M.A. Theses: Donna M. Morrison, 1994

Validity in Lithic Debitage Analysis: An Experimental Assessment Comparing Quartzite to Obsidian

The purpose of this thesis is to assess the validity of three methods of analysis commonly used in the interpretation of lithic debitage. Obsidian and quartzite are used in replicative experiments to explore the nature of raw material variability and the effects that this factor can have on patterning in flaking debris. It is argued that each type of raw material will produce its own unique debitage profile and that methods designed and tested with one lithic material will therefore be unable to accurately identify the technological origins of the debitage from another material type.

The methodology employed includes research into external sources of information such as petrology and fracture mechanics, as well as internal knowledge gained from archaeological experimentation. Replications of various core and biface reduction activities are carried out so that the waste material from these knapping events can be analysed according to the criteria set forth by the three methods under evaluation: 1. "Individual Flake Attribute" analysis based on Magne's flake scar method, 2. Prentiss's modified version of the Sullivan and Rozen flake completeness typology, and 3. Ahler's "Mass Analysis". In order to determine whether these analytical approaches are valid, specific hypotheses are tested regarding their ability to accurately identify the different reduction stages and techniques that produced the experimental flake assemblages.

The results of the analysis support the initial claim that each lithic material will fracture in a unique manner (depending on petrological characteristics and the knapping strategy employed), thereby creating its own individual patterns of debitage. It is shown that these differences between material types are quite pronounced, and that it is not possible to interpret quartzite debitage using methods tested on a developed from obsidian, chert, or basalt. That is, these methods (with the exception of the Mean Flake Weight variable used in Mass Analysis and the Platform Flake/Shatter ratio used in the Individual Flake Attribute analysis) are not valid externally, or, beyond the original parameters from which they were designed. Furthermore, when analysing obsidian flaking debris, it is shown that there is often enough variability even in such an "ideal" material to challenge the internal validity of these approaches, as well. Consequently, these analytical methods require a greater degree of testing before they can be applied to archaeological samples, since raw material variability is shown to have a greater effect on debitage patterning than previously thought.

It is concluded that it is not necessarily safe to assume that the technological origins of debitage assemblages can be consistently or accurately identified regardless of raw material type. This fundamental theoretical construct, which has been one of the basic assumptions in debitage analysis, needs to be more fully examined before we can produce valid inferences about prehistoric technology. Recommendations are provided regarding ways of overcoming this dilemma, and areas for future research are suggested.