Fig 1.
a. Selection of archaeological quartz pebbles from FtJi5 site complex [17, 20]. This site complex is found in a conglomerate at the bottom of Member F and provides a unique abundance of unmodified quartz pebbles in archaeological context at the Shungura formation scale; b. experimental quartz pebbles from the Lower Omo Valley. 1 to 4: medium to low quality quartz pebbles due to internal cracks (1), coarse grain structure (2 and 3), or heterogeneous matrix (4), 5 to 8: good quality quartz pebbles. NB. Scale identical to a.
Fig 2.
a. Free-hand percussion technique performed by a novice knapper; b. Bipolar technique performed by a novice knapper.
Table 1.
Technological composition of the five archaeological assemblages from Member F (Shungura Formation) and experimental assemblage.
(NB. “Cores” includes whole and broken cores, “Pebbles” includes whole and broken pebbles).
Table 2.
Experimental data: Computation of useful variables for each pebble.
Fig 3.
Archaeological and experimental diagnostic bipolar flakes.
1 to 7: archaeological bipolar flakes from OMO 371N –A167; 8 to 12: experimental bipolar flakes. The black triangles indicate the impact points.
Table 3.
Numbers and proportions of diagnostic bipolar flakes in the archaeological and experimental assemblages.
(NB. only whole flakes and broken flakes preserving both their proximal and distal extremities are considered).
Fig 4.
a. Boxplots and stripcharts for the impact of each factor on the proportion of angular fragments; b. Regression tree for the proportion of angular fragments, with knapping technique, raw material quality and knapping expertise as covariates. (In each node, the upper value is the empirical proportion observed for a given combination of factors; the "%" value indicates the percentage of the total sample size in this node).
Fig 5.
a. Boxplots and stripcharts for the impact of each factor on the proportion of sharp-edged flakes; b. Regression tree for the proportion of sharp-edged flakes, with knapping technique, raw material quality and knapping expertise as covariates. (In each node, the upper value is the empirical proportion observed for a given combination of experimental conditions; the "%" value indicates the percentage of the total sample size in this node).
Fig 6.
a. Boxplots and stripcharts for the impact of each factor on the extension of cutting edge; b. Regression tree for the extension of cutting edges, with knapping technique, raw material quality and knapping expertise as covariates. (In each node, the upper value is the empirical proportion observed for a given combination of experimental conditions; the "%" value indicates the percentage of the total sample size in this node).