1Learning involves the mastery and assimilation of the functional parameters of a task. Learning to knap mineral materials in Palaeolithic societies is approached through repeated errors and deviations from an expected standard, expressed in degrees of competence (Pigeot 1983, 1987, 1988, Ploux 1989, Klaric 2018, Anderson 2019). Through the identification of apprenticeship, it becomes possible to address behavioural and social questions, and, more rarely, to approach the processes of assimilation and progression of learners within a group (Pigeot 1987, Ploux 1989, Anderson 2019, Olive et al. 2019). These processes require interactions between learners and tutors, which are based on observation and imitation, as well as verbal interaction, and are therefore difficult for us to perceive for extinct societies. In rare cases, however, on the scale of a site, as identified at Pincevent (Julien and Karlin 2014), and as in the case we propose to develop here, these interactions are perceptible.
2The Bourg-Charente site is a stratified open-air site located in the Charente department (France). The Palaeolithic occupations are located on a limestone slope, on the left bank of the River Charente. Excavated in 2012 as part of a preventive archaeology project, the site includes a Middle Palaeolithic level (Connet et al. 2022) and, the subject of this article, an Aurignacian level preserved over an area of 1,000 m2 (fig. 1). The latter is made up of 12 concentrations of knapped flints in areas of 4 to 13 m2 (noted L1 to n), which are knapping stations separated from each other by areas virtually devoid of remains. They were excavated by hand and systematically wet sieved. The only dating obtained on the Aurignacian level is a TL dating, which gives an age of 38.5 ± 3.0 ka (SIM 3).
Fig. 1 – The gradual acquisition of techno-economic principles during the apprenticeship of laminar debitage (seriation of these principles).
1. Necessity to operate along a dihedral less than or equal to 90°; 2. Elementary organisation of the core (setting up the functional pairing of table and striking plane); 3. Butt faceting of tabular products – setting up and maintaining a permanent oblique plane/reorientation of knapping by the opposite plane; 4. Extraction of laminar products; 5. Spur butts — crest principle – , crest restoration of the table — complex reorientation of the table by functional inversion of the core faces; 6. Better integration of the previous principles; 7. Knowing when to stop.
From Pigeot 1987, p. 79, tabl. XXIII
3In order to tackle the proposed topic, access to primary data is indispensable — in this case, relatively synchronous movable remains — as well as a grasp of the original organisation of the remains. Taphonomic analyses (granulometric analyses, measurements of fabrics, refits and technical connections of lithic remains; Bertran et al. 2019) show that the Upper Palaeolithic levels of Bourg-Charente were significantly disturbed by slope dynamics in a periglacial context (solifluction and possibly runoff), resulting in a diffusion of the remains on the slope, a preferential orientation of elongated elements and a weak sorting of the lithic series. However, the remains were buried rapidly, resulting in the approximative conservation of the original organisation of the site, particularly through the various flint knapping heaps (fig. 1). Before analysing the spatial distribution of the organisation of remains, it is first of all essential to determine the contemporaneity of the flint knapping actions as accurately as possible. To this end, only those concentrations in stratigraphic continuity and linked together by refits of knapped flints were retained in this study. The studied corpus from the Early Aurignacian is known as group F, and comprises seven concentrations of lithic remains spread over an area of 300 m²: L6, L7, L10, L12, L13, L16 inf. and L18.
4Complex F contains 6,209 flint lithic remains, half of which are waste and small fragments (less than 14 mm, whole and fragments). The materials used (Turonian flints) are blocks up to 20 cm in size, with a fine structure and regular morphology (not branched). They mainly come from the various stages of bladelet production and, for concentration L7, from small blade production (narrow, light blades less than 60 mm long). Used and abandoned blades in and around the concentrations also attest to larger blade production (L> 60 mm), and partial associated production phases. The cores from blade production are reused, after fragmentation, for the production of light blades and bladelets.
5Functional analyses reveal relatively varied activities, with light blades used as cutting tools for butchery activities and larger blades and flakes used without retouch or transformed into burins for various types of scraping work.
6At the heart of complex F is concentration L12. It comprises almost 1,200 lithic remains, including the remains of unfinished blade and bladelet production (fragmented, resolved, thick products), i.e., where the given morphological parameters of the tools for different functional purposes were not achieved. The cores were imported to L12, and are prepared along two intersecting planes (table and striking plane) with an angulation of less than 90°. The refits and technical elements show that the same action was repeated on each block and sometimes on several portions of the same block: the removal of flakes from these dihedrals formed by two intersecting secant planes. With one exception, albeit an inconclusive one, the knappers at L12 did not try to correct their errors and, at best, they prepared other dihedrals on the block to enable them to continue “production”. These observations suggest that this concentration of remains was an apprenticeship zone for flintknapping, as shown by technical and methodological errors and deviations from the expected norms. In the case of L12, the aim of the exercise was to master the percussion gesture used to detach elongated products from a dihedron formed by two intersecting planes. According to B. Bril (2020, p. 66) “Understanding learning means (…) understanding how the learner manages to control and modulate the functional parameters of a task (…)”. In the case of Bourg-Charente, this learning concerns the first stages of flint knapping, which can be summed up as the acquisition of mastering the gestures required to extract flakes according to specific methodological principles: correct orientation of the volume to be exploited in relation to the gesture; breadth, strength and restraint of the gesture; mastery of the point of contact. The learners at this stage are referred to as “postulants” by Anderson (2019) in his analysis of Aurignacian learning. The contribution of constructed volumes in L12 supports the idea that the main exercise required of L12 learners was mastery of the percussion gesture.
7From a dynamic point of view, several refits from L12 link this concentration with other knapping zones located at varying distances (from 4 to 20 m). These are either cores and fragments initially knapped outside L12 and then imported to L12 where they were knapped and abandoned, or specific stages of block shaping or repair. In the example of L12.24 (fig. 2, L12.24), the initialisation phase is located in L7, the repair phase in L13 and only the “full production” phases, even if in this case they are of poor quality, take place in L12. In another example linking L12 with L18 and L7 (fig. 2, L12.380), the element found in L7 is the only blade from this refit that was removed after preparation by abrasion of the cornice.
Fig. 2 – Location of the Bourg-Charente site.
Map É. Gaba; background map Wikimedia Commons user, Sting.
8The dynamic nature of these refits shows that the work carried out at L12 was always part of the same exercise (removal of flakes from a dihedral created by two intersecting planes). We observe that the more complex phases, and even the successful production stages, which require perfect mastery of the knapping processes, were carried out in other concentrations. The latter contain the remains of successful production and were therefore undoubtedly carried out by experienced knappers. At Bourg-Charente, the tutors/experienced knappers were called on to correct errors which are not directly linked to the exercise but which would prevent learners from continuing, and to demonstrate the required gestures and processes.
9Based on the functional data and the retouched tools, the activities carried out are centred on the flint concentration areas (fig. 3): butchery activities to the southeast of F (L6, L7, L10) on light blades produced mainly in L7; scraping activities to the north of F (L12, L13, L16 inf. and L18), as well as to the south, in L10, on blades and tools on blades (burins) produced outside the F complex.
Fig. 3 – View of the Bourg-Charente site during excavation.
Photo P. Neury.
10Apart from L12, and with the exception of L7, which is a production site for light blades, the aim of knapping activities in the other concentrations was to produce bladelets. In the absence of used or processed bladelets in area F, it is possible that these bladelets were used elsewhere.
11The apprenticeship area is spatially separate from areas where work was carried out on animal materials and from the lithic production areas of experienced knappers, although it is close to them and physically linked. The fact that the apprentice knappers at L12 were physically isolated from the other knappers could be due to the need to repeat the exercise until they had acquired the gesture, without hindering or interfering with the other activity areas. This achievement of the gesture would then depend mainly, or even essentially, on the learner’s motor skills. Emulation could arise from sharing experiences within the group of learners, under the supervision, from afar, of experienced knappers. It is conceivable that once the gesture had been acquired, the subsequent stages could be carried out in another location, perhaps closer to the experienced knappers, so that they could act as models and guides.
12The aim of the knapping exercises carried out by the learners at L12 was to acquire the percussion gesture, which is the first stage in learning to knap mineral materials (Bril 2020, Anderson, 2019). From this stage onwards, the Bourg-Charente data clearly show spatial interactions and the involvement of tutors in guiding and preparing the volumes to be used by the learners. A few other examples of such social interactions are known during the Aurignacian period (Anderson, 2019; Bordes and Bachellerie 2018; Ortega 2020), but Bourg-Charente reveals how they took place in the area of the group. The learning space is clearly dissociated from the other work areas and, in the case of Bourg-Charente, is in a central space interacting with the rest of the site. Experienced knappers provide the learners with prepared volumes and correct their errors, apparently at the request of the learners. In this way, from the earliest stages, apprenticeship is a guided activity in which all the experienced knappers take part.