1.1. General remarks

4The Klimonas macrolithic tools presented here come from the Communal building (St 10), Building 800 and the building remains of Sector B (Vigne et al. 2017). These different contexts correspond mainly to the PPNA phase of the site. Material contemporaneous with the Neolithic Sotira phase or found out of stratigraphic context is not examined in this report. The macrolithic tools from Klimonas have been categorised into 19 typo-morphological classes: querns, different types of handstones, different types of hammerstones/pounders, anvils, mortars and pestles, work surface slabs and various other types of tools associated with the technical and domestic domain (appendix 16-1).

5The study of macrolithic tools from St 10 and its immediate surroundings, excavated between 2011 and 2012, included a total of 217 whole or fragmentary pieces. This structure is characterised by a succession of different building phases and is composed of stratified deposits and semi-subterranean features (chap. 5).

6The study of macrolithic tools from Building 800 (Sector F), excavated in 2014 and 2015, involved a total of 90 pieces corresponding to various categories in terms of typology and raw materials.

7Finally, in Sector B, excavated between 2015 and 2017 and characterised by a series of buildings, 580 tools (fig. 16-1) were found. The distribution of tools in Sector B varies considerably depending on buildings and their state of preservation. We counted between 1 and 111 tools per building depending on the state of preservation of buildings (for some, only the foundation trench subsisted, appendix 16-1).

Fig. 16-1 – Graph presenting the types of tools found within the different buildings of Klimonas.

8The state of preservation of the tools is relatively poor in general, since it is marked by a very high fragmentation rate: 49.5% of the pieces are fractured. This rate varies from one type of tool to another, but does not concern percussion tools subjected to violent shocks causing them to break. In fact, only a few grinding tools were found whole. As grinding tools (such as querns and handstones) are among the most resistant elements, they tend to have a long service life. The high degree of fragmentation here must thus imply that these tools were used for a prolonged period of time and/or that they were used in percussion posée diffuse (diffuse dormant percussion or passive percussion). The tools often bear concretions, which can impede analyses. Some tools were cleaned with hydrochloric acid diluted with water, in part +/- 2/100, in order to remove concretions from certain stone tools of magmatic origin. This cleaning was necessary for us to understand the initial function of these tools and classify them accordingly.

9At first glance, the macrolithic tools from Klimonas present a wide functional diversity, but this should not mask the over-representation of handstones and hammerstones/pounders. They alone represent more than 63.8% of the assemblage. This dichotomy may therefore correspond to tools used on a daily basis, as opposed to those used for more occasional practices.

1.2. Raw materials

10The raw materials used for the macrolithic tools at Klimonas correspond to siliceous (flint, chalcedony and jasper), sedimentary (limestone and sandstone) and magmatic rocks (diabase, grano-diorite, gabbro and microgabbro). The different petrographic categories making up the corpus are represented in variable proportions, ranging from one to several hundred specimens. Flint dominates, making up more than a third of the assemblage. Five other rock types appear in significant numbers: gabbro, basalt, jasper, chalcedony and microgabbro. We should also mention diabase, a rock used in more than 20 cases. The other petrographic types are present in marginal proportions but underline the diversity of the supply and exploitation of a wide petrographic spectrum. Tools were made from several types of rock forms (blocks, pebbles, cores), which gives an indication of the geological environment.

11At first sight, it appears that the materials utilised by the Klimonas inhabitants come from local sources. The pebbles come from fluvial environments or ancient terraces. Blocks, slabs, and platelets derive from rock outcrops, which, in some cases, may have required quarrying. Lastly, some macrolithic tools were made from cores and flint blocks, reduced beforehand for the manufacture of chipped stone. The raw materials for these macrolithic tools were brought to the site by the inhabitants of Klimonas. Pebbles from fluvial or marine formations are found in similar proportions to materials collected from local rock outcrops (58.1 and 41.9%, respectively). The pebbles incorporate a wide spectrum of raw materials, as a result of the diversity of geological contexts in the foothills of the Troodos Massif and the coastal area. The macrolithic tools made out of magmatic rocks are almost exclusively from Troodos: diabase, granite, gabbro and microgabbro. The Amathous River located less than 4 km to the west of the site, and the sea shore, located less than 1.5 km to the south, are likely to be the main sources of raw materials. Sedimentary rocks, including siliceous materials, are present in the immediate vicinity of the site. The materials used for the grinding and milling tools (querns and handstones) show a rigorous selection, including specific rocks and morphologies. They are exclusively composed of magmatic rocks. This selection is certainly associated with a search for very precise mechanical properties adapted to the intended functions. Some tools used in percussion lancée (handheld or active percussion), such as pecking tools, are also made from specific materials, for the same reason.

1.3. Querns and handstones

12This typological class groups together passive (querns) and active (handstones) tools used for grinding and pulverising different materials, with a longitudinal, circular or random movement. The total number of querns and handstones is 207, including fragments. Many handstone fragments could not be strictly correlated to a typological category. This uncertainty is due to the almost systematic fragmentation of handstones, which unfortunately limits comparisons and detailed morphometric analyses.

1.3.1. Querns

13All 36 querns found at the site were fragmentary, and morphological characterisation was thus difficult. One example, however, is 27 cm wide and 18.2 cm thick. Its length is estimated at over 40 cm. Its active part is 20 cm wide, with a concave longitudinal and transversal curvature, indicating the use of a handstone shorter than the width of the quern (fig. 16-2, A). The preferred rock for querns is gabbro, represented by 34 fragments. Two other quern fragments are in basalt. These materials are granular. For querns, the sought-after raw material properties for optimal efficiency are roughness, which is essential for grinding, medium granular consistency and medium hardness. If the rock is too hard, it tends to smooth out quickly and must be re-pecked more often, which reduces the life span of the tool. On the other hand, if the rock is too soft, numerous rock particles are mixed in with the ground substances (Schoumacker 1993, Santallier et al. 2002).

Fig. 16-2 – A: quern (KL 2015, Bo 37); B: type 1 handstone with ochre residue; C: type 2 handstone (St 120); D: small handstone (rubber) (St 6927).

1.4. Hammerstones/pounders

22A total of 369 hammerstones/pounders were found at Klimonas, subdivided into six types depending on morphology, raw material, the location of percussion marks, and the proportion of the surface used. The group of pounders represents more than 43.6% of all tools.

1.4.1. Irregular flint hammerstones

23These hammerstones (N = 75 whole and 37 fragments) are of flint blocks (N =109) and more rarely of chalcedony (N =3), and are used on their edges. Use-wear can be concentrated in a small area, as limited as less than 25% of the total available surface, or can cover up to 75% of the surface, indicating more pronounced use (fig. 16-3, A). Use-wear traces consist mainly of crushing and small chips in the rock, resulting from impact. These tools were used in punctiform handheld percussion (percussion lancée), on hard materials. The repetitive use of a specific area of the hammerstone modified its morphology, rounding off any angles. The distribution of traces of use on the active parts depends on the type and duration of use. Any impact against hard material causes accidental fractures, and most of the tools bear evidence of scars. Many of these hammerstones were broken at various stages, and in 20 cases only a shard of the hammerstone survived. Average dimensions of the 75 whole specimens vary between 4.1 cm and 10.3 cm in length, with an average length of 6.0 cm; from 3.1 to 7.4 cm in width, with an average width of 5.0; from 2.1 to 7 cm in height, with an average of 3.9 cm. The average weight is 145 g.

Fig. 16-3 – A: irregular flint hammerstone (St 802); B: core-based hammerstone (SU 6105); C: spherical hammerstone (ISOB_200); D: pebble hammerstone (St 6981); E : grooved hammerstone (SU 10.3); F: pestle fragment (SU 10.10).

1.5. Pestles and mortars

1.5.2. Mortars

31At Klimonas, only one type of mortar was found, the cupmark (fig. 16-4, A-B). Four cupmarks were found in total, all made on blocks of rock. One of them is a re-used quern which is also the only fragmentary cupmark; the remaining three are whole. The blocks of rock used for the cupmarks are ovoid to sub-rectangular shape in plan. The three whole blocks are in hard limestone; the fragmentary re-used quern is in gabbro. Measurements are shown in tab. 16-2. The morphology of the cupmarks does not seem to be related to the type of pestles found at the site.

Fig. 16-4 – A: cupmark (Sector B, St 6812); B: cupmark (Sector B - TB2).

Tab. 16-2 – Metrical characteristic of three Klimonas cupmarks.

Length (cm)	55	44	55
Width (cm)	26	35	20
Thickness (cm)	232	21	19
Cup lenght (cm)	12.8	13.4	11.4
Cup width (cm)	13	13.3	11.4
Cup depth (cm)	6.5	6	4.5
Cup volume (ml)	340	350	200

1.6. Spherical tools

1.6.1. Spherical plutonic-igneous tools

32At Klimonas, 21 spherical tools were found, made from diabase, granite, gabbro or microgabbro. Their diameter varies between 3.4 and 8.7 cm, with an average of 5.3 cm. Their weight ranges from 65 to 846 g, with an average of 323.6 g. They were shaped by pecking in punctiform percussion with another hard stone tool (fig. 16-5, A). They bear no other use-wear traces and their function is unclear.

Fig. 16-5 – A, B: spherical basalt tools (SU 6082) (6668); C, D: spherical limestone tools (St 7104); E, F: sling stones (St 6668 and SU 6125); G, H: retouchers/compressors (SU 604 and St 6846).

1.7. Retouchers/compressors

35A total of 114 retouchers/compressors (92 complete and 22 fragments) were found at Klimonas, made of small oblong pebbles, about 10 cm long, 3 to 5 cm wide and 1.1 to 2.5 cm thick. These elongated tools bear one or more working facets. They are characterised by traces of impacts in the form of small striations and, more rarely, traces of light percussion. These facets are always located at the end and on the edge of the object. On a few specimens, we also observe the presence of small scraping and abrasion striations on the periphery of these facets (fig. 16-5, G). The raw material appears to be a fine-grained basalt or gabbro, with a small abrasive grain. The function of these stone implements has been associated with retouching flint tools, in particular arrow heads, which are numerous at Klimonas (chap. 14 and 15; see also Robitaille and Briois 2019). Five pebbles with similar morphology and dimensions to those described here presented no use-wear (fig. 16-5, H), suggesting that they might have been collected for future use.

1.8. Limestone slabs and anvil

36The limestone slabs (N = 2 whole and 21 fragments) are rectangular or square, flat blocks of limestone with an average thickness of 4.6 cm. As most of them are fragments, no morphological typology was identified. Eight of them bear traces of ochre on one of their sides. They were especially abundant in the initial fillings of building B08, where important concentrations of ochre were also found (chap. 7).

37Another four limestone slabs (2 whole; 2 fragments) were clearly used as anvils (fig. 16-6, A). These anvils are characterised by concentrated percussion marks on one or both sides, forming a slightly pitted depression. In the same functional category, we also identified three other anvils made from limestone pebbles, characterised by successive localised marks (fig. 16-6, B). The qualities of the limestone raw material, which appears to be rather friable for use as anvils, suggest that these tools were not used over a long period of time.

Fig. 16-6 – A: limestone slab, B: anvil (St 6766); C: ochre tablets (St 11).

1.9. Ochre tablets

38Ochre tablets (N = 6 fragments) are large pebble fragments of gabbro, with a similar morphology to type 1 handstones, but with larger surfaces. These tools seem to have served as passive supports for the grinding or preparation of ochre (fig. 16-6, C), and all bear traces of ochre on one face. This passive use does not seem to be associated with an active tool such as a handstone; they could have been used as a work surface for processing the ochre into a paste using water.

1.10. General trends in the Klimonas macrolithic tool set

39In summary, it can be seen from figure 16-1 that the numbers of macrolithic tools in 19 typo-morphological categories vary between 2 and 171 specimens per category. Tools used for percussion, such as flint hammerstones and hard stone hammerstones, are predominant, with 368 tools, representing 42.6% of the assemblage. In second place come grinding tools, which include querns, various handstones and undetermined handstone fragments (type 1 or 2 handstones), totalling 222 elements, or 25.6% of the assemblage. These are followed by retouchers/compressors, represented by 114 tools, i.e., 13.1% of the assemblage. The rest of the categories are all represented by numbers of tools varying between 3 and 30.

40The typo-morphological analysis allows us to differentiate tool types and define a classification of the Klimonas tool set adapted to the Cypriot PPNA. Functional analysis will further clarify the tools’ precise functions. The typo-morphological analysis showed that macrolithic tools are relatively standardised across the site. Disparities appear in the acquisition and selection of raw materials: grinding and milling tools (querns and handstones) underwent more rigorous selection for medium- to coarse-grained raw materials. This selection is clearly associated with a search for very specific mechanical properties, adapted to the anticipated use of these tools. For similar reasons, percussion tools, such as spherical hammerstones, are also made of specific materials. Careful shaping further standardises these tools and implies additional investment in tool design. However, only grinding tools and grooved hammerstones underwent an intentional manufacturing process: the tools are pre-formed by knapping and/or regularised by hammering out the original roughness. Moreover, in the case of handstones and querns, the active surface requires preparatory pecking before use. The efficiency of the finished tool is dependent on this technical shaping, which requires precise know-how. In addition, tools used with rough surfaces, for example, retouchers and pebble hammerstones, require rigorous selection as they must correspond to precise morphological or prehensile criteria.

41These different indices of opportunism or standardisation are indicative of a certain organisation of tool types. This assumption is based on the assessment of the efforts involved in the selection and acquisition of raw materials and the skills required for the manufacture, maintenance and use of tools. Grinding tools are technologically enhanced by the highest degree of rigour in all phases of design and the chaîne opératoire. Nonetheless, many of the other less technically complex tools are still carefully designed. Spherical hammerstones, slingshots, retouchers, may have been opportunistically collected, but were still selected for their raw material mechanical properties and/or their natural morphology and were perfectly adapted to the accomplishment of many everyday Neolithic craft tasks.

2.1. Overview of the available data and trends in Southwest Asia

42We compared the macrolithic assemblage from Klimonas with the Near Eastern continental macrolithic industry in order to assess the representativeness of the assemblage. In this region, the PPNA is characterised by a clear increase and diversification of macrolithic tools (fig. 16-7) in comparison to earlier time periods, and especially the Late Natufian (Wright 1994). In 35 Early Natufian sites, 49% contained macrolithic tools (Wright 1991, 1993, 1994), with pestles and mortars being the most common. The pestle-mortar pair represents 76% of the macrolithic tool assemblage from 17 sites, while handstones-querns total 17.6% (Wright 1993). The ratio of grinding tools (querns and handstones) to crushing tools (pestles and mortars) is 1/3.8 at Mallaha II-IV, 1/2.5 at Wadi Hammeh 27 and 1/6.3 at Hayonim Cave B (Wright 1991, 1993). In the Late Natufian, 23 of the 47, 49% of the observed sites included macrolithic tools (Wright 1991, 1992). The pestle and mortar pair is the most frequent and mortars are frequent in the bedrock. Mortars represent 78% of the tools in the macrolithic assemblages. There is also a significant increase in the frequency of handstones and querns compared to the Early Natufian, where they are present at 39% of sites (Wright 1991, 1992). In a comparison of Khiamian and PPNA industries, Wright’s results indicate that 70.8% of the 24 studied sites contained macrolithic tools, representing a 20% increase compared to the Late Natufian. This study shows that stone mortars are less represented: 59% of PPNA assemblages yielded mortars, compared to 70–78% of Natufian and Harifian assemblages (Wright 1994). PPNA assemblages show a clear increase in the frequency of grinding tools – handstones and querns – present in 65% of the sites (Lechevallier 1978, Dorrell 1983, Wright 1993, 2000, Gopher and Orrelle 1995, Commenge 1997, Rosenberg et al. 2008, Rosenberg 2013). Of the 17 PPNA assemblages studied by Wright (1994), handstones were present at 11 sites, representing 65% of the assemblages. This proportion is higher than for any other time period, and the increase is statistically significant compared to Natufian sites in general (Wright 1994).

Fig. 16-7 – Graph presenting percentages of sites which have produced macrolithic tools dating in three different time periods: Natufian, PPNA and PPNB, (red line). Percentage of tools used for grinding (querns, handstones) in different time periods (dark grey). Percentage of tools used in percussion (pestles, mortars) for different time periods (light grey).

43This increase in the frequency of grinding and milling tools is thought to be related to an economic transformation in Neolithic societies. The lithic tools made and used by early farmers reflect important socio-economic changes marking the origins of agriculture. The innovations of this time period include tools designed for woodworking, harvesting and craft production. These changes are accompanied by new social and symbolic values. One of the main characteristics of PPNA tools of the Near East is the reduction in size and weight of grinding tools (querns and handstones) and especially crushing tools (mortars and pestles), compared to earlier time periods (Rosenberg 2004). This phenomenon is associated with a shift from communal to individual tool use (Wright 1991, Belfer-Cohen and Goring-Morris 2002). In the PPNA, tools are no longer used in open communal spaces but rather in individual dwellings (Belfer-Cohen and Hover 2005). It has been suggested that the pestle and mortar pair may be related to processing nuts or acorns, while the quern and handstone pair may be related to processing small seeds (Moore 1985, p. 13, Goring-Morris 1987, p. 439). This argument reflects an assumption that querns and handstones are “better” tools for processing grain (Kraybill 1977, p. 514).

44Other innovations are rather characteristic of the PPNA. The most notable is the location of cupmarks (Mithen et al. 2005, p. 106) also known as the cup-holed mortar. This type of tool is already known from the Natufian and is also found in PPNB assemblages. One of its characteristics in a PPNA context is its presence inside buildings (Noy 1989, Rosenberg and Nadel 2011b), in contrast to previous time periods. In the PPNA, varying frequencies of polished axes, grooved stones, weights, and drilled discs are found (Lechevallier 1978, Dorrell 1983, Wright 1993, 2000, Gopher and Orrelle 1995, Commenge 1997, Rosenberg et al. 2008, Rosenberg, 2013). Mallets, maze heads, chisels and polishers have also been found at some sites. Macrolithic tools are characterised by a greater diversity of tool types and raw materials, which may be related to the development of the craft industry. These trends continue into the Early and Late PPNB (Wright 1993, p. 97-105). Of the 41 Early and Middle PPNB sites observed, 71% contained macrolithic tools (Wright 1993). Grinding tools (querns and handstones) are present at 55% of these sites, while crushing tools (pestles and mortars) are only present at 25% (Wright 1994). For the Late PPNB, 58% of the 44 sites recorded have grinding tools, compared to a 30% for crushing tools (Wright 1994).

2.2. Klimonas tools in their PPNA Context

45We compared the macrolithic assemblage from Klimonas with those from mainland PPNA sites, based on published data. However, it is important to highlight various issues that arose from this comparison. Firstly, there were huge disparities in the quantities of pieces found at different sites. Some sites, such as Çayӧnü and Jericho, yielded an impressive number of tools whereas other sites only yielded a few. Secondly, the available data is very heterogeneous – the tools available for analysis did not generally include all tool types, and percussion tools were rarely described. We must therefore remain cautious and consider that the comparisons proposed here still need to be consolidated by more comprehensive data. Only a few detailed macrolithic tool publications are available, for sites such as Jericho (Kenyon et al. 1981), Mureybet (Nierlé 1983, 2008), Sheikh Hassan (Kozłowsky 2001) and Çayӧnü (Davis 1982), based on a typology specific to every site. Important information is also available for the sites of Gilgal (Rosenberg and Gopher 2010) and Hatula (Noy 1979, Lechevallier and Ronen 1989), and some data is provided for Tell Qaramel (Mazurowski 2002, 2004) and Tzur Natan (Marder et al. 2007). In tab. 16-3 we present the different tools found at several sites.

46At first sight, querns and handstones on one hand, and mortars and cupmarks on the other predominate in the compared assemblages (60 to 90%, appendix 16-1). Mallets, groove stones, axes, limestone tablets, chisels, maze heads and polishers were found in varying frequencies on these sites. Hammerstones are only sparsely represented. Stone retouchers are only found at Klimonas (Robitaille and Briois 2019). Comparing the Klimonas macrolithic assemblage to those from continental PPNA sites reveals significant disparity between the pestles and mortars on the one hand, and the grinding and milling tools (querns and handstones) on the other. The rate of milling and crushing tools at Klimonas (ca. 30%) appears to be low in comparison to the continent. However, at Klimonas, we took into account the very high number of hammerstones in the assemblage. Continental PPNA published assemblages do not generally include hammerstones. This is either because of publication constraints or research bias. If we exclude hammerstones from the Klimonas assemblage the milling and crushing tool percentage rises to 53%, which is comparable to findings on the mainland. We investigated whether the ratio between these pairs of grinding and milling tools, and those used for crushing, could be correlated to the environment, specific use, or the processing of different materials. We propose an ethnographic and experimental approach to their possible functional interpretation.

Tab. 16-3 – Types of tools found at various PPNA sites. Spherical stones* At Çayӧnü different types of tools have been classified in this category (bolas, hammerstones, weights, etc). Hammerstones** In Klimonas 157 tools were interpreted as hammerstones (we have included in this category the flint, chalcedony and quartz hammerstones). Thirty-four tools were classified as limestone and basalt spherical tools.

3.1. Pestles and mortars

47It should be recalled that mortars and pestles were the first stone tools used for daily food production in the Levant and they have long been seen as a central element in the development of early agricultural societies in Southwest Asia. They were first recognised in the Kebarian period, but their numbers clearly increased during the Natufian. Smaller-sized pestles and mortars are then attested to for the various Neolithic periods (Wright 1991, p. 19-22). The heaviest mortars are found among Natufian pounding tools, which have no Neolithic equivalents (Perrot 1966, Bar-Yosef 1983, p. 19, Belfer-Cohen 1988a, Valla et al. 1991, Gopher and Orrelle 1995, Gopher 1996, Dubreuil 2002). The pestle-mortar pair has been reported to have been used for husking, crushing, or pounding foodstuffs and other materials, which are very different activities from grinding in percussion posée (Seeden 1982, p. 57, Marfoe 1986, p. 79).

48In the PPNA, the percentages of grinding tools (querns and handstones) used in percussion posée, and those of crushing tools (pestles, mortars, in situ mortars, and cupmarks), used in percussion lancée, vary from one site to another (tab. 16-3). In the Upper Euphrates, except for Tell Qaramel (excavation 2002, squares L-4B, D/M-4A and C, and excavation 2004, squares K-3 B, K-4 B, D, L-3 A, L-4 A and C), where they only represent 30% of the assemblage, grinding tools are generally well represented. They reach 96% at Abu Hureyra, 72% at Sheikh Hassan and 57% at Mureybet. The same is true for the site of Çayӧnü, in Anatolia, where they represent 90% of the assemblage. In contrast, in the Jordan Valley, at the sites of Bir el-Maksur, Gigal I-III, Jericho and Khiam 5-4, and in Israel, at the sites of Hatoula and Tzur Natan, they represent less than 50% of the assemblages. Compared to the other analysed sites, Klimonas shows a very high percentage of tools used in percussion posée (94% of querns and handstones), as opposed to 5.8% of crushing tools in percussion lancée (pestles and mortars).

49We also observed a disparate ratio between pestles and mortars (mortars, in situ mortars and cupmarks). For example, at Çayӧnü, it is 6:1; at El-Khiam 5-4, 10:1; at Gigal I-III, 4.6:1; and at Jericho, 7:1. At Deir Waraq (El-Khanasry), despite the large number of pestles found, only one mortar was discovered. Ghanimeh and Qadi (2010, p. 149) suggest that this disparity may be related to a high frequency of mortars built directly on the bedrock and possibly not discovered by archaeologists. The use of now perished wooden mortars or bowls is also possible. Ethnographic observations among various Amerindian groups reveal that condiments can be ground using a wooden container and a stone pestle (Fowler and Liljeblad 1986). We also observed the use of wooden mortars operated with a stone or wooden pestle among the Konso in Ethiopia (Robitaille 2021). It was initially suggested that this pestle-mortar pair was related to nut processing (Moore 1985, Goring-Morris 1987). The archaeological evidence suggests, however, that the small pestles found at PPNA sites may have been multifunctional or reused. At Mureybet, some examples showed traces of ochre (Nierlé 2008), and at Abu Hureyra, traces of both ochre and chalk were observed (Moore et al. 2000, p. 171).

50Although the function of pestles from this time period may have been predominantly utilitarian, the stylistic characteristics of some tools found at different PPNA sites suggest that some of them may also have been used for ceremonial purposes. Ceremonial pestles found in burials typically display stylised ends representing the heads of birds or wild goats. Pestles of this type have been found in Anatolia, at Hallan Çemi (Rosenberg 1999), at Çayönü (Davis 1982, Özdoğan 1999), at Nemrik 9 (Kozłowski 1989), and at Körtik Tepe (Özkaya 2009). This type of tool has been typologically classified as follows. At Çayӧnü, Davis (1982, p. 77-80) divides them into 11 types, according to weight, shape and raw material. At Klimonas, no comparable small pestles to those of the mainland PPNA have been found, only those corresponding to type 1 of Davis’ classification (1982, p. 81), namely “large, cylindrical to conical-shaped pestles”. Their morphology corresponds to heavy tools associated with mortars deep enough to grind hard materials or large volumes.

51In a context of cereal husking and crushing various elements, the presence of the pestle-mortar pair should be more represented at Klimonas. Only four cupmarks, and no other kind of mortar have been found so far at Klimonas, along with only eight pestle fragments. The cupmark is found at most PPNA sites (tab. 16-4). At Tzur Natan, mortars are rare, and Marder et al. (2007) suggest that the cupmarks found were used as mortars for pounding activities. According to Rosenberg and Nadel (2011a), PPNA cupmarks are smaller and more regular in shape than those of the Natufian. In the PPNA, these tools were associated with a food processing area. At Dhra’, residential structures with about ten food processing areas have been identified, characterised by standing earthen or stone walls and earthen floors with cupmarks and grinding stones (Kuijt and Mahasneh 1998). The cupmarks are always located in the middle of a flat stone or in the centre of a grinding stone. This is also often the case at Mureybet, which also features the reuse of an old tool as a cupmark. On the examples from this site, the surface of the limestone used for a cupmark was re-pecked. Cupmarks found at other sites generally have a cup placed in the centre of a stone, the surface of which was not worked. Although this tool appeared to be uni-functional (Rosenberg and Nadel 2011a), a quern–mortar and cupmark from Mureybet contained traces of ochre. Another quern–mortar pair contained a shiny black material, suggesting either a preparation of black pigment or resin, or its use as a lantern (Nierlé 2008, p. 549). Mortars in the soil or bedrock have also been found at various sites (Samzun 1994). According to Davis (1982) and Moore et al. (2000), the rarity of large mortars at Çayӧnü and Abu Hureyra may indicate the use of stone pestles on wooden mortars. Ethnographic observations indicate that cereal grains are traditionally dehulled using wooden mortars and pestles (Gast 1968, Hillman 1985, Robitaille 2021).

Tab. 16-4 – Descriptions and types of cupmark found at various PPNA sites.

52Generally, before milling, several successive operations are carried out to free the grains from their various husks (Anderson 1992). According to Hillman (1985), after harvesting, the first task is to detach the stalks from the spikelets by threshing. At this stage, crops are sometimes stored. For glume cereals, such as emmer wheat (Triticum dicoccum/dicoccoides) and barley (Hordeum spontaneum/distichon), identified at Klimonas from impressions on raw earth (Vigne et al. 2012, 2017b; chap. 23), a second threshing is required to separate the grain from the chaff (Hillman 1985). This second operation is usually carried out by pounding, using a pestle and mortar. The limited number of pestles and cupmarks at Klimonas, along with the absence of mortars may either indicate the use of now-perished wooden counterparts, or that the hulling operation was carried out with the same tools as those used for milling, namely the quern and handstone pair. It is also possible to consider other modes of hulling involving the use of other types of tools and techniques. Ethnographic observations in north-western Nepal have shown that shelling can be carried out with a wooden pestle on a slab instead of a mortar (Baudais and Lunström-Baudais 2002, p. 162). The use of bedrock for this purpose should also be considered. It is possible that hulling/pounding areas were common to several families, or even to an entire village. We have also observed the absence of a pestle-mortar among different agropastoral groups in the Omo Valley of Ethiopia, who developed other crushing techniques (Robitaille 2021). For example, among the Mursi of Maki, the back of the handstone on its quern is used for crushing cereals, sorghum and soaked maize. Among the Hamar, grain is crushed by hammering on a skin, using a flat stone striker similar to a small hammerstone (Robitaille 2015). Other ethnographic observations mention the partial hulling of glume cereals. Those of Procopiou (1998), on processing barley rusks in Crete, show that the human consumption of glume cereals with partial hulling is quite feasible.

53In light of the apparent absence of mortars and pestles in different contexts, we experimentally evaluated the efficiency of stone tools, handstones and querns, for grain hulling. According to our experiments (Robitaille 2012) and those carried out by Procopiou (1998, p. 132), hulling with the use of a handstone and quern is more efficient after soaking the seeds in water than after drying or roasting them. The grain tends to remain more intact than when dry hulling, which also requires several sieving operations to remove multiple shells. It is therefore possible to imagine that several methods and tools may have been used for crushing and hulling grains.

3.2. Grinding and milling equipment

54The quern-handstone pair was present at all the PPNA sites (Wright 1991, 1992, 1994) considered in our analysis and, as previously mentioned, their numbers increased significantly at all PPNA sites in comparison to the Natufian (Lechevallier 1978, Dorrell 1983, Wright 1993, 2000, Gopher and Orrelle 1995, Commenge 1997, Rosenberg et al. 2008, Rosenberg 2013). The ratio of querns to handstones, compared to that of the tools used in percussion lancée, varies from site to site. For example, at Gigal I-III and Hatula, querns and handstones represent 22% and 16% of grinding and crushing tools, respectively, while at Abu Hureyra and Çayӧnü their representation is 90% and 90.3% (fig. 16-8). At Jericho and Gigall I-III, handstones represent 89.7% and 88.3% of the quern–handstone pair, respectively.

Fig. 16-8 – Graph presenting the percentage of querns and handstones (dark grey) and pestles and mortars (light grey) from different PPNA sites.

55Ethnographic examples illustrate the primary use of querns and handstones for grinding cereals in a domestic setting, with the presence of querns and handstones of a different type for grinding softer and oilier vegetation (Roux 1986, p. 38, Baudais and Lundström-Baudais 2002; Robitaille 2021). However, this is not always the case. Observations among the Mursi and Dorze in Ethiopia show that various condiments (herbs, salt), wild seeds and tubercles can also be ground with cereal querns (Robitaille 2021). With regard to the continental PPNA, various sites show that querns and handstones may have been multipurpose or reused. Traces of ochre on grinding stones have been identified at Ain Mallaha (Perrot 1966, p. 466), Abu Hureyra (Moore et al. 2000, p. 167) and Mureybet (Nierlé 2008, p. 565-567). At the end of their life cycle, tools were also reused for other functions. At Çayӧnü, fragments of grinding stones were reused to make other tools, for example pestles or small worktables. Other reuses have also been described, such as paving stones, door-socket stones or building materials (Moore 1985, p. 98, Stordeur 2015). Re-use of cereal querns and handstones for other purposes has also been observed in many ethnographic cases. For example, the Dorze and Konso in Ethiopia reuse grinding querns and handstones for crushing and grinding grog and earth for pottery making. Among the Hamar in Ethiopia, old grinding querns are reused for pulverising ochre and grinding incense for body decoration. Among the Mursi in Ethiopia, sorghum and maize are also crushed using querns and handstones (Robitaille 2021). Multifunctional or reused tools are opportunistic. Considering our PPNA examples, it is difficult to distinguish between multifunction or reuse. For example, the different traces of ochre observed on handstones can be characterised as an example of either multifunction or reuse. However, ethnographic examples show that if a tool is in use for a primary function, and is then used for another/second function, this second function must not deform the tool used for the primary function, given that it will be used again for that primary function. In contrast, reused tools change function and transform their active surface, for example anvils from (former) handstones. Ethnographic observations (Robitaille 2021) show that even if food processing equipment is multifunctional (used for crushing and/or grinding), it is only used for food processing. In other words, food processing equipment would not be used for crushing or grinding minerals. Considering therefore the traces of ochre on PPNA handstones, by analogy, we suggest that these handstones were re-used rather than multifunctional.

3.2.1. Querns

56With regard to the Klimonas querns, the microtopography of eight analysed quern surfaces seems to indicate two types of use (fig. 16-9). Firstly, all the querns show a levelling of the grain surface and visible wear polish. At higher magnification, the grains are characterised by micro-fragmentation. The levelling of the grain is quite even and has a matt appearance. For three of the quern fragments, a matt polish and directional striations are perceptible on the levelling of the grains (fig. 16-9, A). For five of the analysed grinding quern surfaces, the grains are less fractured and slight extraction and fragmentation was observed. The grains were slightly blunt and jointed (fig. 16-9, B). The second type of traces seems to result from grinding softer material. No active surfaces bear traces or residues of ochre or other minerals. One quern was reused as a cupmark.

Fig. 16-9 – Different active surfaces of querns presenting diversity of utilisation (enlargement × 10-15).

3.2.2. Handstones

57At Klimonas, handstones appeared to play an important role in daily activities at the site, making up nearly a quarter of the macrolithic tool assemblage. The high fragmentation rate of these tools indicates intensive use and complex management of the materials used, including maintenance and recycling phases. To understand the over-representation of handstones (79%) compared to querns (appendix 16-1), we carried out an initial functional analysis of 48 whole and fragmentary tools in order to assess their function. This over-representation is also observed at different sites in the Near East, where only morpho-functional classification has been carried out so far. At Abu Hureyra, most handstones are made from basalt or greenstone pebbles, and their morphology is generally oblong and, more rarely, rectangular in plan. They are oval or plano-convex in cross-section and vary in length from 9 to 22 cm. According to Moore et al. (2000, p.169), morphological criteria reflect functional differences. At Çayönü, more than one hundred complete handstones, mostly in basalt, were divided into 15 types and 4 subtypes. This typological diversity reflects the variability of functional or stylistic characteristics: weight, size and shape of the working surface. According to Davis (1982), the size of the active surface and weight are important factors for tool efficiency. Lastly, the general shape of the active surface may also reflect functional or stylistic concerns.

58Pecking is visible on nearly 60% of the two types of handstone found at Klimonas. Pecking rejuvenates the abrasiveness of tool surfaces and is associated with grinding cereals. However, these tools may not have been exclusively used for grinding cereals (fig. 16-2). The presence of a crushed mineral material corresponding to ochre was noted on 15 active handstone faces, characterised by striations or deposits, and corresponding to 31% of the studied material. A few examples of pecked handstones bore traces of ochre. One handstone seems to have been used for processing soft material, possibly animal skin. Two were reused as hammerstones. Four were reused as anvils. The active surfaces were prepared for grinding cereals by pecking to regularise them and give them the desired abrasiveness. Traces of this preparation are visible on most of the active surfaces, except where extensive use had erased them. The use-wear produced by grinding cereal, where it survived, is a more or less dense polish (fig. 16-10, A).

Fig. 16-10 – Different active surfaces of querns presenting diversity of utilisation (enlargement × 10-15).

59At Klimonas, a handstone fragment used for processing skins was also identified. The piece is a type 1 ovoid tool with homogeneous invasive wear and completely smoothed edges. The bluntness of the edges suggests a slight rotation during handling or contact with a soft material. The clearly visible lustre on this basalt handstone is the main characteristic of hide-working tools. The worked material is flexible and perfectly follows the shape of the tool and its microreliefs (fig. 16-10, B). This handstone was also used for processing ochre. Two types of ochre can be found only a few kilometres from Klimonas: a hard ochre requiring crushing and grinding, and a chalky ochre that can be easily dissolved and ground with water on a handstone or tablet. This second type seems to have been preferred at Klimonas. Ochre residues have been found in the centre of some handstones and on handstone fragments. Only a small area of the handstones was used at a time, indicating that a limited amount of ochre was generally ground. The handstones used for processing ochre seem to have functioned as passive percussion tools or as small worktables and their use for ochre processing appears to be occasional. Use-wear is not always perceptible, and only the ochre residues attest to this function (fig. 16-10, C). Ochre was also observed on a cereal handstone (fig. 16-10, D). Due to grinding hard ochre, linear traces are visible on the active surface of the tool, where the grains are deformed (fig. 16-10, E). The entire active surface was used, indicating grinding on another hard surface. Here, the handstone was the active tool.

6019.2% of the handstones were used on both faces. The thickness of the tool does not seem to be a determining factor for the use of both faces. Four type 1 handstones in gabbro or basalt were reused as anvils. Anvils were used as percussion tools operating in a passive percussion lancée mode. Traces are located in the centre of the accessible surface, forming a slight pitted depression. These depressions display a rough surface and a circular shape with a diameter of a few centimetres. The transverse morphology of the impacted area is lenticular in shape, and its depth is generally a few millimetres. Reused unifacial or bifacial handstones (some with traces of ochre), comparable in shape to those found in Klimonas have been recorded at Çayӧnü (Davis 1982).

61The multifunctionality and high rate of handstone reuse at Klimonas illustrates the importance of handstones as grinding tools as much for food processing as for craft activities.

3.3. Hammerstones

62Hammerstones form such a general category of tools that a whole range of terms has been adopted by specialists to define this group of percussion tools. According to M.-L. Inizan et al. (1995, p. 155), the hammerstone or striker is a natural hammer (“le percuteur est un marteau naturel”), which can be used to hammer, knap, shape, or retouch hard stone. A striker can be a pebble or a block of stone, a fragment of bone, antler or wood. Inizan’s (1995) terms “hard hammerstone”, for natural mineral hammers, and “soft hammerstone”, for hammers of animal or plant material, are commonly used. According to Archambault de Beaune (1997, 2000), the hammerstone is most often a regularly-shaped ovoid, oblong or spherical pebble. If it is a block, it always has blunt or rounded corners. It is used in percussion lancée and shows the consequent traces: more or less profuse and extensive impacts, crushing, and sometimes the removal of protruding parts (Archambault de Beaune 1997, p. 8, 2000, p. 50-51). Terminological disagreements concerning this category underline deeper classification difficulties. The category of percussion tools is associated with the often occasional and, by extension, random nature of the uses of these tools. Therefore, they are often dismissed, resulting in the frequent under-representation of percussion tools (striker, pecking stone (boucharde), hammerstone) in the analyses of ground stone tools. Additionally, distinctive criteria are often too difficult to perceive for a stricter classification of the archaeological material. Moreover, the random or alternately percussive or passive nature of these tools adds to categorisation difficulties (Hamon 2004, p. 51). Furthermore, the shape of the percussion tool and the traces on the tool depend mainly on the raw material used. As experiments show, strikers used to make chipped stone out of flint blocks show very different marks depending on the type and material of the striker (Leuvrey 1999, p. 39). According to our ethnographic observations (Robitaille 2021), there is a clear distinction between strikers used mainly on hard rocks (e.g., for surface re-pecking); hammering tools for making tools (e.g., to make a quern), and strikers for knapping (e.g., for making blades or arrowheads).

63In our view, a striker (hammerstone) is a pebble or stone used in active percussion, in order to transform another object. This can be flint to be knapped, an active surface of a grinding stone to be rejuvenated (pecked), or a bone to be fragmented. We exclude from this category tools with similar morphologies but used in passive percussion, for example with the aim of reducing ochre to powder or grinding condiments. The type of stone, its shape and size reflect functional differences. For example, crushing tools are larger and generally produce much more violent impacts. It is difficult to confuse pecking stones (boucharde) and shaping tools, as the impact marks on the ends or edges of the latter are much more intense, with clear granule removal modifying the very shape and volume of a striker (hammer). These reflections are also inspired by the diversity of strikers (hammers) used to produce grinding querns in ethnographic contexts (Hayden 1990, Robitaille 2015, 2021). The weight of the tools and the morphology of the active zones vary according to the efficiency of a tool in making another tool. The use of a large striker (hammer) is necessary to shape a block of rock, whereas a smaller tool would be used for finishing and levelling the active surface (Hayden 1990, Robitaille 2021).

64At Klimonas, we divided the hammerstones (strikers) into three sub-categories, based on their morphology, raw material, the proportion of the tool used, and the concentration and location of the impact (use-wear) traces: core and block hammerstones; spheroidal hammerstones in flint or other siliceous rock; and pebble hammerstones.

3.4. Spherical plutonic-igneous and limestone tools

71Spherical basalt tools have been found at several PPNA sites. At Klimonas, one tool was found in association with a conical artefact (“figurine” or “phallic symbol” in Building 800, St 857; Vigne et al. 2017b, fig. 6). Several others were found in built structures (chap. 21). An important feature is the intact nature of these spherical tools: almost all (N = 20 out of 21) are whole, making them the most complete category of artefacts. At Agia Varvara-Asprokremmos, this type of tool was also found in association with a statuette (personal communication C. McCartney). At Çayӧnü, 34 whole tools and 8 fragments, measuring between 2 and 7.1 cm and weighing between 10 and 450 g, were found (Davis 1982). Basalt spherical objects are also reported at Körtik Tepe (Özkaya 2009). No functional or symbolic interpretation has yet been proposed for any of these tools. At Klimonas, some of them bear traces of polish; others only show traces of fabrication by pecking.

72At Klimonas, another 13 small spherical objects, similar to the spherical basalt tool, were made of limestone. Their diameter varies from 2.4 to 5 cm and their weight from 25 to 198 g. This type of object has been found at various PPNA sites. At Mureybet, 13 similar tools have been excavated. Their diameter varies from 1.5 to 4.9 cm and their weight from less than 2 g to 100 g (Nierlé 2008). At Çayӧnü, several spherical objects made of hard or soft limestone were found: 82 are made of soft limestone, ranging from 2 to 6.8 cm in diameter and weighing from 10 to 260 g; 91 were made of hard limestone, ranging from 2.2 to 7.1 cm in diameter and weighing from 15 to 485 g (Davis 1982). At Abu Hureyra, nine of the same type of objects have been described. Their diameter is less than 4 cm and their weight varies from 40 to 101 g (Moore et al. 2000). Similar tools also exist at Gilgal (Rosenberg and Gopher 2010). The shaping of this type of tool appears to be variable from site to site. At Mureybet, these tools were shaped by abrasion and scraping with a sharp tool and traces of use were the result of more or less strong and repeated percussion over the entire surface of the sphere. The only case with no percussion traces shows three non-complete perforation points with no particular orientation (Nierlé 2008). At Çayӧnü, the smaller examples have a smooth surface and the larger ones bear percussion marks (Davis 1982). The regularity in shape and size of these objects points to their use as a unit of measurement, token, game piece or projectile. At Abu Hureyra, these small spherical objects can be divided into three types on the basis of morphology, size and weight: 40 g, 71–80 g, and 101 g. According to Moore et al. (2000), the size of the tools is too small to correspond to hammerstones, and the craftspeople seemed to try to reproduce the same weights, possibly for weighing small quantities of medicinal plants (Moore et al. 2000).

3.5. Retoucher–compressors

3.5.2. First category

75This category comprises 52 small elongated pebbles, with one or two active areas on one or both ends. Length ranges between 4.5 and 8.7 cm, with an average of 6.3 cm; width from 1.4 to 3.6 cm, with an average of 2.5 cm; and thickness from 0.8 and 1.9 cm, with an average of 1.2 cm. Their average weight is 34.2 g. The morphology of these active areas varies from small striations to complete deformation, depending on the extent and intensity of use. On these active surfaces, localised striations were observed, often within a slight depression. The latter is characterised by traces of smaller striations, always located at the end and on the edge of the tool. One or both edges of these pebbles were deformed to the extent that their edge was reduced to a thinner extremity (tip). On some tools, we also noticed the presence of small striations, located on the periphery of these well-defined active areas (fig. 16-11, A - box a, B - box c). Pressure must have been applied on the localised areas to produce these traces. The location of the wear marks seems to indicate use either by a left-handed or a right-handed person, showing that each of these pebbles was utilized by a single user.

Fig. 16-11 – A: Tool (SU 604): percussion flaking and pressure flaking technique, (a) close-up of active areas; B: Tool (St 838), percussion flaking and pressure flaking technique, (b-c) close-up of active areas; C: Tool (SU 10.3), pressure flaking technique, (d) close-up of active areas; D, E: Tools (Sector B), percussion flaking (e) close-up of active area.