Madāʾin Ṣāliḥ [ancient Hegra, al-Ḥijr, Ḥijrā]

Madāʾin Ṣāliḥ is one of the names given to the ancient city of Hegra, in Northwest Arabia, a possibly Liḥyanite and certainly a Nabataean and Roman site. It reached its peak in the first century BCE and in the first two centuries CE. It is famous for its monumental rock-cut tombs with decorated facades, similar to those of the Nabataean capital, Petra.

Introduction

Madāʾin Ṣāliḥ, “the cities of Ṣāliḥ” in Arabic, is the name given by seventeenth-century Ottoman sources to an archaeological site and the associated ancient and modern settlement. It corresponds to present-day al-Ḥijr and to the ancient city of Hegra, named Ḥijrā in the Nabataean inscriptions mentioning it. In the Greek and Latin sources, it appears as Egra and Haegra respectively. In 2008, it was the first Saudi Arabian site registered on the Unesco World Heritage list.

Geography and environment

Madāʾin Ṣāliḥ is located in Northwest Arabia, in the Ḥijāz, ca. 300 km northwest of Medina as the crow flies, and 185 km east of the Red Sea coast at the latitude of al-Wajh. It lies in a large alluvial plain at the foot of and east of a basalt mountain (ḥarrah), Jabal al-ʿUwayriḍ, and is crossed by a northeast–southwest wadi. The closest town, 20 km to the south, is al-ʿUlā (ancient Dadan). The area is characterised by the presence of sandstone outcrops of various dimensions and heights (Fig. 1). They are part of geological formations dating back to the ancient Palaeozoic — beginning of the Primary Era, ca. 500 million years ago — and are all that remains after the severe erosion that affected the region at the end of the Tertiary Era — between 13 and 5 million years ago. The local climate is arid to hyper-arid, with average annual rainfall of around 50 mm. The vegetation consists of shrubs and bushes, particularly Haloxylon salicornicum (rimṯ in Arabic), as well as herbaceous plants and trees such as acacia and tamarisk.

Despite low annual precipitation, groundwater resources are relatively abundant, supplied by underground streams coming from Jabal al-ʿUwayriḍ. This groundwater facilitated the development of an oasis and an urban site. It was accessed through wells (Fig. 2), up to 7 m in diameter. About 130 such wells have been identified, dated to the Nabataean period. Since Antiquity, the local oasis-type agrosystem (Bouchaud 2013) was based on the culture of the date palm in the upper levels, fruit-bearing trees (fig, olive, pomegranate, grape) in the middle levels, cereals (barley, wheat) and, to a lesser extent, legumes (lentil, chick pea, pea and fava bean) in the lower levels. Cotton was also grown locally, as shown by recent discoveries.

History of research

The site’s first western visitor was Charles Doughty, an English writer and traveller, who spent two months there between 1876 and 1878 (Doughty 1888). He was followed, in the early 1880s, by Charles Huber (Facey 2022), a French explorer — murdered in 1884 around Jedda — who was accompanied on his journeys by the Strasburg Orientalist Julius Euting. The accounts of these two pioneers were surpassed by those of two Dominican priests from the École biblique et archéologique française in Jerusalem, Antonin Jaussen and Raphaël Savignac. After three successive journeys to the site, in 1907, 1909, and 1910, they published five volumes containing a documentation that remained unparalleled in the scholarly literature for a long time (Jaussen & Savignac 1909–1922). They were responsible, in particular, for the first typology of Nabataean rock-cut tombs, as well as for the publication of several hundred inscriptions in Nabataean and other scripts (Dadanitic, Ancient South Arabian, Thamudic, etc.). Few visits were undertaken between the two World Wars. From the 1960s onwards, the names of the Canadians Frederick V. Winnett and William L. Reed (1970), and those of British archaeologists Peter Parr and Gerald L. Harding (Parr et al. 1970 and 1972) stand out. The first Saudi excavations took place in 1966 and a 1463-ha-archaeological park was created in 1972. Aerial photographs and photogrammetric restitutions of the Nabataean tomb facades were made by the French National Geographic Institute (IGN) in 1978, who is also responsible for the numbering of the tombs. Some of the latter were cleared by the Saudi Department of Antiquities in 1985 and the site was excavated by the same department between 1986 and 1990, and again in 2003 under the direction of Daifallah al-Talhi (see the reports published in the Saudi archaeological journal, Atlal). Finally, in 2002, a large-scale Franco-Saudi exploration and excavation project, co-directed by the present author, was established, the reports of which are all available online. Over the last twenty years, this project has uncovered significant parts of the site’s history.

The site in the ancient sources

In the second century CE, Hegra is mentioned in the Geography of Ptolemy as a town in Arabia Felix. Earlier, shortly before the Christian era, Strabo (Geog., XVI.4.24) speaks of a village by the sea named ‘Egra’ which is said to be in the territory of the Nabataean king Obodas (30–9 BCE). However, the identification of this place name remains uncertain since the archaeological site of Hegra is not by the sea. In the present state of knowledge, the identification of Strabo’s Egra — Hegra? The port of ancient Hegra? Another site? — remains unclear. It is also possible that Strabo was partly mistaken in his statement. Finally, in his list of towns and peoples from the interior of Arabia, Pliny the Elder (HN, VI.157) mentions among others the towns of Haegra and Domata (modern al-Jawf).

From the ninth century onwards, al-Ḥijr is often mentioned in the Arabic sources. The recurrent theme of these sources is the story of the prophet Ṣāliḥ, after whom the site was named. Indeed, according to the Arab-Muslim tradition, this prophet attempted, in undefined ancient times, to convert the local inhabitants — members of a tribe named Thamūd — to monotheism. As a sign of his prophecy, he sent them a miraculous she-camel which produced as much milk as it drank water. But the Thamūd refused to be converted and killed the animal in a gorge, which the local tradition locates at Mabrak an-Nāqa, ‘the passage of the she-camel’, 16 km northwest of Madāʾin Ṣāliḥ — thus provoking divine anger and the extermination of the Thamūds. This is reported in the Qurʾan (surah al-ʾAʿrāf, 7. 71–73, and surah al-Ḥijr, 15. 80–84), as well as in the works of Arab scholars (Ṭabarī, al-Muqaddasī, Yāqūt, etc.). According to the same tradition, the houses of the Thamūds might have been the rock-cut tombs that can be seen today, but these are without doubt Nabataean tombs. When the Prophet Muhammad passed through al-Ḥijr during the expedition against Tabūk in 630 CE, he forbade his followers to drink the water from the wells and to eat the bread kneaded with this water. In the Arabic sources, the site of al-Ḥijr is described as a village or small fortified town, and according to the tenth-century geographer al-Muqaddasī (Kitāb aḥsan al-taqāsīm, 84), it was a region known for its numerous wells and cultivated fields. In the fourteenth century, Ibn Baṭṭūṭa (Travels, From Damascus to Mecca, 260) reported that “their bones [of the Thamūds] lie crumbling inside these houses [the tombs]”.

Description of the site and results of the archaeological work

Apart from the oasis, which extends into the north-western and western parts of the archaeological park and where the vast majority of the Nabataean wells were dug, the site is composed of three major groups of archaeological remains: in the centre, a 52-ha ancient settlement — or Residential Area (Fig. 3) — surrounded by a ca. 3km-long city wall built mainly of mud bricks in the first century CE; surrounding the settlement, four major necropolises, a number of isolated rock-cut tombs, and pit tombs, all of which belong to the Nabataean period. Finally, northeast of the city is the Jabal Ithlib, where the majority of the religious monuments are located. To these should be added more than three hundred cairns, particularly on top of the western outcrops, which belong to an earlier phase of site occupation. One of them was excavated in 2014 by the Franco-Saudi project (Abu-Azizeh 2019). It is a tower tomb (Fig. 4), connected to two sections of walls, each of which has three aligned internal rectangular compartments. It was radiocarbon dated between 2113 and 1892 BCE (Early Bronze Age). Since the closest parallels are cairns found near Taymāʾ, it is possible that the cairns at both sites belong to a burial tradition specific to Northwest Arabia.

The Residential Area

The city was accessed through at least three, and possibly up to five gates, one of which, in the southeast, was fully excavated. It is well preserved and contains in its walls reused blocks bearing Greek and Latin inscriptions (Fiema et al. 2020). Thirty-eight towers or bastions abutting the city wall were identified and their original number may have reached eighty-one. Within the city wall, the majority of the buildings are domestic and were occupied between the 5th/4th centuries BCE and the 4th/5th CE, the oldest part of the town being in the western part. The houses are built mainly of mud bricks which, from the first century CE onwards, sometimes stand on stone substructures. In the Nabataean and Roman periods—the earliest houses are not well enough preserved to determine their layout—the rooms opened onto courtyards, sometimes equipped with large circular stone basins. The roofs were flat and built of light materials while most of the floors were of beaten earth, exceptionally paved with mud bricks or stone slabs. The street pattern was not rectilinear and showed a dense urban design, while some spaces, particularly on the edges of the city, remained empty.

There were at least two large monuments in the city: a Nabataean sanctuary in the central part and a Roman fort on the southern edge. The Nabataean sanctuary — IGN 132 — is the main religious monument at the site (Fig. 1). It was built at the end of the first century BCE and remained in use throughout the first century CE. In the second century, the soldiers of the Roman garrison stationed at Hegra used it to worship their gods, including Jupiter Damascene, mentioned in a Latin inscription (61279_I01) discovered in the temple. The latter is contained within a large temenos, a Nabataean maḥrametā or ‘sacred space’. It is composed of two parts: installed on top of an outcrop overlooking the ancient city, an upper temple with a paved platform on which stood a tetrapylon. At its foot, the plan of a lower temple dating to the Nabataean period is undetermined because of subsequent occupations. The organisation of the monument, which has no equivalent elsewhere in Nabataea, may be related to the celestial deity it was possibly devoted to: either šmyʾ, i.e., the Heavens, mentioned in a Nabataean inscription carved on a block reused in the masonry of a nearby building, or the sun, whose trajectory can be followed from the top of the hill.

The second large monument is the Roman fort (Fig. 5), built in the first half of the second century CE along the southern section of the city wall, at the foot of a rocky outcrop. A Roman stronghold probably stood on top on the outcrop, but has now been almost entirely obliterated by 20th-century stone-extraction activities. The fort is a large building, 85 x 70 m, with rectangular towers in the outer wall, two projecting corner towers, and a gate flanked by another two towers. The southern section of the wall is irregular as it had to incorporate the — narrower — Nabataean sections of the city wall. The barracks, best preserved in the south-eastern corner of the fort, consist of several two-room deep units best interpreted as contubernia, each double room accommodating a squad of eight soldiers, the front room (arma) serving as storage and living quarters and the rectangular rear room (papilio) containing sleeping quarters. The fort lost its military function at the end of the third century-beginning of the 4th century CE and civilian occupation continued in parts of it into the 5th century. Large quantities of bronze objects were brought to light during the excavations, including armour and horse harness parts, a finger from a larger-than-life-size statue, a figurine of Harpocrates, one of a Priapus, and one of a ram.

Finally, seven areas corresponding to possible pottery workshops were identified in the Residential area by a pedestrian survey and three of them were excavated in 2022. Two are rectangular, one is circular, and they are all characterised by non-permanent cooking chambers built in mud bricks that were dismantled/destroyed after each batch. They may not have been used simultaneously but they all seem to belong to the Nabataean occupation phase of the site.

The necropolises

Four major necropolises surrounded the Residential Area in the Nabataean period. They house 109 funerary monuments inside which a burial chamber was cut. They are rock-cut monuments, ninety-three of which bear a decorated façade (Fig. 6). Their dimensions vary between 1.65 x 2.70 m (IGN 3) to 13.80 x 21.50 m for the largest completed facade (IGN 110, also known as Qaṣr al-Farīd). Thirty-five tombs bear a legal Nabataean inscription on the facade, carved inside a cartouche, which gives the name of the owner(s), identifies the persons who had the right to be buried inside, lists all the forbidden actions — such as selling the tomb or removing a body from it — and contains a cursing formula (Healey 1993). They were all family tombs and the existence of kinship between the individuals buried inside was proved by anthropological studies. Occasionally, the name of the stone carver who built the tomb is mentioned. The dates mentioned in the texts all range between 1 and 75 CE, which is probably the interval during which most of the tombs were carved. The owners of the tombs were men, women, or couples, and they belonged to the city’s important families, in particular to military officers or governors.

Different tomb types bear a decorated facade (Nehmé 2015), typical of Nabataean monumental funerary architecture, comparable, but not entirely similar, to those in Petra: tombs with one row of crowsteps (12); tombs with two rows of crowsteps (14); tombs with half-crowsteps (8); tombs of the so-called proto-Hegra type (36), the façade of which is framed by pilasters, the crown consisting of two opposing symmetrical half-crowsteps and the entablature consisting of an architrave, sometimes a frieze, and an Egyptian cornice; tombs of the so-called Hegra type (14), which differ from the preceding type by the presence of a second entablature separated from the first by an attic; finally, one tomb with a crown consisting of a semi-circular arch.

The burial chambers were cut with less care than the facades. They comprised a door and contained a variable number of funerary structures (average number 6/7), each of which could hold several bodies. These are either rectangular pits dug directly into the bedrock, burial cells cut into the walls — wider than they are deep, or burial niches also cut into the walls — deeper than they are wide.

Ordinary people were buried in simple rectangular pit tombs cut in the rocky substrate, most frequently on top of the sandstone massifs. About 1,600 such tombs have been identified.

Excavations of seven monumental Nabataean tombs at Hegra between 2008 and 2015 by the Madāʾin Ṣāliḥ Archaeological Project have yielded large quantities of human remains, as well as various kinds of objects: fragments of wooden coffins or shrouds, objects from daily life, pieces of jewellery, arrow heads, and ceramic vases containing liquid or solid offerings to the dead. The work of the various specialists who studied this material reconstructed the burial process of a Nabataean individual (Fig. 7), which consisted of six steps from the preparation of the deceased in his/her house to placing the deceased in the funerary structure inside the tomb (Bouchaud et al. 2015, Nehmé 2015: 88–91; Nehmé 2022: 121–127).

The Jabal Ithlib and the Nabataean fraternal societies

Located in the northeast part of the site, the Jabal Ithlib refers to an impressive group of sandstone outcrops and peaks. It has long been assumed that Jabal Ithlib was the religious area of ancient Hegra as that is where most of the religious monuments are found. These consist mainly of rock-cut niches with more or less elaborate decoration (Fig. 8). The majority contain one, two, or three betyls — an upright stone carved in relief and symbolizing the Nabataean deities. These small rock-cut monuments are often associated with meeting places for religious brotherhoods — Nabataean marzeḥā (See Banqueting) — which met at regular intervals and were placed under the protection of the deities they honoured. They are described by Strabo (Geog., XVI.4.26) as comprising thirteen persons, including two musicians. The assemblies — at least six in the Jabal Ithlib — met in triclinia, the largest of which, called the Dīwān (Fig. 9), stands at the entrance of the gorge leading to the interior of the Jabal, which was well protected by the surrounding mountains.

Conclusion

Hegra is one of the major archaeological sites in the Arabian Peninsula, famous primarily for its rock-cut monumental tombs. Excluding the Bronze Age funerary cairns, the first inhabitants of the oasis may have been the Lihyanites, who are suggested to have minted the so-called owl coins, several hundred specimens of which were found in the excavations. The Nabataeans arrived there in the middle of the first century BCE and exercised political authority over the area until 106 (or 107) CE, when the Romans annexed the Nabataean kingdom (See Roman Province of Arabia, Rome [and Arabia]). The Romans controlled the area militarily until the end of the 3rd-beginning of the 4th century, and stationed the Third Legion Cyrenaica and auxiliary troops from the ala Gaetulorum and ala dromedariorum (Gatier 2020) there. It is not clear when the Romans left Hegra but they probably did so along with the military. The civilians who appear later in the sources — for example in a Nabataean 356 CE inscription (Stiehl 1970) — and who had administrative responsibilities in the city, bore Arabian or Jewish names. Arabs and Jews are likely to have played a leading role in the municipalities of Northwest Arabia from the third century onwards.

Laïla Nehmé

References and suggested reading

Sources

• Ibn Baṭṭūṭa / Gibb, H.A.R. 1958. The Travels of Ibn Battuta, A.D. 1325–1354. Translated with revisions and notes from the Arabic text edited by C. Defrémery and B.R. Sanguinetti. Cambridge: The Hakluyt Society, Cambridge University Press.
• Al-Muqaddasī / De Goeje, M.J. 1967 (3rd edition). Kitāb aḥsan al-taqāsīm fī maʿrifat al-ʾaqālīm (Bibliotheca geographorum Arabicorum, 3). Leiden: Brill.

Studies

• Abu-Azizeh, W. 2019. Funerary Cairns in Northern Arabia, in L. Nehme & A. Alsuhaibani, AlUla. Wonder of Arabia: 36–39. Paris: Gallimard, Institut du Monde Arabe.
• Bouchaud, C. 2013. Exploitation végétale des oasis d’Arabie. Production, commerce et utilisation des plantes. L’exemple de Madâ’in Sâlih (Arabie Saoudite) entre le IVe siècle av. J.-C. et le VIIe siècle apr. J.-C. Revue d’ethnoécologie 4. Online: https://journals.openedition.org/ethnoecologie/1217. DOI: 10.4000/ethnoecologie.1217.
• Bouchaud, C., I. Sachet, P. Dal-Prà, N. Delhopital, R. Douaud & M. Leguilloux 2015. New Discoveries in a Nabataean Tomb. Burial Practices and ‘Plant Jewellery’ in Ancient Hegra (Madâ’in Sâlih, Saudi Arabia). AAE 26: 28–42. DOI: 10.1111/aae.12047.
• Doughty, Ch. 1888. Travels in Arabia Deserta. Cambridge.
• Facey, W. 2022. Charles Huber: France’s Greatest Arabian Explorer. With a Translation of Huber’s First Journey in Central Arabia, 1880–1881. Surbiton: Arabian Publishing.
• Fiema, Z.T. & F. Villeneuve 2018. The Roman Military Camp in Ancient Hegra, in C.S. Sommer & S. Matešić (eds) Limes XXIII. Proceedings of the 23rd International Congress of Roman Frontier Studies Ingolstadt 2015 (Beiträge zum welterbe Limes. Sonderband 4/II): 702–711. Mainz: Nünnerich-Asmus Verlag.
• Fiema Z.T., F. Villeneuve & T. Bauzou 2020. New Latin and Greek Inscriptions from Ancient Hegra. ZPE 214: 179–202.
• Gatier, P.-L. 2020. The Rock Graffiti Carved by Roman Auxiliary Troops at Hegra, in L. Nehmé (ed.) Report on the 2018 and 2019 Seasons of the Madâʾin Sâlih Archaeological Project: 79–112. Paris. https://hal.archives-ouvertes.fr/hal-02869017
• Healey, J.F. 1993. The Nabataean Tomb Inscriptions of Mada’in Salih (Journal of Semitic Studies Supplement, 1). Oxford: Oxford University Press.
• Huber, Ch. 1891. Journal d’un voyage en Arabie (1883–1884). Paris: Leroux.
• Jaussen, A. & R. Savignac 1909-1922. Mission archéologique en Arabie (Publications de la Société Française des Fouilles Archéologiques, 2). Paris.
• Nehmé, L. (ed.) 2015. Les tombeaux nabatéens de Hégra (Épigraphie & Archéologie, 2). Paris: Académie des Inscriptions et Belles-Lettres.
• Nehmé, L. 2022. Guide to Hegra. Archaeology in the Land of the Nabataeans of Arabia. Paris: Skira.
• Parr, P.J., G.L. Harding & J.E. Dayton 1970. Preliminary Survey in N.W Arabia, 1968. Bulletin of the Institute of Archaeology 8–9: 193–242.
• Parr, P.J., G.L. Harding & J.E. Dayton 1972. Preliminary Survey in N.W. Arabia, 1968. Bulletin of the Institute of Archaeology 10: 23–61.
• Stiehl, R. 1970. A New Nabatean Inscription, in R. Stiehl & H.E. Stier (eds) Beiträge zur alten Geschichte und deren Nachleben: Festschrift für Franz Altheim zum 6.10.1968. II: 87–90. Berlin: W. de Gruyter.
• Winnett, F.V. & W.L. Reed 1970. Ancient Records from North Arabia. Toronto: University of Toronto Press.

_{Alternate spellings: Madâʾin Ṣâliḥ, Madâʾin Sâlih, Madaʾin Salih, Mada’in Salih, Madain Salih, Mada’in Saleh, Madain Saleh, al-Ḥijr, al-Hijr, Hijr, Hégra, Hegra, Ḥijrā, Ḥijrâ, Ḥijra, Hijrâ, Hijra}