Denis Guilbeau

In Italy, flint daggers first appear probably during the first part of the 4th millennium. These daggers were produced mainly in two areas. In the north, the Lessinian Hills flint outcrops were exploited for a production of bifacial daggers distributed throughout the northern part of Italy till the southern part of Germany. In the south, the Gargano flint outcrops were exploited first for a production of blade daggers, and second for bifacial daggers distributed throughout the southern part of the Peninsula. In central Italy, the Scaglia Rossa flint was used in a minor way to produce bifacial daggers. Other raw materials were exceptionally used for the realization of such artifacts and in the northwestern part of Italy a few blade daggers from Forcalquier (France) were found in Piedmont and Liguria. Sardinia, with only 3 flint daggers, and Sicily with only one dubious flint dagger seem to have shown little interest in such productions. We show here that the strong differences in the frequency, the technique and the morphology of the daggers between the diverse areas correspond at least partly in local and regional cultural choices and in tradition in flint knapping techniques.

In the late fifth and early fourth millennium in the central and northern regions of the Mediterranean, several specialized chipped stone productions have been documented. Following Perlès, specialization is here defined as an activity carried out by a limited number of groups or individuals in order to redistribute the products within a wider community. In southern France, numerous studies have shown the complexity of the production of pressure-knapped bladelets in heated Bedoulian blond flint in Chasséen contexts. The preforms were prepared and heated near the flint outcrops. Some bladelets were made in that area and were distributed as blanks. Some preforms were also distributed, sometimes hundreds of kilometres away from the outcrops, and were knapped there.
In Corsica, the work of L. J. Costa highlighted the presence of types of production that are reminiscent of those of the Chasséen culture. The main difference is that, here, the pressure-knapped bladelets were made from obsidian from the Monte Arci outcrops in Sardinia. Detailed analysis of the obsidian chipped stone industry from Monte Grosso in the northeastern part of the island clarifies the similarities and differences between the two types of production and permits their organization to be more precisely defined. This site was excavated during the late 1960s and early 1970s. The oldest occupation corresponds to the Basian culture, during the first half of the fourth millennium. The site was also occupied during the Bronze Age, at the end of the third millennium. We studied 1762 obsidian chipped stones from both levels. The near absence of obsidian in Corsican Bronze Age sites allows most, if not all, of the obsidian artefacts to be attributed to the Neolithic phase.
The technique, the morphology and the dimensions of the bladelets are very homogenous even if there are two different types of preparation for the platform (plain/linear/punctiform, dihedral/faceted). A small circular crack visible on the platform of a few of them indicates the possible use of a copper tip at the end of the pressure crutch used for the knapping of at least part of the bladelets. Flakes from the first phases of the chaîne opératoire are very scarce in comparison with the bladelets from the full operation. There are only 2 cores, both in an exhausted state, and we identified only 2 flakes made from former bladelet cores. The presence of rejuvenation flakes attests that knapping of the bladelets was carried out at least partly on the site. This suggests that, most of the time, the bladelet cores were introduced to the site already in an advanced phase of knapping, and left the site before the end of bladelet production. The regularity of the knapping, considered among other criteria through the analysis of the operative code, indicates that the knappers were very skilful.
Considering the quantity of bladelets found on the site, we suggest that their working by local knappers is very unlikely. We propose that the bladelets were made by itinerant knappers coming regularly to the site with bladelet cores, producing some bladelets, and taking any incompletely exploited bladelet cores away with them when they left.
Broadening our perspective, this model could also be applied to other Corsican sites like Basi. Beyond this island, only very small amounts of obsidian reached the continent, with the notable exception of a few sites, like Terres Longues (Bouches-du-Rhône) which has yielded 4548 obsidian artefacts. On this site, the obsidian from Monte Arci was probably introduced as preforms and/or as bladelet cores already partly processed. Obsidian bladelets were produced here and were then probably distributed as blanks to other sites in the south of France. In northern Italy, there is no evidence of such intensive obsidian bladelet production. However, the presence of a few cores indicates that at least part of the bladelets found there were made locally. Considering this, it is conceivable that the presumed itinerant knappers in Corsica were able to reach the shores of Italy and France to produce bladelets which were distributed as blanks to the inland regions.
Such a model has not been considered so far because the similarities between obsidian and Bedoulian heated flint knapping in southern France, especially at Terres Longues, were so strong that they were presumed to have been made by the same knappers. However, the analysis of Monte Grosso shows the presence of obsidian knapping very close to that used for Bedoulian heated flint but in an area relatively remote from southern France. These similarities include not only the raw material, since heated flint and obsidian have equivalent properties, but also the morphology and dimensions of the blanks and also the organization of their production.
The productivity of these types of knapping, as testified by experiments, is considerable and the frequency of such bladelets on archaeological sites is relatively moderate. The technical characteristics of such knapping also indicate that the knappers were very skilled and well trained. Consequently, the hypothesis of many knappers making only a few bladelets each year is not sustainable. They would not have been able to acquire, maintain and transmit the skills involved in this knapping activity. These heated Bedoulian flint bladelets and these obsidian bladelets were in both cases made by very small groups of specialists. These specialists were mobile, at least for the obsidian production.
However, analysis of the distribution of both raw materials shows marked differences. The flint bladelets are distributed mostly in inland regions and never reach Corsica or Sardinia. There is no clear geographical difference in their frequency and their management: some sites very remote from the outcrops have yielded as many bladelets as sites nearer to them. The only exception is Catalonia where this material was probably especially valued, as is shown by the presence of cores and bladelets in graves.
The obsidian bladelets are distributed mainly in coastal areas, even if several examples can be found far inland in Southern France and Northern Italy.
There are strong geographical differences in their distribution. Paradoxically, we have not so far found any obsidian pressure knapped bladelets in Sardinia. In contrast, they are abundant in Corsica. We suggest that they are here as important as the heated Bedoulian flint bladelets in Southern France. They are very rare on the continent. Considering their distribution and their presence even on sites very far from the coast, and considering the fact that their rarity cannot be explained by the difficulty in supplying obsidian to sites very far from the sources, we agree with the idea already proposed by some scholars, that obsidian was especially valued on the continent. We suggest that this value was probably linked to the raw material and not to the technique or morphology of the artefacts.