Minuscule flakes made from recycled flint were identified at the late Lower Palaeolithic site of Qesem Cave in Israel (Figure 1), dated to 400–200 thousand years ago (kya) (Barkai et al. 2003; Gopher et al. In press). Our ongoing research at this exceptionally well-preserved site indicates that it was repeatedly occupied by early hominins, ancestral to Homo sapiens and/or Neanderthals (Hershkovitz et al. In press), who left ample evidence of their lifestyle. Our analysis of the tiny flakes (Figure 2) suggests that they were used to cut meat.
The occupants of Qesem Cave produced innovative flint tools, and in particular sharp flint blade knives, using cutting-edge technology of the time (Barkai et al. 2009). They hunted cooperatively, bringing body-parts of fallow deer back to the cave, which were then butchered, shared (Stiner et al. 2009), and — as suggested by fire usage throughout the cave's 7.5m-deep stratigraphy — eventually barbecued. Microscopic use-wear analysis of flint artefacts from Qesem Cave, complemented by experimental replication work, revealed that a diversified assemblage of flint blades was manufactured and used. Thick-edged blades, shaped through retouch, were used for scraping semi-hard materials such as wood or hide, whereas blades with straight, sharp working edges were used to cut soft tissues (Lemorini et al. 2006). Functional studies point to short-lived usage of these meat-cutting blades, which were hardly ever re-sharpened for further use. This behaviour was evidently sustained by an innovative systematic blade production technology, providing a constant supply of fresh cutting edges (Barkai et al. 2009).
Blades, however, were not the only meat-cutting tools used. We recently identified a specific category of minuscule flakes, typically removed from available parent-flakes (Figure 3). Parent-flakes were selected from already available flakes that were, most probably, produced originally for other purposes than as cores for the production of tiny flakes. In many cases parent-flakes were made on former tools such as scrapers and retouched flakes. This ongoing production, in a mode of behaviour reminiscent of recycling, attests to another significant aspect of hominin behaviour. Very small flakes have also been reported in Lower and Middle Palaeolithic sites worldwide (e.g. Goren-Inbar 1988; Dibble & McPherron 2006; Ashton 2007), but the ultimate aim of their production remained obscure owing to the lack of technological, functional and experimental studies. Here we summarise the results of a functional study of a sample of 86 tiny ‘recycled’ flakes from Qesem Cave and 192 parent-flakes from which such flakes were removed. This sample was collected according to techno-typological criteria and originates from all strata of the cave.
The tiny flakes were removed from the ventral face of the parent-flake ('core-on-flake') with little or no preparation. Consequently, they differ from ordinary flakes as they have two ventral, flat and smooth, faces. These minuscule flakes are not longer than 1–3 cm and, by convention, not thoroughly studied. At Qesem Cave all artefacts, regardless of size, were collected and studied, thus giving fresh insights into aspects of prehistoric human behaviour.
We produced a reference sample of minuscule double-ventral flake replicas and parent-flake replicas for use on different materials (Figures 4 & 5). The replicas were made from archaeological flint flakes collected from the surface of the cave and from damaged contexts. For functional analysis of the small double-ventral flakes and the parent-flakes we used a light-reflecting stereoscope (Nikon SMZ) at 3.5× to 35× magnification and a light-reflecting metallographic microscope (Nikon Eclipse) at 100×, 200× and 500× magnification.
Comparison of use-wear traces on the replicas to those identified on the authentic artefacts disclosed use traces on 34 (39.5 per cent) of the 86 minuscule authentic flakes but only 12 (6 per cent) of the 192 parents (traces at the specific location where the tiny flake was removed), confirming that the tiny flakes were the desired object. Examination of these minuscule flakes revealed that most were used to cut soft materials such as meat. Some showed traces of contact with fleshy tissues and bone, suggesting disarticulation or the separation of flesh or muscle tissues from bones. Experiments with replicas verified that the flakes were razor-like implements, sharp enough for users to easily cut muscles, tendons or skin, while their size restricted their efficiency in cutting deep muscles. Absence of hafting traces on the minuscule flakes indicated that they were probably hand-held (Figure 6). Our experiments suggested that these sharp flakes might facilitate butchering of small animals or tasks such as cutting of skin or sinew strips in larger animals.
Many of these minuscule implements had two lateral sharp edges and a typically dull (hinged) distal end, making them 'safe' for use near the mouth. Thus, they might well also have been used as hand-held knives while eating. Moreover, for instant and rapid actions requiring sharp implements they would have presented an excellent solution, easily achieved via a simple knapping technique based on recycling of old discarded flakes.
These findings shed new light on meat consumption behaviours of early hominins (also Stiner et al. 2009) and draw attention to a little-explored aspect of lithic studies, namely the recycling of old flakes as cores in the production of specific, sharp, tiny flakes for delicate and precise meat-cutting activities. Tiny recycled flakes were used as hand-held cutting tools as part of a diversified meat-processing Palaeolithic tool-kit.
Field work in Qesem Cave is supported by the CARE Archaeological Foundation, the Leakey Foundation and the Thyssen Foundation. Laboratory work and dating are supported by the Israel Science Foundation and the Wenner Gren Foundation. We thank R. Yerkes, S. Lev-Yadun and L. Maul for reading and improving this paper. We would like to thank K. Drechsel for skilfully producing Figure 1.
* Author for correspondence.