Version classiqueVersion mobile

Chalcis/Qinnasrin (Syrie)

Marie-Odile Rousset

The cities of Qinnasrin and Chalcis from the Bronze Age to the medieval period

Marie-Odile Rousset

Texte intégral

1The site of Chalcis ad Belum/Qinnasrin lies under the modern village of al‑‘Iss, 25 km south-west of Aleppo. It is located in the foothills of the easternmost part of the calcareous massif of northern Syria, thus in a zone between agricultural areas in the north-west and steppe lands in the south-east. The archaeological research programme at Qinnasrin (Syria) has been implemented since 2003 in collaboration with the Syrian Directorate of Antiquities and Museums, and the Archaeological Museum of Aleppo. Work conducted in 2008-2010 at al‑‘Iss enabled us to determine the extension of Qinnasrin from the beginning of its settlement (Bronze Age) up to the last phases of Islamic occupation. The vast corpus of written sources and aerial photographs, 3 excavated trenches, geomagnetic surveys, topographical maps, and the surveys by foot all enabled us to confirm the location of historical Qinnasrin, and to discover that the site was much larger than previously believed: it extends more than 2.7 km from north to south by 1.7 km from east to west (Chapter 6, fig. 3). Depending on the period, different parts of the site were settled.

2The earliest archaeologically attested period is that of the transition from the Early Bronze Age to the Middle Bronze Age (Chapter 16; Chapter 19, fig. 18a). It corresponds to the establishment of a vast necropolis on the summit and slopes of the Jabal al‑‘Iss (Chapter 6, fig. 70‑77). At the north-western end of the massif, the geophysical map shows an oval construction, 15x12 m (Chapter 6, fig. 64), which could be a collective tomb similar to the royal tombs of Bahrain where rulers of Amorite origin are buried (Laursen 2017, p. 132, fig. 180). These tombs can be related to small settlements located near the river and to the ancient occupation of the tell (Albright, Dougherty 1926, p. 8).

  • 1 Geographical and archaeological survey programme directed by Bernard Geyer (CNRS, Archéorient) and (...)

3On the top of the mountain, a major strategic location, a fortress was discovered. It is about 700 m wide with an interior area of about 35 hectares. Two sections, in P409 (Chapter 6, fig. 48‑49) and P86, indicate that this fortification was built in mud brick and reinforced with calibrated rubble blocks bound with earth. Inside the fortress, the geomagnetic survey showed several buildings that were quite distinct from one other. At least some of them were built during the Bronze Age, as pottery from this period is abundant on the surface and in holes dug for looting along two walls (Chapter 6, fig. 61). One building stands out clearly from the others, in the centre of Qinnasrin’s fortification on the highest point. The magnetic response to geophysical survey is very strong and suggests a construction of basalt blocks measuring 80 m per side (Chapter 6, fig. 62). The presence of a large Middle Bronze fortress at Qinnasrin corresponds to a type of settlement found in the Arid Margins area. 1 Qinnasrin fits perfectly into the defence network of the Middle Bronze Age (1850‑1600 BC), observed at the eastern edge of the Jabal al-‘Ala (Rousset et al. 2017). Analysis of regional settlement in the Middle Bronze Age shows that Qinnasrin occupies a hierarchical position within this network just below that of the capitals of the kingdoms of Aleppo and Ebla, from which the site is equidistant. One of the roles of Qinnasrin was to maintain surveillance over the major military route that passed at its feet. The identification of the site in the written sources remains to be found.

  • 2 Bagg 2007, pp. 197‑198 ; Tadmor, Yamada, Novotny 2011 (Tiglath-Pileser III, 35, ii 8’).

4The earliest mention of the toponym “Qinnasrin” appears in two inscriptions by Tiglat-Pileser III, who recounts his expeditions to Syria between 745 and 727 BC. The stele mentions “the city of Qinasrina [URU.qi‑na‑as‑ri‑na] which is in the steppe”. 2 As in many other cases, this name, transmitted from Aramaic to Arabic, has persisted up to modern times, in parallel with the Greek name, with or without continuity of occupation of the site.

5The settlement of Qinnasrin in the 8th century BC was therefore considered important enough to be mentioned in these commemorative inscriptions. However, no pottery has been found that corresponds to this period. The Iron Age settlement may correspond to an early phase of the tell (Chapter 6, fig. 10f).

6The name “Chalcis” itself implies a Greek foundation as it refers to the great city of Chalcis in Euboea. However, the thesis of the foundation of the city by Seleucos I (358‑281 BC), mentioned by Appian (Todt, Vest 2014, p. 1048), has not been confirmed by the pottery discovered. On the other hand, it is an unambiguous evidence for an occupation at the end of the Hellenistic period, in the 2nd century BC (Chapter 18). This dating is consistent with evidence of occupation at that time: in 145‑144 BC, Diodotus Tryphon, an usurper claiming to represent the Seleucid king Antiochos VI, quartered his army near Chalcis (Todt, Vest 2014, p. 1049). The city would then have been at the centre of an Arab principality (Chapter 1).

7According to observations made on the site, the occupation was then essentially located on the tell and to the south-west (Chapter 20, fig. 1). It corresponds to the building of a fortress on the pre-existing tell, with the probable construction of a glacis and, at the top, a wall reinforced with towers, having an entrance to the south-west. Several communication routes follow the orientation of the traces revealed by geophysical survey on the tell (Chapter 6, fig. 10e). They are perceptible in the current grid of the village and in the aerial images (Chapter 6, fig. 1 and 2).

8About 1 km east of the tell, on the banks of the Quwayq, site P36 also produced late Hellenistic pottery, including many luxurious imports (Chapter 18). This site could have been a residence by the river.

9These results confirm the principle of development observed elsewhere in Syria (Will 1989; Leriche 2000), according to which the Hellenistic foundations are above all military in nature. There would first have been a citadel/fortress, and only later a city of Hippodamian plan.

10According to J. Aliquot, the “kingdom of Chalcis” must indeed be located at Chalcis ad Belum, the Chalcis of northern Syria. This kingdom existed from at least 41 AD until Vespasian’s takeover and reorganization of the regional territories by 75 AD at the latest. The coinage of Chalcis, with a special production in the name of Flavia Chalcis, continued until 169 (Chapter 1). Chalcis may have had the status of a colony in the first half of the 3rd century AD, as suggested by the text of the inscribed basin studied by P.‑L. Gatier (Chapter 2).

11The urban structure in Roman times reflects the importance of the city and was a continuation of the Hellenistic occupation (Chapter 20, fig. 2). The distribution of the pottery indicates that the tell continued to be occupied, probably as an acropolis of the city, while the district at its feet was expanding (Chapter 19, fig. 18d). Alignments identified by geophysical survey on the tell follow the same orientation as its southern rampart and the main streets of the city (Chapter 6, fig. 10d). Vestiges of a large fortified city appear to have been preserved in the later constructions and even in the present street pattern and the layout of the recent canals. In P408, the digging of a modern canal led to the discovery of a wall 2.70 m wide (Chapter 8). The location and orientation of this wall correspond to a clear line, perfectly visible on the aerial images, which extends for a total of approximately 880 m (Chapter 8, fig. 2). It could have been a rampart of the Roman city, destroyed at a time when the surface area of the city was reduced. Other evidence enables us to complete the plan of the fortification on the other sides (Chapter 20, fig. 2).

12Between the modern town and the mountain is an area of quarries, and a necropolis. These marked the northern limit of the classical city. The quarries were used from the late Hellenistic-Roman period (Chapter 13) up to the Islamic period. Some of the quarrying activity cut into and destroyed ancient hypogea bearing inscriptions and decoration (Chapter 14). These tombs were dug into the calcareous mountain in a typical layout, that is, with a dromos, a central chamber, and 3 alcoves, each with loculi (Monceaux, Brossé 1925, fig. 2). A statue from the Roman period was found in a cave in 1919 (Monceaux, Brossé 1925, p. 348; Seyrig 1934, pl. XXI; Seyrig 1970, pp. 94‑95; Chapter 20, fig. 3).

13Several ancient blocks inventoried in the present village are ascribable to this phase of the town of Chalcis. These vestiges show that at least one large monument was built in Roman times at Chalcis, perhaps a temple in the central part of the town (Chapter 15, fig. 16a), as well as a monument at the crossing of the cardo and the decumanus (Chapter 15, fig. 14f, 20c and 20d) as at Palmyra (Balty 2000, p. 228). On the aerial image of 1960 (Chapter 6, fig. 2), material recovery trenches reveal the plan of a large rectangular construction on the eastern edge of the modern village – could this be the forum? In P83, flattened walls built in cut stone could belong to a large construction (Chapter 6, fig. 18). The oval depression, 118x86 m, is in the south-west corner of the city (Chapter 6, fig. 17) – could this correspond to the remains of an amphitheatre? Its dimensions and layout (on the outskirts of the city) are comparable to those of the amphitheatres of Anavarza/Anazarbus (Posamentir, Sayar 2006, pp. 350‑351), Palmyra (Hammad 2006‑2007, p. 88), Bosra (Dentzer-Feydy 2007, p. 207) and Caesarea (Weiss 2014, p. 112). Finally, a possible location for a theatre would be the southern flank of the north mound, where the steep slope would have provided a good base for the location of such a building.

14In Roman times, Chalcis was an important road junction: it is mentioned in the Table of Peutinger (between 200 and 450 AD) and in the Itinerary of Antoninus (around 300 AD). The city lies at the crossroads of the road from Beroia/Alep to Apamea of Syria, as well as that from Antioch to Litarba/Atharib, Gabbula/Jabul/Zebed and Barbalissos/Balis on the Euphrates (Todt, Vest 2014, p. 1049). In the field, these roads can be located as passing through urban gates, which in turn can be located by the presence of large dumps outside the city.

15The plan of Chalcis in Roman times, based on the elements described above, would correspond to a slightly deformed rectangle, approximately 1,320 m by 1,200 m, having 8 gates, the north and west corners being truncated. Its surface area would correspond to about 145 hectares, comparable to the largest of the Hellenistic cities that developed in Roman times (Will 1989, p. 474): larger than Aleppo (Sauvaget 1941a, pl. LII: about 110 ha) and Damascus (135 ha) but comparable to Antioch (150 ha). However, Chalcis would have been much less extensive than Apamea, with its 250 hectares (Balty 1991, p. 16), Laodicea, estimated at 225 hectares but measuring 170 hectares without the harbour (Sauvaget 1934, p. 128), or Baalbek, whose ancient imperial city (late 1st century BC to late 1st century AD) also covers 170 hectares (Lehmann 2015, fig. 154 and 156). This is perhaps not surprising for a city that would have had the status of a colony.

16The end of this period could have been around 253 AD, after the capture of the city and that of Antioch by the Sassanid Persians of Shapur I (Todt, Vest 2014, p. 1049). Indeed, there appears to have been a decline of the city in the 3rd‑4th centuries. (Chapter 19, fig. 19a). The end of this period would have corresponded to the relocation to Chalcis by Emperor Julian (360‑363 AD) of all the inhabitants of Anatha, now “Ana on the Euphrates” (Bowersock 2002, p. 49). This sudden increase in population would explain the higher density of the city observed beginning with the end of the 4th century.

17During the Byzantine era, Chalcis was of major strategic importance in the defence system of northern Syria. But several hagiographical accounts indicate that it was also a centre of traditional Hellenism, which played a fundamental role in the emergence of Christianity (Chapter 4). According to the fieldwork, the period of the 5th‑6th centuries corresponds to the maximum extension of the occupation as seen in the abundance of pottery (Chapter 19, fig. 19b).

18Chalcis was refortified by Justinian in 550‑551 AD. The reconstruction of the wall is documented by the inscriptions on the lintels of the city gates, including the one still in place on the western gate, as well as by Procopius of Caesarea (Fourdrin 1994, p. 305; Feissel 2009, p. 84; Chapter 2). Our observations made it possible to complete the plan which had been drawn up by J. Lauffray (Mouterde, Poidebard 1945, pl. I; Introduction, fig. 6) and thus to clarify certain elements found in the texts. The northern part of the wall could have been a reconstruction along the lines of the Roman rampart, restored or used as a foundation for the new wall (Chapter 20, fig. 4). The western, eastern and southern walls would have been entirely constructed.

19Procopius, in his description of the work carried out at Chalcis under Justinian, also mentions the reinforcement of the ramparts by “external constructions” (II, 11, 1 and 8). The survey at Chalcis led to the discovery of elements of this proteichisma. The sections studied in the schoolyard, west of the Byzantine rampart, show that a large trench was dug through the Roman layers (Chapter 7). In our trench A (Chapter 9), we revealed a large ditch, parallel to the north-western rampart, which corresponds to a geomagnetic anomaly (Chapter 6, fig. 38) as well as a trace on the satellite images, about 200 m long (Chapter 6, fig. 1). Its profile (Chapter 9, fig. 22) is very similar to that of the ditch outside the rampart of the city of Sura (Othman 2018, pp. 534‑535).

20In addition to the western gate still preserved today, there was another gate at the northern corner, visible on the geophysical map (Chapter 6, fig. 34a). The 6 Byzantine gates are arranged along the same main lines as in the previous period. All these constructions could have been built by reusing as much as possible the building materials of the ancient rampart.

21On the tell, the acropolis was remodelled with the construction, to the east, of a monumental entrance overlooking the city. The geomagnetic map (Chapter 6, fig. 9, 10b and 10c) shows a road network aligned with the street leading from the new entranceway to a fairly large construction that could have been an official building. Another street, perpendicular to the first one, leads to a circular building with an apse in the south-east corner of the tell (Chapter 6, fig. 9). This is a church whose plan consists of two concentric circles of 35 m and 22 m in diameter and a semicircular eastern apse, about 14 m wide. Although rare, this circular plan was known in Syria in the 6th century and is thought to be related to the cult of the martyrs (Lassus 1947, pp. 150‑155). Among the plans strictly comparable to that of Chalcis, the “rotunda” of Apamea and the church of Beisan/Scythopolis are larger (Balty 2013, pp. 204‑205), while the building discovered in Germia has the same proportions (Niewöhner et al. 2013, pp. 122 and 127). It is perhaps this construction on the acropolis that is the subject of an episode in the lives of the monks Alexander (late 4th to early 5th century) and Rabbula (412‑435 AD; Gatier 1995, pp. 445‑446; Gatier, Rousset 2016).

22In today’s village, the inventory of the architectural blocks indicated the presence of a church at the central point of the Byzantine town (Chapter 15). On one of them, a Syriac inscription, mentions the name of a priest (Chapter 3). At least three other churches may also have existed intra-muros, although only two of them are known in the literature (Todt, Vest 2014, p. 1056). Finally, a block of basalt indicates that public latrines and probably baths existed in the Byzantine city (Chapter 15, fig. 34a).

23The structures identified by geophysical survey in the northern sector correspond to large buildings. In trench A, there are walls associated with Byzantine levels (Chapter 9); in trench B (Chapter 10), a bath already existed at that time.

24An occupation of an artisanal nature is attested by the presence of smaller buildings and by more than a hundred kilns, most of them potters’ kilns (Chapter 6, fig. 34a). Some of them produced imitations of Phocean late terra sigillata pottery and imported amphorae (Chapter 19). Trench A (Chapter 9) produced evidence for an industry related to the exploitation of plant soda ash, for the production of ash for soap fabrication or for the manufacture of primary glass. This activity corresponds to the Byzantine and early Islamic periods. These districts outside the walls, to the north and south of the city, may also have had an economic purpose, as areas for unloading goods brought by caravans (hadirs).

25An isolated building (P63) located close to the route to Antioch, in which rich material and a mosaic with a religious representation were found, could correspond to a monastery. Georges Tchalenko (1953‑1958, III, p. 103) mentions a monastery of Mar David in Chalcis and five other Monophysite establishments. Could this supposed monastery be identified with that of Mar ‘Aqiba, located on the outskirts of the city and mentioned from 475 onwards (Todt, Vest 2014, p. 1055)?

26On the hill overlooking the site to the north are the remains of a settlement in the form of caves (Chapter 6). One lintel above a cave entrance is decorated with an engraved cross. A group of hermits may have lived in these “caves” linked to a monastery on the mountain. This type of settlement is known, for example, in the Jerusalem area, where there are monasteries with cells scattered along the wadis (Patrich 1994). Evidence for the presence of monks living in caves in the “Desert of Chalcis” from the end of the 4th century and the existence of several monasteries in Chalcis itself support this hypothesis.

27The urban plan of the town of Chalcis in Justinian’s time presents strong parallels with several other contemporary towns. As at Cyrrhus and Dibsi Faraj, the ramparts partly follow the layout of the earlier ramparts (Abdul Massih 2018; Harper 1980, p. 328 and fig. F). The square projecting towers on the outside and their spacing, the system of gates, the citadel with its monumental city gate and postern opening directly to the outside and its location on one side of the urban fortification, as well as the profile of the moat, appear to correspond to Justinian’s fortification standard (Feissel 2000b; Ulbert 2000).

  • 3 To be comparable, the surfaces were determined inside the ramparts and calculated using the Google (...)

28The area enclosed by the ramparts is 78 hectares, i.e. half of the ancient city (Chapter 20, fig. 2). It is equivalent to that of Cyrrhus in the same period (72 ha; Abdul Massih 2018, p. 399) but larger than those of the cities on the Euphrates restored by Justinian: Balis/Barbalissos (13 ha), Dibsi Faraj/Neocaesaria (27 ha), Sura (43 ha), Resafa/Sergiopolis (21 ha), Raqqa/Callinicum (55 ha after Gaborit 2012, II, p. 445; al‑Khalaf, Kohlmeyer 1985), Halabiyya/Zenobia (10 ha). 3

29There does not appear to have been any radical change in the form of the city since the Muslim conquest. The accounts of the conquest emphasize the Treaty of Chalcis, signed in 636‑637; some of the inhabitants took refuge in Antioch at that time. According to al‑Sarakhsi (died in 899), the fortifications of Qinnasrin at the end of the 9th century still followed the same plan as in the Byzantine period (Eddé 1984, p. 26). The city continued to play an important military role as a garrison town, and beginning with the reign of Yazid b. Mu‘awiyya (680‑684), as the capital of the newly created “Qinnasrin Jund” governorate of northern Syria. The city was prosperous and exploited a rich agricultural region (Chapter 4).

30The occupation of Qinnasrin in the early Islamic period appears to have been concentrated within the Byzantine walls and in the northernmost parts of the northern sector and the mountain (Chapter 19, fig. 19c‑d; Chapter 20, fig. 5). On the old acropolis, a new road, deriving from the main road linked to the gate, led from the central part of the tell to building P4 located on the highest part of the acropolis (Chapter 6, fig. 10b). Could this be an official building or the governor’s residence? In the central part of the tell, geophysical survey revealed another building (P2), measuring about 25 m on each side, which could be a mosque (Chapter 6, fig. 10b).

31The continued occupation of the northern part of the town is marked by the continued use of the P91 and P406 dumps. The inhabited districts spread both within the fortifications, with the rampart and the ditch still active, as well as outside, in the northern district, with the westward extension of the residential area (Chapter 6, fig. 33b). The activity related to the exploitation of plant soda ash continued (Chapter 9). The use of ash in the manufacture of glass has been revealed by studies of the composition of Umayyad and Abbasid glass at al‑Hadir, 4 km east of Qinnasrin (Gratuze, Foy 2012). It is also one of the essential components of Aleppo soap, the production of which is known for the early Middle Ages (Grandin 1984, p. 145).

32Necropolises continued to be used, which confirms textual information that the inhabitants of the town remained Christian, as Islamicisation did not occur before the end of the 8th century (Chapter 4).

33The relative urban decline observed in the 7th and 8th centuries may have begun earlier and been one of the consequences of the 6th-century crises which affected the whole of northern Syria, with in particular the seizure of the city by the Persians in 611 (Chapter 4), but this remains difficult to determine without excavation.

34The prosperity of Qinnasrin and its region, mentioned in written sources, continued at least until the end of the 9th century. The city still played an important economic and military role, with fortifications and an explicitly named governor until the middle of the 10th century (Chapter 4). This period is well represented in the sectors surveyed (Chapter 19, fig. 19d; Chapter 20, fig. 6).

35Residential housing continued in the northern sector: the bath discovered in trench B (P20), which was associated with one of the previous large residences, represents the last occupation in the 9th century (Rousset, Rochette 2014; Chapter 10). The town ditch, at least in its northern part, was no longer used as an element of fortification but rather as a dump where rubbish was burned (sounding A; Chapter 9).

36A major discovery was made on the mountain above the city, a remarkable observation point that rises between agricultural lands and the steppe (Rousset 2013a, 2015; Chapter 6, fig. 44‑46). The remains found are hardly perceptible: trench C has shown that ashlars of the walls were systematically removed for reuse in the Middle Ages (Chapter 12). The ashlars were extracted from an extensive open area; in the north-eastern corner, four huge tunnels are the remains of ancient chamber quarries, from which the crushed stones for building most of the walls of the fortress were extracted (Chapter 6, fig. 51‑53; Chapter 13). The walled area measures approximately 900x700 m and is largely based on the plan of the Middle Bronze Age fortress. It includes three small tells, which were probably towers (P49, P8 and “Nabi ‘Is”; Chapter 6, fig. 54‑57). The latter are surmounted by a mausoleum. According to the relative chronology (including the destruction of cavities attributed to the 5th­‑7th centuries), the pottery and the construction technique (cutting of and building with stone), the construction of this fortification can be dated to between the end of the 7th and the 9th century. It is not possible to determine whether the imposing building in the centre and at the highest point of the fortification (Chapter 6, fig. 62) could be contemporary to the Islamic fortification. We suppose that this fortress could have belonged to the refortification of the marching zones that were in contact with the Byzantine domains, around the middle of the 8th century. It may have been built by Salih b. ‘Ali (died in 769), one of the first Abbasid governors of Syria, who was in charge of the fortification of the border. It may have been destroyed, at the earliest, by al‑Ma’mun (813‑833) or abandoned together with the city, at the latest in the mid 10th century (Chapter 4).

37The decline of the city seems to have been gradual; mentions of Qinnasrin at the beginning of the 10th century refer to its past greatness and indicate its decline in favour of Aleppo (Chapter 4). At the end of the 10th century and beginning of the 11th, it was several times devastated by the Byzantines.

38From the 11th century to the first quarter of the 12th century, under the rule of the Mirdasids, Seljuks and Zengids, the function of Qinnasrin appears to have been essentially military. The image that emerges from the texts relating to this period is that the site was still of strategic interest, being mentioned as a refuge, a place of battle, and a base for military expeditions. It would seem that the fortifications of the town were limited to those of the ancient acropolis and were restored several times (Chapter 4). There may have been a fortress or fortified village within the pre-existing fortification. The walls recorded by J. Lauffray (Mouterde, Poidebard 1945, pl. I) and a large part of those appearing on the geomagnetic map correspond to this construction, which is fairly dense (Chapter 6, fig. 10a). The size of the village of Qinnasrin would then have been reduced to 8 hectares (Chapter 20, fig. 7).

39Approximately 500 m to the north, a necropolis has provided inscriptions that are contemporary to the occupation of the tell (Chapter 5). Other funerary inscriptions have been found farther north, in fields at P22 and P34, and a small necropolis was excavated in area P20 (Chapter 11). For the 11th and early 12th centuries, the proportion of land occupied by necropolises in relation to constructed areas appears to have been unusually large. This would tend to confirm the use of the area as a military camp and also as a burial place for soldiers who died in battle.

40Qinnasrin hardly appears in the texts of the Ayyubid period (mid 12th to mid 13th century). Archaeology confirms that the settlement, if it existed, must have been very small (Chapter 19, fig. 20b). A few houses may have been built towards the northern rampart of the town, near cavities in the rock. This type of construction is evidence for an occupation of rural character, more oriented towards animal husbandry (Rousset 2011, pp. 127‑129). The mention by Yaqut (who died in 1229) of a small herd that was subject to tithing confirms the pastoral orientation of the site at that time (Gaudefroy-Demombynes 1923, p. 30).

41The occupation of the Nabi ‘Iss hill (P54, P41; Chapter 6, fig. 55‑57) is confirmed by both archaeology and texts, which provide evidence for pilgrimage activity around the mausoleum of Salih (Sourdel 1957, pp. 13‑14).

42Evidence in the texts and in the field demonstrate that another part of economic activity at the time consisted of the recovery of building materials (Chapter 4). This was revealed by the excavations at al-Hadir and sectors B (P20) and C at al‑‘Iss, where the walls were removed (Rousset, Bovagne, Rochette 2012, p. 66‑67; Rousset 2013a; Chapter 10; Chapter 12). This use of the ruins of Qinnasrin as a stone quarry has been intense. The materials were used to build the monuments and fortifications of Aleppo in the Zangid and Ayyubid periods.

43Occupation during the Mameluke period (Chapter 19, fig. 20c) has been revealed in only a few areas. Three of these (P16, P40 and P42) correspond to traces of isolated quadrangular buildings, all located on ancient routes of circulation. At the beginning of the 13th century Yaqut mentions a caravanserai at Qinnasrin where caravans stopped (Gaudefroy-Demombynes 1923, p. 30). One of the buildings found during the survey may correspond to this caravanserai.

44Mameluke and Ottoman sherds found in other areas (Chapter 19, fig. 20d) show a continuity in the occupation of the caves connected to pastoralism.

45At a point of contact between nomads and sedentary people, the settlement of Chalcis/Qinnasrin was one of the landmarks of the road networks of the caravan trade. In addition to this economic vocation, during Late Antiquity a cultural and spiritual dimension was added to the area, with the development of hermitages and monasteries. However, the main reason for the existence of Chalcis/Qinnasrin throughout its history was its military role as a garrison town, as an outpost facing the steppe until Late Antiquity, and then, after the Arab conquest, as the rear base of Islam against the Byzantines and then the Crusaders.


1 Geographical and archaeological survey programme directed by Bernard Geyer (CNRS, Archéorient) and Nazir Awad (DGAMS), in which I participated as an archaeologist. The area surveyed corresponds to Salamya 1/200,000 map. Qinnasrin is located in the north-western angle of this map.

2 Bagg 2007, pp. 197‑198 ; Tadmor, Yamada, Novotny 2011 (Tiglath-Pileser III, 35, ii 8’).

3 To be comparable, the surfaces were determined inside the ramparts and calculated using the Google Earth tool.

Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search