Stone Bead Technologies and Early Craft Specialization: Insights from Two Neolithic Sites in Eastern Jordan Katherine I. Wright, Pat Critchley and Andrew Garrard With contributions by Douglas Baird, Roseleen Bains and Simon Groom What social groups were involved in Neolithic craft production? What was the nature of early forms of craft specialization, long before urban economies evolved? One way to look at this is to investigate manufacture of Neolithic prestige goods. Seasonal camps in Wadi Jilat (eastern Jordan) revealed unusually detailed evidence for manufacture of stone beads: debris, blanks, finished beads, and tools for drilling, sawing and abrasion. The material is ‘Dabba Marble’, a metamorphic rock of which the major source is nearby. This article describes lapidary technology at Jilat 13 and Jilat 25, equivalent in age to the Pre-Pottery Neolithic C (PPNC). Mineral-chemical characterization data on Dabba Marble are presented. These sites raise issues about early craft specialization. These beadmakers seem to have been master craftsmen/women. We suggest that these sites illustrate a particular form of ‘site specialization’, namely sites located in remote territories and focused on special materials and intensive production of prestige goods. However, these craft activities were also embedded in hunting, herding and, perhaps, ritual, as suggested by figurines and pillars. Keywords: technology, specialization, Neolithic, stone beads, Levant, Jordan Introduction The Neolithic in the Near East involved a technological revolution which included expansions in the use of stone, clay and plaster. The social significance of this expansion is still not understood. One example is stone bead-making, which began to expand in the Late Epipalaeolithic (cf. Natufian) and expanded much further in the Neolithic. From the perspective of use, personal ornaments have much potential for questions about social identity and the body as a medium of enculturation (Bourdieu 1977, 94). Dress expresses and reproduces social identities such as gender, age and group affiliation (Barnes and Eicher 1992; Eicher 1995; Meskell 2001; Sciama and Eicher 1998; Sorenson 1997; Treherne 1995). However, from the perspective of manufacture, key technical questions are still unanswered. Detailed data on Neolithic stone bead-making may permit us to identify individual artisans, skills, choices, chaines operatoires (Roux et al. 1995; Vidale et al. 1992), and help us to understand craft specialization better in its earlier stages, as opposed to the later version seen in urban societies (Kenoyer 1992a; Roux and Matarasso 1999; Vidale 1989). One goal of this paper is to present the lapidary techniques revealed in two Jordanian manufacturing sites of the Pre-Pottery Neolithic C / early Late Neolithic (PPNC/ELN). A second goal is to consider these data in terms of early forms of craft specialization. We consider issues of individual specialists, specialist households and regional specialist sites. Finally, we consider how these data relate to other information on Neolithic stone beadmaking in the southern Levant. Stone Beads and the Near Eastern Neolithic Katherine I. Wright (corresponding author), Pat Critchley, Andrew Garrard, Roseleen Bains and Simon Groom, Institute of Archaeology, University College London, 31–34 Gordon Square, London WC1H 0PY, UK; email: [email protected]. Douglas Baird, School of Archaeology, Classics and Egyptology, University of Liverpool, Abercromby Square, Liverpool, UK ß Council for British Research in the Levant 2008 Published by Maney DOI 10.1179/175638008X348016 In the Near East, stone beads are exceedingly rare in pre-Natufian sites. Even in the Natufian, ornaments made of shell, bone and teeth greatly outnumber Levant 2008 VOL 40 NO 2 131 Wright et al. Stone Bead Technologies Figure 1 Map of the southern Levant showing sites and raw material sources mentioned in the text stone beads, and stone beads are known only from selected sites (e.g., Bar-Yosef-Mayer 1991; Larson 1978; Maréchal 1991; Moore 2000; Reese 1991; Valla et al. 2004). In the Pre-Pottery Neolithic A and B, stone ornaments expand widely in numbers and forms (e.g., Critchley 2007; Gopher 1997; Talbot 1983; Wheeler 1983). From the PPNC onward, there are hints of expanding trade networks in stone beads (e.g., Wright and Garrard 2003; cf. Bar-Yosef Mayer et al. 2004; Diamanti 2003; Dubin 1995; Hamilton 2005; Jackson 2005; Wright 2008; Wright 2006). Neolithic stone beads are often found in contexts of use or discard (houses, middens, graves). Of manufacturing areas and production techniques, we have mostly brief reports (Berna 1995; Fabiano et al. 2004; Finlayson and Betts 1990; Garfinkel 1987; Gorelick and Gwinnett 1990; Hauptmann 2004; Jensen 2004; Kaliszan et al. 2002; Rollefson 2002; Rollefson and Parker 2002). We know more about other regions (Barthelmy de Saizieu and Bouquillon 1994) or later periods (e.g., Bar-Yosef Mayer et al. 2004; Calley 1989). How were Neolithic stone beads made? What variations of material and technique do we see? Substantial evidence on one tradition comes from sites of the Jilat-Azraq project, eastern Jordan (Fig. 1). A preliminary description of bead assemblages and typology has appeared (Wright and 132 Levant 2008 VOL 40 NO 2 Garrard 2003). Here we present details of manufacturing technology from two production sites, Jilat 13 and Jilat 25 (Fig. 2), equivalent in age to the PPNC or ELN (c. 6950–6400 cal BC). The Sites Wadi Jilat lies in limestone steppe 30–40 km east of the present-day margins of the Levantine Corridor, where rain-fed cultivation is possible and where large Neolithic villages emerged (Fig. 1). Azraq Oasis is 50 km north-east of Jilat, between limestone and basalt steppe-desert. In this work, 18 Palaeolithic and Neolithic sites were excavated (Baird et al. 1992; Garrard et al. 1986; 1987; 1994a; 1994b; 1996; Garrard 1998; Garrard in preparation; Garrard and Byrd 1992). Bone and shell beads were found at the majority of the 10 late Upper Palaeolithic and Epipalaeolithic sites investigated, but no stone beads were discovered — even at Azraq 18, which is a Natufian site. This may be an accident of sampling/discovery; as some Natufian sites have significant numbers of stone beads (D. Bar-Yosef Mayer, personal communication; Cauvin 1974; Maréchal 1991; Moore 2000; Valla et al. 2004). However, at the time of writing, no stone beads have been found at pre-PPNB sites in Wright et al. Stone Bead Technologies Figure 2 (a) Plan of Wadi Jilat 13 Late Phase. Note location of workslab in the centre-left area of the oval structure; (b) Plan of Wadi Jilat 25 Area A structure eastern Jordan (Richter et al. 2008; T. Richter, personal communication). All eight Neolithic sites revealed stone bead production, along with shell and bone beads. The sites date to the PPNB (e.g., Jilat 7, 26, 32; Azraq 31) and PPNC/ELN (Azraq 31, Jilat 13 and 25). The other two are ‘burin sites’ of either LPPNB or PPNC/ELN age (Jilat 23, 24) (Garrard et al. 1994b). All of these sites were seasonal camps of small groups engaged in varying combinations of hunting, trapping, foraging, cultivating, herding (Garrard et al. 1996; Martin 1999). They lived in shelters constructed of upright limestone slab foundations. The Neolithic sites yielded 10,547 artefacts of stone beads or related debris from secure contexts. About 88% of these came from Jilat 13 and Jilat 25 (Table 1). They include finished ornaments, unfinished roughouts and bead blanks, and debitage (Tables 2–5). Densities of beads, blanks and debris, Levant 2008 VOL 40 NO 2 133 Wright et al. Stone Bead Technologies Table 1 Summary of Dabba Marble artefacts at Jilat 13 and Jilat 25 (see also Wright and Garrard 2003, 279–80) Jilat 13 Jilat 13 Jilat 25 Jilat 25 Artefact Category Raw frequency (N) Weight (grams) Raw frequency (N) Weight (grams) Finished beads (secure contexts) Finished beads (surface/mixed contexts) Roughouts and blanks (secure contexts) Roughouts and blanks (surface/mixed contexts) Debitage (secure contexts) (5 all debitage from nodules to microflakes and shatter) TOTAL (secure contexts) TOTAL (secure contexts z beads and blanks from surface/mixed) 144 60 115 62 12 no data 4 no data 180 222 89 82 43 no data 4 no data 7369 6905 1381 976 7693 7748 7187 1585 1593 1120 and other data, indicate intensification of stone beadmaking in PPNC/ELN sites in the Jilat-Azraq region, relative to the PPNB (Wright and Garrard 2003). It is of interest that this coincides with the first appearance of domestic sheep-goat in the Jilat sites (see Garrard et al. 1996; Martin 1999). Stratigraphy, site formation, and chronology of Jilat 13 and 25 are discussed in more depth elsewhere (Garrard et al. 1994a; Wright and Garrard 2003). The sites were excavated as part of a regional programme and areas of excavation vary. At Jilat 25, surface artefacts extended across an area of 3200 sq m. From the surface, one oval upright-slab structure (7 x 4.5 m) was visible (Fig. 2b). About 50% of this was excavated, an area of 21 sq m and a total volume of 7.8 cu m. Three occupation phases were identified. The early phase included an occupation fill (Aa19) rich in primary refuse; this yielded a date of 8020 ¡ 80 uncal BP (OxA-2408). Above this, the middle phase involved addition of bins and hearths and accumulation of an ashy occupation fill (Aa15), also rich in primary refuse. A final occupation fill (Aa7) accumulated and the building was filled with rubble followed by a deposit of sand, which sealed the layers below (Table 6). In each phase, artefact clusters on occupational surfaces and within their fills were identified, via 1 x 1 m horizontal units (i, ii, etc.); we discuss these clusters selectively here. The chipped stone assemblage from Jilat 25 is characterized by debitage dominated by flakes, although both flake and blade cores were found. However, 84–85% of tools were made on blades. Prominent tools include Nizzanim points (the sole point type), burins, drill bits made from burin spalls, and other types (Baird 1993, 469, 521; Baird in Garrard et al. 1994a, 85, table 1). Further details on the chipped stone 134 Levant 2008 VOL 40 NO 2 assemblages are available in a number of works (Baird 1993; 1994; 1995; 2001a; 2001b). Surface artefacts at Jilat 13 extended across an 800 sq m area. One large oval upright-slab building (10 x 6.5 m) was visible at surface. Almost all of this was excavated, an area of 73.5 sq m (Fig. 2a). Three phases were identified. In the early phase, the excavated volume of deposit was 14.5 cu m; the structure was built, and primary occupation deposits accumulated, then a pavement was laid down in the western area and hearths were constructed in the south and east. Dates from this phase were 7920 ¡ 100 uncal BP (OxA-1800) and 7870 ¡ 100 uncal BP (OxA-1801). The middle phase exposed 8.7 cu m of deposit and yielded no C14 dates. An interior partition wall separated the western end of the building and pits and stone-lined hearths were added in the eastern area. The late phase, of which 17.4 cu m of deposit was excavated, yielded two dates of 7900 ¡ 80 uncal BP (OxA-2411) and 7830 ¡ 90 uncal BP (UB-3462). At this time a new upper pavement was added above a rubble foundation. In the western area a large workbench, with evidence of drilling, abrasion and flaking, is associated with this phase (Figs 2a, 14c). Primary refuse was found in all phases (selected contexts are shown in Table 7), but the stratigraphy is more complex than at Jilat 25. In the chipped stone assemblage at Jilat 13, each phase revealed a bladebased assemblage with blades or bladelets making up 57% of debitage. Most tools are made on blades (78–81% of all tools are blade based). Major tool types include projectile points, burins, piercers and drills, scapers and endscrapers, notches and denticulates, and bifacial tools such as tile knives. In the early phase, for which we have most detailed information, burins and points are the most TOTAL TOTAL TOTAL TOTAL - Beads - N Beads - % Blanks - N Blanks - % Disc beads Ring beads Oval beads Cylinder beads Barrel beads Irregular beads Indeterminate or fragment Pendants - triangular Pendants trapezoidal Pendants - oval Pendants rectangular Pendants - square Pendants - teardrop Pendants - other or indeterminate Bracelets 6 1 24 20?2 1 6 1 9 38 40?9 4 1 21 Blanks -N Beads -N 12 Green Dabba Marble Green Dabba Marble 70 58?8 3 67 Beads -N Red Dabba Marble Table 2 Beads and blanks: Jilat 25 (all contexts) 40 43?0 1 39 Blanks -N Red Dabba Marble 6 5?0 4 2 Beads -N Black Dabba Marble 10 10?8 10 Blanks -N Black Dabba Marble 14 11?8 7 7 Beads -N White Chalk 4 4?3 4 Blanks -N White Chalk 5 4?2 5 Beads -N White Quartzite 1 1?1 1 Blanks -N White Quartzite 0 0?0 Beads -N Other Blanks -N Other 119 100?0 7 1 1 1 12 95 2 Beads -N Total 100?0 5?9 0?8 0?8 0?8 10?1 79?8 1?7 Beads -% Total 93 100?0 4 1 6 7 75 Blanks -N Total 100?0 4?3 1?1 6?5 7?5 80?6 Blanks -% Total Wright et al. Levant 2008 Stone Bead Technologies VOL 40 NO 2 135 136 Levant 2008 VOL 40 NO 2 Disc beads Ring beads Oval beads Cylinder beads Barrel beads Irregular beads Indeterminate / fragment Pendants - triangular Pendants - trapezoidal Pendants - oval Pendants - rectangular Pendants - square Pendants - teardrop Pendants - other / indeterminate Bracelets TOTAL - Beads - N TOTAL - Beads - % TOTAL - Blanks - N TOTAL - Blanks - % 94 1 1 7 1 2 1 1 1 2 62 40.4 1 1 57 1 36 202 90.6 11 1 36 Blanks -N Beads -N 7 2 Green Dabba Marble Green Dabba Marble 68 43.6 3 13 40 12 Beads -N Red Dabba Marble Table 3 Beads and blanks: Jilat 13 (all contexts) 11 4.9 2 1 8 Blanks -N Red Dabba Marble 20 12.8 1 4 10 5 Beads -N Black Dabba Marble 4 1.8 1 1 2 Blanks -N Black Dabba Marble 1 0.6 1 Beads -N White Chalk 2 0.9 2 Blanks -N White Chalk 1 0.6 1 Beads -N White Quartzite 0 0.0 Blanks -N White Quartzite 3 1.9 1 2 Beads -N Other 4 1.8 1 2 1 Blanks -N Other 0 155 100.0 2 1 2 2 1 1 2 7 59 20 0 4 54 0 Beads -N Total 0.0 100.0 0.6 1.3 1.3 0.6 0.6 1.3 1.3 38.1 12.9 0.0 2.6 34.8 0.0 4.5 Beads -% Total 223 100.0 0 11 1 1 0 0 1 0 96 49 0 0 1 62 1 Blanks -N Total 100.0 0.0 0.4 0.4 0.0 0.0 0.4 0.0 4.9 22.0 0.0 0.0 0.4 27.8 0.4 43.0 Blanks -% Total Wright et al. Stone Bead Technologies Wright et al. Stone Bead Technologies Table 4 Finished disc beads: size data (complete or measurable beads) (na 5 not applicable) Diameter (mm) Height/Thickness (mm) Perforation Diameter (mm) Site N Mean SD Mean SD Mean SD Jilat 7 PPNB Jilat 26 PPNB Jilat 32 PPNB Azraq 31 PPNB Jilat 25 PPNC/ELN Jilat 13 PPNC/ELN Azraq 31 PPNC/ELN 13 0 0 2 66 47 34 8.7 na na 9.8 8.3 6.9 6.8 3.6 na na 4.6 1.6 2.4 1.6 3.0 na na 2.5 2.6 2.5 2.6 2.8 na na 1.4 0.8 0.9 1.1 2.8 na na 2.3 2.1 1.9 1.8 1.0 na na 0.4 0.5 0.6 0.4 frequently occurring types: burins constitute 29% of tools and points 16%. Piercers and drills occur in much lower numbers (4% of tools), as do other tool types. Of projectile points, Nizzanim points are the most frequent (39.7% of points), followed by Byblos (31.3%), Herzeliya (12%), Amuq (8.4%), Transverse (6%) and Haparsah (2.4%) points (Baird 1993, 469, 500–17, 625; Baird in Garrard et al. 1994a, 85, table 1). The 14C dates from these two sites have been recalibrated using IntCal 2004 and the date ranges at one standard deviation are as follows: Jilat 25 (early phase) context Aa19a (OxA2408) 5 9020–8760 cal BP; Jilat 13 (early phase) context A21a (OxA1800) 5 8980–8600 cal BP; Jilat 13 (early phase) context A15a (OxA1801) 5 8980–8550 cal BP; Jilat 13 (late phase) context C24 (OxA2411) 5 8980–8590 cal BP; Jilat 13 (late phase) context C22 (UB3462) 5 8780–8460 cal BP. Thus, stratigraphy and radiocarbon dates suggest that we are dealing with two sites of reasonably good temporal resolution, followed by abandonment and sealing of primary refuse deposits. Five C14 dates indicate occupation between about 7830 and 8020 uncal BP, with low standard deviations for each date. In radiocarbon terms, this is about as precise as it gets, and the two sites may overlap in time. However, projectile points do suggest we are dealing with a somewhat wider time span for the Jilat 13 sample than the Jilat 25 sample. Given different extents of excavation, comparisons of these and other stone bead-making sites entail challenges. Density data, however — as measured by numbers of beads, blanks and debris per cu m volume of excavated deposit — can be revealing. For example, the PPNB occupations in Wadi Jilat (3 sites, 11 occupations) had low densities — an average of 10.72 artefacts per cu m, and a maximum of 31.3 per cu m (Wright and Garrard 2003, table 2). In contrast, the density data for Jilat 25 are 222.3 per cu m (early phase), 331.2 per cu m (middle phase) and 56.2 per cu m (late phase). Density data for Jilat 13 are 310.2 stone bead artefacts per cu m (early phase), 161.5 per cu m (middle phase) and 109.9 per cu m (late phase). This suggests greater intensity of beadmaking in Jilat 13 and 25. Sources, Quarrying and Raw Materials Most Jilat beads were made of Dabba Marble, which occurs in green, pink/red, and black (Appendix A). This is our focus here. A few other materials were also used: other local sedimentary rocks and non-local turquoise (nearest source Sinai), malachite (nearest sources Faynan and Timna) and carnelian (nearest source unknown). Non-local stones formed only 0.15% of the materials (Wright and Garrard 2003). The largest known sources of Dabba Marble lie 15–25 km west of Jilat 13 and 25; some may be closer (Fig. 1 and Appendix A). These are bodies of limestones, chalks and cherts, lightly metamorphosed Table 5 Finished barrel beads: size data (complete or measurable beads) (na 5 not applicable) Diameter (mm) Height/Thickness (mm) Perforation Diameter (mm) Site N Mean SD Mean SD Mean SD Jilat 7 PPNB Jilat 26 PPNB Jilat 32 PPNB Azraq 31 PPNB Jilat 25 PPNC/ELN Jilat 13 PPNC/ELN Azraq 31 PPNC/ELN 1 2 0 1 6 28 13 4.5 7.5 na 11.5 6.3 6.8 10.9 na na na na 1.4 2.8 5.6 4.0 6.5 na 10.0 6.4 7.1 13.3 na na na na 1.7 3.9 9.1 1.5 2.5 0.0 4.0 2.1 2.2 2.9 na na na na 0.9 0.9 0.9 Levant 2008 VOL 40 NO 2 137 Wright et al. Stone Bead Technologies and injected with various minerals, e.g. apatites, red iron oxides (Fig. 3a). Outcrops show substantial variations in mineralization even over small areas; in Fig. 3b, the left side of the outcrop is soft green Dabba Marble; the right side is red Dabba Marble. The geology and mineralogy of Dabba Marble are presented in Appendices A–B. The Neolithic Jilat beads are consistent with this source (Appendix B). Methods of Analysis Comprehensive recovery of small beads, debitage and micro-artefacts was possible due to intensive, finescale sieving. All excavated contexts were sieved through a 5 mm mesh; many samples were dry sieved or wet sieved through a 1.5 mm mesh (the latter after flotation). Artefacts from floors were collected from 1 sq m horizontal grid units, to permit identification of activity areas. Fine-grained spatial data and microartefacts are important in understanding lithic technologies (Dunnell and Stein 1989; Cessford and Mitrovic 2005). This is borne out by the Jilat data, since micro-flakes are one byproduct of stone bead retouch. In cases of intense housecleaning, microartefacts may reveal bead-making where macroartefacts do not (Wright and Bains 2007). For each context, artefacts were separated by raw material and classified into major groups: nodules and debitage (Fig. 4), roughouts (Fig. 5), unfinished blanks, and finished ornaments (Figs 6–10). We counted and weighed each group to determine relationships between debitage and finished beads. Measurements (diameter, height, perforation diameter) were taken on beads and blanks, to assess standardization and drilling techniques. Finished ornaments were classified into 8 basic types (Wright and Garrard 2003). Circular disc beads are the most numerous, smallest, and most standardized (Figs 6d, 7e–f). They occur in the widest range of materials; most red Dabba Marble beads were discs. Barrelshaped beads are larger, more variable and mostly made of green Dabba Marble (Fig. 9e–f). Pendants are the largest, rarest and most diverse items, shaped as triangles, rectangles and ovals; most are of green Dabba Marble (Fig. 10d, f). Bracelets were made of white chalk (Tables 2–3). Unfinished beads (blanks) were classified according to the same typology as finished beads, when the intended final product could be ascertained (e.g., disc blank) (Tables 2–3; Figs 6a–c, 7a–d, 8, 9a–d, 10c, e). Within these categories, blanks were further classified according to traces of flaking, grinding, perforation. Figure 8 shows these stages for 3 sequences of disc bead manufacture. Differences between Sequences A, B and C are differences in the original blank (Stage 1: thin flake, thick flake, tabular roughout) and presence or absence of flaking retouch on edges (Stage 2). Further analyses of sequences for these and other bead types are still in progress. Debitage was sorted into nodules, cores, roughouts, flakes, angular shatter, micro-flakes and Table 6 Stone beads, blanks and debris, Jilat 25: all contexts 138 Phase Context Description Beads Blanks Debris Comments / Selected associated artefacts Early Early Early Aa24 Aa27 Aa19 Fill of bedrock cut Fill of bedrock cut First occupation fill 0 0 29 0 0 23 2 1 384 Early Early Early Early Middle Aa20 Aa21 Aa18 A 44 Aa15 First occupation fill First occupation fill Early fill in entrance area Early fill in entrance area Second occupation fill 1 4 1 0 47 5 1 5 0 31 6 4 8 3 544 Middle Middle Middle Middle Middle Aa16 Aa13 Aa10 Aa11 Aa7 Hearth Fill of stone feature Fill of stone feature Ash lense Third occupation fill 1 1 0 3 17 1 0 0 0 15 0 81 3 17 109 Middle Aa8 Middle Aa14 Middle A 42 Third occupation fill Fill in entrance area Fill in entrance area 1 7 0 1 3 0 1 4 1 Late Late Late Late Late Sandy late fill Rubbly late fill Late fill in entrance area Late fill in entrance area Fill of late intrusive feature 1 7 0 0 1 0 4 0 0 0 55 72 1 6 0 Levant Aa 2 A6 A 33 A 36 A 34 2008 VOL 40 NO 2 1 cutmarked slab, 3 sandstone fragments, probably handstones; 1 limestone handstone 1 limestone capstone; 1 grooved stone (basalt); 1 sandstone abrader 2 grooved stones (limestone and basalt); 2 basalt handstones; 1 sandstone fragment 1 sandstone abrader. Bead: 1 RDM disc. Debris: all GDM Wright et al. micro-shatter (Fig. 4). Definitions of these categories are broadly similar to those established in chipped stone analysis (Andrefsky 1998). Flaking and Initial Reduction: Nodules, Cores, Debitage and Roughouts The soft limestone and hard chert in Dabba Marble permits it to be worked via chipping, flaking, grinding, sawing and drilling. To varying degrees, the material has conchoidal fracture. Where chert content is high, conchoidal fracture is excellent. Limestone itself also has conchoidal fracture, particularly when fine-grained, as Dabba Marble is. The tabular structure of the laminated limestones also makes it possible to create flat faces easily. The difficulty of shaping beads would have varied, depending on specific material. Most green Dabba Stone Bead Technologies marble is fairly homogeneous, composed of calcite-rich soft limestone (Mohs 5 3) and apatite (Mohs 5 5). Red Dabba marble occurs in a soft pale pink variety (Mohs 5 3–4); a dark pink variety of medium hardness; and a dark red siliceous variety, essentially red chert (Mohs 5 7). This red chert variant (Mohs 7) will have been more difficult to modify. Flaking and chipping were particularly important in working this material, and abrasion will have been more difficult. This may be why so many beads of the red, cherty Dabba Marble were disc beads made on flakes (Fig. 7). Flaking figured prominently in the making of beads from softer materials. However, sawing and abrasion played a greater role in modification of these materials. Comparable variations in technique, depending on material hardness, are seen at other prehistoric sites (Gorelick and Gwinnett 1990). Table 7 Stone beads, blanks and debris, Jilat 13: selected contexts Phase Context Description Beads Blanks Debris Comments / Selected associated artefacts Early Early Early Early C106 B77 B79 B71 Fill of bedrock cut First occupation fill Fill of hearth Second occupation fill 0 1 0 10 0 5 0 11 1 752 6 750 1 circular limestone disk Early Early Early Early A17 A15 A16 A18 Lower pavement Third occupation fill Third occupation fill Third occupation fill 0 1 2 0 0 8 1 0 0 8 17 11 Early Early Early Early B44 B45 B69 C56 Third Third Third Third 5 0 6 8 9 2 5 5 74 139 383 313 Middle Middle A3 A5 Fourth occupation fill Fourth occupation fill 3 7 4 9 60 139 Middle Middle Middle Middle Middle B31 B38 B54 B56 C39 Fourth Fourth Fourth Fourth Fourth 4 12 1 1 10 2 6 0 1 12 33 520 9 14 254 1 limestone figurine (bird?) 1 basalt grinding slab fragment 1 flint cutmarked slab; flaked limestone ’scraper’ Late C14 0 5 65 1 basalt ground fragment Late Late Late B3 C4 A2 Foundation for upper pavement Upper pavement Upper pavement Fifth occupation fill 0 0 4 0 0 2 0 0 154 Late B4 Fifth occupation fill 0 2 7 Late B7 Fifth occupation fill 2 5 150 Late Late Late B9 B21 C3 Fifth occupation fill Fifth occupation fill Fifth occupation fill 0 5 2 0 7 4 3 193 61 1 cutmarked limestone slab; 1 post-socket 1 drilling and abrading bench (Fig? 14c) 10 figurines or figurine preforms suggesting animals, phalluses, arrows (flint, limestone, travertine). 1 pebble mortar or capstone with ochre (limestone); 1 perforated object, 1 handstone on flake (flint); 2 vessel fragments, 1 ochred pebble, burnt pebble (limestone); 1 ground indeterminate stone (basalt). Beads are 2 barrels and 2 pendants; blanks are 1 barrel and 1 indeterminate 1 grooved limestone abrader (Fig 15c) Blanks are both pendant blanks 1 cutmarked limestone; 1 basalt handstone/pestle fragment 1 perforated pumice abrader 1 basalt handstone; 2 basalt fragments occupation occupation occupation occupation fill fill fill fill occupation occupation occupation occupation occupation fill fill fill fill fill 1 limestone handstone 1 drilled and grooved limestone object; 1 flaked limestone ’scraper’; 6 basalt fragments 1 cutmarked slab 1 worked limestone object; snake-shaped flint pebble 1 miniature pestle, basalt Basalt fragments: 3 from vessels, 1 handstone/pestle fragment, 6 unidentifiable 1 basalt handstone fragment No ground stone 1 double-grooved sandstone abrader (Fig? 15a) 1 limestone capstone (? 2 surfaces); 5 basalt fragments 1 cutmarked flint slab; 1 grooved limestone object; ochred flints; figurines; pillars Levant 2008 VOL 40 NO 2 139 Wright et al. Stone Bead Technologies Figure 3 (a) View of the modern Dabba Marble quarry, west of Wadi Jilat, from which raw material samples were obtained (see Appendices A–B); (b) Close-up view of metamorphosed deposits in situ, with interdigitating green (left) and red (right) Dabba Marble Five main stages of manufacture were identified, although reduction sequences vary. These are (1) reduction of raw nodules to cores, roughouts and flakes, via flaking and chipping; (2) shaping of roughouts and flakes into blanks, via further flaking, chipping, sawing and rough grinding; (3) perforation, via boring and/or drilling; (4) further grinding to produce the final shape; (5) final polishing. Nodules, cores and debitage were recovered in large amounts at Jilat 13 (7369 artefacts, 6905 grams) and Jilat 25 (1381 artefacts, 976 grams). The largest unworked blocks are of green Dabba Marble. They are about 15 cm in diameter, but most are about the size of an adult human fist. Some nodules were weathered, suggesting that they were picked up from surface rather than quarried from bedrock layers (Fig. 4, top). Reduction of nodules by flaking resulted in (1) cores, (2) tabular roughouts, (3) large flakes, (4) large angular shatter fragments, (5) micro-flakes, and (6) micro-shatter (Fig. 4). Cores, defined as nodules with two or more flake scars, are not numerous or consistent in form. Shatter, micro-flakes and microshatter were normally discarded as byproducts. 140 Levant 2008 VOL 40 NO 2 Tabular roughouts and larger flakes were used as the basis for further reduction into bead blanks. Roughouts are early stages in bead reduction (Kenoyer 2003, 16). At Jilat 13 and 25, roughouts are tabular, reflecting the bedded structure of the material. They were flaked and chipped around the edges into roughly symmetrical shapes (Figs 5a, 10a). Roughouts were sometimes subjected to initial drilling or sawing at an early stage, prior to any intense abrasion (Fig. 5b). In other cases, roughouts were abraded first (Fig. 5d) and then sawn or drilled (Fig. 5c). Tabular roughouts were the basis for making larger, thicker ornaments, such as most pendants and barrel beads (Figs 9–10). Small subcircular flakes were the basis of most disc beads (Figs 6–7). The flakes have platforms, bulbs of percussion on ventral surfaces, and often scars from previous removals on dorsal surfaces. A certain consistency in size, shape and morphology suggests that a prepared-core technology was used to predict and produce these flakes. Since cores are rare, we do not yet know precisely how this was achieved. Experiments are still needed, but we suspect that chipping and flaking was accomplished by varying Wright et al. Stone Bead Technologies Clear indications of heat treatment were rare, but burnt pieces do occur. Sawing, Drilling and Abrasion: Bead Blanks and Finished Beads Figure 4 Example of raw material nodule (top), angular shatter (upper centre), flakes (lower centre) and micro-debitage (lower row) from a single context at Jilat 13 uses of indirect percussion, soft-hammer percussion and/or pressure flaking, since the small size of the beads required precision. The use of antler or horn for pressure flaking of soft stones is widely seen in the ethnographic record and experimentally (Foreman 1978, 19). Even hard stones can be flaked with soft hammers made of animal horn (Kenoyer 1986; 1994; 2003; Kenoyer et al. 1991). Cores and flakes do not clearly indicate use of the Cambay technique of inverse indirect percussion, as seen in carnelian beadmaking (Kenoyer 2003; Possehl 1981). Many blanks display micro-retouch on the edges. On thicker disc bead blanks, micro-flakes were removed by striking downward at the edge of each bead face (Fig. 7c). The retouch scars show that striking was bipolar, i.e. from opposite directions. Products of this procedure were micro-flakes (Fig. 4, bottom). Bead blanks are the further reduction of a roughout into a form closer to the final bead shape (Kenoyer 2003, 16). At Jilat 13 and 25, blanks were abandoned at many different stages. The most extensive evidence for lithic reduction concerns disc beads; it is possible to identify some chaines operatoires (Fig. 8). Figure 6 illustrates two of these paths, for green Dabba Marble disc beads made on flakes. Sometimes, circular flakes were perforated early, before any abrasion (Fig. 6c). More often, flakes were abraded slightly on ventral and dorsal surfaces, before any drilling (Fig. 6a–b). At this stage, edges were still rough. Perforation was added later (Fig. 6d). In such cases, a number of disc beads were probably then strung together and abraded on the edges, by rolling the string back and forth on abrasive stones of varying textures, such as coarse sandstone, fine sandstone, or limestone. The final products were evenly smoothed disc beads, relatively standardized in size. The procedure also resulted in edges that are sharp, perpendicular to the bead faces, and flat rather than convex (Figs 6d, 7e–f). Experiments indicate that this procedure also contributes to the polishing of faces and edges of beads (cf. Foreman 1978). Figure 7 shows artefacts from one context at Jilat 25, indicating a similar sequence for red Dabba Marble disc beads, from subcircular flake to final disc. For barrel beads, the starting point was typically not a flake but a tabular roughout. A roughout was often flaked into an approximately cylindrical form, sometimes with a hexagonal transverse cross section (that is, the section perpendicular to the perforation) (Fig. 9a). Scars indicate that the hexagonal form was accomplished by flaking. Sometimes, hexagonal blanks were perforated before much abrasion (Fig. 9b). In many cases, hexagonal blanks were heavily abraded to obliterate sharp angles before any perforation was begun (Fig. 9c). Resulting bead forms varied in cross section, from circular to elliptical or lenticular (see Wright and Garrard 2003, fig. 3). Perforation of barrel beads was from opposite directions, resulting in hourglass perforations (Fig. 9e) and, occasionally, perforation errors (Fig. 9d). Most pendants were begun as tabular roughouts chipped into shapes anticipating the final form, such as an asymmetrical triangular pendant (Fig. 10a, cf. Fig. 5a for a rectangular pendant). Roughouts were Levant 2008 VOL 40 NO 2 141 Wright et al. Stone Bead Technologies a b c d Figure 5 (a) Tabular roughout, rectangular; (b) Tabular roughout, subcircular, with drill mark; (c) Abraded roughout with sawing mark at bottom; note that the sawing is incomplete; the groove and snap technique has resulted in both the saw mark and an irregular protrusion of stone; (d) Abraded tabular roughout, subcircular then abraded and sawing was sometimes applied (Fig. 10b). One unfinished pendant shows that perforation preceded final abrasion (Fig. 10c). The most common form is the asymmetrical triangular pendant (Fig. 10d), but there are also rectangular, square and oval pendants. Not all pendants were made on tabular roughouts; some were made on thin flakes, e.g. rectangular and ‘teardrop’ forms (Fig. 10e–f). Sawing Disc beads were made individually, one by one, on flakes, as opposed to being sawn from a perforated cylinder, which is another possible way. However, sawing was central to the production of pendants and barrel beads (cf. Fig. 5c, 10b). In some cases, the roughout was sawn only to a shallow depth, permitting unwanted material to be snapped off, leaving a protruding piece of stone — a kind of groove and snap technique (Fig. 5c). The protruding ‘boss’ was then abraded. 142 Levant 2008 VOL 40 NO 2 A range of chipped stone tools in the sites might be suitable for sawing. These include tabular chert knives, sometimes also called tile knives (Fig. 11). These are bifacially retouched cutting tools characteristic of the eastern Jordanian Neolithic (Baird in Garrard et al. 1994a, 89). However, typically the bifacial retouch produces an edge that is robust but somewhat sinuous in profile (Fig. 11). In addition, many tile knife edges have significant curvature in plan. Use of tile knives would probably generate relatively wide and slightly sinuous cut marks but this is an area worthy of experimentation and use wear study. Alternative tools that perhaps better match the relative scale and precision of cut marks on bead blanks are a range of relatively robust non-formal tools on blade or elongated flake blanks, with edges that are relatively straight in profile and plan. These are found in some numbers in conjunction with the bead making debris in Jilat 13 and 25. It is also notable that despite the purported dominance of flakes in many broadly contemporary PPNC/ELN assemblages, significant Wright et al. a b c d Stone Bead Technologies Figure 6 Green Dabba Marble disc bead blanks and finished beads. (a–b) Circular flakes, slightly abraded; (c) Unabraded perforated flake; (d) Finished disc bead proportions of Jilat 13 and Jilat 25 tool blanks are blades (Baird 1993, 469 and figs 8.18–8.21) even though the debitage assemblages are mostly flake dominated. Other possible sawing tools include flaked limestone artefacts with a thin edge (Fig. 12c). Several cutmarked limestone slabs were found at Jilat 13 and Jilat 25 (Fig. 12a–b). The cutmarks are shallow or deep incisions. Interestingly, the cutmarked slabs do not display evidence of drilling. Conversely, the Jilat 13 bench, with marks probably resulting from drilling (Fig. 14c; see below), has no cutmarks. This suggests segregation of drilling and sawing activities. Some variability in spatial distribution of different stages in the bead making process may be further borne out by the discrete distribution of drills in Jilat 25 contexts where substantial bead making debris was recovered. For example, in Jilat 25 Context A15, drills were found clustered in spatial units towards the northern end of the structure. evidence for drilling from bead perforations and stone tools. Other experiments show that perforation morphology reveals drilling methods and shape, diameter and length of the drill used (e.g., Gwinnett and Gorelick 1999). At Jilat, perforations indicate that drilling methods varied and that choices were probably affected by material hardness. Drilling Rotary Drilling From Two Directions, with Hafted Drills Experiments in drilling of Dabba Marble are in progress (Bains forthcoming). Here, we present The most common method involved rotary drilling from two directions. Drilling was almost always Hand Drilling with Large Piercers and Borers The simplest method used appears to have been hand drilling. Soft Dabba Marble could have been drilled using a borer or piercer held in the hand and not hafted to a drilling stick. Drilling by this method would produce rough, irregular and relatively large perforations (Gorelick and Gwinnett 1990); we see examples of this in broken beads. Borers and piercers were found at the sites. They are relatively large perforation tools made on blades, suitable for manipulation by hand without a haft (Fig. 13, nos 9–11). Levant 2008 VOL 40 NO 2 143 Wright et al. Stone Bead Technologies a b c d e f Figure 7 Red Dabba Marble disc bead blanks and finished beads from Jilat 25, Context Aa15. (a–b) Circular flakes, slightly abraded; (c) Heavily abraded and perforated disc bead blank, edge abrasion incomplete; (d) Abraded perforated blank, abrasion not complete on face; (e–f) Finished disc beads bipolar. That is, the blank was drilled to roughly halfway through, then turned and the perforation completed from the opposite direction. As the two perforations converged, the result was an hourglassshaped perforation (Fig. 9e). Experiments indicate that this prevents chipping and flaking of the alternate face which can occur if a blank is perforated completely from only one direction (Possehl 1981). Most hourglass perforations, on both soft and hard stones, appear to have been produced by rotary drilling. Rotary drilling results in regular perforations and concentric striations on them (Gorelick and Gwinnett 1990). We observed both traits on many broken beads and blanks (Fig. 9b). Perforations of hourglass form included very small drillholes, indicating that the drilling tools were smaller than piercers and borers made on blades. The probable drills in this case were small drills on bladelets, and especially drill bits on burin spalls (Fig. 13, nos 1, 4, 5, 7). Microscopic examination of perforations is in progress (Bains forthcoming). 144 Levant 2008 VOL 40 NO 2 Oddly, piercing and drilling tools were found in relatively low absolute numbers at the two sites (e.g., Baird 1993, 505–06, 625; for Jilat 13 early phase, about 4% of tools were piercers or drills; for the Jilat 25 which have concentrations of bead-making debris, just under 3% of tools were spall drills). However, specific contextual data, other technological considerations, and comparisons with other sites give us confidence that these were the main tools involved in the drilling, and that the low absolute numbers of drills abandoned may sometimes relate to systematic removal or discard of these tools by the artisans. For example, burin spall drills were closely associated spatially with bead-making debris in specific activity areas in the Jilat 25 structure, e.g. Context Aa15, in the building’s northern area (Baird 1993, 521; see Table 6). In addition, Jilat 13 and 25 produced large numbers of angle burins, including truncation burins, from which burin spalls were detached (e.g., about 29% of all tools at J13’s early phase were burins) (Baird 1993, 500, 516, 520–26, 625). Burins are Wright et al. Stone Bead Technologies Figure 8 Châines opératoires for disc beads. Sequence A refers to a disc bead made on a thin flake. Sequence B refers to a disc bead made on a thick flake. Sequence C refers to a disc bead made on a tabular roughout, not a flake contextually closely associated with drills and beadmaking debris in occupation fills at Jilat 25 (c. 30% of tools from relevant Jilat 25 contexts were burins) — a situation also seen at Jebel Naja in the Basalt Desert — where both experiments and microwear studies suggest that drill bits on burin spalls were used for beadmaking (Baird 1993, 521; Finlayson and Betts 1990). Burin spall drills would have had some technological advantages over bladelet drills. One advantage is that the triangular or rectangular cross-sections of burin spall drills make them stronger and less likely to break during drilling. In addition, the manner of their retouch from the same direction on each edge creates a prismatic cross-section that probably served to produce a neater hole and to make drilling more efficient and possibly faster, and reduce breakage in drilling. The tips of burin spall drills are blunted, which is an advantage in drilling hard stone (Gorelick and Gwinnett 1990). Predictability and miniaturization of burin spall drills may be relevant to an Levant 2008 VOL 40 NO 2 145 Wright et al. Stone Bead Technologies a b c d e f Figure 9 Green Dabba Marble barrel bead blanks and finished beads. (a) Hexagonal blank, flaked but not abraded; (b) Abraded hexagonal blank; (c) Abraded blank, not perforated; (d) Perforation error on abraded blank; (e) Finished bead, broken, showing bipolar perforation and hourglass perforation shape; (f) Perforated barrel bead, almost finished except for final abrasion to smooth out last surface irregularities apparent increased standardization in perforation size observed at Jilat 25 from an early phase to a later phase (Critchley 2000). Since we have evidence of rotary drilling from bead perforations, we believe that drill bits were hafted to sticks (probably made of wood). Drill bits on burin spalls would have been too small to manipulate effectively by hand alone. To work efficiently, a drill needs to be weighted and stabilized, and there must be some means of protecting the artisan’s hand. Ground stone artefacts support the idea that rotary drilling, with drill bits hafted to drilling sticks, was in use. They include a probable capstone, made of limestone, from Jilat 25. The capstone would have been held in the hand, enabling the beadmaker to manipulate the drilling stick from the top (Fig. 14a). This artefact resembles a miniature bowl and fits easily into one hand. The interior surface displays use wear: circular striations parallel to the rim and vertical striations perpendicular to the rim. The base of the interior converges to a flat surface about 1 cm in diameter. 146 Levant 2008 VOL 40 NO 2 Some Levantine Neolithic sites have revealed similar objects with similar wear (Banning 1998; Wright 1992, fig. 5–15b), and a comparable item was found at Jarmo (Moholy-Nagy 1983, fig. 129.13; Gorelick and Gwinnett 1990, 30). Capstones are characteristic of the use of bow drills (Banning 1998; Foreman 1978), but strictly speaking, they are not necessarily diagnostic of the bow drill (theoretically, simpler drilling techniques involving hafting might have been facilitated by the use of a capstone). For the moment, we can say that at Jilat 13 and 15, the evidence for rotary drilling is unequivocal. However, the evidence for the use of the bow drill, while suggestive, is not quite definitive. Size data, considered carefully, support a link between burin-spall drills and beads. To interpret dimensions of drills and perforations, we must keep in mind that stone bead drilling was overwhelmingly bipolar and converging. Thus, drills did not have to penetrate all the way through the full height of the bead — only about half of it. Therefore, what matters Wright et al. a b d e Stone Bead Technologies c f Figure 10 Green Dabba Marble pendant blanks and finished pendants. (a) Tabular roughout, flaked into approximately trapezoidal shape, but not abraded; (b) Abraded trapezoidal blank with sawing mark at bottom; (c) Perforated asymmetrical triangular pendant blank, incompletely abraded (note striations on face); (d) Finished asymmetrical triangular pendant; (e) Pendant blank made on a flake, unabraded; (f) Nearly-finished pendant made on a flake is the very tips of the drill bits — the upper few millimetres — not the full length of the drills. Since disc beads are consistently about 2.5 to . 2 6 mm in height, only the upper 1.5 mm or so of the drills — the most distal ends of the tips — had to have been about 2.1 mm or less in thickness to ‘match’ the perforation diameters of disc beads (Table 4). In the case of barrel beads, since barrel beads were between 6.4 and 7.1 mm in height, about 3.2 to 3.6 mm of the most distal ends of the drill bits have to have been about 2.1 mm or less in thickness to ‘match’ the perforations of barrel beads (Table 5). Measurements of the drill bit tips within these small uppermost reaches show that the maximal thicknesses of the drill bits’ distal tips fall within the range of sizes indicated by the perforations (Fig. 13). Multi-stage Drilling Sometimes, drilling may have been accomplished in several stages. Some beads, particularly those made of harder stones, have smooth, cylindrical perforations with parallel walls, displaying no hint of the hourglass shape. One way to achieve this is to create the hourglass perforation, and then to turn it into a perfect cylindrical perforation via the use of a second drilling procedure. An example of this was cited by Calley (1989), who proposed that two tools found in carnelian bead workshops at Early Bronze Age Kumartepe (Turkey) had complementary functions: a drill bit for making the initial hourglass perforation, and a borer for finishing and smoothing it into a cylindrical perforation. Since both drill bits and borers were found on the Jilat sites, this method of finishing perforations could have been used. At Jilat 13 and 25, evidence for multi-stage drilling can be seen in some beads which have a depression on one or both faces. The depression may be a result of drilling with a larger drill bit first, in order to provide a guide for the final perforation with a finer drill bit. Such a technique is seen in India-Pakistan (Possehl 1981, 39; Kenoyer 1992b, 73). Levant 2008 VOL 40 NO 2 147 Wright et al. Stone Bead Technologies wooden frame with cup-holes is sometimes used for stabilizing beads (Kenoyer 1986; P. Wright 1982). The holes are filled with beeswax and clay; the bead is inserted into the wax-clay mixture, which hardens, holding the bead steady. After drilling, beads are released by breaking the wax-clay mixture (Allchin 1979; Stocks 1989). As wood was probably scarce in Jilat, and as there are no indications of the use of clay, we can probably rule out this technique. However, stone items with cup-holes are seen in Neolithic sites (Kirkbride 1966; Wright 1992, fig. 5.27b). Abrasion Figure 11 Chipped stone artefacts from Jilat Neolithic sites. Tile knives, or possible saws for beadmaking. From Baird 1993, fig. 8.5 Stabilizing Beads: Anvils and Drilling Benches What method was used to stabilize a bead during drilling? The simplest technique is to fix a bead to a stone anvil with an adhesive. Ethnography and experiments show that such anvils can be small (the size of handstones) (Foreman 1978, figs 6–7). Use of such an anvil should result in small, shallow depressions, a form of use-damage resulting from force exerted from above. Figure 14b shows a small flat stone from Jilat 25, with a central depression of the expected size and depth. Figure 14c shows the limestone bench from Jilat 13 (previously discussed), with a number of such pits in the centre. The marks are evenly spaced and are concentrated in one small well-defined area on the widest part of the bench; this wide area was also finely abraded. We see this bench as a multifunctional worktable/ anvil. We suspect that the beadmaker sat on the narrow part of the bench (straddling it), placed bead blanks on the pitted work surface, fixed them with adhesive, and then drilled them. The bench also displays evidence of other activities, including flaking and grinding. Adhesives for hafting chipped stone tools could have been adapted for use in fixing beads to anvils. Bitumen would be one candidate. In some cultures, a 148 Levant 2008 VOL 40 NO 2 Ground stone artefacts are crucial to bead production in traditional technologies (Foreman 1978; Inizan et al. 1992; Kenoyer et al. 1991, 53; MoholyNagy 1983, 294; Roux and Matarasso 1999, 57–58). Grinding of beads can be achieved by abrading each bead individually, with a hand held abrader, or by rubbing them on a large grinding slab. Either method will produce linear striations of the type we see on many unfinished beads (cf. Figs 7c–d and 10c). However, final shaping of disc and cylindrical beads was most likely achieved by stringing them or loading them on to a thin rod capable of penetrating the perforation (individually or in groups), and then rolling them on a grinding slab, adding abrasives (such as sand) and water to produce a smoother finish (Foreman 1978; Moholy-Nagy 1983, 298). Some disc beads, with parallel edges perpendicular to the faces, would probably have been rolled on a flat, smooth stone. Other beads, with facetted, sharply angled or bevelled edges (e.g. many barrel beads), suggest that they were rolled or abraded along a stone with grooves and slanting sides (e.g., Fig. 15a, c). Abrasion of bead blanks may have been a multistage process, involving a number of materials of different textures and hardness. In an early stage, grinding of bead materials on vesicular basalt would have been the easiest way to abrade a large nodule quickly. The pores of vesicular basalt and pumice form natural cutting ‘teeth’, comparable to the metal rasp used by present-day sculptors. At Jilat 13 and Jilat 25, we found broken fragments of vesicular basalt grinding slabs, but not large complete examples. The small fragments suggest two possibilities: either there were large, complete slabs at one time, which were then broken, worn out and discarded; or only small fragments of these heavy duty grinding tools were needed. For finer abrasion, sandstone grinding slabs might be expected. None were found. However, small Wright et al. Stone Bead Technologies a b c Figure 12 Ground stone artefacts probably for beadmaking in Jilat Neolithic sites. (a) Cutmarked slab (limestone), Jilat 13. The cutmarks are shallow and are concentrated in the left area, running parallel to the main axes of the slab; (b) Cutmarked slab fragment (limestone), Jilat 13, showing deep cutmarks; (c) Flaked and ground abrasive cutting tool made of limestone, Jilat 13 handheld abraders made of sandstones with coarse, hard quartz crystals were found at both sites (Fig. 15a–b). Sandstone is not local to Wadi Jilat and these items were imported and probably used extensively (resulting in the small size). For even finer abrasion, limestone could have been used. Such needs could have been met by the large bench of Jilat 13 (Fig. 14c), which was finely ground over a very large area (within which the possible drilling marks were placed). Some might see the grooved items made of sandstone (Fig. 15a) and limestone (Fig. 15c) as ‘shaft straighteners’. One or two shaft straightener fragments made of basalt occur at Jilat 13. Like other basalt shaft straighteners in Neolithic Jilat (e.g. Jilat 7) (Garrard et al. 1994a; Wright 1993), the usesurfaces of these have U-shaped cross-sections. By contrast, the cross sections of the use surfaces on the sandstone and limestone grooved items are not Ushaped, but have sharp angles. We see these tools mainly as abraders for grinding bead facets, especially on barrel beads. The widths of the use surfaces are 11 mm for the sandstone tool and 13.8 mm for the limestone one. These dimensions correspond to the average diameter (5 width) of 31 measured green Dabba Marble barrel bead blanks from Jilat 13 (N 5 31, Mean diameter 5 9.2 mm, standard deviation 5 2.8). Grooved stones are used in bead grinding in Levant 2008 VOL 40 NO 2 149 Wright et al. Stone Bead Technologies Figure 13 Chipped stone artefacts from Jilat Neolithic sites. Drilling and piercing tools. From Baird 1993, fig. 8.2 modern India (Roux and Matarasso 1999, 57–58); a grooved sandstone tool was found with an unfinished bead at Çatalhöyük (Wright and Bains 2007). Tosi and Vidale (1990) regard failure resulting from breakage or human error as more common in the grinding stage than during perforation, although this depends on material and opinion (Kenoyer 2003, 18). In some cases, it might be more efficient to shape the bead first before attempting perforation. This may explain why barrel beads tend to be shaped and abraded prior to perforation. Finished beads were always finely abraded and sometimes polished, to the extent of reflecting light. Ethnographic observations indicate that polishing is sometimes done by placing beads en masse in a leather bag, along with an abrasive, and shaking and rolling the bags for long periods (Allchin 1979; Kenoyer 2003). This simple method would have been more likely than the method of rolling beads and abrasives in a wooden barrel, which is also documented ethnographically (Kenoyer 2003). 150 Levant 2008 VOL 40 NO 2 However, shaking beads in a bag with an abrasive tends to produce beads with rounded, convex edges (Gwinnett and Gorelick 1989). Some beads at Jilat have this form, but many — especially those made of the hard cherty red Dabba Marble — have sharp edges perpendicular to bead faces, indicating that they were individually polished, or polished during rolling on a string or stick. Alternative methods of polishing can include rubbing with leather or wood, with the addition of fine sand or chalk and water, or with animal hair/wool and animal fat (Kenoyer 1986, 20). These methods would have been possible with Jilat technology. Craft Specialization and Comparisons with Other Sites Technology needs to be studied not only in terms of materials and techniques, but also in terms of social groups involved in artefact production, individual artisans and skill (Dobres and Hoffmann 1999, 2–12). In egalitarian societies, craft skill and knowledge of Wright et al. a Stone Bead Technologies b c Figure 14 Ground stone artefacts probably used for beadmaking in Jilat Neolithic sites. (a) Capstone made of limestone, Jilat 25; (b) Handstone probably used as an anvil or hammer, Jilat 25; note depression in face and flake scars at edges; (c) Limestone drilling bench, Jilat 13. Note evidence of use-wear on the left and widest area of the bench: smooth abraded surface; drill marks in the centre-left face; small flake scars with negative bulbs of percussion close to the edge of the bench special materials can enhance the power of an individual or a group, encouraging hierarchy (Dobres 2000, 119–20). These questions force us to reconsider issues of craft specialization, a term often used loosely or in widely different ways. Three concepts of specialization are considered here: individual specialists; specialization between households or domestic groups within a given site; and site specialization within a regional or social network of craft production and exchange. Costin (1991) draws a contrast between independent specialization and attached specialization. Independent specialization involves production of utilitarian goods for nonélites, and can be householdbased or based in special workshops. Attached specialization refers to production of prestige goods under control of élites. Whilst Costin notes that attached specialization can also be based in households or workshops, we often think of attached specialization in connection with central institutions and urban or state-level societies. It is of considerable interest that in early urban societies, the making of stone personal ornaments in association with central institutions figures prominently (Stein 1996; Vidale 1989). How earlier forms of craft specialization may have evolved into large-scale attached specialization is poorly understood. Most research has concentrated on later stages in this process (e.g. Stein 1996); in discussions of earlier forms of specialization, definitions and criteria for diagnosis could sometimes be clearer (e.g., Quintero and Wilke 1998; Rollefson and Parker 2002; Rosen 1997). If we are dealing with Levant 2008 VOL 40 NO 2 151 Wright et al. Stone Bead Technologies a b c Figure 15 Ground stone artefacts possibly used for beadmaking in Jilat Neolithic sites. (a) Grooved sandstone abrading tool, Jilat 13; (b) Sandstone hand-held abrader, Jilat 25; (c) Grooved limestone abrading tool, Jilat 13 specialization in Neolithic societies, it is probably independent specialization. Still, Neolithic societies may fall between Costin’s two categories. Neolithic groups were clearly producing prestige goods; personal ornaments are not strictly utilitarian; and there is at least a possibility that élites were emerging (Byrd 1994; Kuijt and Goring-Morris 2002; Verhoeven 2002). Individual Specialists First, there is the issue of identifying individual specialists. Some use ‘specialization’ to refer to any situation in which some artisans have high levels of skill, a view that has been criticized (Rice 1991). We agree with Rice that identifying a high level of technical skill in a craft is not on its own sufficient for a diagnosis of specialization, since it robs the concept of force in discussing long term change (by this criterion, the Upper Palaeolithic was riddled with craft specialists). However, we would argue 152 Levant 2008 VOL 40 NO 2 that if there are other indications of craft specialization, a search for levels of ability in different individuals is appropriate — if the data permit. Such data would be primary refuse from well-dated sites of high temporal resolution, in which technical processes (e.g., errors in chaines operatoires) can reveal experts vs. novices (Pigeot 1990; Stout 2002; Vanzetti and Vidale 1994). In the case of Jilat 13 and 25, we are dealing with much primary refuse and two closely dated sites (see above and Tables 6–7) — with periodic use of the sites over a time range on the order of 200 years at most, and possibly much less. Jilat 25 yielded activity areas with clear spatial clusters, as do a number of contexts at Jilat 13. Space does not permit us to present all details (this awaits the final site report), but we feel confident that aggregate consideration of the data from each site allows us to discuss general questions of individual skills — master craftsmen vs. novices. Wright et al. To explore this, we analysed unfinished blanks and the frequency of perforation errors. Drilling is a sensitive moment in bead production when many things can go wrong (Kenoyer 2003). A low incidence in perforation errors, combined with high numbers of successfully finished beads and evidence of standardization, should be reasonably good indicators of an artisan’s mastery. This is the situation seen at Jilat 25 (Fig. 7). In all, of 93 blanks from Jilat 25, only 2 were perforation errors, a failure rate of only 2%. On the other hand, a high incidence of perforation errors would suggest that a beadmaker was a relative novice. Jilat 13 produced more blanks with perforation errors than Jilat 25. We thought we might find individual contexts with a high rate of error, suggesting apprentices. The data did not bear this out. Of 223 bead blanks from Jilat 13 (all phases), 22 were perforation errors — a failure rate of 9.8 %. All but two errors were on barrel bead blanks of green Dabba Marble. The others were one pendant blank and one irregular bead blank. These 22 blanks were distributed across 20 different contexts, and no more than 2 perforation errors were found in any one context. A close look at those contexts reveals low error rates. In Context B21, 2 green barrel bead blanks with perforation errors were found with 5 other blanks. The others included green and red barrel and disc blanks. This context also yielded 5 successfully finished beads, including black and red barrel beads and a green disc. Black and green debitage and a discoidal handstone were also found here. In Context B71, 2 blanks with perforation errors were found along with 9 other blanks and 10 successfully finished beads, including green Dabba Marble barrel beads but also black and red disc beads. Much debitage was found, along with a flaked limestone cutting tool (cf. Fig. 12c) and a limestone object with a drill mark. The higher rate of error at Jilat 13 compared to Jilat 25 could be because artisans at Jilat 13 were making more of the larger, more complex barrel beads (as opposed to simple discs; see Tables 2–3), which offer greater chances of failure due to the longer drilling distance. However, at Jilat 13 the overall failure rate was still quite low, particularly considering the overall volume of material recovered (Table 1). We conclude that the beadmakers at Jilat 13 included few trainees. These appear to have been skilled artisans, a small task force composed of relative experts. This raises the question of whether these beadmakers were examples of ‘producer Stone Bead Technologies specialists’ as traditionally defined, that is, people involved in production of a good by a relatively small number of individuals (compared to the total output and number of consumers) ... individuals who are, as a result of this selectivity and the routinization or repetition of their tasks, particularly skilled in manufacture. (Rice 1991, 263) Jilat 13 and 25 do suggest small numbers of individuals and a certain selectivity in who was involved in stone bead making. The quantity of debitage/debris relative to (1) the number of finished beads, and (2) the modest sizes of these structures, collectively suggest that these craftsmen/women served a larger number of consumers relative to the number of craftsmen involved here. If so, who were these consumers — other residents of Jilat, or groups beyond? Intra-site Specialization Between Domestic Groups Several studies suggest ways of identifying different forms of craft specialization between domestic groups in a site. For later sites, Stein (1996) has explored spatial data, particularly how craft production facilities are dispersed among domestic units; or located near special (or central) institutions. In Neolithic sites, questions about dispersal between households, and possible links between craft production and special buildings or special sites have been underexplored. For specialization, Costin (1991) suggests that we would expect: (1) variability between production units (e.g., households, occupations, regions) in relevant artefacts; (2) high densities of craft production debris relative to some other generally used item in some production units; (3) high ratios of unfinished goods to finished goods in some production units. Kenoyer et al.’s (1991) ethnoarchaeological studies of modern stone bead-making in Khambat (India) suggest something similar, in the case of households as production units. There, nonspecialist households, engaging in casual, opportunistic bead production, revealed low quantities of bead-making debris, few unfinished beads, and little standardization in finished beads. By contrast, households specializing in bead production were characterized by stockpiling of large quantities of raw materials, many unfinished beads, and more standardized finished products. In looking at whole occupations (i.e., phases within sites) as a unit of observation, we argue elsewhere (Wright and Garrard 2003) that artefact densities and other data from 13 PPNB and 8 PPNC/ELN Levant 2008 VOL 40 NO 2 153 Wright et al. Stone Bead Technologies occupations in the Jilat-Azraq region suggest that PPNC/ELN producers were more specialized than those of the PPNB — at least in the Jilat-Azraq region. Unlike the PPNB sites, Jilat 13 and 25 have very high densities of bead-making debris; specialized tools for drilling and sawing; there is an abundance of unfinished materials relative to finished products; bead-making tools and debris are abundant relative to normal domestic food processing equipment (ground stone artefacts) (Wright and Garrard 2003; cf. Costin 1991). That argument is chronological, comparing the PPNB to the PPNC/ELN occupations, and we stand behind it for the Jilat-Azraq case (different trajectories may apply to south Jordan and the Negev, see below). Still, it leaves open the issue of house-tohouse variation within Jilat 25 or Jilat 13. As we have only one excavated structure per site, we cannot, on present evidence, argue that each building housed a group of specialists serving other households within the same site. In fact, we have no direct evidence that there were other structures on each site — the ones excavated were the only ones visible on the surface. There are nuances of material and typological choices between the Jilat 25 structure (red disc beads) and the Jilat 13 structure (larger, complex green ornaments), but these nuances are beyond the scope of this paper. In all, we suspect that stone bead production at Jilat 13 and 25 was geared not for other domestic units ‘in camp’, but for use by these quite mobile communities themselves — as well as export to other sites, within the steppe and beyond (see below). In Costin’s terms (Costin 1991), Jilat 13 and 25 might each be an example of an independent, non-centralized, dispersed workshop, possibly ‘family run’ (Rice 1991, 262) — in which handling of Dabba Marble was not routinely entrusted to novices. There are hints of house-to-house variations in bead-making in major villages, with larger samples of domestic structures, as in PPNB Beidha. At PPNB Beidha (earlier than Jilat 13 and 25), Building 14 stands out: Other forms of production besides food preparation were less common. Building 14 represented the best example of artefact production in these basements. A wide range of artefacts was recovered apparently associated with building 14’s basement floor. Particularly widespread were stone polishers, bone tools, raw material (including hematite, malachite and magnesium), more than 60 unmodified marine shells, and at least two dozen beads of shell, stone and bone, some of which were in various stages of 154 Levant 2008 VOL 40 NO 2 production. Some artifacts lay on stone slabs, indicating that at least some slabs may have functioned as work surfaces. (Byrd 2005, 117) Such hints are intriguing. Still, macro-artefacts from house floors are affected by problems of site formation and house abandonment. Micro-artefact data can help (Rainville 2005). House-by-house exploration, with attention to micro-artefacts, is needed (Wright and Bains 2007). Site Specialization Within a Regional Network of Craft Production and Exchange Both sites, and other Neolithic sites in Jilat and Azraq, exhibit resource specialization, the selective use of particular resources in craft manufacture, such as certain clays that are repeatedly used ... (Rice 1991, 262) The Jilat artisans emphasized Dabba Marble from a wider array of available stone types. The JilatAzraq sites collectively show that green Dabba Marble was a special target. This is revealed in both frequencies and weights (for debris data and weight data see Wright and Garrard 2003; for frequency data on beads and blanks see Tables 2 and 3). At both sites, debris is greatly dominated by green Dabba Marble — 80–95% of debris, by both weight (grams) and frequency. Green Dabba Marble also dominates blanks in terms of weight, at both sites — indicating the use of green for making larger, heavier and more conspicuous ornaments. Turning to frequencies, at Jilat 13, about 90% of blanks are green, and only 5% are red. However, finished beads include both green and red forms in comparable percentages (40.4% and 43.6% respectively) (Table 3), suggesting that finished red beads were brought into the building from a place of manufacture located somewhere else. At Jilat 25, blanks of red and green Dabba Marble occur in comparable proportions (40.9% green, 43% red), whilst finished beads are predominantly red (58.8%) (Table 2). If Jilat 13 and 25 were absolute contemporaries, this could suggest variations in choices and exchange between residents of different structures. Rice (1991) has defined site specialization as a situation in which a site is located near a special resource and the inhabitants emphasize production of a craft based on that resource. Relative to other sites, not so located, the range of site functions and activities would be expected to be more limited. Thus, site specialization involves individual localities or sites having evidence of limited functions or intensive productive activity, Wright et al. often determined by fortuitous environmental factors ... This includes proximity to mineral deposits ... and so forth. (Rice 1991, 262) Do Jilat 13 and 25 fit this concept? In both sites, 15 km from the source, we have a wealth of evidence for intensive stone bead production and scanty evidence for routine food processing tools (Wright and Garrard 2003). However, Jilat 25 may be closer to the limited function model envisioned by Rice. In Jilat 13, bead-making is clearly intensive, but we are also dealing with a chipped stone assemblage of not only drills and burins, but also projectile points and other tools. Jilat 13 also reveals production of figurines (of animals, phalluses and arrow-shaped forms), and sculpted pillars (Baird 1993; Garrard et al. 1994a; Martin 1999; Wright 1992, 1993). In all, Jilat 13 and 25 suggest that we can apply Rice’s ‘site specialization’, on the grounds of intensive productive activity and proximity to source — but with the qualification that this specialization was also embedded in hunting-herding activities and, at Jilat 13, production of other special crafts, unique among the sites, of a symbolic nature (figurines, pillars). One possibility is that we are dealing with a camp of hunter-herder corporate groups — part of a wider regional network — engaged in special activities, in remote areas, involving art, personal ornaments and ritual. Such groups are suggested by other data from the PPNB–PPNC time span, for example Nahal Hemar (ornaments, masks, hunting tools), Dhuweila (carving of hunters with dress variations), Çatalhöyük (images of hunters with dress variations, as in the bull hunt mural) and Gobeklitepe (sculpture production, including a quarry; wild-animal iconography, hunting) (Bar-Yosef and Alon 1988; Betts 1998; Mellaart 1967; Schmidt 2000). In all, these data hint that craft production of special items was embedded in hunting and/or herding activities in remote areas; this may have been a key means by which exotic materials reached major villages. In a very broad way, this supports a sketch outlined by Bar Yosef (Bar-Yosef 2001), although we do not necessarily agree with specifics of his model, e.g. the territories, boundaries, social groups and mechanisms which he suggests (cf. Asouti 2006). Instead, we see this as one example of diverse strategies of social networking linking lineages, households and communities in the Neolithic (Wright forthcoming). Returning to the stone bead data from Jilat 13 and 25, the apparent stockpiling of raw material and intensive production activity hints at production of the green material for export (see above and Wright Stone Bead Technologies and Garrard 2003). New data now add weight to this idea, and suggest the following provisional picture of regional specializations and exchanges in ornaments: (1) Based on typology, technology and materials, there appears to be a distinctive eastern Jordanian koine or network of stone bead manufacture and exchange, featuring Dabba Marble, and a probable regional specialization in it. Other Neolithic bead production sites are known from eastern Jordan. Indications are that most of the green bead materials are Dabba Marble, probably from Jilat (further work is needed). In these sites, diversity of materials used seems low, with few imports from beyond the steppe. Shells, however, testify to wider networks (Bar-Yosef-Mayer 1989; Cooke and Reese in Betts 1998; Garrard et al. 1994a; Reese 1991). Azraq 31 (PPNB and ELN) is in our suite of study materials (Baird et al. 1992; Garrard et al. 1994a; Wright and Garrard 2003) and debris, blanks and beads were found. Preliminary XRF analysis of green samples (by Groom) indicates that the material is consistent with Jilat Dabba Marble. Typology and technology are also similar, and as at Jilat 25, burin spall drills were closely associated with bead-making in specific activity areas (Baird 1993, 521; Baird et al. 1992, 25). However, the ground stone revealed no capstones, drilling benches or sandstone abraders (Wright 1992, 1993; Wright in Baird et al. 1992). Material used for beads at Dhuweila (PPNB and ELN) has been called Dabba Marble (Cooke and Reese in Betts 1998). It is consistent in appearance with our material, although we have no details on mineralogy or source. Typologically, these beads are similar to those of Jilat 13 and 25. The ground stone is very different from the suite at the Jilat sites (Wright in Betts 1998); for comparisons of chipped stone assemblages, see Baird (1993). At Bawwab al-Ghazal (Late PPNB), Rollefson et al. found drills bits on burin spalls, tile knives, beads, bead blanks, and bead debris, some of which they called green Dabba Marble (Rollefson et al. 1999; Quintero et al. 2004). They argued that there is a possible source of this material outcropping near the site — ‘For the last raw material [green Dabba Marble], small nodules of the green limestone were embedded in the local bedrock’ (Rollefson et al. 1999, 3). However, as shown in Appendix A, the true Dabba Marble only occurs in a restricted area west of Wadi Jilat. The ‘bedrock’ which outcrops near Azraq 31 and Bawwab al-Ghazal is travertine, and some of this formed during the late Pleistocene and Holocene. The nearby Epipalaeolithic site of Azraq 17 was Levant 2008 VOL 40 NO 2 155 Wright et al. Stone Bead Technologies found partially embedded in travertine, and artefacts including nodules of imported basalt were found in the travertine at this site. It is possible that the nodules of green limestone found at Bawwab alGhazal were imported Dabba Marble which subsequently became incorporated in a travertine near the site. An alternative possibility is that the local travertine is infused with green minerals. In Wadi Jilat, travertines were noted with traces of green minerals, and in the Siwaqa area to the south-west of Jilat, there are extensive travertines, some of which are infused with pale green minerals including volkonskoite (Barjous 1986 — see also Appendix A). In eastern Jordan, some Neolithic sites are greatly dominated by burins (‘burin sites’). At a number of these there is evidence for stone bead-making (e.g., Jebel Naja). These ‘burin sites’ typically reveal few ground stone tools; we know of no examples of capstones, drilling benches, or sandstone abraders from them. It was argued that the main role of the burins was production of spalls for drills used mainly for beads (Betts 1987; 1998; Finlayson and Betts 1990). However, burin sites are not universally associated with stone bead making. At Jilat 23, 24 and 26, burins were numerous, but other evidence for bead-making is scanty (Baird 1993, 520). Therefore, some have questioned whether burins and burin spalls in such sites were solely or primarily intended for making drills and beads, noting that other functions are possible (Quintero et al. 2004, 209). Given that these sites typically represent short-term occupations, it is possible that many sites might be spall manufacture sites, with spalls transported to other settings where drill manufacture and bead manufacture took place, such as Jilat 13 and 25. Rollefson expressed doubt that either bladelet or burin spall drills would have been used in bow drills partly because of their fragility and also because most of them have a pronounced curve on the long axis which would have made high speed drilling impossible (Rollefson and Parker 2002). But Finlayson and Betts (1990) and Berna (1995) successfully experimented using burin spalls with bow drills, and Kenoyer (personal communication 2007) notes that drills made on burin spalls are the most common type of drill in diverse prehistoric stone bead-making sites (in India, Pakistan and elsewhere). In the case of Jilat 13 and Jilat 25, we believe burinspall drills were part of the bead-making repertoire, for reasons noted above. This does not rule out other uses of burin spalls. Nor does it rule out the use of other perforating techniques. Drills do not have to be 156 Levant 2008 VOL 40 NO 2 made of stone when soft stones are being perforated — wood, bone and abrasives can be used (Foreman 1978). (2) There are hints of export of Dabba Marble to at least some villages in the Levantine corridor. However, major Neolithic villages display much diversity in materials and especially more importing of exotic stones than is seen in the eastern Jordanian sites. Many writers report Neolithic beads made of unspecified ‘greenstone’. This term is unfortunate, since it can refer to metabasalt and similar altered igneous rocks (Schumann 1992, 248). Some of these ‘greenstones’ may be limestone with apatite, or other materials (e.g., Garfinkel 1987, 81–82; Hauptmann 2004; Talbot 1983, 789). Beads of green limestones rich in fluorapatite were found at late PPNB Basta in southern Jordan (Hauptmann 2004), as confirmed by XRF and XRD. Green beads identified as Dabba Marble were also reported from Ain Ghazal, a site with both PPNB and PPNC occupations (Rollefson et al. 1990, 103–04). Whether these are Dabba Marble from the Jilat source (Appendices A–B) is not yet fully clear. Ongoing research will sort out these mysteries (e.g., D. Bar-Yosef Mayer personal communication; Hauptmann 2004). Major villages in the Levantine Corridor reveal different degrees of emphasis in styles and materials; often we see a certain emphasis on local materials and preferences. However, considerable exchange networks are indicated in major villages, which display much wider ranges of stone ornament materials compared to the eastern Jordanian sites (such as Ain Ghazal, Yiftahel, Basta, Ba’ja, Ghuwayr 1, Fidan 1) — along with, of course, shells (Affonso and Pernicka 2004; Bienert and Gebel 1998, 84–86; Garfinkel 1987; Hauptmann 2004; Rollefson and Simmons 1984; 1986; Rollefson et al. 1990; Simmons and Najjar 1998; Starck 1988). As the Neolithic evolved, stone bead-making entailed increasingly diverse techniques and materials. One example may be drills. In early/middle PPNB sites, most drills are made on bladelets. In the late PPNB, PPNC and Late Neolithic there may have been a gradual increase in drills made on burin spalls, although perhaps not in all regions (Jensen 2004; Rollefson and Parker 2002). Underlining increasing diversity of technological practice between groups, Baird (2001b) pointed out geographic variation in this technology, with spall drills dominating in the north-eastern Jordanian steppe from MPPNB through Early Late Neolithic, whilst contemporary sites to west and south use bladelet blanks for their piercing tools. Wright et al. (3) South Jordanian stone bead-making sites involved very different materials and different technologies local to those regions — and may reveal a more intensive exploitation of stone bead resources in the PPNB, i.e., earlier than in eastern Jordan. In south Jordan, Jebel Arqa, Jebel Rabigh and Jebel Salaqa revealed Middle PPNB beadmaking sites. Published analyses are based on surface collections. Around huts of upright stone foundations were found large quantities of amazonite (microcline; Mohs 5 6.5) debris, blanks and finished beads. Thin borers constitute 60–80% of the chipped stone. Awls are also numerous, as are ground stone tools (not described). Beads and blanks are dominated by disc beads made on thin flakes struck from the tabular material (Fabiano et al. 2004, 266–72; Vianello 1985). Fabiano et al. interpret these sites as seasonal camps exploiting mineral resources in a somewhat specialized way and that bead production was geared largely to export (2004, 265). Although materials and technologies differ from the Jilat sites, certain elements (flake-based beads, architecture) broadly mirror the Jilat situation, although Jilat bead-making is modest until the PPNC/ELN. Jebel Ragref also revealed amazonite bead production (Berna 1995). In experiments, Berna found that drills of only 1.0–1.5 mm in diameter could produce 3 mm diameter perforations in amazonite beads. Berna found that hand-drilling of hard amazonite was difficult, whilst use of an experimental version of a bow drill made this easier. However, drill bits frequently became exhausted and worn out and had to be replaced. Exhausted drill bits mimic different types of borers, raising the possibility that many borers and drill bits found on sites are in fact exhausted, discarded tools (Fabiano et al. 2004, 272). (We wonder if the low frequency of drill bits found in the Jilat sites is due to careful discard of exhausted drill bits.) Al Basit, Late PPNB site near Petra (Fino 1998), revealed drills made on bladelets, along with limestone and sandstone objects with drilling marks (Rollefson 2002). Bladelet drills dominate a chipped stone collection gathered from a spoil heap in one area; the spoil heap also included shell beads and a possible malachite fragment. Rollefson sees the material from this spoil heap as indicating specialization in bead-making, in contrast to a more generalized chipped stone toolkit from an in situ domestic area (Rollefson 2002, 5; Rollefson and Parker 2002, 22). Bead-making debris was found in secondary refuse (pits, middens) at Shkarat Msaied (Middle PPNB) Stone Bead Technologies (Jensen 2004; Kaliszan et al. 2002). Borers and green beads were found in a concentration, and grinding stones not far away. The material was provisionally identified as turquoise and malachite. Asymmetrical drills/borers made on bladelets dominate the chipped stone, although a few drills/borers were also made on burins and blades. (4) Data from PPNB and ELN sites in the Negev and Sinai indicate yet other specialities, and a wider array of imported materials than seems to obtain in eastern Jordan. Some sites in Negev-Sinai are impressive in the variety of stone bead materials (e.g., Nahal Issaron, Negev) (Goring-Morris and Gopher 1983, 156) and specialist shell bead-making sites have been described in Sinai (Bar-Yosef-Mayer 1989; 1997). Nahal Issaron revealed piercers with long thick bits; similar tools were found at Jilat 13 (Baird 1993, 515–17). Conclusions Much work on early stone beads remains to be done. The Jilat sites illustrate an expansion of stone bead-making in the PPNC/ELN and suggest that an early version of craft specialization emerged at that time. In other regions, different processes may have been at work. One question is the role of sheep-goat herding in procurement of prestige minerals in remote areas. Something like this is seen in the Early Bronze Age (Rosen 2003). Its roots were earlier, though views vary on timing, processes and goods involved (cf. Bar-Yosef- Mayer 1997; Fabiano et al. 2004; Garrard et al. 1996; Martin 1999; Quintero et al. 2004; Rollefson et al. 1999). It is worth investigating whether there was a general expansion of trade networks in the PPNC/ELN, when herding may also have become more extensive. Appendix A: Distribution, Nature and Origins of Dabba Marble Andrew Garrard Dabba Marble (also spelled Dab’a, Dab’ah) occurs in a restricted area of the Maestrichtian to Paleocene (late Cretaceous–early Tertiary) Muwaqqar ChalkMarl (MCM) Formation in central Jordan. The MCM formation outcrops in an arc from Sahab east of Amman to the Wadi Sirhan east of the El Jafr Basin in southern Jordan. However, the marbles themselves are restricted to a north-west/south-east trending fault zone lying to the east of Khan es Zabib and Siwaqa stations on the former Amman to Ma’an Railway line. The area covers about 25 km (NW-SE) by 13 km (SW-NE) and lies between c. 36u06’–36u17’/ 36u22’ E and 31u21’–31u34’ N (Barjous 1986; Jaser Levant 2008 VOL 40 NO 2 157 Wright et al. Stone Bead Technologies 1986; Powell 2006). The Neolithic sites of Wadi Jilat 13 and 25 lie adjacent to each other at 36u25’ E and 31u30’ N, some 7–15 km east of the nearest marble outcrops (the exact location of the nearest sources needs to be checked in the field). The modern quarry source of marble used for comparative analysis lies at c. 36u15’ E and 31u29’ N. The Dabba Marbles occur within the upper unit of the MCM Formation, which consists of chalky limestones, chalks and micritic limestones, interbedded with dark grey to brownish chert (Bender 1974; Jaser 1986; Powell 1988). The MCM Formation was deposited in moderate to deep water on the southern margin of the Tethys Sea and bituminous chalky marls (or oilshales) occur as lenses in structural depressions within the formation. The marbles formed as a result of metamorphosis and recrystallization of the parent rocks along cracks and fissures in this formation and range in colour between black, violet, brown, red, pink, green and yellow (Fig. 3). The coloration results from infusion with bitumen and iron oxides, chromites and apatites. The marbles are cross-cut with veins containing diverse minerals (more than 50 were identified) (Jaser 1986; Nassir and Khoury 1982). At present there are many small quarries in the marble formations, as the multi-coloured but fractured rocks (particularly the green apatitic varieties) are popular for setting in plaster for floor and wall tiles. Much debate surrounds the origins of the marbles, but there are no high level intrusive or volcanic bodies in the area and thus no possibility of metamorphism through contact with igneous formations. The marbles appear to have formed during the late Tertiary — a major period of tectonic activity — with the uplift of the Jordanian Plateau and the formation of the Rift Valley. It is thought that the high temperatures required for their formation was the result of oxygen reaching bituminous lenses in the MCM formation through fissures caused by tectonic movements, and that this may have led to the localized combustion of hydrocarbons. Subsequently a rich array of minerals were circulated through the faultlines and fissures in hot alkali-rich ground waters (Jaser 1986; Khoury and Nassir 1982). There are also extensive travertines in this area (particularly to the east of Siwaqa) some of which contain pale green volkonskoite. These suggest hydrothermal activity (Barjous 1986). Analogous marble formations are known elsewhere in the southern Levant. They include the Maqarin Formation in the Yarmouk River Valley north of Irbid, and the Hatrurim Formation 158 Levant 2008 VOL 40 NO 2 (formerly known as the Mottled Zone), west of the Rift Valley. These are contained in facies containing bituminous shales which are similar and timeequivalent with the MCM Formation in central Jordan. In both areas there are localized occurrences of lightly metamorphosed rocks with a rich array of minerals resulting from what were probably similar processes (Gross 1977; Moh’d 2000). For the location of the two main exposures of the Hatrurim Formation, see Fig. 1. Appendix B: Mineralogy and Chemical Composition of Dabba Marble Simon Groom and Roseleen Bains Six samples of Dabba Marble from a modern quarry in the heart of the marble zone 15 km to the west of the Neolithic sites of Jilat 13 and 25 were analysed to determine mineral/chemical composition (Appendix A and Fig. 3). In order to determine if the material described as Dabba Marble from the prehistoric sites in Wadi Jilat were from a similar source, 5 finished bead fragments were analysed from the surface levels at Jilat 13. Surface material was chosen for this pilot study so as not to damage material from intact levels. As a result of fine grain size and thin laminations, thin section analysis proved problematic. Thus X-ray fluorescence (XRF) and X-Ray diffraction (XRD) were used in combination for the quarry samples and Scanning Electron Microscopy (SEM) for the finished beads. Samples from the Quarry: XRF-XRD Analysis (Simon Groom) Materials Seven surface nodules were collected by Andrew Garrard from the scree slope at the modern quarry (Fig. 3) which were regarded as being broadly representative of the range of marbles exposed in the quarry-face. Six of these were subjected to XRF and XRD analysis. The quarry samples were categorized into 3 groups before analysis: green (G1-2), red (R1-2), and laminated green (LG1-3). All but Sample R2 (a bright red chert clearly composed of silica and iron) were subjected to XRF and XRD. G1 and G2 are fine grained materials with a similar green colour but significantly different textures. G1 is a rich green colour with a mottled appearance due to the presence of fine white and maroon veins. G2 is paler green and more uniform with very little macroscopic variation. R1 is texturally similar to G2, but pale pink in colour and relatively uniform. R2 is a brecciated Wright et al. material consisting of large chunks of red chert trapped inside a buff matrix. The macroscopic pitted appearance of the matrix is sufficiently distinctive to identify the material as a travertine or tufa. The chert brecciated within this is mottled with a variation between pale pink and typical white. LG1-3, the laminated green materials, are distinctive, commonly showing extremely fine scale parallel layers of maroon through to a rich green. While some layers of green material are several millimetres thick, this layering is commonly at less than a millimetre scale. All layered samples show non-parallel white veining. Methods Methods of analysis for quarry samples were as follows. For homogenous samples, 20 g of representative material were removed using a diamond-coated tile cutter blade and subjected to standard XRF and XRD sample preparation procedures. Large heterogeneities, such as veins and chert fragments, were ground from the sample surface using a hand sander. The resulting bulk material was crushed using a steel percussion mortar and pestle, then ground to a grain size of ,60 mm using an agate planetary ball mill. For quantitative XRF analysis, a sample of the resulting powder was initially dried, then mixed with a wax binding agent at a ratio of 8 g sample to 0.9 g wax. This mixture was then pressed in a hydraulic press and analysed in triplicate using the industrial standard TurboQuant (TQ0261a) method on a Spectro X-Lab 2000 (P)ED-XRF Spectrometer. Pure chert samples (e.g., Sample R2) were analysed Stone Bead Technologies qualitatively as objects in the same instrument for categorization. Due to its qualitative nature, these data are not presented here. XRD analysis was performed on the remaining powder. The instrument used was a Philips 1720 diffractometer fitted with a curved graphite crystal monochromator. Philips PC-APD software (version 1.6) was used to interpret the data. Results In terms of major elements, the 6 quarry samples are all composed predominantly of calcium oxides varying from 48 wt% to 58 wt% (Table 8). The most variable component is phosphate, with concentrations between 13 wt% and 7 wt%. Minor element concentrations are present of silica, iron oxides and alumina in the 1 wt% to 7 wt% range. Within the trace element data (Table 9), the material shows unusually rich concentrations of transition metals for a calcium carbonate based unit: in particular chromium, nickel, zinc and cadmium. Also notable are the low concentrations of alkali and alkali earth metals for a calcium carbonate based unit: with rubidium, potassium, sodium and magnesium all near the limits of detection. The XRD data show that the samples contain the major components calcite (CaCO3) and fluoroapatite (Ca5(PO4)3F), with the LG1 sample containing traces of margarite (CaAl2(Al2Si2)O10(OH)2): a less common member of the mica mineral group. The major components fit exactly with what would be expected from the chemical data. The margarite also fits the Table 8 Proportional weights of major elements, obtained by XRF Sample Na2O MgO Al2O3 SiO2 P2O5 SO3 Cl K2O CaO TiO2 MnO Fe2O3 Sum Unit % % % % % % % % % % % % % G1 G2 LG1 LG2 LG3 R1 0.8 0.5 0.1 0.1 0.4 0.9 0 0 0.3 0.2 2.5 0.4 0.4 0 1.7 0.6 1.7 1.2 3.8 1.8 7.5 13.03 11.9 3.99 8.68 6.28 7 0 0 0 0 0 0 0 0 0 0 0 0.4 0 0 0 0 0 0 55.6 57.7 51.3 56.3 48.5 49.7 0.052 0.03 0.079 0.049 0.134 0.085 0.006 0 0.024 0.005 0.008 0.006 0.85 0.14 1.72 0.47 1.67 1.17 74.7 72.2 66.9 70.5 69.1 67.4 4 7.6 6.4 Table 9 Trace elements in parts per million (ppm), obtained by XRF Sample V Cr Co Ni Cu Zn Ga As Rb Sr Y Zr Cd Sb Ba La Pb Th U Unit ppm ppm ppm ppm ppm ppm ppm ppm ppm ppm ppm ppm ppm ppm ppm ppm ppm ppm ppm G1 G2 LG1 LG2 LG3 R1 342 158 512 225 538 0 1424 5 409 5 695 35 810 2 1239 22 311 10 220 75 425 131 417 256 93 33 97 105 193 184 1311 644 2365 795 1893 2077 2 2 2 3 2 3 77 15 68 44 119 73 3 1 6 3 2 2 1239 112 1422 56 1141 29 1223 61 1290 108 953 95 19 4 16 5 32 18 52 63 64 18 64 210 10 1 9 1 10 9 69 312 1084 158 610 72 Levant 15 6 2 11 12 18 11 6 6 14 10 6 2008 3 1 3 2 4 2 VOL 42 9 18 30 39 28 40 NO 2 159 Wright et al. Stone Bead Technologies chemistry and is likely to be the mineral state in which LG1’s slightly elevated alumina and silica levels are resident. By comparison to the macroscopic variation, these analytical data are unexpectedly simple, suggesting that the material is structurally varied but chemically very similar. The categorization cited in Appendix A, classifying these materials as hydrothermal deposits and travertines, closely fits the composition of the samples. The elevated levels of transition metals are likely to have been dropped from solution by changing fluid chemical conditions, and suggest that the macroscopic variation is structural rather than mineralogical. The rich green colour appears to correlate closely with increasing proportions of phosphate, likely to be present in the form of fluoroapatite, a mineral which is green in colour. The coloration in the red sample is indicative of the iron oxides present at the 1% range. This sample contains amongst the lowest concentrations of phosphate. Manufactured Beads: SEM Analysis (Roseleen Bains) For this pilot study, five bead fragments from surface levels of Jilat 13 were examined using a Scanning Electron Microscope (SEM). SEM was used instead of XRF and XRD, because (a) the size of each bead sample was small and (b) we wanted to do as little damage as possible to the beads. Materials The fragments encompass a range of materials as identified with the naked eye, but do not necessarily represent all variants of Dabba Marble used at Jilat 13. Four beads were green and one was pale pink/red. At a macroscopic level, each bead appeared homogenous in colour. Methods The five samples were mounted in epoxy-resin and polished and coated with carbon to create an electrically conductive surface. They were analysed under SEM (Model S-3400N Hitachi, with a Backscattered Electron Detector and an Oxford Instruments Energy Dispersive Spectrometer (EDS) for semi-quantitative compositional spectrum) using an accelerating voltage of 20 kV. For each sample, images were captured under set magnifications, using the secondary electron detector and backscattered electron detector, to acquire topographical information on texture, surface features, and crystallography. The backscattered detector revealed images in which minerals or compounds with elements higher in atomic number appeared lighter and brighter 160 Levant 2008 VOL 40 NO 2 Figure 16 Backscattered image of sample S6-S. The scale is 50 microns long in comparison to the other minerals or compounds also present in the image, helping to differentiate the minerals for better analysis. These data, combined with data from EDS concerning elemental distributions, were used to compare with results obtained by XRF and XRD on material from the modern quarry. Each sample underwent repeated trials of bulk, box and spot analysis. Results The four green beads revealed very similar results in both surface studies and elemental analyses. Any differences in micrographs were the result of varying degrees of polishing or the type of rock formed by the similar compounds present. We present detailed results for one green and one red bead below. Sample S6 (Green Bead) SEM and backscattered analysis Initial low magnification inspection, using a backscattered detector, reveals two distinct areas, one light and the other darker in shade. The darker area is represented by an anhedral crystal structure and the lighter area has a hexagonal, euhedral, lathe crystal structure. The former can be identified as calcium carbonate (CaCO3), and the latter fluorapatite (Ca5(PO4)3F). At high magnification we can see that the fluorapatite is forming bands adjacent to the fissures in the calcium carbonate (Fig. 16). Both components appear in relatively equivalent ratios. As with the other green bead samples, the mineral fluorapatite has metamorphosed and recrystallized with the calcium carbonate. Elemental analysis and identification Trials of box and spot analysis for sample S6 confirmed that the darker shaded area can be identified as the compound calcium carbonate and the lighter shaded hexagonal crystal structures as the Wright et al. Stone Bead Technologies Figure 17 Backscattered image of sample S4-C0. The scale is 10 microns long mineral fluorapatite (Table 10). Fluorapatite’s chemistry, occurrence, and physical and optical properties, as listed in The Handbook of Mineralogy Vol. 4 (Anthony 2000) are consistent with those observed in all the green bead samples. Sample S4 (Red Bead) SEM and backscattered analysis Sample S4 was the only red Dabba Marble bead to be analysed. It is a relatively soft stone and was more difficult to polish during preparation. Under lower magnifications, the surface appears to be more compact and using a backscattered detector two or more different components become visible, once again represented as lighter crystals within a darker, more compact matrix (Fig. 17). The lighter euhedral crystal habit has formed within an anhedral crystal structure. Both minerals are much more intricately fused in comparison to the green Dabba samples and, consequently, SEM micrographs reveal a finer grain in contrast to the green samples analysed (Fig. 17). Each of two inclusions (Areas) present in sample S4 was analysed under SEM and an elemental spectrum was produced. The crystals which appear lighter (Area 1) are dissimilar to the apatite crystals which are found in the beads manufactured from the green Dabba Marble. These crystals can be identified as calcium carbonate with the addition of some magnesium and silicon (Table 11). Analysis of Area 2, or the darker anhedral mineral, revealed a mineral with a significant presence of calcium and silicon, and other elements in lesser quantities such as iron, aluminum, magnesium and chlorine (Table 11). In contrast to the samples of beads made from green Dabba Marble, the red beads have a greater amount of silica. The identification on the basis of these elements is calcium carbonate within a medium of silica and calcium. Summary SEM surface studies of the four green beads revealed two components: (1) an euhedral hexagonal prismatic Table 10 Table of elements present in sample S6 Elemental weight % Compound/Mineral C O Calcium carbonate 19.1 43.6 Fluorapatite 20.6 30.3 Number of analyses for each compound/mineral: 6 Ca I F Si P S 37.0 30.8 0.5 0.3 0.0 4.0 0.0 1.0 0.0 12.6 0.0 0.3 Table 11 Table of elements present in sample S4 Elemental weight % Compound/Mineral C O Area 1 - Calcium carbonate 20.1 46.3 Area 2 - Matrix 26.2 33.5 Number of analyses for each compound/mineral: 3 Mg Al Si S Cl K Ca Fe 0.3 1.9 0.0 0.8 0.4 9.7 0.0 0.1 0.0 0.6 0.0 0.2 32.7 26.0 0.0 0.7 Levant 2008 VOL 40 NO 2 161 Wright et al. Stone Bead Technologies crystal structure within (2) an anhedral crystal structure. The EDS detector later identified these as (1) the mineral fluorapatite (Ca5(PO4)3F) and (2) the compound calcium carbonate (CaCO3) respectively. Topographical studies of the red bead also exposed two components: (1) an euhedral crystal structure within (2) an anhedral crystal habit. Elemental analyses revealed the euhedral crystal structures to be calcium carbonate with two additional elements: magnesium and silicon. The anhedral crystal habit was more diverse in regards to the elements present, the major difference being the high silicon and calcium content and the presence of iron. The minerals comprising the red Dabba Marble can be identified as calcium carbonate within a silicon-rich matrix. The green Dabba Marbles used in bead production were formed as a result of the fluorapatite metamorphosing and recrystallizing with the calcium carbonate along cracks and fissures. A similar process could be found in regard to the red Dabba Marble bead, but in this instance, it was the calcium carbonate which formed the euhedral crystalline structure. Conclusions The SEM results on the small sample of beads from Jilat 13, indicate that they are very similar in terms of mineralogy and geochemistry to the modern quarry samples obtained from the heart of the Dabba marble exposures 15 km to the west of the site, and are probably from a source in the same general area. However, to tie down possible sources further, it would be useful to obtain samples from a wider range of outcrops of raw material across the Dabba marble region. Similarly, it would be helpful to undertake SEM studies on a more varied selection of beads from Neolithic sites in the Wadi Jilat. In the past, comparison of stone bead materials from Levantine prehistoric sites tended to be based on macroscopic visual examination. This is now changing (D. Bar-Yosef Mayer personal communication; Hauptmann 2004). Wider application of such techniques would be enormously valuable in defining the range of raw materials used and their likely sources. This in turn would be of great benefit in the reconstruction of group mobility patterns and of wider exchange and social interactions in the region. Acknowledgements For supporting this research, we are grateful to the Department of Antiquities of Jordan, the British Institute at Amman (now Council for British 162 Levant 2008 VOL 40 NO 2 Research in the Levant), the British Academy and the Wainwright Fund for Near Eastern Archaeological Research. For photographs of artefacts we thank Ken Walton, Stuart Laidlaw and Helena Coffey. For help and/or useful discussions, we thank Kevin Reeves, Dafydd Griffiths, Thilo Rehren, James Lankton, John Powell and Tobias Richter. Any errors of fact or interpretation are our own. Bibliography Affonso, M.T.C. and Pernicka, E. (2004) Mineralogical analysis of Late PPNB rings. Pp. 155–67 in H. J. Nissen, M. Muheisen and H. G. Gebel (eds), Basta I: The Human Ecology. Berlin: Ex Oriente. Allchin, B. (1979) The agate and carnelian industry of Western India and Pakistan. Pp. 91–105 in J.E. van Lohuizen de Leeuw (ed.), South Asian Archaeology: Papers from the Third International Congress of the Association of South Asian Archaeology. Leiden: E.S. Brill. Andrefsky, W. (1998) Lithics: Macroscopic Approaches to Analysis. Cambridge: Cambridge University Press. Anthony, J.W. (2000) Handbook of Mineralogy, Volume 4. Tucson, Arizona: Mineral Data Publications. Asouti, E. (2006) Beyond the Pre-Pottery Neolithic B interaction sphere. Journal of World Prehistory 20, 87–126. Bains, R. (forthcoming) Procurement to Production: Manufacturing Technologies of Stone Beads at Catalhoyuk, Turkey. PhD dissertation, Institute of Archaeology, University College London. Baird, D. (1993) Neolithic Chipped Stone Assemblages from the Azraq Basin. PhD dissertation, Department of Archaeology, University of Edinburgh. — (1994) Chipped stone production technology from the Azraq Project Neolithic sites. Pp. 525–41 in H.G. Gebel and S. K. Kozlowski (eds), Neolithic Chipped Stone Industries of the Fertile Crescent. Berlin: Ex Oriente. — (1995) Chipped stone raw material procurement and selection in the Neolithic Azraq Basin: implications for Levantine Neolithic cultural development. Pp. 505–14 in K. ’Amr, F. Zayadine and M. Zaghloul (eds), Studies in the History and Archaeology of Jordan V. Amman: Department of Antiquities of Jordan. — (2001a) Explaining technological change from the 7th to the 6th millennium BC in the southern Levant. Pp. 319–31, in I. Caneva, C. Lemorini and D. Zampetti (eds), Beyond Tools. Berlin: Ex Oriente. — (2001b) The analysis of chipped stone in Jordanian Archaeology. Pp. 693–706 in R. Adams, P. Bienkowski and B. MacDonald (eds), The Archaeology of Jordan. Sheffield: Sheffield Archaeology Monographs. — Garrard, A.N., Martin, L. and Wright, K.I. (1992) Prehistoric environment and settlement in the Azraq Basin: an interim report on the 1989 excavation season. Levant 24, 1–31. Banning, E.B. (1998) The Neolithic period: triumphs of architecture, agriculture and art. Near Eastern Archaeology 61, 188–237. Bar-Yosef, O. (2001) Lithics and the social geographic configurations identifying Neolithic tribes in the Levant. Pp. 437–48, in I. Caneva, C. Lemorini, D. Zampetti and P. Biagi (eds), Beyond Tools. Berlin: Ex Oriente. — and Alon, D. (1988) Nahal Hemar Cave. Jerusalem: Atiqot 18, Israel Department of Antiquities. Bar-Yosef-Mayer, D. (1989) Late Paleolithic and Neolithic marine shells in the southern Levant as cultural markers. Pp. 169–74 in C.F. Hayes (ed.), Shell Bead Conference. Rochester, New York: Rochester Museum and Science Center. — (1991) Changes in the selection of marine shells from the Natufian to the Neolithic. Pp. 629–36 in O. Bar-Yosef and F. Valla (eds), The Natufian Culture in the Levant. Ann Arbor: Prehistory Press. — (1997) Neolithic shell bead production in Sinai. Journal of Archaeological Science 24, 97–111. — Porat, N., Gal, Z., Shalem, D. and Smithline, H. (2004) Steatite beads at Peqi’in: long distance trade and pyro-technology during the Chalcolithic of the Levant. Journal of Archaeological Science 31, 493–502. Wright et al. Barjous, M.O. (1986) The Geology of Siwaqa. Map Sheet 3252 IV (1:50,000). Amman, Jordan: Jordan Natural Resources Authority, Geology Division, Geological Bulletin 4. Barnes, R. and Eicher, J.B. (1992) Dress And Gender: Making and Meaning in Cultural Contexts. Oxford: Berg. Barthelmy de Saizieu, B. and Bouquillon, A. (1994) Steatite working at Mehrgarh during the Neolithic and Chalcolithic periods: quantitative distribution, characterization of material and manufacturing processes. Pp. 47–59 in A. Parpola and P. Koskikallio (eds), South Asian Archaeology 1993, Volume I. Proceedings of the Twelfth International Conference of the European Association of South Asian Archaeologists, Helsinki University 5–9 July 1993. Helsinki: Suomalainen Tiedeakatemia. Bender, F. (1974) Geology of Jordan. Berlin: Gebrüder Borntraeger. Berna, F. (1995) La lavarazione dell’amazzonite nel villagio neolitico di Jebel Ragref (Giordania meridionale). L’ecologia del Quaternario 17, 41–54. Betts, A.V.G. (1987) Recent discoveries relating to the Neolithic periods in eastern Jordan. Pp. 225–30 in A. Hadidi (ed.), Studies in the History and Archaeology of Jordan III. Amman: Department of Antiquities of Jordan. — (1998) The Harra and the Hamad. Sheffield: Sheffield Academic Press. Bienert, H. and Gebel, H. (1998) Archaeological investigations at Late PPNB Ba’ja: a preliminary report on the 1997 season. Annual of the Department of Antiquities of Jordan 42, 75–90. Bourdieu, P. (1977) Outline of a Theory of Practice. Cambridge: Cambridge University Press. Byrd, B.F. (1994) Public & private, domestic & corporate: the emergence of the village in southwest Asia. American Antiquity 59, 639–66. — (2005) Early Village Life at Beidha, Jordan: Neolithic Spatial Organization and Vernacular Architecture. Oxford: Oxford University Press and the Council for British Research in the Levant. Calley, S. (1989) L’atelier de fabrication de perles de Kumartepe: quelques observations technologiques. Anatolica 16, 157–84. Cauvin, J. (1974) Troisième campagne de fouilles à Mureybet (Syrie), 1973: rapport préliminaire. Annales Archéologiques Arabes Syriennes 24, 47–58. Cessford, C. and Mitrovic, S. (2005) Heavy residue analysis. Pp. 45–64 in I. Hodder (ed.), Excavations at Çatalhöyük, Volume 5. Changing Materialities at Çatalhöyük: reports from the 1995–1999 seasons. Cambridge and London: Monographs of the McDonald Institute for Archaeological Research, University of Cambridge; British Institute for Archaeology at Ankara. Costin, C.L. (1991) Craft specialization: issues in defining, documenting, and explaining the organization of production. Archaeological Method and Theory 3, 1–56. Critchley, P. (2000) Stone Bead Production at Wadi Jilat 25, a Neolithic Site in Eastern Jordan. MA dissertation, Institute of Archaeology, University College London. — (2007) The stone beads. In B. Finlayson and S. Mithen (eds), The early Prehistory of Wadi Faynan, southern Jordan: Excavations at the Pre-Pottery Neolithic A site of Wadi Faynan 16 and Archaeological Survey of Wadis Faynan, Ghuweir and al-Bustan. Oxford: Oxbow Press for the Council for British Research in the Levant. Diamanti, J. (2003) Beads, trade and cultural change. Pp. 8–13 in J. Lankton (ed.), A Bead Timeline, Volume I: Prehistory to 1200 CE. Washington, DC: The Bead Society of Greater Washington. Dobres, M.A. (2000) Technology and Social Agency: Outlining a Practice Framework for Archaeology. Oxford: Blackwell. and Hoffmann, C.R. (1999) The Social Dynamics of Technology. Oxford: Blackwell. Dubin, L. (1995) The History of Beads, 30,000 B.C. to the Present. London: Thames and Hudson. Dunnell, R.C. and Stein, J.K. (1989) Theoretical issues in the interpretation of microartifacts. Geoarchaeology 4, 31–42. Eicher, J. (1995) Dress and Ethnicity: Change Across Space and Time. Oxford: Berg. Fabiano, M., Berna, F. and Borzatti von Löwenstern, E. (2004) PrePottery Neolithic amazonite bead workshops in southern Jordan. Pp. 265–73 in Sécretariat du Congrès (ed.), Le Néolithique au Proche-Orient et en Europe. The Neolithic in the Near East and Europe. Acts of the XIVth UNESCO International Congress of Stone Bead Technologies Prehistoric and Protohistoric Sciences, 208 September 2001, Université de Liège. Oxford: BAR International Series 1303, Archaeopress. Finlayson, B. and Betts, A.V.G. (1990) Functional analysis of chipped stone artifacts from the Late Neolithic site of Jebel Na’ja, eastern Jordan. Paléorient 16/2, 13–20. Fino, N. (1998) Al-Basit Neolithic site in southern Jordan. Annual of the Department of Antiquities of Jordan 42, 103–11. Foreman, R. (1978) Disc beads: production by primitive techniques. Bead Journal 3, 17–22. Garfinkel, Y. (1987) Bead manufacture on the Pre-Pottery Neolithic B site of Yiftahel. Mitekufat Haeven, Journal of the Israel Prehistoric Society 20, 79–90. Garrard, A.N. (1998) Environment and cultural adaptations in the Azraq Basin, 24,000–7000 B.P. Pp. 139–48 in D.O. Henry (ed.), The Prehistoric Archaeology of Jordan. Oxford: BAR International Series 705. — (in preparation) Beyond the Fertile Crescent: Late Palaeolithic and Neolithic Communities of the Jordanian Steppe (4 volumes). Oxford: Council for British Research in the Levant, Levant Supplementary Series, Oxbow Books. — Baird, D., Colledge, S., Martin, L. and Wright, K.I. (1994a) Prehistoric environment and settlement in the Azraq Basin: an interim report on the 1987 and 1988 excavation seasons. Levant 26, 73–109. — Betts, A., Byrd, B.F. and Hunt, C. (1987) Prehistoric environment and settlement in the Azraq Basin: an interim report on the 1985 excavation season. Levant 19, 5–25. — and Byrd, B.F. (1992) New dimensions to the Epipalaeolithic of the Wadi el-Jilat in Central Jordan. Paléorient 18/1, 47–62. — Byrd, B.F. and Baird, D. (1994b) The chronological basis and significance of the late Palaeolithic and Neolithic sequence in the Azraq Basin, Jordan. Pp. 177–200, in O. Bar-Yosef and R.S. Kra (eds), Late Quaternary Chronology and Palaeoclimates in the Eastern Mediterranean. Tucson: Radiocarbon. — Byrd, B.F. and Betts, A. (1986) Prehistoric environment and settlement in the Azraq Basin: an interim report on the 1984 excavation season. Levant 18, 5–24. — Colledge, S. and Martin, L. (1996) The emergence of cultivation and pastoralism in the ’marginal zone’ of the southern Levant. Pp. 204–26 in D. Harris (ed.), The Origins and Spread of Agriculture and Pastoralism in Eurasia. London: UCL Press. Gopher, A. (1997) Ground stone tools and other stone objects from Netiv Hagdud. Pp. 151–76, in O. Bar-Yosef and A. Gopher (eds), An Early Neolithic Village in the Jordan Valley. Cambridge, Massachusetts: Peabody Museum of Archaeology and Ethnology, Harvard University. Gorelick, L. and Gwinnett, A.J. (1990) Innovative lapidary craft techniques in Neolithic Jarmo. Archaeomaterials 4, 25–32. Goring-Morris, A.N. and Gopher, A. (1983) Nahal Issaron: a Neolithic settlement in the southern Negev. Israel Exploration Journal 33, 149–61. Gross, S. (1977) The Geology of the Hatrurim Formation, Israel. Jerusalem: Geological Survey of Israel, Bulletin 70. Gwinnett, A.J. and Gorelick, L. (1989) Evidence for mass production polishing in ancient bead manufacture. Archaeomaterials 3, 163– 68. — (1999) A brief history of drills and drilling. Beads: Journal of the Society of Bead Researchers 10–11, 49–56. Hamilton, N. (2005) The beads. Pp. 325–32 in I. Hodder (ed.), Excavations at Çatalhöyük, Volume 5. Changing Materialities at Çatalhöyük: reports from the 1995–1999 seasons. Cambridge and London: Monographs of the McDonald Institute for Archaeological Research, University of Cambridge; British Institute for Archaeology at Ankara. Hauptmann, A. (2004) ’Greenstones’ from Basta: their mineralogical composition and possible provenance. Pp. 169–76 in H.J. Nissen, M. Muheisen and H.G. Gebel (eds), Basta I: The Human Ecology. Berlin: Ex Oriente. Inizan, M., Jazim, M. and Mermier, F. (1992) L’artisanat de la cornaline au Yémen: premières données. Techniques et Culture 20, 155–74. Jackson, B. (2005) Bead material identification. Pp. 373–76 in I. Hodder (ed.), Excavations at Çatalhöyük, Volume 5. Changing Materialities at Çatalhöyük: reports from the 1995–1999 seasons. Levant 2008 VOL 40 NO 2 163 Wright et al. Stone Bead Technologies Cambridge and London: Monographs of the McDonald Institute for Archaeological Research, University of Cambridge; British Institute for Archaeology at Ankara. Jaser, D.A. (1986) The Geology of Khan ez-Zabib. Map Sheet 3253 III (1:50,000). Amman, Jordan: Jordan Natural Resources Authority, Geology Division, Bulletin No. 3. Jensen, C.H. (2004) Production areas at MPPNB Shkarat Msaied, southern Jordan. Neo-Lithics 2/04, 22–25. Kaliszan, L., Hermansen, B.D., Jensen, C.H., Skuldbol, T., Bille, M., Bangsgaard, P., Ihr, A., Sorensen, M.L. and Markussen, B. (2002) Shaqarat Mazyad: the village on the edge. Neo-Lithics 1/02, 16–19. Kenoyer, J.M. (1986) The Indus bead industry: contributions to bead technology. Ornament 10, 18–21. — (1992a) Craft specialization and the question of urban segregation and stratification. Eastern Anthropologist 45, 39–54. — (1992b) Lapis lazuli beadmaking in Afghanistan and Pakistan. Ornament 15, 70–87. — (1994) Experimental studies of Indus Valley technology at Harappa. Pp. 345–62 in A. Parpola and P. Koskikallio (eds), South Asian Archaeology 1993, Volume I. Proceedings of the Twelfth International Conference of the European Association of South Asian Archaeologists, Helsinki University 5–9 July 1993. Helsinki: Suomalainen Tiedeakatemia. — (2003) The technology of stone beads: bead and pendant making techniques. Pp. 14–19 in J. Lankton (ed.), A Bead Timeline, Volume I: Prehistory to 1200 CE. Washington, DC: The Bead Society of Greater Washington. — Vidale, M. and Bhan, K. (1991) Contemporary stone beadmaking in Khambat, India: patterns of craft specialization and organization of production as reflected in the archaeological record. World Archaeology 23, 44–63. Khoury, H. and Nassir, S. (1982) A discussion on the origin of Daba’ Marble. Dirasaat, Engineering Sciences 9, 54–66. Kirkbride, D. (1966) Five seasons at the Pre-Pottery Neolithic village of Beidha in Jordan: a summary. Palestine Exploration Quarterly 98, 5–72. Kuijt, I. and Goring-Morris, A.N. (2002) Foraging, farming and social complexity in the pre-pottery Neolithic of the southern Levant: a review and synthesis. Journal of World Prehistory 16, 361–439. Larson, P. (1978) Ornamental beads from the Late Natufian of Southern Israel. Journal of Field Archaeology 5, 120–21. Maréchal, C. (1991) Elements de parure de la fin du Natoufien. Pp. 589–612 in O. Bar-Yosef and F. Valla (eds), The Natufian Culture in the Levant. Ann Arbor: Prehistory Press. Martin, L. (1999) Mammal remains from the eastern Jordanian Neolithic and the nature of caprine herding in the steppe. Paléorient 25 (2), 87–104. Mellaart, J. (1967) Catal Huyuk: A Neolithic Town in Anatolia. London: Thames and Hudson. Meskell, L. (2001) Archaeologies of identity. Pp. 187–213 in I. Hodder (ed.), Archaeological Theory Today. Cambridge: Polity Press. Moh’d, B.K. (2000) The Geology of Irbid and Ash Shuna ash Shamaliyya (Waqqas). Map Sheets 3155 II and 3155 III (1:50,000). Amman, Jordan: Jordan Natural Resources Authority, Geology Division, Geological Bulletin 46. Moholy-Nagy, H. (1983) Jarmo artefacts of pecked and ground stone and of shell. Pp. 290–346 in L. Braidwood (ed.), Prehistoric Archaeology along the Zagros Flanks. Chicago: University of Chicago Press, Oriental Institute Publications 105. Moore, A.M.T. (2000) Stone and other artefacts. Pp. 165–87 in A.M.T. Moore, G.C. Hillman and A. Legge (eds), Village on the Euphrates: From Foraging to Farming at Abu Hureyra. Oxford: Oxford University Press. Nassir, S. and Khoury, H. (1982) Geology, mineralogy and petrology of Daba’ Marble, Jordan. Dirasaat, Engineering Sciences 9, 107– 40. Pigeot, N. (1990) Technical & social actors: flintknapping specialists & apprentices at Magdalenian Etiolles. Archaeological Review from Cambridge 9, 126–41. Possehl, G. (1981) Cambay beadmaking: an ancient craft in modern India. Expedition 23, 39. Powell, J.H. (1988) Stratigraphy and Sedimentation of the Phanerozoic Rocks in Central and Southern Jordan. Part B Kurnub, Ajlun and Belqa Groups. Amman, Jordan: Jordan Natural Resources Authority, Geology Division, Geological Bulletin 11. 164 Levant 2008 VOL 40 NO 2 — (2006) Personal communication, at British Geological Survey, Keyworth, Nottingham, United Kingdom. Quintero, L., Rollefson, G.O. and Wilke, P. (2004) Highland towns and desert settlements: origins of nomadic pastoralism in the Jordanian Neolithic. Pp. 201–13 in H. Bienert, H. Gebel and R. Neef (eds), Central Settlements in Neolithic Jordan. Berlin: Ex Oriente. Quintero, L. and Wilke, P. (1998) Evolution and economic significance of naviform core and blade technology in the southern Levant. Paléorient 21 (1), 17–33. Rainville, L. (2005) The Organization of Domestic Activities in Upper Mesopotamian Households and Neighborhoods during the Early Bronze Age: A Micro-Archaeological and Architectural Approach. Oxford: British Archaeological Reports International Series 1368. Reese, D. (1991) Marine shells in the Levant: Upper Palaeolithic, Epipalaeolithic and Neolithic. Pp. 613–28 in O. Bar-Yosef and F. Valla (eds), The Natufian Culture in the Levant. Ann Arbor: Prehistory Press. Rice, P. (1991) Specialization, standardization and diversity: a retrospect. Pp. 257–84 in R. Bishop and F. Lange (eds), The Ceramic Legacy of Anna O. Shepard. Boulder: University of Colorado Press. Richter, T., Colledge, S., Luddy, S., Jones, D., Jones, M., Maher, L. and Kelly, R. (2008) (in press) Preliminary report on the 2006 season at Epipalaeolithic ’Ain Qasiyah, Azraq esh- Shishan. Annual of the Department of Antiquities of Jordan 51. Rollefson, G.O. (2002) Beadmaking tools from LPPNB al-Basit, Jordan. Neo-Lithics 2/02, 5–7. — Kafafi, Z. and Simmons, A. (1990) The Neolithic village of Ain Ghazal, Jordan: preliminary report on the 1988 season. Bulletin of the American Schools of Oriental Research Supplement 27, 95–116. — and Parker, M.C. (2002) Craft specialization at al-Basit, Wadi Musa, southern Jordan. Neo-Lithics 1/02, 21–23. — Quintero, L. and Wilke, P. (1999) Bawwab al-Ghazal: preliminary report on the 1998 testing season. Neo-Lithics 1/99, 2–4. — and Simmons, A. (1984) The 1983 Season at ’Ain Ghazal: Preliminary Report. Annual of the Department of Antiquities of Jordan 28, 13–30. — and Simmons, A. (1986) The 1985 Season at ’Ain Ghazal: Preliminary Report. Annual of the Department of Antiquities of Jordan 30, 41–56. Rosen, S.A. (1997) Lithics after the Stone Age: A Handbook of Stone Tools from the Levant. Walnut Creek, CA: AltaMira Press. — (2003) Early multi-resource nomadism: excavations at the Camel Site in the central Negev. Antiquity 77 (298), 749–60. Roux, V., Bril, B. and Dietrich, G. (1995) Skills and learning difficulties involved in stoneknapping: the case of stone bead knapping in Khambat, India. World Archaeology 27, 63–87. Roux, V. and Matarasso, P. (1999) Crafts and the evolution of complex societies: new methodologies for modelling the organization of production, a Harappan example. Pp. 46–70 in M.A. Dobres and C.R. Hoffmann (eds), The Social Dynamics of Technology. Oxford: Blackwell. Schmidt, K. (2000) Göbeklitepe, Southeastern Turkey: A Preliminary Report on the 1995–1999 Excavations. Paléorient 26, 45–54. Schumann, W. (1992) Rocks, Minerals and Gemstones. London: HarperCollins. Sciama, L.D. and Eicher, J. (1998) Beads and Bead Makers: Gender, Material Culture and Meaning. Oxford and New York: Berg. Simmons, A. and Najjar, M. (1998) Al-Ghuwayr 1, a pre-pottery Neolithic village in WadiFaynan, southern Jordan: a preliminary report on the 1996–1997 seasons. Annual of the Department of Antiquities of Jordan 42, 91–102. Sorensen, M.L.S. (1997) Reading dress: the construction of social categories and identities in Bronze Age Europe. Journal of European Archaeology 5, 93–114. Starck, J. (1988) Stone rings from Baga and Basta: geographical and chronological implications. Pp. 137–73 in A. Garrard and H. Gebel (eds), The Prehistory of Jordan: The State of Research in 1986. Oxford: British Archaeological Reports International Series 396. Stein, G.J. (1996) Producers, patrons, and prestige: craft specialists and emergent élites in Mesopotamia from 5500–3100 BC. In V.G. Childe and B. Wailes (eds), Craft Specialization and Social Evolution: in Memory of V. Gordon Childe. Philadelphia: University of Pennsylvania Museum of Archaeology and Anthroplogy. Wright et al. Stocks, D. (1989) Ancient factory mass-production techniques: indications of large-scale stone bead manufacture during the Egyptian New Kingdom Period. Antiquity 63, 526–31. Stout, D. (2002) Skill and cognition in stone tool production: a case study from Irian Jaya. Current Anthropology 45, 693–722. Talbot, G. (1983) Appendix K: beads and pendants from the tell and tombs. Pp. 788–801 in K.M. Kenyon and T. Holland (eds), Jericho V. London: British School of Archaeology Jerusalem. Tosi, M. and Vidale, M. (1990) Fourth millennium BC lapis lazuli working at Mehrgarh, Pakistan. Paléorient 16/2, 89–99. Treherne, P. (1995) The warrior’s beauty: the masculine body and self identity in Bronze Age Europe. Journal of European Archaeology 3, 105–44. Valla, F., Khalaily, H., Valladas, H., Tinerat-Laborde, N., Samuelian, N., Bocquentin, F., Rabinovich, R., Bridault, A., Simmons, T., Dosseur, G., Miller-Rosen, A., DuBreuil, L., Bar-Yosef Mayer, D. and Belfer-Cohen, A. (2004) Les fouilles à Mallaha en 2000 et 2001: troisième rapport préliminaire. Journal of the Israel Prehistoric Society 34, 49–244. Vanzetti, A. and Vidale, M. (1994) Formation processes of beads: defining different levels of craft skill among the early beadmakers of Mehrgarh. Pp. 763–76 in A. Parpola and P. Koskikallio (eds), South Asian Archaeology 1993, Volume II. Proceedings of the Twelfth International Conference of the European Association of South Asian Archaeologists, Helsinki University 5–9 July 1993. Helsinki: Suomalainen Tiedeakatemia. Verhoeven, M. (2002) Transformations of society: the changing role of ritual and symbolism in the PPNB and the PN in the Levant, Syria and southeast Anatolia. Paléorient 28, 5–14. Vianello, F. (1985) La lavorazione dell’amazzonite a Wadi Hafir, Giordania meridionale. Studi per l’Ecologia del Quaternario 7, 77–122. Vidale, M. (1989) Specialized producers and urban élites: on the role of craft industries in mature Harappan urban centres. Pp. 171–81 in J.M. Kenoyer (ed.), Old Problems and New Perspectives in the Archaeology of South Asia. Madison: University of Wisconsin. Stone Bead Technologies — Kenoyer, J.M. and Bhan, K. (1992) A discussion of the concept of the chaı̂ne opératoire in the study of complex societies. Pp. 181–94 in A. Gallay (ed.), Ethnoarchéologie: justification, problèmes, limites. Juan-les-Pins. Wheeler, M. (1983) Appendix J: greenstone amulets. Pp. 781–87 in K.M. Kenyon and T. Holland (eds), Jericho V. London: British School of Archaeology Jerusalem. Wright, K.I. (1992) Ground Stone Assemblage Variations and Subsistence Strategies in the Levant, 22,000 to 5500 b.p. PhD dissertation, Department of Anthropology (University Microfilms International), Yale University. — (1993) Early Holocene ground stone assemblages in the Levant. Levant 25, 93–111. — (2006) Stone bead technology studies at Çatalhöyük, summer 2006. http://www.catalhoyuk.com/archive_reports/2006/. Çatalhöyük Website Archive Reports 2006. — (forthcoming) Households, lineages and networks: a framework for understanding Near Eastern social organization. In K.I. Wright (ed.), The Ancient Levant: A Social Archaeology. London: in preparation. — (in preparation) Beads and the body: ornament technologies of Building 3 at Çatalhöyük. In R. Tringham and M. Stevanovic (eds), Çatalhöyük Building 3: the Excavations of the University of California at Berkeley at Çatalhöyük (BACH). Los Angeles: Monographs of the Cotsen Institute of Archaeology, University of California at Los Angeles. — and Bains, R. (2007) Stone bead technology at Çatalhöyük, summer 2007. http://www.catalhoyuk.com/archive_reports/2007/. Çatalhöyük Website Archive Reports 2007. and Garrard, A.N. (2003) Social identities and the expansion of stone beadmaking in Neolithic western Asia: new evidence from Jordan. Antiquity 77 (296), 267–84. Wright, P. (1982) The bow-drill and the drilling of beads, Kabul, 1981. Afghan Studies 3–4, 95–101. Levant 2008 VOL 40 NO 2 165
© Copyright 2026 Paperzz