An Experimental Approach to Quantifying Reduction Intensity of Pre-European New Zealand Flaked Stone Adze Manufacture Using a Scar Density Index (SDI)

Abstract

This thesis uses experimental replication to examine the application of the Scar Density Index (SDI) to Aotearoa New Zealand Pre-European flaked stone adze manufacturing assemblages by comparing reduction intensity between two adze blank types. SDI is a standardised index for quantification of reduction intensity. Previously New Zealand adze manufacture has been quantified by stage models or flake typologies (e.g., Leach and Leach 1980; Turner 1992, 2000). However, application and comparability may be biased by interobserver subjectivity, discard fragmentation, raw material and/or form variation. Experimental core and biface studies indicate SDI has little variation across raw materials and manufacturing methods (Clarkson 2013; Shipton and Clarkson 2015). This suggests utility for standardised quantification of adze manufacturing assemblages. Blank types were compared because New Zealand adzes have a variety of “functional types” (Turner 2000), with manufacture influenced by raw materials (blank form, raw material properties) and functional requirements. Secondly, previous SDI research demonstrated potential issues with initial blank size or type, which are explored in the thesis: (1) scar erasure due to smaller initial size and surface area constriction; (2) differences in how close the blank form is initially to final form in bifaces resulted in differential mass loss, but no difference between blank types in the prediction of SDI by mass lost (Ditchfield 2016; Shipton and Clarkson 2015). Two greywacke adze blank types (Cobble (n = 6) or Split cobble (n = 9); N = 15), were reduced to the point of completion (n = 6) or discard (n = 7); two blanks were not completed. Completion was defined as suitability of hammerdressing (shaping with hammer percussion after flaking). As a percentage of the previous stage, measurements of blanks and detached flakes were taken in intervals of mass removal (“stages”). SDI and percentage mass lost were highly correlated (R² = 0.894) (log-transformed), supporting the use of SDI as a reduction intensity index. There was no difference between blank types. Blanks with lower initial mass were generally completed sooner than blanks with higher initial mass. During early stages, blanks with high initial mass had a larger proportion of high mass flakes than small initial mass blanks. Flake size across all blank sizes decreased as reduction intensity increased. An indicator of scar erasure, the Flake Recovery Rate (FRR) (Braun et al. 2005), decreased as the reduction stage increased – indicating less flakes removed than visible scars i.e., there was no scar erasure in later stages; this may be due to decreases in flake size, increases in edge exploitation or because intervals of mass removal decreased as reduction increased.