Traceological analysis of lithics from the Camel Site, al-Jawf, Saudi Arabia: an experimental approach to identifying mineral processing activities using silcrete tools

The Camel Site is in the north of Saudi Arabia in the province of al-Jawf. It is characterised by three decaying sandstone hillocks with life-sized 3D engravings (or reliefs) of camels and equids likely carved during later prehistory. A survey in the central area of the site identified clusters of flakes and other flintknapping remains in the lower areas between the sandstone spurs and larger silcrete tools directly underneath the animal depictions. Some of these tools presented abraded edges, possibly from prolonged contact with the soft and abrasive sandstone that constitutes the rock spurs where the animals were carved. Experiments were performed to test this hypothesis and have a reference collection for further traceological analysis. The chaine opératoire of the experimental engraving tools, from raw material procurement, tool manufacture and use, reuse and discard, was conducted with locally available materials comparable to the archaeological specimens. Specific experimental variables, including how the force was applied, in what direction the movement took place and the orientation of the stone tool during the experiment, were also recorded. Macro-and microscopic analyses of the experimental collection and a sample of archaeological artefacts seem to show that the ancient tools found on the surface were probably used to make the camelid and equid reliefs at the site.

Traceological analysis of lithics from the Camel Site, al-Jawf, Saudi Arabia: an experimental approach to identifying mineral processing activities using silcrete tools Yamandú Hilbert, Ignacio Clemente-Conte, Rémy Crassard, Guillaume

Introduction
This study describes an exploratory experiment aiming to identify wear traces on different silcrete stone tools used to process sandstone.We will present the results of the subsequent traceological analysis conducted on the experimental tools and compare the traces with an archaeological sample from the Camel Site et al.-Jawf in northern Saudi Arabia.The rationale behind this experiment and subsequent traceological analysis involves the use of stone tools to create rock art engravings.The use of stone tools in the carving of engravings, portable motifs in the form of plaquettes or other symbolic expressions, have been theorised and described (Leroi-Gourhan and Allain 1979;Mélard et al. 2016;Moro Abadía and González Morales 2013;Plisson 2007).We will focus our discussion on the carving of sandstone cliffs and immovable boulders, albeit mobile art made on hard organic and inorganic materials such as antler, bone, clay and ivory along with engravings on schist, jet stone (and other artefacts types), are found throughout prehistory across all continents.
1 3 Analyses designed to identify mineral processing activities have been conducted sporadically since the advent of the traceological methods, and some researchers have included limited numbers of experimental pieces used to modify varying types of minerals in their experimental reference collections (e.g.Clemente-Conte 1997;González-Urquijo and Ibañez-Estéves 1994;Keeley 1980;Mansur-Franchomme 1986).Specific experimental and traceological studies aiming to identify modification caused by mineral processing, however, are rare (Alvarez et al. 2001;Alvarez and Fiore 1995;de Beaune and Pinçon 2001;Hamon et al. 2021;López-Tascón et al. 2018;2020).These previous studies have focused on either limestone or schist supports transformed by pebbles, quartzite or flint experimental tools.
Here, we present the first study concentrated on sandstone carving using tools made from silcretes; two minerals commonly used throughout prehistory across the Arabian Peninsula and beyond.Additionally, our experimental protocol focuses on the use of transverse motions instead of typical longitudinal movements used in classic engraving activities.This focus is a function of the specific archaeological case study addressed herein, namely, the 3D engravings depicting life-size camels and equids from the al-Jawf province in Saudi Arabia.

Background
The al-Jawf province and the immediate surroundings of the Camel Site have undergone surveys and are the subject of different archaeological missions operating in northwestern Saudi Arabia since the second half of the twentieth century (Adams et al., 1977;Parr et al., 1978, for a list see Charloux 2018).The oldest human occupations of the region are attributed to the Lower and Middle Palaeolithic, with some rare occurrences of Upper Palaeolithic materials found throughout the plateaus flanking the Wadi Sirhan fault and even within the fringes of the northern Nafud (Hilbert et al. 2017;Hilbert and Crassard 2020).Neolithic occurrences were found in the vicinity of the Camel Site and throughout the sandstone plateau west of Dumat al-Jandal.
The Camel Site (DAJ155) is characterised by three sandstone spurs that present a series of façades of different magnitudes at varying altitudes upon which zoomorphic reliefs of naturalistic proportions have been carved (Fig. 1); predominantly camels and occasionally wild equids are shown in realistic anatomical proportions with great attention to detail (Charloux et al. 2018;Guagnin et al. 2021).A recent literature review of the still burgeoning field of Arabian rock art study has identified over 50 sites located across the Arabian Peninsula, from Najrân in southern Saudi Arabia, up to Petra in Jordan, that present large-sized camel depictions (Charloux et al. 2020).These sites span a considerable chronological depth as well as a wide geographical area and were recently structured into different traditions of which the reliefs from the Camel Site, which depicts mostly male wild camels, are likely the oldest and possibly of a Late Neolithic age (Andreae et al. 2020;Guagnin et al. 2021).
The geology of the Camel Site and its immediate surroundings within the Sakaka basin is characterised by regolith surfaces, inselbergs and sandstone rock spurs of the late Cretaceous Wasia Group, also known as Sakaka sandstone (Powers et al. 1966;Wallace et al. 1997).The Sakaka sandstone has a primarily marine origin across northwestern Saudi Arabia.It is relatively soft and weathers rapidly by wind abrasion, causing exfoliation, subsequent pitting and the formation of vertical fissures, which instigate the decay of the rock spurs and the reliefs.The sandstone rock spurs on which the camel reliefs and other graphic depictions (encompassing recent graffiti and tribal symbols) have been made are severely deteriorated and eroded by wind abrasion leading to the decay and destruction of the sandstone spurs.The camel reliefs are in most cases only partially preserved, and fragments of the reliefs themselves or even larger blocks that have been carved into near to life-size naturalistic camel reliefs are found collapsed, laying on the low talus slopes; recent bulldozing of the area around the spurs has further deteriorated the archaeological site.
The analysis of the marks of the production of the panels from the Camel Site by Guagnin et al. (2021) indicated the use of percussive and abrasive actions, including scraping and pecking, possibly by using stone tools.The life-sized camel depictions show different preservation conditions, some parts being highly eroded while others maintain some forms of fine details and carvings.The reliefs identified in 30 individual panels found throughout the site were likely made in two separate stages of production: the first stage of manufacture would have encompassed the subtraction of a considerable amount of sandstone volume, particularly around the ventral and dorsal portions of the animals as well as around the extremities.The presence of unfinished reliefs that have been abandoned during the first stage (panel 4) provides a glimpse into how this phase was undertaken.Deep pecked holes have been identified that were likely made using stone tools (Fig. 2).The second phase incorporated the smoothing of the contours on the animal reliefs and the addition of further detail by adding fine lines to the eye, nostrils and mouth.
To shed additional light on the creation of this emblematic site, experimental and traceological analyses were conducted.Here, we report the results of our experiment designed to identify the carving activities using stone tools on sandstone supports.We use this experimental reference collection to identify and classify specific wear patterns of sandstone abrasion to bolster the traceological interpretation of a sample of silcrete artefacts we classify as engraving tools.

Material
A sample of 11 engraving tools found on the talus slope of the sandstone rock spur C that make up the Camel Site was selected for the analysis.These have been collected from the base of panels 11 and 30 and presented bevelled and abraded surfaces indicating their possible use in abrasive work resultant from scraping activities (Fig. 3).In addition to the artefacts collected from the base of the panels, excavations on the talus slope on the southwestern side of rock spurs B were undertaken.A test trench of 5 m 2 (Trench 1) was excavated, and lithics material, faunal remains and stone beads were recovered.Lithic artefacts are diminutive, and to a larger part of flake proportions, tools include different ad hoc retouched implements and a series of pressure-retouched transverse projectile armatures of triangular shape.Used raw materials include different variants of chert, flint and quartz.Based on typology and known parallels from the Levant and Southeastern Jordan, the transverse arrowheads fall between the late sixth to mid-fifth millennium BC (Betts et al. 2013;Rosen 1997).This chronology is supported by three radiocarbon samples taken on bioapatite providing ages falling into the mid-sixth millennium BC (Guagnin et al. 2021).Of the 706 lithic artefacts excavated from trench 1, a single unmodified piece of dark ferruginous sandstone, excavated from square 4 at a depth between 5 and 10 cm, presented what appeared to be lightly abraded edges and was therefore included in the traceological analysis.Archaeological remains from later periods are found in the vicinity of the site, and additional, younger engraved showing light weathered symbols attest to later occupations of the Camel Site.
The lack of suitable raw materials in the direct vicinity of the Camel Site is noteworthy; however, excavations and surveys activities have identified a considerable number of flaking remains, indicating an in situ production and potential use of lithics, that took place at the site (Fig. 4).Raw materials found were therefore brought to the site by human action.The analysed sample consists of three types of raw materials, broadly classified as silcrete, that can be differentiated macro-and microscopically from each other by colour, grains size and sedimentary matrix.Some of the artefacts can be classified as thick discoid cores showing a successive reduction of flakes from either core surface; others bear minimal amounts of modification leading to the formation of what could be interpreted as an active edge, while others were used without any modification.

Method
Excavation and survey methods followed modern standards of archaeological field praxis, excavations were conducted in 5-cm spits, all sediments were sieved, and the piece selected for traceological analysis was briefly rinsed with water and bagged separately.The position of the engraving tools found on the eastern slopes of spur C was recorded by GPS, each artefact was bagged separately to avoid damage, and all artefacts were rinsed with water to remove any loose sediment from their surfaces.Artefacts were exported to Spain with the permission of the local authorities and analysed at the microscopy laboratory of the working group: Archaeology An exploratory experiment was conducted to create a frame of reference for the modifications deriving from the transformative actions using specific variations of silcrete raw materials on the Wasia Group sandstone to classify tool motion and narrow the use alterations and exclude taphonomic origins of the observed traces on the archaeological samples.Raw materials used for the experiment were comparable to the archaeological specimens.They included 1 3 a dark silicified sandstone (RM1) variation, a red-orange and yellow-banded silcrete variant (RM2) and a dark ferruginous sandstone (RM3).These occur in different outcrops and show different properties.The raw materials for the experiment were collected at specific outcrops found within a 15-km radius of the site.In total, 13 experimental tools were made with the three raw materials using a direct hard hammer knapping technique; the flake and core tools were used during different intervals of either 1, 3 or 5 min averaging between 50 and 70 strokes per minute (Table 1).
The experimental tools were produced by RC and YHH using direct hard hammer percussion and included flakes, cores and natural chunks that were retouched to create a suitable working edge.Two large slabs of Sakaka sandstone collected from the immediate surroundings of the Camel Site were used to conduct the experiment.The specific kinetics of the movements used while wielding the experimental tools are summarised in Fig. 5.The applied force during the experiment was either by pressure or using direct or indirect percussion action using a stone hammer.Movements included transverse and longitudinal motions administered in either unidirectional or bidirectional fashion.All experiments were conducted without the addition of water to the sandstone.The experiment helped to identify abrasive tracks, directional markers, micro-chipping, micro-polish and sandstone residues on the archaeological tools.
Traceology builds on the use of microscopy to identify technological and functional related traces, a principle   1 3 quantitative methods of functional analysis are being explored (Caux et al. 2018;Galland et al. 2019;Ibáñez and Mazzucco 2021;Pedergnana et al. 2020).By examining the surfaces and edges on prehistoric stone tools using different magnifications, micro-negatives, striations, micro-polishes and other traces of abrasion are plotted on the artefacts.The different roughness of the surfaces, regularisation of surfaces by abrasion and their intensity are the function of time, materials transformed and specific kinetics of the productive activities undertaken.The use of different ways to apply force with the wielded tool, either by percussion or by pressure, leads to different patterns of micro-and macro-chipping on the active tool edges; here, specific working angles play a considerable role.For this particular study, low power magnification analyses were conducted using an Olympus SZX7 stereomicroscope to scan the edges and surfaces of the tool while micrographs were taken using a HIROX digital microscope.High power analysis was performed using an Olympus BH50 metallurgical upright microscope equipped with Nomarski prisms, and photographs were taken with a Canon EOS camera; micrographs were then combined using HeliconFocus software.

Taphonomic alterations of archaeological remains
Taphonomy greatly influences the quality and reliability of the identified functional patterns that lead to the interpretation of tool function and consequently play an essential role in the traceological analysis of archaeological samples (Levi Sala 1986;Levi-Sala 1988;Mansur-Franchomme 1986;Schiffer 1983).It particularly affects the quality of the results from lithics that have suffered from prolonged exposure to wind and sand abrasion while resting on exposed and deflated surfaces across arid to semi-arid environments.
The most common alteration observed on lithic artefacts from such contexts is labelled 'desert varnish' (Hunt 1954;Perry et al. 2005), which can be differentiated from gloss patination, dark gloss patination and so-called river patina (Howard 2002;Vaughan 1985).Different theories have been postulated as to the nature of these alterations (adsorption theory vs dissolution theory), incidentally, however, as Rottländer (1975) pointed out, the difficulties in experimentally replicating these alterations and the lack of a unified terminology of 'patination', hamper understanding the origins of these alterations.Recently, Schmidt et al. (2020) demonstrated using reflectance infrared microscopy that usewear polish on chert has its origins in a physical process (abrasion), making it possible to infer a similar origin for the formation of taphonomic surface modifications; nonetheless, additional research into the subject remains to be undertaken.
Artefacts from the Camel Site collected from the surface underneath the panels found on rock spur C show an ephemeral polish that coats all surfaces, possibly due to the mechanical abrasion of wind-born particles on all surfaces.These alterations form indiscriminately over the entirety of the surfaces of the lithics.They are seen on the elevated portions of the microtopography of the surfaces and within the depressions forming an evenly distributed light polish that covers both surfaces and edges.The primary source of taphonomic alteration is sand and wind; the productive properties this study focuses on are the transformation of a sandstone surface by extracting material using different modalities of abrasion and percussion.The experimental sample used to work and modify the sandstone surface allows us to better discriminate between the taphonomic alterations and the use related to modifications of the active edges of the tools (Fig. 6).
Due to the natural origin of these alterations, a chaotic distribution of both polished surfaces, scars and scratches can be differentiated from the traces (striations, polish micro-negatives, etc.) on the controlled and deliberately conducted anthropogenic actions of productive and transformative activities associated with tool use.This is seen when the alterations on the interior portions of the tools, namely, the ventral or dorsal surfaces of CS.ET#1 in Fig. 6.The taphonomic action upon the degradation of the crystals is seen on the unmodified natural surface and knapped surfaces of CS.ET#5 is evident by the damage and scarring on the micro-surfaces showing the different levels of alteration (see unaltered crystal grains on the fresh knapped surface of experimental tool CS.EXP.2 for comparison) (Fig. 6 top).

The experimental reference collection
The patterns of alteration on the experimental collection show a specific characteristic that makes it possible to distinguish the use-wear from the taphonomic alterations.The type of motion and kinetics used during the experiment, including percussion force and pressure, transverse and longitudinal movements, serve different purposes and show other various traces on both the worked material and the tools themselves.The main characteristic of all experimental tools used in the reference collections is the widening and abrasion of the working edges, which becomes highly regularised as work duration advances (Figs. 7, 8 and 9).The abrasive nature of the worked material (sandstone) causes a relatively fast regularisation of the working edges,   of contact with the sandstone.All types of motion and delivered force led to the fast development of these specific regularisations and broadening of the working edges on the tools.The formation of typical flat mineral polish, striations and abrasive tracks associated with the contact with the mineral material, however, was rare and related to elevated portions of the tools' working edge topography, as these can be best observed on the CS.EXP.04F1 and CS.EXP.04F2 in Fig. 8.While the used edges of the experimental samples are quickly abraded, specific micro-polish are mostly constrained to the eminences on the peripheries of the working edges' microtopography and rarely reach high levels of development due to the continuous abrasion.This, however, remains to be tested by prolonging the duration of use of the experimental specimens.
In addition to the mechanically induced wear traces on the tools' working edge, the depressions between the individual quartz crystal grains exhibited a high amount of mineral residues associated with the abrasive nature of the task performed.The red iron oxide-stained powder became adhered to the working edge and further catalysed the abrasion of the edges of the tools.These were found concentrated in the areas that functioned as working edges or active edges.

Traceological results of the archaeological material
Most engraving tools collected from the surface below rock spur C of the Camel Site have been classified as cores, which is to say that these served a different purpose at some stage and have been reused at different times leading to different weathering patterns on the assemblage.The engraving tools classified as cores, however, show different levels of development, some showing multiple centripetal removals on both dorsal and ventral faces, while other show minimal removal of flakes and appear to be specially manufactured to be used as tools.Some pieces showed only limited or no removals and were used without considerable modification.Except for two pieces that may have been used in percussive activities, longitudinal and transverse motions by the use of pressure prevail.Of the 12 analysed engraving tools, four presented macroscopic modifications comparable to the traces found on the experimental tools, by either transverse or longitudinal abrasion; the dimensions and specific traits of the archaeological engraving tools are summarised in Table 2.The four archaeological pieces presenting wear traces will be briefly discussed.
CS.ET#1 is made of red and yellow-banded silcrete (RM2) and shows a highly elaborated centripetal scar pattern, suggesting that the tool was used as a core to produce normalised flakes with predetermined shapes and sturdy straight edges.Pecking marks on the ventral convex surface may indicate the expedient use of the core in percussive activities; the low incidence of this feature, however,  may also suggest that the core was set on some form of anvil during its knapping.The artefact shows two abraded flat surfaces on the dorsal face, indicating that this particular face was facing the contacted material during the use of the tool (Fig. 9).A series of longitudinal striations on the active portion of the tool on the left side indicates a combination of longitudinal and transverse motions.The exertion of pressure during the work is evident by the presence of micro-negatives seen on the bottom of micrograph F2 in Fig. 10, located at the bottom of the image, while parallel scars are evident on the macro-images.
Tool CS.ET#5 is made on a thick brick-shaped piece of coarse-grained dark ferruginous sandstone (RM1) that shows two removals that formed the active edge.Additional removals are seen on the bottom part of the tool, but no use traces were identified in this area.Macroscopic abrasions are seen on different portions of the active edge, which has become broader where the tool made contact with the worked material.The rounded morphology of the abraded areas is comparable to the abrasion observed on the experimental pieces used in bidirectional and longitudinal motions, only much less developed.Signs of damage and crushing can be observed on the macro-image M1b in Fig. 11 that can be distinguished from the non-altered quartz grains found away from the active edge of the tool.Additional weakly developed flat micro-polish and orientation markers are observed on the high areas of the microtopography, while red iron oxide-rich dust residues are found in the depressions.Further flat polished areas are seen on the surfaces of the quartz crystals in micrograph F3.
CS.ET#4 is made on a dark ferruginous sandstone chunk, attributed to the RM1 type, that shows minimal modifications by direct percussion on one side.The active working edge is evident in two separate areas that show abrasion on both ventral and dorsal sides of the tool.The difference between the unmodified wind-abraded and rounded quartz grains and the crushed and damaged grains resultant from the abrasive work is visible in the M1a and M1b macroimages in Fig. 12.Similar to the other engraving tools that show traces of use, notably abrasive tracks, orientation markers indicating a longitudinal motion and mineral polish is seen on the elevated areas of the heavily regularised microtopography of the active edge, while the depressions tend to be filed with the mineral residues as seen in F1.
The only artefact from the excavated test trench that presented possible use-wear resultant from the contact with an abrasive mineral material was CS.ET.SQ4.5-10cm, which incidentally would not be considered an artefact if it was not for the abraded area on the distal portion of an unmodified chunk of fine-grained dark ferruginous sandstone (RM3).Compared to the other engraving tools found on the surface, the piece is relatively small and has been used without any modification to its natural shape by either percussion, abrasion or any other form of force.Yet, as can be seen on the macro-images showing the detail of the crushed and abraded portion of the active edge in Fig. 13, the piece shows traces that suggest its use in longitudinal movements on an abrasive mineral material.Similar to the experimental tools, regularisation and broadening of the otherwise tapered working edge can be observed in M2 and the crushed surfaces of the quartz grains and the presence of striations.Additional traces supporting the interpretation of the tool as being used in mineral processing activities are the abrasive tracks, striations and flat polished eminences on the microtopography of the tool's working edge, seen in F1 and F2 of Fig. 13.

Discussion
The combined results from the traceological analysis of the experimental and archaeological tools from the Camel Site provide additional data on this important and iconic archaeological site in northern Saudi Arabia.The identified patterns on the experimental tools indicate that the abrasive nature of the Sakaka sandstone impends the development of large areas of flat regular polish, with homogenous distributions due to the granular properties of the raw materials used by the makers of the archaeological tools and the experimental replicas.However, the relatively rapid development of the abraded working edge indicates that even a few strokes are sufficient to leave characteristic macro-traces on the working edges.Unfortunately, the preservation of the polished areas on the archaeological samples is additionally impeded by the severe taphonomic modifications resulting from the suspension of sand grains and particles by wind and weather on the tools and panels themselves.Albeit a high amount of debitage attributed to the RM2 has been identified at the site, it is currently difficult to establish whether the human groups that used the tools to process the sandstone were the same that created, quarried and knapped them.This is especially difficult for the pieces that display a high amount of elaboration in the form of prepared core reduction (e.g.CS.ET#1, CS.ET#3 and CS.ET#8).Additionally, we cannot rule out the possibility that the engraving tools found on the surface were used to create some of the more recent petroglyphs.However, the taphonomic modifications that formed on the abraded edges of the tools may indicate that these have been exposed on the surface for a considerable amount of time.
Despite these limitations, the following interpretations can be made regarding the tools found at the Camel Site concerning their possible function and association with the carving activities, which took place at the site.No modification made by retouch to the active edges of the tools was identified, suggesting that the tools have not been resharpened, which is remarkable given the relatively fast regularisation of the used edges.Furthermore, the lack of distinctive wear traces and scaring associated with the implementation of hafting technology (e.g.Moss 1987;Rots 2010) indicates the engraving tools were likely handheld.The limited sample size and taphonomic constraints acting on the material, however, imped a conclusive statement on the subject.The prevalence of traces associated with the longitudinal motion and the weak development of the abrasions on the active edges of the tools may be indicative that these were used during the second stage of the production of the reliefs, namely, the making of the fine details on the head of the animals and the sharpening of the outlines.Alternatively, these were used to carve out the deep wedges seen on some of the panels; unfortunately, it is difficult to positively correlate individual tools with specific panels.What can be said with some certainty is that the engraving tools were used to process abrasive mineral materials, likely sandstone and that their location underneath the panels on rock spur C may hint at a possible correlation.
The majority of the graphic activities known throughout the Arabian Peninsula, North Africa and the greater Middle East are characterised by engravings using distinctive techniques including pecking, engraving and smoothing or polishing large surfaces delimited by linear carvings.These extractive modifications are conducted by using friction as the main force, pecking either with direct or indirect percussion or drilling.These different techniques appear in different constellations throughout the known ensembles of prehistoric rock art sites across Arabia (e.g.Angás et al. 2021;Guagnin et al. 2015;Jennings et al. 2013;Khan 2013;Zerboni et al. 2021;Ziolkowski 2007).The large carved 3D engravings from the Camel Site, however, have, so far, no known equivalent throughout the Neolithic of the Arabian Peninsula, making a comparison of both carving techniques and the characteristics of stone tools used difficult.
A comparison with traceological studies conducted on Palaeolithic sites in Western Europe, however, may present some, albeit tenuous comparison.Of specific interest is the study conducted by López-Tascón et al. (2020), who have conducted experiments with the friction of stone against stone.The focus of their experimental protocol has been laid on the longitudinal engraving motions, a specific function of the archaeological case study addressed by the authors, the Palaeolithic site of La Viña Rockshelter in central Asturias (Spain).López-Tascón et al. report regularisation, rounding and abrasion of the active edges after only a short period.The authors, however, uphold that the essential characteristic of the mineral processing activities is the formation of the distinctive smooth and flat micro-polish; incidentally, due to the abrasive nature of its origin, the polish appears cracked with fissures resulting from material fatigue.A characteristic lacking in the experimental collection and archaeological assemblage is presented here.This may have different reasons including the variance in mineral processed (Carboniferous period limestone and Sakaka sandstone), and differences in use intervals (López-Tascón et al. experiments used intervals of 60 and 120 min), and tool raw materials (flint) come to mind.
Furthermore, the La Viña Rockshelter parietal ensembles are composed of engravings instead of large 3D carvings of natural proportions.Such occurrences are known from France, most notably at the Solutrean site of Roc-de-Sers (Delport 1984) located in the Charant, the Middle Magdalenian sites of Cap Blanc in the Dordogne and the site of Le Roc-Aux-Sorciers (Rousseau 1929)  The main categories of tools found in the excavations at the base of the murals included massive quartz pebbles with traces of violent impact and negatives, pick shaped flint nodules with traces of impacts and crushing at the pointed end, split quartz pebbles with traces of crushing on the active edge and sandstone pebbles with polished surfaces.The experimental reconstruction of the panels undertaken by the authors mainly using pebbles and split cobles included the regularisation of the carved area by percussive action, the creation of an outline, the reduction of volume by scraping and the addition of fine anatomical details.Traces on the experimental tools included battered surfaces and abraded edges, which have been found analogous to the archaeological material by the authors; they conclude that mostly pebbles and cobbles have been used to fashion the panels (de Beaune and Pinçon 2001: 74).In a later study of 300 multicomponent lithic tools from the Le Roc-Aux-Sorciers assemblage excavated from the domestic layers found underneath the panels of the rock shelter, Beyries and Cattin (2015) Fig. 11 CS.ET#5.Photograph and schematic illustration of the tool showing the position of macro-and micro-traces.M1a and M1b show the abraded and churched working surface that became abraded through contact with an abrasive mineral material during longitudinal motion.F1 micrograph was taken at 100 × showing the rounded and abraded higher areas of the microtopography and the striations, while red iron oxide dust residues are seen on the depressions; detail micrograph magnification 200 × showing the orientation markers; F2 polished and abraded areas on the elevated areas of the crystals and additional residues in the depressions at magnification 150 × ; F3 flat polished areas resultant from contact with abrasive mineral material at 100 × .(Image YHH, photos ICC and YHH) ◂ 1 3 conclude that part of the tools were used in other domestic activities such as hide working prior to the mineral processing activities.They note the low incidence of retouching on the tools that show traces of mineral processing.The use of unmodified or only slightly modified blocks of raw material in abrasive mineral activities was observed at the lithic assemblage found at the base of the panels from the Camel Site.The low incidence of artefacts with traces resulting from percussive actions, however, stands out especially given the tool marks seen on panel 2 (Fig. 2).The low incidence of retouch after use on the engraving tools from the Camel Site is also noted.
The complex and long-lasting development of the Camel Site, attested by multiple surface lithic scatters attributed to different chronologies, the stratified occupation and the carving activities, are being revealed by scientific research.The significance of the Camel Site for the al-Jawf province during the end of the Neolithic as a unique landmark and point of congregation for hunter, herder and gatherer communities across the millennia may be inferred.The Wadi Sirhan fault connects the Sakaka basin with the southern Levant and Jordan towards the northwest.The eastward draining wadis of which Wadi Badana is the most prominent connect the area to ancient Mesopotamia (Charloux et al. 2018).

Conclusions
Here, we have presented the results of the traceological analysis of a sample of stone tools coming from surface and stratified contexts, used to process highly abrasive mineral substances, most likely Sakaka sandstone.We have bolstered 1 3 our analysis by incorporating an experimental reference collection, used to process this specific raw material, and found a positive correlation between the traces observed on the archaeological material and the replicas.These traces have been associated with both transverse and longitudinal motions.The low levels of trace development, particularly on the specimen from the stratified context, indicate that stone tools were possibly used to create the delicate outlines and details on the animal reliefs during the final stages of manufacture of the panels at the Camel Site.The rapid development of the traces of mineral processing on the tools, the lack of tool curation and the fast reduction of sandstone mass by the scraping activities on the panels themselves may indicate that the carving activities were conducted by a small number of specialists in relatively short period of time.The amount of planning including raw material acquisition and the likely construction of scaffoldings to reach some of the engraved panes testifies for high level of foresight and determination of the prehistoric artisans responsible for the graphic activities at the site.

Fig. 1
Fig. 1 Map and pictures of the Camel Site.Top, oblique view of a 3D model of the site with the location of the panels (modified after Guagnin et al. 2021); bottom, panoramic photo of the rock spurs with detail of carved camels.(Photo MG and YHH)

Fig. 2
Fig. 2 Tool marks on the camel reliefs.A Close up of camel muzzle from panel 2 showing pecking signs in the background as well as smooth surface and thin lines; B deep longitudinal grooves and smoother surfaces forming the legs in panel 6; C short parallel grooves in panel 3; D fine lines forming the mouth and lips in panel 8. (Photographs GC and MG)

Fig. 3
Fig. 3 Map of the Camel Site and location of the engraving tools found under the panel on the C spur of the site.(GC and YHH)

Fig. 4
Fig. 4 Lithics on the surface of the site.A and B Large engraving tools (cores) found underneath the panels; C and D remnants of flaking activities found between the rock spurs C and B. (Photo YHH)

Fig. 5
Fig. 5 Experimental protocol for the creation of traces related to sandstone processing.(Image YHH, photos GC and YHH)

Fig. 6
Fig.6Micrograph of the alterations caused by taphonomy located on the cortical portions of the tools and located in the knapped surfaces of the archaeological samples, as well as unmodified areas on the experimental reference sample.(Image YHH, photos ICC and YHH) ◂

Fig. 7
Fig. 7 CS.EXP.03.Flake tool used in transverse bidirectional motion during a 5-min interval on Sakaka sandstone.M1 showing the bevelled and regularised working edge of the tool with the crushed grains and the depressions filled-in with mineral residues.F1 micrograph

Fig. 8
Fig. 8 CS.EXP.04.Flake tool used in longitudinal bidirectional motion during a 5-min interval on Sakaka sandstone.M1 showing the rounded and regularised working edge of the tool.F1 abrasive tracks

Fig. 9
Fig. 9 CS.EXP.02.Flake tool used in transverse unidirectional motion during a 5-min interval on Sakaka sandstone.M1 regularised and bevelled working edge with parallel transverse scratches.F1 light

Fig. 10
Fig. 10 CS.ET#1.Schematic drawing and photo of the tool showing the location of the abraded edges and the position where the individual micrographs showing the traces resultant from the use were taken.M1a and M1b show the abraded area with the parallel striations indicating the transverse motion during use.F1 showing the distribution of the polish and striations on the eminences of the microtopography at magnifications 50 × and 200 × ; F2 micrograph showing the abraded and rounded topography and the micro-negatives resultant from the pressure exerted during the abrasive task at magnifications 100 × and 200 × .(Image YHH, photos ICC and YHH) ◂ located in the Vienne.The study of the lithic material from Le Roc-Aux-Sorciers provides essential information on the techniques used to manufacture the panels.The site is located in the Anglin valley and is characterised by a 50-m long rock shelter composed of soft Jurassic bioclastic limestone.Excavations at the base of the panels have exposed occupation layers dating to 14,500 BP.The initial investigation on the techniques used to create the murals depicting animal reliefs with naturalistic anatomical proportions conducted by de Beaune and Pinçon (2001) has established seven phases of production identifiable by specific traces: (a) the regularisation of the limestone surface by percussion; (b) extraction of mineral material to create the outline (or preform) of the animal by pecking using a pointed tool; (c) trimming of the animal's silhouette by transverse scraping action; (d) shaping the surface of the sculpture by additional subtraction of mineral material by transverse scraping action; (e) regularisation of the surface of the sculpted parts of the animal by abrasion or polish; (f) engraving of anatomical details by incision and longitudinal action and finally (g) painting of the sculpted animals by using red and dark pigments.

Fig. 12
Fig. 12 CS.ET#4.Photograph and schematic illustration of the tool showing the position of macro-and micro-traces.M1 shows the active edge on the ventral surface.F1 shows the polish and striations

Fig. 13
Fig. 13 CS.SQ4.5-10cm.Photograph and schematic illustration of the tool showing the position of macro-and micro-traces.M1 macroimage on the dorsal side of the active edge of the tool; M2 macroimage of the top view of the working edge showing the crushed and

Table 2
Specific traits and results of the functional analysis of the engraving tools from the Camel Site Artefact