Table 1.
Mass values of the bipolar, multiplatform, discoidal, biface, Levallois, prismatic blade, punch blade and pressure blade cores, waste and blanks from each reduction sequence.
Initial nodule masses in bold refer to reduction sequences conducted by the expert knapper.
Fig 1.
Demonstration of the method used to measure cutting edge length, showing a photograph of an original blank (a), and two stages in the process of reducing the photograph to a measurable polygon (b and c).
Note the platform is excluded in the polygon measurement so as to measure possible cutting edge only.
Table 2.
Mean mass, counts and cutting edge values for each reduction strategy.
Kruskal-Wallis tests were conducted for each variable based on the values of the three or seven repetitions of each reduction method. Variables containing significant differences among the eight different technologies at the α = 0.05 level are represented in bold. *The bipolar values were not included in the first five statistical comparisons, as significantly smaller cores were used owing to the typically small size of bipolar cores.
Fig 2.
Bar chart, with one standard error bars and each data point superimposed, showing the cutting edge per gram values for each repetition of bipolar (N = 3; μ = 10.36), multiplatform (N = 7; μ = 16.90), discoidal (N = 7; μ = 17.69), biface (N = 7; μ = 19.39), Levallois (N = 7; μ = 21.83), prismatic blade (N = 7; μ = 22.46), punch blade (N = 3; μ = 22.95) and pressure blade (N = 3; μ = 25.49) knapping.
Open circles represent the results from the intermediate knapper and closed circles represent the expert knapper.
Fig 3.
Bar chart, with one standard error bars and each data point superimposed, of the eight technologies grouped into their corresponding time periods, showing the Oldowan, consisting of bipolar, multiplatform and discoidal technologies (N = 17; μ = 16.07), the Middle Palaeolithic, consisting of biface, Levallois and prismatic blade technologies (N = 21; μ = 21.23), and the Upper Palaeolithic and onwards, consisting of punch blade and pressure blade (N = 6; μ = 24.22).
Open circles represent the results from the intermediate knapper and closed circles represent the expert knapper.
Fig 4.
Scatter plots with both axes transformed using the natural log (ln) examining the influence of mass, thickness, bulb thickness, length, width, elongation, platform depth, platform width and exterior platform angle (EPA) on the cutting edge length per gram of core for individual flakes.
The sample size of each scatter plot is 488, except for the platform depth, platform width and EPA scatter plots, which had sample sizes of 460 owing to the presence of some crushed platforms.
Fig 5.
Boxplots of qualitative variables.
Platform type plot (a) compares dihedral (N = 178; Mdn = 25.99), plain (N = 182; Mdn = 30.52) and focalised (N = 100; Mdn = 65.81) platforms. Termination type plot (b) compares plunging (N = 22; Mdn = 17.45), step or hinge (N = 58; Mdn = 23.02) and feather (N = 408; Mdn = 36.45) terminations. Platform preparation plot (c) compares no preparation (N = 105; Mdn = 19.49), either overhang or faceting (N = 202; Mdn = 26.44) and both overhang and faceting (N = 181; Mdn = 44.67).
Table 3.
Number of blanks, mass of waste and cutting edge per gram values between the expert knapper (N = 2 for all five technologies) and intermediate knapper (N = 5 for all five technologies) for each technology.