relatively large human groups or simply compounded palimpsests of frequently repeated visits by much smaller groups to the same location. As noted above, one of the most conspicuous features of these sites is that they can often be shown to have been occupied at many different periods of the Palaeolithic, ranging from the Acheulian through to the Aurignacian or Upper Peri-gordian (for example at Le Dau, Coursac, Plateaux Baillard, Plateau Cabrol, La Plane, Metayer etc.: Rigaud 1969, 1982; Le Tensorer 1973, 1981; Turq 1977a, 1978, 1988a; Duchadeau-Kervazo 1986; Geneste 1985). Clearly, many of these sites were highly favourable locations, which attracted human occupation repeatedly throughout the Palaeolithic succession. Equally significant is the fact that that when systematically surveyed, they can usually be seen to break down into smaller and discrete concentrations. This has been recorded, for example, for several sites on the heavily occupied Meyrals plateau in the Sar-lat region, on the Plateau Cabrol and Plateau Baillard areas in Lot-et-Garonne and at the large open-air site of La Croix-Guémard (Fig. 8.14) in Deux-Sèvres (Rigaud 1982; Turq 1978; Ricard 1980). Such indications reinforce the impression of frequent, repeated visits to the same location by relatively small human groups. The only indication that these sites might represent rather different activities from those represented at the archaeolog-ically poorer sites is provided by the composition of some lithic assemblages recovered from the sites, as discussed further below.

3. The most thorough attempt to analyse these patterns of the varying character and composition of lithic assemblages from open-air sites has been published recently by Turq (1988a, 1989a, 1992b) based on his surveys in

Figure 8.14 Distribution of surface finds of Acheulian, Mousterian, Upper Palaeolithic and Neolithic artefacts at the site of La Croix-Guémard (Deux-Sèvres), after Ricard 1980. This site illustrates the common tendency for open-air sites to show occupation at many different periods - especially when (as in this case) located in proximity to a source of abundant flint. The Mousterian finds are represented mainly by large quantities of cordiform hand-axes.

Figure 8.14 Distribution of surface finds of Acheulian, Mousterian, Upper Palaeolithic and Neolithic artefacts at the site of La Croix-Guémard (Deux-Sèvres), after Ricard 1980. This site illustrates the common tendency for open-air sites to show occupation at many different periods - especially when (as in this case) located in proximity to a source of abundant flint. The Mousterian finds are represented mainly by large quantities of cordiform hand-axes.

□ Acheulian o Mousterian

• Post-Palaeolithic a a Upper Palaeolithic Neolithic axes the Lot and Lot-et-Garonne regions between the Dordogne and the Lot. Turq suggests that the majority of lithic assemblages recovered from open-air locations in this region can be categorized into four basic facies reflecting, predominantly, the varying representation of lithic extraction and production activities, as opposed to more general use and discard of tools on the different sites. Briefly, the main elements of his scheme can be summarized as follows (Turq 1988a: 98-9; 1989a; see also Geneste 1985: 515-7):

'Extraction and exploitation' sites As discussed in Chapter 5, these are the sites which in other contexts have usually been described as 'quarry' sites, devoted primarily to the extraction and initial working of flint resources from local, usually high quality, raw material outcrops. Turq (1988a: 98; 1989a: 188) describes one such major site from this region - that of Lascabannes, located immediately adjacent to a rich flint source on the Senonian outcrop in Lot-et-Garonne. The assemblage is characterized predominantly by a concentration of flint nodules, most of which exhibit signs of deliberate scratching or scoring of the surface (to assess the quality of the flint) followed by preliminary testing of the nodules by occasional flake removals, or more systematic removal of outer (cortical) flakes to reduce the weight of the nodules for transportation from the site. An apparent example of a similar extraction site has been reported briefly by Geneste (1985: 515) from the site of Campsegret in the Bergerac region, but has not yet been described in any detail.

7Extraction and production' sites Turq (1988a: 98) categorizes these sites as essentially 'workshop7 (atelier) locations, devoted partly to the extraction of raw materials from local flint outcrops, and partly to the systematic working down of nodules into fully prepared primary flakes (such as Leval-lois or related flakes) or to certain specific forms of retouched tools. He quotes two major sites of this type, those of Barbecornio and Lagrave, both in the Lot Department and both located immediately adjacent to rich and high-quality flint supplies (on the San-tonian and Bajocian/Bathonian outcrops respectively) (Turq 1988a: 98; 1989a: 188). The former site seems to have been devoted primarily to the production of simple, non-Levallois flakes while the latter was focused on the production of hand axes manufactured mainly from large flake blanks. The distinguishing features of these assemblages, in Turq's terms, consist of high frequencies of debitage flakes, in which flakes retaining large amounts of cortex (from initial trimming of nodules) usually represent over 50 percent of the total flake component (Fig. 5.16). The majority of finished flakes and tools are assumed to have been removed from the sites for use in other locations, so that frequencies of major primary flakes and finished tool forms are low. Significantly, most of the hand-axe forms from the Lagrave site are said to be unfinished or broken specimens, apparently abandoned in various stages of manufacture (Turq 1989a: 188).

'Mixed strategy' sites

These sites account for around 20 percent of those documented in Turq's surveys and are assumed to represent the widest and most varied economic and technological activities (Turq 1988a: 99). He emphasizes that all the sites in this category include some substantial component of basic flaking and reduction of raw flint nodules carried out wTithin the sites themselves, and that most are located very close to good-quality flint outcrops.

Their main distinguishing feature is the occurrence not only of these primary stages of flake and tool production but also high frequencies of retouched tools and, in most cases, large numbers of heavily worked and reduced cores. As examples, Turq quotes the sites of La Plane, La Burlade, Les Ardailloux and Plateau Cabrol, all represented by large and well documented flint assemblages and all apparently belonging to the MTA variant (Turq 1988a: 104; 1989a: 189).

Turq points out that in terms of overall frequencies of different lithic forms (nodules, cores, primary flakes, retouched tools etc.) it is these assemblages which appear to correspond most closely to those documented from the great majority of cave and rock-shelter sites within the region (Turq 1988a: 104; 1989a; see also Geneste 1985: 516). They seem, in other words, to represent what would normally be thought of in the context of cave or rock-shelter sites as essentially 'domestic' or 'residential' sites, reflecting a relatively wide and diverse range of activities carried out at the same location. Significantly, Turq (1988a: 104) points out that it is these 'mixed strategy' sites which occur most commonly on the higher plateaux of the region and which usually command extensive views over the surrounding habitats and ecological zones. They also generally correspond with the most archaeologically rich sites, suggesting either single periods of prolonged activity on the sites or at least frequent and repeated visits to the same location.

7Episodic' occupations

Finally, Turq (1988a: 99) describes a more heterogeneous group of sites which he refers to collectively as 'episodic' occupations or temporary stopover (halte) sites. This grouping overlaps in some respects with the other site types discussed above, but is characterized in his terms by evidence for much more ephemeral occupations suggesting very brief periods of activity on the sites. Regretta bly, few examples of specific sites are quoted, but he claims that all the sites appear to reflect some specialized, short-term activities, related either to the localized working of a few flint nodules, or to the use of a restricted range of retouched tool forms or unretouched flakes within small, spatially restricted areas (Turq 1988a: 99). His comments on the general character and distribution of the sites are similar to those of Duchadeau-Kervazo (1982, 1986) and Gen-este (1985) for sites further north in the northern Dordogne area. These small, short-term occupations are said to account for most of the documented occurrences at open-air sites (estimated at approximately 80 percent of his sample: Turq 1988a: 99) and the sites are distributed fairly widely over plateau-top anu river-valley locations as opposed to selectively on the higher plateaux (Turq 1988a: 104). It is unfortunate that so few of these sites have as yet been published. Potentially they could provide evidence for the kinds of discrete individual activities or occupation episodes which are notoriously difficult to identify at more intensively occupied sites.

How far Turq's analysis can be extrapolated to other regions of the Perigord remains to be seen. There seem to be some general similarities with the distribution patterns reported by Duchadeau-Kervazo (1982,1986) for the northern Perigord, and also with those documented by Geneste (1985) for sites in the Isle and Dronne valleys. In common with Turq, Geneste (1985: 515-6) has postulated the existence of specialized extraction and production (atelier) sites, though admitting that available information on these sites in the northern and western Perigord remains rather limited. Geneste has, however, documented a series of small-scale sites located at various points in the Euche valley (Fig. 5.13), which may correspond broadly with Turq's category of episodic or stopover sites (1985: 76-105). None of the sites has so far been investigated in great detail and their overall extent and total content of lithic material remains uncertain. Nevertheless Geneste points out that most of these sites seem to be characterized by relatively high proportions of retouched tools and usually high frequencies of heavily reduced cores (1985: Table 16) (Fig. 5.14). One of the most intriguing sites was recorded by Geneste (1985: 84-8) in the Carrière Thomasson, where a small group of artefacts (comprising four racloirs, two naturally backed knives, a single denticulate, two flint cores, and a group of 26 flakes) was found in close association with the bones of a single mammoth. In this case we would seem to have a rare glimpse of a localized butchery site, apparently focused on a single carcase of either a hunted or possibly scavenged animal (Geneste 1985: 517). This remarkable and as yet unique discovery may provide a clue to the kinds of ephemeral activities represented at some of the smaller Middle Palaeolithic open-air sites in southwestern France.

4. Finally both Geneste (1985, 1989a) and Turq (1989a) have reported some interesting patterns in the types and frequencies of raw material supplies employed on open-air sites. Geneste in particular emphasizes that open-air sites in general tend to show a stronger emphasis on the exploitation of local flint resources than those documented in the majority of cave and rock-shelter sites. Thus he points out (1985: 504; 1989a: 79) that in the case of open-air sites the frequencies of purely local raw materials (derived from distances of at most 4-5 km) range between 82 and 98 percent (with a mean value of 94 percent), whereas comparable percentages recorded in various cave and rock-shelter sites show generally lower values ranging from 66 to 89 percent (with an average of 78 percent). To some extent this may reflect simply that most open-air sites are located directly on or very close to major sources of raw materials. He also points out, however, that this pattern is equally consistent with the idea that most occupations at open-air sites were restricted either to very brief periods of time or to the exploitation of food and other resources within a short distance from the sites. These patterns tend to reinforce the notion that a high proportion of open-air sites were relatively short-term, special-activity sites, reflecting more ephemeral activities than those documented in the majority of the cave and rock-shelter occupations.

This is not, however, an invariable pattern at open-air sites. At the site of La Plane, for example, Turq (1989a: 191) has documented that whilst approximately 89 percent of the raw material was derived from local sources

(up to 2 km from the site) there is evidence for the utilization of at least ten other flint sources, located at distances ranging between 10 and 80 km to the north, east and west of the site location (Fig. 5.6). Similar patterns have been documented at the site of Le Dau (with at least six separate flint sources up to 80 km from the site) and at Tour-de-Faure (with at least four sources, up to 40 km away) (Gen-este 1985: 406-7; Turq & Dolse 1988). The evidence seems to reinforce other indications discussed above that some of the larger open-air sites may well represent the focus of relatively intensive and wide-ranging economic activities, extending over substantial areas of the surrounding habitats.

Was this article helpful?

0 0

Post a comment