Recurrent techniques

The definition of 'recurrent' Levallois techniques in Boëda's terms lies in the clear intention, from the initial stages of core preparation, to produce not one but a repeated succession of flakes of predetermined shape and size from the same, carefully prepared upper face of the core (Boëda 1988a) (e.g. Fig. 3.9). In this sense they are more economical in terms of flaking effort than the lineal techniques described above, requiring much less systematic reshaping or modification of the core between the successive flake removals. Boëda uses two main sites to illustrate these techniques - those of Biache-Saint-Vaast in the Pas-de-Calais (dating from around the middle of the penultimate glaciation, with thermoluminescence dates of ca 175,000 BP) and the later site of Corbehem (also in the Pas-de-Calais) dating from an early stage of the last glaciation (Boëda 1982, 1986, 1988c; Tuffreau 1979; Tuffreau & Sommé 1988). The detailed character of the flaking and core-preparation strategies recorded in these two sites show some major differences which led, inevitably, to corresponding differences in the resulting flake and core forms.

The principal flaking strategy employed at Biache-Saint-Vaast was intended for the production of long, narrow flakes and therefore depended on the production of elongated cores, usually of roughly rectangular form (Boëda 1988c). The initial stages in the prepa ration and shaping of these cores were effectively identical to those involved in the lineal or classical Levallois techniques and required the initial preparation of a continuous striking platform extending around the perimeter of the core, followed by a succession of centripetal flake removals to block out the domed, upper surface (Fig. 3.9). Once this initial phase of centripetal flaking was completed, however, the subsequent sequence of Levallois flake removals was carried out exclusively from either one or both ends of the core with the aim of producing a series of clearly elongated flakes. The principal feature which distinguishes it from the lineal procedures was that for each of the major sequences of flake removals a succession of either two or three flakes was removed from immediately adjacent points on the core surface without any intervening phase of deliberate preparation (or re-preparation) of the core surface (Fig. 3.10). In other words, whilst the first major flake removed in this sequence (i.e. the 'primary' flake) might be closely similar if not identical to that produced in the lineal techniques, the ensuing sequence of flake removals from the same surface will inevitably show the negative scars of the earlier removals along one or more of the margins of the resulting flake (see Fig. 3.11). In this way Boeda has been able to define a hierarchy of three separate types of adjacent, overlapping flake removals (which he refers to simply as first, second and third flakes) all removed in direct succession from immediately adjacent points on the core surface, and all showing distinctive scar patterns on their dorsal surfaces (Fig. 3.11). Where these sequences of adjacent flakes were removed from both ends of the core (i.e. in his 'bipolar' as opposed to 'unipolar' sequences) the resulting dorsal scar patterns are more complex but again can be categorized into a succession of primary, secondary and tertiary removals (see Fig. 3.10). The whole procedure seems to have been designed to produce a complex succession of elongated flakes of

Art Lineal

Figure 3.9 Schematic representation of the 'recurrent unipolar' Levallois technique as defined by Boeda. In this case the upper face of the core is prepared to produce two immediately successive, overlapping Levallois flakes, with distinctive scar patterns on the surfaces of both the detached flakes and the residual core. After Boeda 1988a.

Figure 3.9 Schematic representation of the 'recurrent unipolar' Levallois technique as defined by Boeda. In this case the upper face of the core is prepared to produce two immediately successive, overlapping Levallois flakes, with distinctive scar patterns on the surfaces of both the detached flakes and the residual core. After Boeda 1988a.

variable, but predictable outlines which eliminated the need for individual preparation of each of these intended flakes. In this sense the technique is more economical in terms of time and flaking effort than the lineal. Classified in morphological terms, the flakes thus produced can vary from essentially classic, oval Levallois flakes to other forms normally described as either Levallois blades, or elongated forms of Levallois points (Boeda 1988a,c).

Although grouped together by Boeda under the same general heading of recurrent Levallois methods, the flaking technique documented at Corbehem is much simpler than that represented at Biache-Saint-Vaast

BIPOLAR RECURRENT LEVALLOIS METHOD UNIPOLAR RECURRENT LEVALLOIS METHOD

Figure 3.10 Illustrations of the 'recurrent bipolar' and 'recurrent unipolar' Levallois techniques as represented at the site of Biache-Saint-Vaast in northern France. In both strategies the upper face of the core can be repeatedly re-prepared to produce a succession of up to three adjacent and overlapping flakes, detached from either one end of the core (in the 'unipolar recurrent' technique) or from both ends (in the 'bipolar' technique). The arroivs indicate directions of flake removals. After Boëda 1988a.

Levantine MousterianLevantine Mousterian

Figure 3.10 Illustrations of the 'recurrent bipolar' and 'recurrent unipolar' Levallois techniques as represented at the site of Biache-Saint-Vaast in northern France. In both strategies the upper face of the core can be repeatedly re-prepared to produce a succession of up to three adjacent and overlapping flakes, detached from either one end of the core (in the 'unipolar recurrent' technique) or from both ends (in the 'bipolar' technique). The arroivs indicate directions of flake removals. After Boëda 1988a.

and is, in essence, more like that of the classical Levallois (lineal) techniques, described earlier (Boëda 1982, 1986, 1988a, 1993b). Here, the whole of the flaking strategy seems to have been organized in a radial fashion; flakes were removed successively from all directions on the core perimeter (Fig. 3.12), in contrast to the exclusively unipolar or bipolar patterns of flaking documented at Biache. The critical feature which distinguishes this strategy from that documented in the lineal Levallois methods is simply that the removal of a succession of radial or centripetal flakes was carried out, apparently in close succession, without any major, intervening phase of deliberate reshaping or remodelling of the main face of the core. Thus, the form of the flakes detached in this strategy was controlled almost entirely by the patterns of earlier flake removals from the same surface, without any intermediate phase of trimming or re-shaping of the core surface (see Fig. 3.12).

One might ask how far the technique documented at Corbehem differs from the conventional notion of 'disc-core' technique as

Figure 3.11 Examples of flakes produced by the 'recurrent unipolar and 'recurrent bipolar' techniques at Biache-Saint-Vaast. Nos 1-3 are 'first-order removals; nos 4-6 'second order'; and 7-9 'third order - each characterized by distinctive scar patterns on the dorsal surfaces. After Boeda 1988a.

Was this article helpful?

0 0

Post a comment