3. Lithic Production

3.1 Raw material selection

Several rock types were exploited for many tasks during the Courbet-Marine Iberomaurusian occupation. Flint was almost exclusively chosen for the production of microlithic bladelets. Moreover, homogeneous green quartzite was also knapped, but in a lower proportion. This under-representation is probably due to the fact that part of the archaeological deposit was removed from the site during previous explorations, prior to the excavation reported herein. Other rocks, like liparite and sandstone, locally available, were exploited for the manufacture of hammers and anvils. In addition, two ochre nodules with utilisation patterns were found in the archaeological deposit, in association with graphite, which is lacking in the region.

The brown flint cores are less numerous than the others (Table 1), but this fact does not influence the production of blanks. The examination of cores and lithic transported pieces indicates that the Eocene brown flint was brought to the Courbet-Marine site from Middle Miocene conglomerates, while small flint pebbles were collected from Quaternary alluvial terraces.

Table 1: Lithic raw material categories

Lithic production Brown flint Flint pebbles Quartz Quartzite Total
Cores 210 711 1 - 922
Crested pieces 362 227 1 - 590
Rejuvenation flakes 95 83 - - 178
Unretouched pieces 6043 9325 23 52 15443
Retouched pieces 3936 3383 3 9 7331

The brown flint nodules have an oxidised calcareous cortex and its distribution among cores and transported lithic pieces reveals the absence of cortex removal sequence determined by decorticating products (Fig. 7). This suggests that the preliminary flaking of brown flint nodules was initiated at the outcrops.

Figure 7

Figure 7: Cortex distribution among lithic shipped pieces

3.2 Knapping related to raw material morphology

Flint cores from Courbet-Marine site were knapped according to a wide register of shaping-out modalities to produce pointed bladelets (parallel-sided blanks measuring less than 12mm in width and twice as long as they were wide). Removal scars of these products are more important in the larger brown flint cores, while those of pebbles show limited exploitation linked to their small size. The operational modes implemented were varied, depending on raw material shapes and sizes. Indeed, two main schemes are recognised for the production of bladelets relating to the technological examination. The first is a knapping scheme that is primarily orientated to produce bladelet blanks from large flaking surfaces and which is applied both to pebbles and nodules (Fig. 8A). The second is an alternative knapping technique which consists of the exploitation of reduced blade cores (Fig. 8B). This latter, less frequent, technique was only applied to brown flint cores.

Figure 8

Figure 8: Two knapped core samples (drawings by L. Sari)

Many of the pebbles with pyramidal shapes and single striking platforms were knapped as bladelet cores, with minimal shaping that retained considerable areas of cortex, while the brown flint cores were either prismatic or resulted from the recycling of reduced or broken thicker blanks. The examination of cores and rejuvenation products of both striking platform and flaked surface indicates that shaping crested blades was more frequent in the brown flint. This reflects a more elaborate shaping-out sequence applied to this imported flint, whereas this is not the case for pebbles that have natural convexities and in which a crested blade is replaced by a first-cortical elongated blade. However, the creation of a 'néocrête' (Pélegrin 1995), which occurs during the 'plein débitage' sequence, is a frequent operation applied to both large nodules and pebbles.

Cores with single striking platforms are common. These platforms, usually located at the transverse axis of the core, owe their position to the initial raw material shape. A few examples of refitting operations, although incomplete, confirms this fact (Fig. 9). A second striking platform core was sometimes used to correct knapping errors, mainly to the main flaked surface, by plunging or hinging.

Figure 9

Figure 9: Partial refitting schema of flint pebble

3.3 Blank management

The lithic production defined at Courbet-Marine is characterised by bladelet production oriented primarly to the manufacture of backed bladelets (Table 2).

Table 2: Tool group distribution by blank type

Raw material End scrapers Burins Backed blades Backed bladelets Notches/denticulates Truncated pieces Geometric microliths Microburin technique Miscellaneous Total
Brown flint 53 5 18 3081 167 50 10 153 274 3811
Pebbles 47 0 5 6028 134 28 18 44 249 6553

Scrapers of brown flint are made into thick and wide laminar blades belonging to the earliest stages of the core reduction sequence, while those of pebbles are shaped on first flakes. They were produced essentially with hard-hammer percussion and frequently exhibit a cortical reserved zone. A few backed blades, only known in the brown flint, were made on thin, rectilinear and regular blanks obtained with soft-hammer percussion. Notches, denticulates, and truncated pieces are made on various blanks. In addition, some bladelets were transformed into truncated pieces, shouldered bladelets and perforators.

The main purpose of bladelet production was the preparation of backed bladelets. Produced by soft-hammer, bladelets were transformed mainly into straight or arch-backed pieces (Fig. 10). It is not without interest to know that the typical bladelet types such as Mouillah point and piquant-trièdre (Tixier 1963) were shaped from brown flint. Another use of this blank was in the production of geometric microliths, mainly segments and microburins. For this purpose, the chosen blanks were bladelets, selected for their regularity and straightness (Fig. 10).

Figure 10

Figure 10: Courbet-Marine lithic assemblage (Brahimi 1970)

The occurrence of broken backed bladelets is about 30%. The importance of fragmentation reflects intense activity and includes breakages resulting from both post-depositional alteration and accidents that arose during backing or repairing. These accidents are attested by the presence of cone fractures and Krukowsky microburins. Ten fragments of backed bladelets exhibit diagnostic step or hinge-terminated bending fractures, both facially and laterally generated. Moreover, these pieces show a distinctive edge-wear pattern, appearing as scars of symmetrical, crushed and deep removals (Fig. 11).

Figure 11

Figure 11: Backed bladelets showing impact fractures (photos and drawings by L. Sari)

As reported in many studies based on the technological examination of fracture patterns in connection with experimental tests (Fischer et al. 1984; Pelegrin and O'Farell 2005; Shea 2006), these impact fractures are likely to relate to a violent axial impact, specific to the projectile points. Although similar patterns have already been reported and described for the Iberomaurusian of Tamar Hat zone I (Merzoug and Sari 2008), the game-hunting modalities as well as food acquisition strategies are unknown for the Courbet-Marine site. Moreover, the techno-functional analysis requires the establishment of a specific experimental framework related to the Iberomaurusian studied sites.


© Internet Archaeology/Author(s) URL:
Last updated: Wed Jul 29 2009