Mechanical Basis of Percussion Flaking

Abstract The total variability in archaeological material derives from at least 4 different sources: technological, functional, stylistic, and random variation. Different attributes may be needed to quantify each aspect of the total variability, and the particular attributes that the archaeologist singles out for analysis will determine, to a large extent, the utility and validity of any subsequent typology. At present, there is relatively little theory to aid in the selection of the attributes most appropriate for a particular archaeological problem. In the analysis of prehistoric flake stone tools, the technology of flake production falls largely within the realm of the natural sciences; thus, technological variability is perhaps that aspect of the total variability for which suitable theory may be most readily developed. Flake production can be divided into 2 distinct processes, percussion flaking and pressure flaking, on the basis of the duration of the applied force. The fracture processes involved in these 2 types of flake production are sufficiently different to warrant a separate treatment of each. Attention is focused specifically on percussion flaking because this process appears to have been the predominant mode of flake production throughout most of human prehistory and, at present, it remains poorly understood. It is found that a percussion flake is detached from a core largely by a process known as spalling, the fracture produced by the reflection of a stress wave at the free surface of the core adjacent to the striking platform. The thickness of a flake is a function largely of 4 impact parameters and 2 constants of the material: the shape, intensity, and wavelength of the stress wave emanating from the point of impact, the angle of incidence of the stress wave on the free face of the core, the critical impact strength of the material in tension, and an elastic property of the material. Other attributes of flake size such as the length, width, area, weight, and volume are also influenced by these parameters. Several other factors affecting the size of a flake are discussed, and areas where further research is needed are outlined. It is concluded that the various attributes of flake size are highly and predictably interrelated.