Levallois techniques were only one of the major strategies of primary flake production employed by Middle Palaeolithic groups. The existence of many alternative flaking strategies has been recognized since the earliest stages of research into the Middle Palaeolithic (e.g. Bourlon 1906, 1910, 1911), and it is now evident that these alternative techniques were often employed alongside Levallois methods or, in some cases, to the total exclusion of Levallois techniques. The special case of blade production techniques will be discussed in a later section. Other techniques, however, were clearly designed, as in the Levallois method, for the production of relatively broad, substantial flakes but apparently with much less conscious effort at systematic shaping of the core prior to the main sequence of flake removals. Despite the obvious variety of these techniques, they are usually grouped together collectively under the heading of non-Levallois techniques.
The question of how far disc-core techniques (Bordes 1961a: 72-3; Pigeot 1991; Boëda 1993a) can be separated from the broader grouping of Levallois methods has been raised in the preceding section. The point to emphasize is that these disc-core techniques were reliant on precisely the same basic sequence of core preparation as that in the classic Levallois techniques, involving the initial preparation of a continuous striking platform around the perimeter of this nodule, followed by successive removals of flakes from the upper surface of this nodule (Bordes 1950a, 1961a). The only criterion for differentiating between the two techniques (Levallois on the one hand, and disc-core on the other) seems to lie in the varying degrees of special preparation applied to the upper face of the core. My own inclination is to see this distinction as a matter of degree rather than of kind (see Boëda 1993a for a contrasting view). Even authors who support this basic distinction (e.g. Bordes 1961a: 85) have pointed out that the effective operation of disc-core techniques frequently involved the removal of occasional flakes to regularize the main flaking face of the cores - and thereby to allow better control over the shape and size of subsequent flake removals. In many contexts, examples of cores have been found which are apparently intermediate between typical Levallois and conventional disc-core forms (Boeda 1986; Mellars 1964). Many typical disc cores are likely to represent the heavily reduced end products of relatively intensive flaking strategies in which cores were progressively worked down from larger and more complex to smaller and simpler forms (Bordes 1961a: 16, 73) (Fig. 3.13). This would explain the relatively frequent cases where typical Levallois flakes have been recovered from sites where associated parent cores are apparently totally lacking.
More interesting technologically are the strategies which have little in common with the Levallois techniques and which reflect entirely different methodological and conceptual approaches to primary flake production. These have attracted less attention than the more impressive forms of Levallois technology and are less well documented in the literature. The most detailed studies of these strategies have been undertaken recently by Alain Turq (1988b, 1989b, 1992a,b), in the course of his studies of a number of classic Quina-Mousterian industries. As an illustration of the potential complexity and in some ways sophistication of these non-Levallois techniques, the results of Turq's recent work are worth examining closely.
Quina Mousterian assemblages are by definition non-Levallois in character and depended for almost all aspects of tool manufacture on the production of simple forms of flakes, most of which are comparatively thick in cross-section and often retain large amounts of cortex on their dorsal surfaces (Bordes 1953a, 1968a, 1981,1984: 158-60; Turq 1989b, 1992a). These pieces were used predominantly for the production of various side-
scraper forms, usually characterized by fairly heavy, invasive retouch along one edge of the flake which defines a steep, heavily convex working edge. In some cases there is evidence that tools were successively resharp-ened during use, in the course of which the worked edge bit progressively deeper into the central and thicker parts of the flakes (Lenoir 1986; Meignen 1988; see below).
Turq (1989b, 1992a) has argued that the apparently simple or even crude forms of the flakes employed in Quina industries were not simply the result of unstructured or haphazard flaking strategies but were deliberately selected as the most appropriate forms of flake blanks for producing distinctively Quina tools. He argues that the primary hallmark of these Quina flakes was the location of the maximum thickness of the flake directly opposed to the main working edge of the eventual tool, partly to assist in the handling and grip of the tool and partly to allow the maximum scope for successive resharpening in the course of use (Fig. 3.14).
By selecting these thick, triangular-section flakes it was possible to continue the process of progressively resharpening the tool edges through several different phases of tool use and still maintain a substantial rear edge to the piece which could be either held in the hand or, conceivably, attached to a wooden haft. Turq describes a number of flaking strategies by which these thick, triangular-section flakes could be produced - based partly on his studies of material from a number of typical Quina-Mousterian assemblages in the Dordogne region, and partly on his own experiments in replicating similar techniques (Turq 1988b, 1989b, 1992a, b). As he points out, the range of possible alternative strategies is flexible, and appears to have varied in response to such factors as the different flaking qualities of each raw material, and the specific forms of the nodules that were locally available. In essence, however, two basic strategies of core reduction can be identified:
1. The first has generally been referred to in the earlier literature as the 'salami slice' or 'citrus slice7 technique and has been recognized as one of the distinctive hallmarks of typical Quina-type industries since the early years of this century (e.g. Bourlon 1906,1910; see also Cheynier 1953; Bordes 1981: 79 etc.). As the name suggests, the technique consists of removing a sequence of successive, transverse slices through an elongated flint nodule, much as one might slice through a lemon or a link of salami (Fig. 3.15). The resulting flakes inevitably retain a strip of cortex around a large part of the edge, and could be transformed into characteristic Quina-type racloirs simply by applying heavy retouch to one of the longer margins of the flake (Turq 1989b, 1992a). In some of the later stages of flaking the patterns of negative scars on the dorsal surfaces of the flakes might become slightly more complex if some of the earlier salami-slice removals extended through only part of the section of the nodule. Turq stresses, however, (1989b) that this technique is only really practicable where the available raw material occurs in the form of elongated, narrow nodules and appears to be lacking in certain Quina assemblages where other forms of raw materials (e.g. tabular flint or more irregular nodules) were employed.
Was this article helpful?