Knapping tools in Magdalenian contexts: New evidence from Gough's Cave (Somerset, UK)

Abstract

Our knowledge of the recolonization of north-west Europe at the end of the Last Glacial Maximum depends to a large extent on finds from Gough's Cave (Somerset, UK). Ultra-high resolution radiocarbon determinations suggest that the cave was occupied seasonally by Magdalenian hunters for perhaps no more than two or three human generations, centred on 12,600 BP (~14,950-14,750 cal BP). They left behind a rich and diverse assemblage of Magdalenian lithic and osseous artefacts, butchered animal bones, and cannibalised human remains. The faunal assemblage from Gough's Cave is one of the most comprehensively studied from any Magdalenian site, yet new and unexpected discoveries continue to be made. Here, we record previously unrecognized flint-knapping tools that were identified during a survey of the Gough's Cave faunal collection at the Natural History Museum (London). We identified bones used as hammers and teeth manipulated as pressure-flakers to manufacture flint tools. Most of the pieces appear to be ad hoc (single-use?) tools, but a horse molar was almost certainly a curated object that was used over an extended period to work many stone tools. This paper explores how these knapping tools were used to support a more nuanced understanding of Magdalenian stone-tool manufacturing processes. Moreover, we provide a standard for identifying minimally-used knapping tools that will help to establish whether retouchers and other organic stone-working tools are as rare in the Magdalenian archaeological record as current studies suggest.

Fig 1

A. Location of Gough’s Cave. B. Plan of the outer part of Gough’s Cave showing where the knapping tools were recovered. Five of the eight knapping tools from Parry’s excavation were found in November 1927, and the other specimens recorded for that year, as well as the phalanx from the 1987 excavation (Area I), were found in the same area of the ‘entrance’ (bounded by the red oval). Plan based on [–15].

Fig 2

Photographs (A) and drawings (B) of horse molar M50064, showing knapping damage (pink) and scratches (green) in labial, mesial, lingual, and distal views (Scale = 50 mm). SEM micrographs, Alicona images and macro-photograph detailing: (C-D) chipped enamel along the mesostyle; (E) transverse parallel abrasions associated with flaked enamel; (F) macro-photograph of the working area on the mesial surface, showing minute punctiform features, transverse micro-abrasion and a large depression where the focus of pressure-flaking has abraded the enamel to form a bowl-shaped depression. This area has many pits and scores with two dominant alignments at right angles to each other; (G) detail of transverse micro-striations associated with the areas of concentrated knapping features on the mesial surface; (H) SEM image and (I) EDX element plot highlighting fragments of flint embedded in the enamel. Red = silicon (flint and sediment), blue = calcium (enamel); (J) example of random striations from contact with a stone tool. Although these marks occur over much of the surface of the tooth, they are more easily discernible on approximal wear facet; (K-L) gouge with flaked enamel and internal parallel striations; (M-N) indentations that appear to be compression features or shallow abrasions.

Fig 3

Photographs (A) and drawings (B) of horse incisor M49811, showing knapping damage (pink) in mesial, labial, distal, and lingual views (Scale = 10 mm). SEM micrographs, Alicona images and macro-photograph detailing: (C) close-up of knapping marks near the crown-root junction on the buccal surface; (D-E) isolated pit; (F-G) score with associated tool-edge scratches; (H-I) tool-edge scratches and micro-pits. Pits and scores are concentrated near the crown-root junction, whereas vertically-aligned tool-edge scratches are more prominent and extensive on the surface of the crown.

Fig 4

Photographs (A) and drawings (B) of horse distal metacarpal M49934, showing knapping damage (pink), cut marks (blue), and post-excavation flaking (light grey) in dorsal, medial, palmar, and lateral views (Scale = 50 mm). SEM micrographs and Alicona images detailing: (C-E) impact area associated with ancient transverse breaks and flaking with enlargements of the same; (F-H) cluster of large knapping pits and SEM enlargements showing internal features (partly obscured by cave-earth). Cracks are post-excavation features, possibly following lines of weakness created during knapping.

Fig 5

Photographs (A) and drawings (B) of horse metacarpal M50024, showing knapping damage (pink), cut marks (blue), sample hole (dark grey) and post-excavation flaking (light grey) in dorsal, lateral, palmar, and medial views (Scale = 50 mm). SEM micrographs and Alicona images detailing: (C-D) remnant of displaced bone adjacent to the pit; (E-H) scores; (I-J) knapping pit with an irregular profile from the chipping-away of bone matrix.

Fig 6

Photographs (A) and drawings (B) of horse distal metacarpal M50000, showing knapping marks (pink) and cut marks (blue) in dorsal, medial, palmar, and lateral views (Scale = 50 mm). SEM micrographs and Alicona images detailing: (C) close-up of impact (white arrow) at the apex of the shaft on the dorsal face and incipient crack that extends distally along the shaft; (D) close-up of knapping damage with areas of isolated and overlapping scores and pits figured in images E-K.

Fig 7

Photographs (A) and drawings (B) of horse metatarsal M49873, showing knapping marks (pink) and rectangular sample hole (dark grey) in dorsal, medial, plantar, and lateral views (Scale = 50 mm). Alicona images detailing wedge-shaped gouges (C-E) with concentric cracks demarcating depressed cortical bone (C), and shallower superficial scores (F).

Fig 8

Photographs (A) and drawings (B) of horse metatarsal M50025, showing knapping marks (pink), cut marks (blue), carnivore chewing marks (brown) and sample hole filled with plaster of Paris (dark grey) in dorsal, medial, plantar. and lateral views (Scale = 50 mm). Cut marks (white arrows) and carnivore chewing marks clustering at the distal end (C-D) include furrows (black arrows) and pits (grey arrows). The gouged morphology of the knapping marks (E-H) distinguishing them from the compressed features of the carnivore tooth pits. Note the superficial root-etched channels in E and G (open arrows); these indicate the bone was buried in the daylight zone of the cave.

Fig 9

Photographs (A) and drawings (B) of red deer distal metatarsal M49847, showing knapping marks (pink), cut marks (blue) and samples (drilled hole = dark grey, cut piece = red oval) in dorsal, medial, plantar, and lateral views (Scale = 50 mm). SEM micrographs and Alicona images detail a point of impact with associated flaking (C), superficial slicing cut marks made with the tool held at an oblique angle to the bone surface (D), and knapping marks on the plantar and lateral faces (E-G).

Fig 10

Photographs (A) and drawings (B) of horse phalanx I Unreg. 3482, showing knapping marks (pink), cut marks (blue) and concreted sediment (brown) in dorsal, medial, plantar, and lateral views (Scale = 50 mm). Close-up of knapping area (C) and Alicona (D) and SEM micrographs (E-G) of knapping pits and scores on the dorsal surface and cut marks on the side of the bone (H).

Fig 11. Schematic outlines of postcranial bones from Gough’s Cave used as knapping tools.

Red indicates knapping damage; arrows mark the location of impact features on the side of the bone. Note the consistent pattern of breakage of the metapodial shafts close to the distal articular end.

Fig 12

Plans of the outer part of Gough’s Cave showing stages in the removal of the Pleistocene sediments: A 1892–1903; B, 1927–1931; C, 1934–1952; D, 1986–1992 with dashed line indicating the grill gates that separate the Entrance Chamber from the Vestibule (plans based on [–15]). Area with knapping tools bounded by the red oval.

Fig 13. Gough’s Cave: Vertical distribution of flint and chert artefacts (based on [13]), Pleistocene horse and red deer bones [41, Table 1] and knapping tools recovered during excavations supervised by Parry in 1927–1931 (Fig 1).

The deposits were excavated in six-inch (~15 cm) layers (or spits). Note that pottery and Holocene bones are confined to spits 1–9 and the horizontal spits cutting through the natural slope of the deposits gives the illusion of stratified sequence [13, 41].

Fig 14

A-D. Magdalenian techniques of knapping flint tools at Gough’s Cave. A. Horse metapodial (e.g. M50025) used as a hammer to detach blades/bladelets from a core; B. horse phalanx (UNREG 3482) used as a hammer to remove small blades or flakes, or for retouching an edge; C. horse incisor (M49811) used as an active retoucher (sensu [85]) in pressure-flaking; D. horse molar (M50064) used as a passive retoucher (sensu [85]) in pressure-flaking.

Fig 15. Integrating the lithic chaîne opératoire with the knapping tools at Gough’s Cave.

Fig 16. Location of Gough’s Cave and other Magdalenian sites with organic knapping tools in relation to the general distribution of Magdalenian sites (shown in green, after [–93] and [, Fig 8.3, p.439]).

Key: 1. Oelknitz; 2. Andernach; 3. Gönnersdorf; 4 Pincevent; 5. La Garenne; 6. Rochereil; 7. Roc-de-Marcamps; 8. Saint-Germain-de-la-Rivière; 9. Moulin-Neuf; 10. Laugerie-Haute; 11. Isturitz; 12. Laa 2 Cave; 13. Labastide; 14. Enlène; 15. La Vache. Map made with Natural Earth. Free vector and raster map data @ naturalearthdata.com.