The area distribution of the artefacts in the flint workshop of the San Bartolomeo Rock-shelter (Abruzzo, Central-Italy)

The S. Bartolomeo rockshelter is situated at 800 m a.s.l. in a narrow valley on the northern side of the Maiella mountain, on the eastern side of the Appennine ridge in Central Italy. The place was a good campsite because the shelter faces to the South, water was available from the nearby stream and chert nodules crop out from the bottom wall. The shelter was frequented during the Late Glacial, when the climatic amelioration improved the mountain area environment. Though 14C datings are still unavailable, the typological characteristics of the Epigravettian tool assemblage suggest an age of slightly younger than 12’000 BP; moreover, a thin level of primary tephra that could be ascribed to the Tufo Giallo Napoletano (~11’900 BP) volcanic event is interfingered in the shelter sequence and gives support to this hypothesis. A sequence of at least four narrowly spaced hearths testifies to several cyclic phases of short occupation of the site; after this rather short period the shelter was abandoned. The whole operational sequence from chert extraction to tool manufacturing is well documented in the site; the most outstanding features connected to this activity are a flint workshop layer and a quarry-debris heap that partially covers it. The chert nodules were extracted from the bottom wall of the shelter using limestone cobbles as hammers, flint picks and flint wedges; several of these tools were found in the workshop level, near the bottom wall of the shelter. The limestone debris was piled up in a wedge-shaped heap lying against the bottom wall and overlying an older section of the flint workshop layer. Then, the flint nodules were tested for knapping by removing a cap and a few flakes; unsuitable nodules were dumped in the debris heap while the good ones were flaked in the flint workshop area. The flint workshop layer is about 20 cm thick and 8-10 m2 wide; the distribution of the refittings shows that only minor postdepositional disturbance and reworking – and possibly also ancient trampling – acted on this level. Several thousands of flakes of chipped debris were found in this area and their space distribution can suggest some inferences about space use in the area; the most outstanding characteristic is that flakes and blades refitting on a core are usually concentrated in a small fan-like area near the core itself. Some peculiarities of the space distributions of flint debris type and size and of the tools within the workshop and in the other features suggest that various specialised activities took place in different areas of the flint quarry and workshop. Various types of chert raw material were used in addition to the local brown one: whitish-greyish chert cobbles were collected in the stream bed, a few tenths of metres far from the shelter; lower Palaeolithic flint debris, cores and tools with a typical amber-brown patina were collected in the nearby site of Valle Giumentina - 500-1000 m far – and re-flaked in the shelter workshop. A few other flint types, mainly supports of finished tools, testify to far off procurement or exchange areas, at least 40-50 km far to the North. The cores shaped on flint quarried in the shelter are usually small because of the relatively small size of the original nodules, but are mainly thoroughly exploited residues. Conversely, the cores fashioned on flint coming from the stream bed are rather rare, usually very large and their shape is often rectangular, as if they were “prepared” for some other purpose. All these data suggest that the S. Bartolomeo shelter was not frequented only for raw material procurement, even if some “specialised” quarrying and pre-shaping cannot be excluded. As shown by a complex pattern of other secondary sites in the same valley, it was probably a campsite selected for several reasons – not least the availability of chert – along a path followed by seasonal hunters during their movements.