Rock art landscapes beside the Jubbah palaeolake, Saudi Arabia.

Jennings, Richard P. ; Shipton, Ceri ; Al-Omari, Abdulaziz 等

[ILLUSTRATION OMITTED]

Introduction

In a recent synthesis on rock art interpretation, Chippindale and Nash (2004) emphasised that images and depictions must be interpreted within their landscape setting in order to appreciate why a particular setting was chosen. Such an approach, they surmise, is applicable at different scales--that of the rock surface on which a particular panel is painted or engraved, and the scale of the wider environment, which itself may change while the rock art endures. Previous research on the rock art of Arabia has largely focused on the imagery itself, with systematic surveys being rare. (1) Here we present the first combined systematic survey and quantitative study of rock art distributions undertaken in the Arabian peninsula.

[FIGURE 1 OMITTED]

The current study is part of a comprehensive programme of Late Pleistocene and Holocene archaeological research at Jubbah oasis, where palaeolake deposits have been identified (Petraglia et al. 2011, 2012). Our approach involved the systematic archaeological survey of four jebels (hills): Jebel Qattar, Jebel Gattar A, Jebel Gattar B and Jebel Katefeh. These are located south-west and east of a major rock art complex called Jebel Umm Sanman, which was surveyed in 1976 and 1977 (see Parr et al. 1978; Figures 1 & 2). The current survey covered 39[km.sup.2] and documented numerous animal and human depictions and inscriptions--for background see Parr et al. (1978), Khan (1993) and Aldowsari (2009). These fall into different chronological divisions which we refer to below as late prehistoric (Neolithic, Chalcolithic and Bronze Age), Thamudic and recent.

[FIGURE 2 OMITTED]

The first aim of our research was to explore the spatial relationships between late prehistoric rock art sites and known palaeolakes in the area. Parr et al. (1978) hypothesised that late prehistoric rock art coincided with periods of high rainfall at Jebel Umm Sanman, where grazing bovids (i.e. wild or domesticated cattle), equids, ibex and caprids dominate the engravings. Radiocarbon dating of the main palaeolake at Jubbah, reported by Garrard et al. (1981), yielded a date of 6685 [+ or -] 50 BP (Q-3118), but this age should be treated with caution given the era in which it was obtained. We recently determined that palaeolake deposits beside Jebel Qattar had formed in the early Holocene (Crassard et al. in press). This fits well with other palaeoenvironmental evidence in the region for the early Holocene being more humid than its present day arid environment: a perennial lake existed 240km west of Jubbah at Tayma oasis between 10 000-9000 cal BP (Engels et al. 2012); an early Holocene humid phase (9250-7250 cal BP) is reported in cores taken from the Red Sea (Arz et al. 2003); and speleothem records at Soreq Cave in the southern Levant show an overall trend of increased precipitation from the onset of the Holocene to 7500 BP, with peaks at 8500 BP and 7500 BP (Bar-Matthews et al. 1997).

The second aim of our research was to interpret the distribution of the Thamudic inscriptions and animal imagery around the jebels. Thamudic rock art is found primarily in desert environments of northern and central Arabia and to a lesser extent in southern Arabia, the Transjordan Plateau, the Negev Desert and Egypt (Al-Theeb 1999, Anati 1999; MacDonald 2010). The writing belongs to a branch of scripts known as South Semitic, which were written in Arabia from about the middle of the first millennium BC to the arrival of Islam (Beeston 1981; Shah 2008). They are usually divided into Groups A-E but these divisions are widely accepted as needing revision (see Al-Theeb 1999; MacDonald 2010). The scripts are poorly dated. A text bearing the name of the mid sixth century BC Babylonian king Nabonidus is the earliest known, while the latest is an inscription dated to AD 267 (MacDonald 2010). MacDonald (2010) believes Thamudic B, C and D scripts were written by nomadic peoples who had learned how to write from merchant traders at oasis towns such as Dedan, Tayma or Dumah (Dumat al Jundal). Such merchants crossed central Arabia en route from southern Arabia to the Near East during the first millennium BC (Figure 1). Camels were the main pack animal of the caravan routes. The dromedary camel arrived in south-eastern Arabia about 5000-6000 years ago (Uerpmann & Uerpmann 2002). It is absent in late prehistoric art but is often depicted with Thamudic inscriptions (Parr et al. 1978; Khan 2007). Thamudic imagery at Jubbah also includes ibex and other species of goat, felids, ostrich, human figures, horse riders and palm trees (Parr et al. 1978).

Survey area and methods

Three of the sandstone jebels surveyed lie 16km east of Jebel Umm Sanman. Jebel Qattar (elevation: 800m base, 892m top) is c. 600m long x 400m wide, with a north-south orientation. Many boulders lie on the base of the jebel, and ancient lake deposits are visible immediately to the north and east. Jebel Gattar A (elevation: 830m base, 930m top) is of a similar size to Jebel Qattar but with an east-west orientation and with dune sand running up the middle on both sides and breaching its centre. Jebel Gattar B (elevation: 840m base, 870m top) is one third of the size of its neighbours. The main palaeolake of Jubbah is less than a 30-minute walk away over the dunes to the west and is visible from the upper western slopes of these jebels. The fourth, Jebel Katefeh (elevation: 830m base, 1020m top), offers a useful point of contrast as it is 26km west-southwest of the other three and is 14.5km south-west of Jebel Umm Sanman. It is 1km long x 500m wide, oriented north-south, and overlooks its own palaeolake to the east.

The rock art was surveyed using handheld GPS and a total station. The lower reaches of the four jebels were systematically surveyed on foot while the middle and upper reaches were surveyed where it could be done safely. Each site was recorded and photographed and the data entered into a spatial database. Attribute data included coordinates, condition, visibility, method and style of application, density and type of content depicted, writing style and orientation, and association with other forms of cultural evidence. Google Earth photography and Aster 30m digital terrain models were used to make the maps, and analysis was undertaken in ESRI ArcGIS 9.3 software.

Survey outcome

A total of 107 tock art sites were recorded (recent: 23, Thamudic: 60, late prehistoric: 45) (Table 1). A few sites contained multiple phases. Late prehistoric and Thamudic feature in similar numbers at Jebel Katefeh and Jebel Qattar with Thamudic styles dominant at Jebel Gattar A and B. Eight sites contain richly decorated panels, 16 are of medium density, and 73 are of low density. The survey also documented 19 lithic scatters, 13 sites with one or more cairns, and the remnants of seven walled structures.

The rock art is predominantly found along the base of the jebels, although variation was detected between phases (Table 2). A chi-square test revealed that fewer occurrences of Thamudic rock art than expected are located up the jebels in comparison to late prehistoric sites (n = 90, 3 d.f. = 16.23, p> 0.037). Instead, Thamudic sites are on boulders scattered on the base of jebels or on bedrock. 13 late prehistoric sites are higher up the jebels, notably JQ-34, 42, 43, 44 and 45 on Jebel Qattar.

Recent petroglyphs

A total of 23 recent rock art sites were recorded. These comprise unpatinated engravings of Arabic script that were most often carved with a metal object. The script is commonly a person's name with a date in the 1400s of the Hijri calendar, meaning the words were inscribed within the last 30 years. A few of the undated Arabic inscriptions were pecked with another stone rather than carved using metal, providing a link with Thamudic inscriptions and suggesting greater antiquity. There are depictions of mounted camels and occasionally fight or battle scenes with 'stick-figure' people using lances to fight on horses. There are also scenes that resemble Thamudic styles of people hunting ostriches with rifles, and an image of a motor vehicle.

Thamudic petroglyphs

Thamudic rock art is comprised exclusively of pecked engravings with low levels of patination. It is dominated by Thamudic inscriptions and depictions of camels (Figures 3 & 4). There are 42 sites with inscriptions. Most are written vertically, while some rarer longer ones are written horizontally. Site JQ-6 on Jebel Qattar has an exceptional series of short horizontal inscriptions on a boulder on the base of the jebel (Figure 5). It is written in Thamudic B, in contrast to five other sites with vertical inscriptions on the same jebel (JQ-3, 6, 23, 31 & 40) which are written in Thamudic C or D. These sites contain lines written in pairs of 4-6 Thamudic characters (Figure 6). Other examples contain only 1-4 characters. Some may be wusum signs--tribal signs left throughout the ages (Khan 2000).

Thamudic-style camels were found at 36 sites (Table 3). The majority (25 out of 36) are associated with Thamudic inscriptions, especially on Jebel Gattar A and B, where all but one camel have an associated inscription. Sometimes an image of a camel and a short vertical inscription are contained within a circle (Figure 6). Other images in Thamudic style are depictions of ostriches, dogs, date palms and ibex. The date palms may be indicative that date cultivation was practised.

There is no clear association between Thamudic writing and material remains, although two scatters of quartz lithics with Thamudic scripts beside them at the base of Jebel Qattar and Jebel Gattar A and B (JQ-40 and JG-19) maybe candidates. The lithics in question are small cores and flakes that demonstrate seemingly expedient bipolar reduction of whitish pebble quartz, but which lack technologically diagnostic features that would make links with Thamudic engravings conclusive. The remnants of four structures were also found in the vicinity of Thamudic writing. These are: a small hearth or possible water collection structure at JQ-22; a dug-out shelter 4m in diameter at JQ-31; a linear structure 7m long of unknown use at JQ-38; and a linear windbreak structure 5m long and 0.5m wide at JG-B1.

Late prehistoric petroglyphs

Late prehistoric petroglyphs feature at three of the four jebels (Figures 7 & 8). The engravings are larger and denser compared to Thamudic rock art: seven of eight panels with a high density rank belonged to this phase. The engravings are pecked, heavily patinated and are overlain by unpatinated Thamudic script and camels at a few locations, as was noted earlier by Parr et al. (1978) and Khan (1993) at Jebel Umm Sanman. Two of the more elaborate sites at Jebel Qattar (JQ-31 and JQ-34) are rockshelters that overlook an extensive, multi-period, late prehistoric surface site (JQ-101). A short climb is required to view the art at JQ-34, which is prominently placed high in the landscape (Figure 9-11).

[FIGURE 3 OMITTED]

The style of the engravings, limited skills of the engravers and difficult rock surfaces mean it is not always a straightforward process to identify the animals depicted to species level. This is apparent with some of the cattle and goat depictions, and in particular whether they are wild or domesticated forms. Overall, the range of species observed broadly matches the observations of Parr et al. (1978) at Jebel Umm Sanman. However, while they reported cattle as the dominant animal depicted, the majority of animals recorded in this study area are Nubian ibex (Capra nubiana), which feature at 29 sites. The species is identified by its huge swept-back horns and beard (Figure 11). It is commonly shown with a stripey coat, although piebald and fully engraved plain bodies are also known. The ibex images range from around 0.15m to over lm in length.

[FIGURE 4 OMITTED]

Cattle are the next most common animals, with examples from 25 sites. These are among the most elaborate engravings in the survey. They display large flaring horns and have a large body size in comparison to ibex--the largest is 1.5m long x 1.5m tall. The cattle are usually shown with their heads tilted to the side, so that both their horns and ears are visible (Figure 9). This is referred to as the 'Jubbah style' by Parr et al. (1978). It is not clear if wild or domesticated species are represented, or both. If wild, they are likely to be the extinct wild aurochs (Bos primigenius) or the giant long-horned buffalo (Pelorovis antiquus). The presence of some piebald coats (Figure 12) may indicate that they are a domesticated species, as this is a known trait among domesticated animals (McCorriston & Martin 2009).

Images of other animals with beards and parallel horns that curve in opposite directions at the top appear to be wild rather than domesticated goats (Capra aegagrus) (sites JKF-39, 40,41, 43, 47 and 51). Sites JQ-14, JKF-25 and JKF-34 contain images of an ungulate with a large head, probably a wild ass (Equus africanus). The asses have two short forward-facing appendages on the head which are probably the ears, while two images were found with manes on the back of the neck. JKF-25 also contains a depiction of a horse. One of the asses and one of the aurochs appear to be pregnant. Some of the ibex are juveniles, based on their relative size. Canids were also noted at JKF-22.

[FIGURE 5 OMITTED]

[FIGURE 6 OMITTED]

Late prehistoric human figures were found at 13 sites. They tend to be tall and elongated (some were lm high but only 0.1m wide). Some very elaborate elongated human figures were found in rockshelter JQ-43. They are male (they have erect penises) and appear to be wearing grass skirts and some kind of head-dress. Two appear to be wielding hooked implements, while one has a bow and arrow. Other examples of human figures are seen at JQ-34 (Figure 9), JQ-31 and JKF-49. Several smaller, less elongated human figures were found in JQ-34 and JQ-43. They are also armed with bows and arrows. Similar examples were recorded at Jebel Umm Sanman by Parr et al. (1978) and Khan (1993).

Late prehistoric pictographs

In the largest rockshelter (JQ-34; Figure 10), which has a small cave at the back, there is a fourth type of rock art painted in red ochre. This is the most sheltered of all the rock art localities, so painted aft may originally have been more widespread but only survived here. The art includes three bovids, one of which was possibly an aurochs, hence it may be late prehistoric. However, there is also a series of abstract designs that do not occur in the engravings. The designs consist of square-filled dots, approximately 100mm in size. These occur five times on one panel. In one case, an ibex has been painted around the square, but it is unclear which was painted first. The ochre occurs in two different shades: a brownish-red in which the bovids are painted and a purplish-red in which the squares and dots are painted. In another instance, the brownish-red ochre is clearly overlying the purplish-red ochre. Behind this panel are six sequences of parallel lines in red ochre. These parallel lines also do not occur in the engraved art.

Cupules, symbols and grinding slicks

Cupules were found at six sites but their phasing is unclear. A cluster was found with grinding slicks at rockshelters JQ-31 and JQ-43. These cupules are approximately 0.25m in diameter. A large, red-stained cupule around 0.35m in diameter is visible on a boulder between JQ-101 and the JQ-31 rockshelter (JQ-32). Two possible cupules are associated with Thamudic and Arabic writing at JQ-38. Cupules also feature at JKF-22 and JKF-28 on Jebel Katefeh. These are associated with both late prehistoric and Thamudic art. Abstract symbols were recorded at 11 sites. Symbols accompany late prehistoric panels at JKF-34, 39 and 43 while two late prehistoric vulvic symbols are engraved at JQ-43. Other symbols are two Thamudic geometric shapes (JQ-5 and JG-B1), two of recent age (two crosses associated with Arabic script at JQ-37 and JGA-20) and two of uncertain age (JQ-8 and JQ-18).

[FIGURE 7 OMITTED]

[FIGURE 8 OMITTED]

Spatial analyses

Rock art and palaeolakes

The results show that 37 out of 45 (82.2 per cent) late prehistoric rock art sites overlook palaeolakes (Figures 2, 7 & 8). This is convincing evidence that lakes were present at the time the rock art was depicted. If the palaeolakes were dry when the art was drawn, one would expect the art to be randomly distributed around the jebels, but this is not the case. Recent dating of the Jebel Qattar palaeolake deposits suggests that the lake was present in the early Holocene, when the climate was more humid than in subsequent phases of the Holocene (Crassard et al. in press). The palaeolake at Jebel Qattar is visible from 12 out of 16 late prehistoric rock art sites on the jebel, and from three sites on neighbouring Jebel Gattar A. Two sites on the opposite side of Jebel Qattar overlook the main palaeolake at Jubbah, which is 1km to the west of the jebel (Figure 7). The pattern is striking at Jebel Katefeh, where 22 out of 25 sites overlook a palaeolake to the east; only three sites are not on the lake-facing slopes of this jebel (Figure 8).

[FIGURE 9 OMITTED]

The relationship between the animals depicted in the rock art and the presence of palaeolakes is unlikely to be a coincidence. A humid climate would have allowed grassland habitats to develop in the vicinity of the palaeolakes, probably on a seasonal basis. Such habitats would have supported the cattle and ibex that feature in the rock art. Further information comes from other aspects of the archaeological survey. JQ-101, a multi-period archaeological site, was identified beside the Jebel Qattar palaeolake. Its lithic assemblage included Pre-Pottery Neolithic A El-Khiam points, early Pre-Pottery Neolithic B Helwan points and Chalcolithic points. The limited evidence for other phases of the lithic reduction sequence and the absence of structures or storage pits suggest that the site was occupied seasonally. The El-Khiam and Helwan points correspond very well with the early Holocene age for the palaeolake at Jebel Qattar (Crassard et al. in press). It seems highly likely, therefore, that at least some of the late prehistoric art was made by early Holocene populations on seasonal visits to Jubbah.

[FIGURE 10 OMITTED]

[FIGURE 11 OMITTED]

[FIGURE 12 OMITTED]

Elsewhere, the cairns in the study area await classification and investigation. Cairns are ubiquitous across Arabia and typically date to the Chalcolithic and Bronze Age (Parr et al. 1978). A total of 28 cairns were recorded at 13 sites. The highest number is at Jebel Katefeh (18 cairns at six sites) followed by Jebel Qattar (seven cairns at five sites) and Jebel Gattar A (three cairns at one site). Eight are large (>5m in diameter, e.g. JKF-20) and the rest medium-sized (2m x 2m to 5m x 5m). Only one cairn is linked to a rock art site (JQ-31), where a cairn is located in front of a collapsed rockshelter. No cairns were inscribed with late prehistoric rock art and no patterns were apparent in the distribution of cairns and rock art at Jebel Qattar. At Jebel Katefeh, cairns cluster at the south-eastern base of the jebel, over an area measuring 500m x 150m (Figure 8).

Thamudic caravan routes

The expedient nature of Thamudic rock art and its focus at the jebel bases suggest that it was made by transhumant people. The lack of material remains, the limited themes presented in the rock art, and the profusion of camel images, which are occasionally mounted, all suggest that the people who created this rock art were nomadic. This supports MacDonald s (2010) hypothesis that many nomadic societies in ancient Arabia were literate and were profligate at marking graffiti on rocks in the desert. Nomadic literacy, he suggests, came about from nomadic peoples coming into contact with merchant traders.

The distribution of Thamudic rock art around the bases of Jebel Qattar and Jebel Gattar A and B exhibits a clear difference from late prehistoric rock art, in that the focus on palaeolakes was lost. Instead, the art is more evenly dispersed around the jebels but with an emphasis on their western sides, mainly on boulders on the base. One explanation for why Thamudic art is not focused on the palaeolakes is the possibility that these had dried up by the first millennium BC. The location of Thamudic art on the western sides of the jebels may be due to these areas receiving sufficiently high seasonal humidity to support grazing land (consider Lancaster & Lancaster, 1999: 108-109). However, no palaeoenvironmental evidence exists to support this premise, and the absence of Bos depictions in the Thamudic rock art would suggest that cattle were not grazed here at this time.

A plausible hypothesis for the location of Thamudic art at Jebel Qattar and Jebel Gattar A and B is that it may reflect the route in which nomadic peoples, merchant traders and other travellers moved through the landscape as they traversed between Jubbah and neighbouring settlements to the south of the Nefud, such as the town of Ha'il. This town is 90km southeast of Jubbah and was an important node on the southern Arabia to southern Mesopotamia trade route in the first millennium BC (MacDonald 2010). There are no recorded ancient caravan routes between Ha'il and Jubbah to support this hypothesis, but given that the jebels are highly visible in the landscape and are situated near the narrowest crossing point of the desert, it is not inconceivable that merchant traders, caravanserai or nomadic peoples passed through and encountered or made the rock art (Figure 3). The predominance of Thamudic rock art on the east side of Jebel Katefeh could also be an indicator of a caravan route (Figure 4). This route would have linked Jubbah to the major southern Arabia to northern Mesopotamia/Levant trade route (Figure 1). The likely destination was Tayma, a trading town where numerous Thamudic inscriptions have been recovered (Eichmann et al. 2006; MacDonald 2010).

Conclusions

The information presented above concerning rock art at four jebels in Jubbah shows the merits of interpreting rock art from a landscape perspective. Our results indicate that 82 per cent of late prehistoric rock art overlooks palaeolakes. This suggests that occupation took place during wet phases of the early Holocene. The elaborate nature of the late prehistoric art, along with its restricted horizontal distribution in the landscape, and its extensive vertical distribution at prime locations, suggest relatively long-term occupation. The absence of settlement structures indicates that this occupation stopped short of permanent settlement. However, the discovery of stone points and the common depiction of wild animals such as ibex, as well as humans with bows and arrows, suggest that the localities were used as seasonal hunting grounds. Spatial analysis of Thamudic rock art identified possible trade routes through the Jubbah landscape, based on the prevalence of camel depictions and inscriptions at the eastern base of Jebel Katefeh and on the western bases of Jebel Qattar and Jebel Gattar A. Variations in cultural adaptations and landscape-use behaviours therefore appear to be linked to changes in environments. Future research at Jubbah will further assess the relationships between rock art distribution, ecological settings and landscape behaviours.

Acknowledgements

We thank HRH Prince Sultan bin Salman, President of the General Commission for Tourism and Antiquities, and Professor Ali I. Al-Ghabban, Vice President for Antiquities and Museums, for permission to carry out this study. We also thank Dr Hussain Abu Al Hassan, Habeeb Turki, Abdalrahman Al-Thobiti, Abdalrahman Almansour, Jamal S. Omar and the people of Jubbah for their support and assistance with the field investigations. We acknowledge the financial support of the National Geographic Society, the Leakey Foundation, the European Research Council (grant no. 295719) and the Saudi Commission for Tourism and Antiquities. Thanks also go to the anonymous reviewers for their helpful comments.

References

ALDOWSARI, S. 2009. Rock art at Jebel Umm Sanman in Hail region: an archaeological study. Unpublished MA dissertation, King Saud University.

AL TALHI, D. 2012. Almulihiah: a rock art site in the Hail region, Saudi Arabia. Arabian Archaeology and Epigraphy 23(1): 92-98.

AL-THEEB, S. 1999. Thamudic inscriptions from the Kingdom of Saudi Arabia. Riyadh: King Fahad National Library (in Arabic).

ANATI, E. 1968a. Rock art in central Arabia I: the 'oval-headed' people of Arabia. Louvain: Bibliotheque du Museon.

--1968b. Rock art in central Arabia 2 (parts I & II). Louvain: Institute Orientaliste, Universite Catholique de Louvain.

--1972. Rock art in central Arabia 3: corpus of the rock engravings (parts I & II). Louvain: Institute Orientaliste, Universite Catholique de Louvain.

--1974. Rock art in central Arabia 4: corpus of the rock engravings (parts III & IV) (Institute Orientaliste publication 4). Louvain: Institute Orientaliste, Universite Catholique de Louvain.

--1999. The rock art of the Negev Desert. Near Eastern Archaeology 62(1): 22-34.

ARZ, W.H., F. LAMY, J. PATZOLD, P.J. MULLER & M.A. PRINS. 2003. Mediterranean moisture source for an early-Holocene humid period in the northern Red Sea. Science 300: 118-21.

BAR-MATTHEWS, M., A. AYALON & A. KAUFMAN. 1997. Late Quaternary paleoclimate in the eastern Mediterranean region from stable isotope analysis of speleothems at Soreq Cave, Israel. Quaternary Research AT: 155-68.

BEESTON, A.F.L. 1981. Languages of pre-Islamic Arabia. Arabica 28(2-3): 178-86.

BRAEMER, F., P. BODU, R. CRASSARD & M. MANQUSH. 2007. Jarf al-Ibil et Jarf al-Nabirah: deux sites rupestres de la region d'al-Dali, in M.-L. Inizan & M. Rachad (ed.) Art rupestre et peuplements prehistoriques au Yemen: 95-100. Sana'a: CEFAS.

CHIPPINDALE, C. & G. NASH. 2004. Pictures in place: approaches to the figured landscapes of rock art, in C. Chippindale & G. Nash (ed.) The figured landscapes of rock art: looking at pictures in place: 1-36. Cambridge: Cambridge University Press.

CLARK, C. 1975. The rock art of Oman. Journal of Oman Studies 1: 113-22.

CRASSARD, R. 2006. [TEXT NOT REPRODUCIBLE IN ASCII] [Preliminary report on the discovery of ALI-1 site in Wadi bin Ali, Hadramawt, Yemen]. Chroniques Yemenites en langue arabe 3: 3-10. Sana'a: CEFAS (in Arabic).

CRASSARD, R., M. PETRAGLIA, A.G. PARKER, A. PARTON, R.G. ROBERTS, Z. JACOBS, A. AL-OMARI, A. ALSHAREKH, P. BREEZE, N.A. DRAKE, H.S. GROUCUTT, R.P. JENNINGS, E. REGAGNON & C. SHIPTON. In press. Beyond the Levant: first evidence of a Pre-Pottery Neolithic incursion into the Nefud Desert, Saudi Arabia. PLoS ONE. 8: e68061.

EICHMANN, R., H. SCHAUDIG & A. HAUSLEITER. 2006. Archaeology and epigraphy at Tayma (Saudi Arabia). Arabian Archaeology and Epigraphy 17: 163-276.

ENGELS M., H. BRUCKNER, A. PINT, K. WELLBROCK, A. GINAU, P. VOSS, M. GROTTKER, N. KLASEN & P. FRENZEL 2012. The early Holocene humid period in NW Saudi Arabia--sediments, microfossils and palaeo-hydrological modelling. Quaternary International 266: 131-41.

GARCIA, M. & M. RACHAD. 1997. L'art prehistorique, in C.H. Robin & B. Vogt (ed.) Yemen: au pays de la reine de Saba, catalogue de l'exposition presente a l'Institut du Monde Arabe: 26-29. Paris: Flammarion.

GARCIA, M., M. RACHAD, D. HADJOUIS, M.-L. INIZAN & M. FONTUGNES. 1991. Decouvertes prehistoriques au Yemen, le contexte archeologique de l'art rupestre de la region de Saada. Comptes Rendus de l'Academie des Sciences de Paris 313 series II: 1201-16.

GARRARD, A.N., C.P.D. HARVEY & V.R. SWITSUR. 1981. Environment and settlement during the Upper Pleistocene and Holocene at Jubba in the Great Nefud, northern Arabia. Adal5: 137-48.

HASSIBA, R., G.B. CIESLINSKI, B. CHANCE, F.A. AL-NAIMI, M. PILANT & M.W. ROWE. 2012. Determining the age of Qatari Jabal Jassasiyah petroglyphs, QScience Connect 2012: 4. doi: 10.5339/connect.2012.4

INIZAN, M.-L. & M. RACHAD (ed.). 2007. Art rupestre et peuplements prehistoriques au Yemen. Sana'a: CEFAS.

INSALL, D. 1999. The petroglyphs of Shenah. Arabian Archaeology and Epigraphy 10: 225-45.

KHAN, M. 1993. Prehistoric rock art of northern Saudi Arabia. Riyadh: Ministry of Education, Department of Antiquities and Museums.

--1998. A critical review of rock art studies in Saudi Arabia. East and West 48(3/4): 427-37.

--2000. Wusum, the tribal symbols of Saudi Arabia. Riyadh: Ministry of Education.

--2007. Rock art of Saudi Arabia across twelve thousand years. Riyadh: Deputy Ministry of Antiquities & Museums.

LANCASTER, W. & F. LANCASTER. 1999. Land and water in the Arab Middle East. Amsterdam: Harwood Academic.

--2011. A discussion of rock carvings in Ra's al Khaimah Emirate, UAE, and Musandam province, Sultanate of Oman, using local considerations. Arabian Archaeology and Epigraphy 22(2): 166-95.

LIVINGSTONE, A. & M. KHAN. 1985. Epigraphic survey 1404-1984. Atlal 9: 128-44.

MACDONALD, M.C.A. 2010. Ancient Arabia and the written word, in M.C.A. MacDonald (ed.) The development of Arabic as a written language (Supplement to the Proceedings of the Seminar for Arabian Studies, vol. 40): 5-27. Oxford: Archaeopress.

MCCORRISTON, J. & L. MARTIN. 2009. Southern Arabia's early pastoral population history: some recent evidence, in M.D. Petraglia & J.I. Rose (ed.) The evolution of human populations in Arabia: 237-50. New York: Springer.

NAYEEM, M.A. 1998. Qatar: prehistory and protohistory from the most ancient times (ca. 1 000 000 to end of BC era). Hyderabad: Hyderabad Publishers.

--2000. The rock art of Arabia: Saudi Arabia, Oman, Qatar, the Emirates and Yemen. Hyderabad: Hyderabad Publishers.

PARR, P.J., J. ZARINS, M. IBRAHIM, J. WAECHTER, A. GARRARD, C CLARKE, M. BIDMEAD & H. AL-BADR. 1978. Preliminary report on the second phase of the northern province survey 1397/1977. Atlal 2: 29-50.

PETRAGLIA, M.D., A.M. ALSHAREKH, R. CRASSARD, N.A. DRAKE, H. GROUCUTT, A.G. PARKER & R.G. ROBERTS. 2011. Middle Paleolithic occupation on a marine isotope stage 5 lakeshore in the Nefud Desert, Saudi Arabia. Quaternary Science Reviews 30: 1555-59.

PETRAGLIA, M.D., ALSHAREKH, P. BREEZE, C. CLARKSON, R. CRASSARD, N.A. DRAKE, H.S. GROUCUTT, R. JENNINGS, A. PARKER, A. PARTON, R.G. ROBERTS, C. SHIPTON, C. MATHESON, A. AL-OMARI & M.-A. VEALL. 2012. Hominin dispersal into the Nefud Desert and Middle Palaeolithic settlement along the Jubbah palaeolake, northern Arabia. 2012. PLoS ONE7(11): e49840. doi: 10.1371/journal.pone.0049840

PRESTON, K. 1976. An introduction to the anthropomorphic content of the rock art of Jebel Akhdar. Journal of Oman Studies 2: 17-38.

SHAH, M. 2008. The Arabic language, in A. Rippin (ed.) The Islamic world: 261-77. London: Routledge.

UERPMANN, H.P. & M. UERPMANN. 2002. The appearance of the domestic camel in south-east Arabia. The Journal of Oman Studies 12: 235-60.

ZIOLKOWSKI, M.C. 2007. Rock on art: petroglyph sites in the United Arab Emirates. Arabian Archaeology and Epigraphy 18(2): 208-38.

Received: 11 July 2012; Accepted: 22 August 2012; Revised: 26 October 2012

(1) Anati 1968a & b, 1972, 1974; Livingstone & Khan 1985; Khan 1993, 1998, 2000, 2007; Nayeem 2000; Aldowsari 2009; Al Talhi 2012 (Saudi Arabia); Garcia et al. 1991; Garcia & Rachad 1997; Crassard 2006; Braemer et al. 2007; Inizan & Rachad 2007 (Yemen); Clark 1975; Preston 1976; Insall 1999 (Oman); Ziolkowski 2007; Lancaster & Lancaster 2011 (United Arab Emirates); and Nayeem 1998; Hassiba et al. 2012 (Qatar).

Richard P. Jennings [1], Ceri Shipton [2], Abdulaziz Al-Omari [3], Abdullah M. Alsharekh [4], Remy Crassard [5], Huw Groucutt [1] & Michael D. Petraglia [1]

[1] School of Archaeology, Research Laboratory for Archaeology and the History of Art, University of Oxford, Oxford 0X1 2HU, UK

[2] School of Social Science, University of Queensland, Brisbane, QLD 4072, Australia

[3] Taif Antiquities Office, Taif, Makka, Saudi Arabia

[4] Department of Archaeology, College of Tourism & Archaeology, King Saud University, PO Box 2454, 11451 Riyadh, Saudi Arabia

[5] CNRS, UMR5133, Maison de l'Orient et de la Mediterranee, 517 rue Raulin, 39365 Lyon cedex 07, France

Table 1. Breakdown of the frequency of rock art styles at the four
jebels in the study area, showing that Jebel Qattar and Jebel
Katefeh, the two jebels with associated palaeolakes, contain the
highest proportion of late prehistoric art.

 Styles represented

Area surveyed Number of rock Late prehistoric Thamudic Recent
 art sites

Jebel Katefeh 42 25 22 3
Jebel Qattar 37 16 19 11
Jebel Gattar A 21 4 14 6
Jebel Gattar B 7 0 5 3
Total 107 45 60 23

Table 2. Location of rock art within the study area. Thamudic rock art
is located exclusively on the jebel base, often on boulders, whereas
late prehistoric art occurred both on the base and higher up the
jebels, particularly in rockshelters.

 Late Late
Physical location prehistoric prehistoric %

Jebel base and/or boulder on jebel base 30 66.7
Rockshelter at jebel base 1 2.2
Combination of 1 and 2 1 2.2
Total at jebel base 32 71.1
Jebel slope and/or boulder on jebel slope 9 20.0
Rockshelter elevated on jebel 4 8.9
Total up jebel 13 28.9
Total 45

Physical location Thamudic Thamudic %

Jebel base and/or boulder on jebel base 45 75.0
Rockshelter at jebel base 8 13.3
Combination of 1 and 2 1 1.6
Total at jebel base 54 90.0
Jebel slope and/or boulder on jebel slope 6 10.0
Rockshelter elevated on jebel 0 0.0
Total up jebel 6 10.0
Total 60

Table 3. The number of times that inscriptions and camel depictions,
the main components of Thamudic art, occur together on the same
panels. This occurs most often at Jebel Gattar A.

 Main representations

Area surveyed Thamudic sites Scripts Camels Script and camels
 at same site

Jebel Katefeh 22 14 13 7
Jebel Qattar 19 14 9 5
Jebel Gattar A 14 11 11 10
Jebel Gattar B 5 3 3 3
Total 60 42 36 25