General overview

1The experimental toolkit of bifacial pieces made in flint was created between December 2004 and June 2008 under the direction of É. Claud. It is comprised of 94 pieces made by three experimenters: S. Maury (78 pieces), M. Brenet (12 pieces), V. Mourre (2 pieces) and M. Lenoir (one piece) (figure 1).

2The principal aim was to create bifaces to be used in the acquisition and transformation of different materials under different functional constraints. The second objective was to provide the products and by-products of complete knapping sequences to respond to techno-economic research questions. and the third was to test the effect of resharpening bifaces on the formation of traces of use (table 1). 81 bifaces were used on their edges on organic materials, or in some cases transported in sacks, and the resulting use-wear patterns have been described in Part I, chapter 2.10. Of these 81 bifaces, two were unfortunately not recovered at the time this inventory was compiled: one of them was used on wood and the other for skinning, 13 of the bifaces inventoried were not used experimentally and were reserved for studies of the traces related to manufacture.

Table 1 - Recording of the manufacturing sequences for the 92 bifacial tools

Experimenters	Sequences of knapping not conserved	Sequences of knapping conserved	Sequences of resharpening conserved	Total
Michel Brenet		12		12
Michel Lenoir	1			1
Serge Maury	64	7	6	77
Vincent Mourre	2			2
Total	67	19	6	92

3Information pertaining to the hammers, the techniques and the gestures of percussion and abrasion, and the order in which they were employed during manufacture were systematically considered in 21 tests. For the 71 others, only the type of hammer – mineral or organic – was recorded.

Materials and methods

4The experimental assemblage was inventoried on September 29, 2015 and includes 81 bifaces, four bifacial points and seven bifacial side scrapers in flint from different sources (primarily: Santonian flint from Charente or “Grain de Mil”, Maastrichtian flint of the Bergeracois type, Senonian flint from Dordogne). The bifaces were mostly made on flakes (59 of 92), though some were made on blocks (table 2, figures 2^a-3^b).

Table 2 - Flint types and blanks used for biface manufacturing

	Type de silex
	Grain de mil	Bergeracois	Senonian	Sault	Pressigny	Undetermined	Total
Plaques and blocks		5	1				6
Non-cortical flakes	29	14		1		1	45
Cortical flakes	10	4					14
Cores	2						2
Undetermined	15	4	3	1	1	1	25
Total	56	27	4	2	1	2	92

5Two thirds of the pieces are asymmetrical in cross section: planoconvex or convex / planoconvex. The rest do not show any differential surface treatment and are more or less regular and biconvex in cross section. It is worth noting that 62 of the 64 asymmetrical pieces were made on flakes while five of the seven biconvex pieces were made on plaques (table 3, figures 3^a, 5^b).

Table 3 - Cross section of bifaces according to the blanks used

	Cross section
	Biconvex	Convex / planoconvex	Convex / plane	Total
Nodule or plaque	5		1	6
Non-cortical flake	7	18	20	45
Cortical flake	4	5	5	14
Core	1	1		2
Undetermined	11	10	4	25
Total	28	34	30	92

Typology and techno-typology of bifaces

6The experimental pieces were sorted into four primary techno-typological groups. based on structure (outline. convergence. angle. and pointedness) of their lateral edges and their point or their lateral edges and their transverse distal edges (table 4). These techno-types correspond to two primary techno-morpho-functional groups of archaeological bifaces known for the period concerned.

Table 4 - Techno-types et types de pièces bifaciales façonnées

	Biface techno-types
	TLCAP	TLCTDO	TLCAC	TLCTDT	TC	Total
Cordiform biface	17	3				20
Elongated cordiform biface	20	2				22
Cordiform biface with cortical butt	2					2
Pointed cordiform biface	1					1
Sub-cordiform biface	4	2	1			7
Triangular biface	4	1				5
Sub-triangular biface	8	1				9
Oval biface		1	2	2		5
Sub-oval biface	1			2		3
Lanceolate biface	3					3
Irregular biface	2	1				3
Irregular biface with cortical base			1			1
Bifacial Mousterian point	3					3
Atypical point			1			1
Bifacial convergent scraper	1					1
Bifacial oval scraper		2	3			5
Bifacial convex scraper					1	1
Total	65	14	8	4	1	92

7The most common techno-type is that of bifaces with convergent lateral cutting edges and a pointed apex (TLCAP, n=65) (figures 2^a-3^a). The pieces within this type display variation in terms of outline, cross section, morphology of lateral cutting-edges (rectilinear, convex, irregular or denticulate) and in term s of the cross section and angle of the apex. These have been distinguished from pieces with convergent lateral cutting edges and a convex non-pointed apex (TLCAC, n=8) (figures 2^b, 4^a).

8The non-pointed pieces with a distal edge that is continuation of the lateral edges were divided into two groups:

• bifaces with lateral convergent or convex cutting edges and an oblique distal cutting edges (TLCTDO, n=14) (figure 3^b);

• bifaces with lateral convergent or convex cutting edges and a transverse distal cutting edge (TLCTDT, n=4) (figure 4^b).

9The pieces were also compared to the classical bifacial types for the late Middle Palaeolithic (Bordes. 1988); cordiform bifaces were the most numerous (n=52) followed by triangular bifaces (n=14) and oval bifaces (n=8). It must be noted that nine of the pieces are not typologically bifaces but bifacial side scrapers or Mousterian points (table 4, figure 4^a); such is the case of the only piece with convex cutting edge (TC), which is typologically a bifacial convex side scraper.

Techno-types and cross sections of the pieces

10Amongst the largest group of pieces – that of bifaces with convergent and pointed cutting edges (TLCAP) – pieces that are asymmetrical in cross section are the most prevalent (49 of 65). This is also true for the bifaces with convergent cutting edges and convex apical cutting edges (TLCAC; 6 of 8). For the bifaces with oblique distal cutting edge (TLCTDO), there is an equal number of symmetrical and asymmetrical pieces (7 and 7), while three out of four pieces with transverse cutting edge (TLCTDT) are biconvex in cross section (table 5).

Table 5 - Biface cross sections according to their techno-types

	Volumetric form
	Biconvex	Convex / planoconvex	Convex / plane	Total
Lateral convergent cutting edges and pointed apices	16	30	19	65
Lateral convergent cutting edges and convex apices	2	2	4	8
Lateral convergent or convex cutting edges and oblique distal cutting edge	7	2	5	14
Lateral convergent or convex cutting edges and transverse distal cutting edge	3		1	4
Convex cutting edge			1	1
Total	28	34	30	92

Techno-types and use

11Ten bifaces and three bifacial side scrapers of the 92 experimental pieces were not used. The 79 pieces that were used were primarily employed in woodworking (n=34) and various cutting activities in butchery (n=20), while eleven pieces were used in the transformation of animal materials (hide, bone, and antler), three were used to cut herbaceous plants, and four were used to dig in a clay-sand soil; finally, two bifacial points were used experimentally as hunting projectiles, and five bifaces (with or without wrapping) were transported in a sack. The numbers presented here are slightly different than those presented in Part I, chapter 2.10 because they are based on the pieces as a whole, while the numbers in chapter 2.10 are based on active zones, of which some pieces have more than one.

1261 of the 65 pointed bifaces (TLCAP) were used, mostly in woodworking (28) and butchery (14) (table 6).

Table 6 - Practised activities according to the biface techno-types

	Simplified techno-type
	TLCAP	TLCTAC	TLCDO	TLCTDT	TC	Total
Woodworking	28	2	2	2		34
Butchery	14	1	3	2		20
Hide working	6				1	7
Bone working	3					3
Antler working	1					1
Cutting herbaceous plants	2	1				3
Projectile use	2					2
Digging soil	2	1	1			4
Transport	3		2			5
Not used	4	3	6			13
Total	65	8	14	4	1	92

Hafting of the pieces

13Over the course of their use, only 14 pieces were used with a wooden handle (n=7) or a wrap made of strips of hide or plant fibers (n=7), the 65 others were held in hand (table 7, figure 5). It must be noted that nine of the handles or wraps were applied to bifaces that had been used to strike wood.

Table 7 - Prehension mode of bifaces during use, according to activities

	In a wrap	Hafted	Bare hand	Total
Woodworking	5	4	25	34
Butchery		1	19	20
Hide working			7	7
Bone working	1		2	3
Antler working			1	1
Cutting of herbaceous plants	1		2	3
Projectile use		2	0	2
Digging soil			4	4
Transport			5	5
Total	7	7	65	79

Duration of use of the pieces

14With regard to the duration of use of the pieces, in the case of the two most treated materials – wood and meat – the bifaces were used for an average of 46 minutes each in butchery, compared to an average of 37 minutes per tool in woodworking (table 8). Use-time varied for the other materials worked (table 8):

• cutting herbaceous plants, average of 47 minutes;

• hide working, average of 33 minutes;

• antler working, 30 minutes for a single biface;

• digging soil, an average of 26 minutes;

• bone working, an average of 13 minutes.

15These durations, which were generally adequate for the completion of the respective tasks (except in the case of the working osseous materials, to which the cutting edges of the bifaces proved ill‑suited) are very different from the durations of experimental transport and propulsion. For the pieces used as projectiles, the experiments terminated as soon as the point penetrated the animal, which was achieved with one shot for the first piece and with two shots for the second, the transport-times in the sack were around 6 h 20 min per piece (table 8).

Table 8 - Number of bifaces used for each activity, and experiment durations

	Number of pieces	Average duration of use (min.)	Cumulative durations of use (min)
Woodworking	34	3 685	1253
Butchery	20	4 568	868
Hide working	7	3 286	230
Bone working	3	1 333	40
Antler working	1	30	30
Cutting of herbaceous plants	3	4 667	140
Projectile use	2	15	3
Digging soil	4	2 625	105
Transport	5	379	1895
Total	79		4564

Dimensions of the pieces

16The dimensions of the pieces are variable, ranging from 50 to 158 mm in length, from 38.5 to 101.9 mm for width, and between 9.8 mm and 39.8 mm in thickness. The majority of the pieces (69 of 92) fall between 80 and 120 mm in length. With regard to mass, this variable shows the greatest diversity, from the smallest pieces at 28 and 32 g, used as projectiles, to the bifaces weighing more than 350 g that were hafted for woodworking (table 9, figures 6‑7).

17The pieces with the greatest average volume are those used in butchery (104 mm in length and weighing 160 g on average) and in woodworking (99 mm in length and weighing 134 g on average). The two points used as projectiles were the smallest (62 mm and 29.8 g on average). The other pieces, used to work bone or hide, cut herbaceous plants, dig soil, or test the effects of transport are of average lengths ranging from 90.2 to 99.8 mm and of average mass ranging from 88 to 99.6 g (table 9). The sole piece used to work antler measured 104 mm in length and weighed 98 g.

Table 9 - Average dimensions and masses of bifaces according to the praticed activities

	Number of pieces	Average length (mm)	Average width (mm)	Average thickness (mm)	Average masses (g)
Woodworking	34	991	646	201	1 342
Butchery	20	1 037	699	228	1 595
Hide working	7	975	585	175	984
Bone working	3	902	575	182	883
Antler working	1	104	593	184	98
Cutting of herbaceous plants	3	911	57	182	918
Projectile use	2	617	458	104	298
Digging soil	4	916	604	161	973
Transport	5	998	601	177	996
Total	79

Angles of the cutting edges

18The average edge angle was calculated with a goniometre for four zones on each piece:

• on the basal edge (three measurements);

• on the two lateral cutting edges at the apical third of the piece (two to three measurements);

• at the point or distal cutting edge (two to three measurements).

19For all techno-types, the edge angles of the pieces decrease regularly from the bases to the cutting edges and then the apices.

20The pieces with convergent pointed cutting edges or convex apices (TLCAP and TLCAC) possess edge angles and cutting edge angles that are on average very narrow: 62° to 63° on the basal edge, 42° on the lateral cutting edges and 35° to 37° at the apices.

21The pieces with distal oblique cutting edges (TLCDO) have wider average edge angles than the preceding group on all edges: 68.4° on the basal edge. 46° to 46.4° on the lateral cutting edges. and 42.5° at on the apical cutting edge.

22Finally, the pieces with transverse distal cutting edges (TLCTDT) present even wider average edge angles on the bases and lateral edges – 79.5° to 53° – while the average angle of the transverse cutting edges was similar to that of the pointed pieces: 35.5° (table 10).

23Within each activity, the angles of the lateral and distal or apical cutting edges of the pieces are relatively varied. If we exclude the eleven experiments of projectiles, digging, and transport, the most acute lateral and distal or apical cutting edges (from 41.7° to 42.6° and from 34.9° to 35°) were preferentially used for butchery and hide working (table 11). It is worth noting that the average of 36.6° for the distal sections of the 34 pieces used on wood would not be taken into account because the distal cutting edges of 10 of these pieces were either too irregular to be measured or were fractured during use.

24Finally, we note that the experiments in working bone and antler were conducted with bifaces whose lateral and distal or apical cutting edges were among the widest, with averages between 43° and 46.3°, but the angles were measured after use. and it is possible that the measured angle was augmented by the scarring of the edges related to use on these particularly hard materials (table 11).

Table 10 - Average angles of bifaces cutting edges according to techno-types

	Code	Average angles
		base	left side	right side	distal part
Lateral convergent cutting edges and apical point	TLCAP	62	424	438	354
Lateral convergent cutting edges and convex apex	TLCAC	629	421	433	368
Lateral convergent cutting edges and oblique distal edge	TLCTDO	684	464	46	425
Lateral convergent cutting edges and transverse distal edge	TLCTDT	795	533	53	355

Table 11 - Average angles of bifaces cutting edges according to the practised activities

	Average angle
	left side	right side	lateral edges	distal part
Projectile use	395	40	398	28
Digging soil	393	415	404	36
Transport	404	412	408	37
Butchery	428	407	417	349
Hide working	447	404	426	35
Woodworking	44	476	458	366
Cutting of herbaceous plants	403	45	427	38
Antler working	42	44	43	43
Bone working	46	467	463	43