Three excavation seasons were carried out in al-Sinnabra, or better known as the Early Bronze age site of al-karak/Bet Yerah, during February 2018, and February and September of 2019. These renewed excavations were undertaken after the publication of our report, Bet Yeraḥ III: Hellenistic Philoteria and Islamic al-Sinnabra, in which we published the findings of the previous excavations carried in the 1950s. The renewed excavations aimed to touch unexcavated soil in the hope of finding some clear stratigraphy as well as some ceramics and other small finds. A third aim was the exposure of new units. Since remains of a bath and an audience hall were already exposed on site, we thought that there might be a third element that constituted an essential part in any Umayyad place, namely a mosque. Despite the challenges of recent disturbance, undocumented excavation, erosion and landscaping, which often frustrated our attempts to identify the remains of the dismantled palace, preserved only at its foundations, we ultimately succeeded in tracing a significant, completely unexpected annex to the palace situated outside its northern fortification wall.


Report on the third season


Arab historians name al-Ṣinnabra, located at the southern end of the Sea of Galilee, as the seasonal residence of several Umayyad caliphs, among them the first caliph Muʿāwiya.  Despite many citations, its location long eluded certain identification. In 2002, a brief communication by Donald Whitcomb pinpointed al-Ṣinnabra at the well-known Early Bronze Age mound of Bet Yeraḥ or Khirbat al-Karak, in an area in the north of the mound excavated extensively between 1945 and 1953 (Whitcomb 2002).  Tel Bet Yeraḥ/Khirbat al-Karak is situated six kilometres south of Ṭabariyya, the Umayyad capital of Jund al-Urdunn, the provincial district that replaced Scythopolis (Baysān/Beth Shean) the previous provincial capital of the Byzantine Palaestina Secunda.  Ṣinnabra lies on the southwestern shore of the Sea of Galilee, where the Jordan River exits the lake (see fig.1).  The principal post-Bronze Age structure exposed on the site comprises a fort enclosing a basilical building, with a bathhouse attached to the fort’s southern wall.  The fort was initially dated to the Roman or Byzantine period, while the basilica was identified as a synagogue, due to its southward-facing apse and to the discovery of a column base with a menorah carved on it (Applebaum 1988).  North of this complex another building was excavated in 1952–1953 by Pierre Pinhas Delougaz and Richard C. Haines on behalf of the Oriental Institute of the University of Chicago.  This was an ‘Arab building’ built above a Byzantine church, about 50 meters north of the fort (Delougaz and Haines 1960).

Daadli 2019 20photo1 1

Fig. 1. Location of al-Ṣinnabra and neighbouring localities

Whitcomb suggested that the building originally identified as a synagogue, and later as a Roman or Byzantine enclosure, is actually the Umayyad palace referred to in historical sources, while the ‘Arab building’ to its north was a manor of the same period.  Based on Whitcomb’s suggestion, the Tel-Aviv University expedition at Tel Bet Yeraḥ set out in 2009 to recover possible palace contexts not compromised by the massive excavations conducted in 1950–1953 by Pesach Bar-Adon and P.L.O. Guy.  Coins discovered in sealed contexts during the 2009 excavations established a seventh-century terminus post quem for the central structure and an eighth-century terminus post quem for the bathhouse (Da‘adli 2017; see fig. 2).  Subsequent research on the earlier excavations corroborated the date for the fortified enclosure, and further Umayyad remains related to the water system and the approaches to the mound from the west were revealed in recent salvage excavations (Alexandre 2013; 2014), thus adding decisive weight to Whitcomb’s identification.

According to the medieval geographer Yāqūt al-Hamawī (1179–1229), al-Ṣinnabra was ‘a place in the Jordan district, opposite ‘Aqabat Afīq, at a distance of three miles from Ṭabariyyah. Mu‘āwiya spent the winters there’ (Yāqūt, Mu‘jam III: 482).  Afīq/Fīq is identified on the eastern side of the southern part of the Sea of Galilee.  Therefore al-Ṣinnabra should be located on the opposite, southwestern side of the lake, as suggested by Whitcomb (Whitcomb 2002:1-6).  Furthermore, it seems that the main road which once connected between the district capital, Tiberias (Ṭabariyyah), and the main capital of the Umayyad dynasty, Damascus, went around the southern tip of the Sea of Galilee, ascended to ‘Aqabat Fīq on the eastern side of the lake and continued to the north, to Damascus (Elad 1999: 78-79).  Caliph Mu‘āwiya ibn Abī Sufyān (r. 660–680 C.E.), referred to by Yāqūt, succeeded his brother Yazīd b. Abī Sufyān in his dual role as military commander of the Muslims and governor of the province of Damascus after the latter’s death in the plague of ‘Amwās (Humphreys 2006: 45-50).  Thus, ‘the place where Mu‘āwiya used to spend the winter’ was built either prior to his appointment as caliph, when he was governor, or during his caliphate.  The royal residence at al-Ṣinnabra could therefore have been built any time between the years 639 and 680 CE.

Mu‘āwiya established his palace on what was probably a largely abandoned and secluded mound (marked only by the ruins of the tri-apsidal church), at a pleasant spot, with fresh water on hand, a short distance away from what would become the provincial capital, Ṭabariyya.  The style of the architecture was formal and imperial, with deep foundations and levelling operations that completely disregarded ancient construction, creating an imposing platform with an excellent view in every direction.  Although the palatial complex stood by the lake-shore, Mu‘āwiya and his successors invested considerable effort in delivering spring water to the palace, tapping the main aqueduct to Tiberias and conveying the water through channels, siphons, and pipes (Alexandre 2017) to the palace and baths.  ‘Abd al-Malik seems to be responsible for completing the main components of the complex, including the fortifications and baths.  In view of his extensive regional building projects, including, among many others, the grading of the mountain pass at Fīq and the paving of the highway from Jerusalem to Damascus, it is reasonable to attribute the construction of the basalt siphon to his rule as well; the more so, as it appears to have been constituted from blocks of the dismantled Sussita-Hippos siphon, transported across the lake in what must have been a major engineering project.



The three excavation seasons concentrated on revealing the remains of an edifice to the north of the fortifications.  The remains of the structure revealed consist of three west-east rows of circular bases, each containing six bases, an eastern boundary wall (w1) and the northern fortification wall serving also as the southern boundary of the structure.  On the same lines as the bases, three west-east walls were also exposed (fig.2, w2-4), emerging from a north-south wall (fig.2. w1) that extended from the northern tower of the qasr.  A fifth wall, running from east to west, was also exposed (fig.2, w5), damaged at three points at the column bases; this wall probably predates the hall or at least the column bases.  Marble column bases are inserted in three of the six bases of the southern row and one in the western edge of the middle row.  As the northern part was damaged through erosion and modern interventions, we do not have any evidence from this side.  As related to the western area, traces of gravel pavement were exposed with pitches of white plaster.

Daadli 2019 20 photo 2

Fig.2. The remains revealed in the three excavation seasons.

 The remains raise some challenging questions.  The southern row differs from the two rows further to the north; and the nature of wall 5, running through the whole width of the structure, is not yet clear.  The bases of the southern row are built out of rubble forming the base and supporting marble bases.  We assume that there were also marble bases inserted in the other rubble bases.  The two northern rows are built out of solid moulded mortar mixed with fieldstones.  A similar method of mixed mortar and fieldstone forms the core of walls 1–5.  It appears that there were some changes in plan and the southern base row was added.  Perhaps wall 5 meant to support a row of bases or columns, and then some changes were made and it was cancelled and the southern row was built.


Hypostyle hall

In either case, the reconstruction of the remains can be shaped to a hypostyle hall composed out of, at least, three aisles parallel to the northern wall of the qasir.  Wall 1 can stand as the eastern border of this hall.  Fifteen bays are enclosed in this hall supported by eighteen columns and paved by plaster laid on gravel.  This hall lacks the essential niche situated in the middle of the southern wall that may be used as the mihrab of the mosque.  However, the order of the construction, 4-m-bays, and the open facades may support the assumption that it was used as a mosque.

Whatever the use of the building may have been, huge efforts were made to level the whole northern area of the qasir.  After levelling, gravel pavement was laid, which probably supported a plaster floor and perhaps some stone slabs, forming a courtyard to the west of the hypostyle hall that probably served as a mosque.  Access to the levelled courtyard from the west was made by a staircase that was attached to the north-western qasir tower.  A door situated in the middle of the northern wall provided access to the mosque from the qasir.



The main dating finds are several copper coins found in the gravel layer.  The date as Arabo-Byzantine indicates that the levelling, paving and construction were made during the early Umayyad period.



The bases were cutting walls and domestic insulations from the Hellenistic and Early Bronze eras.  At least one Hellenistic structure can be reconstructed on the western side (fig.3, yellow).  This enclosed at least two terra cotta ovens.  Some remains of the Early Bronze era were found in between the bases in the eastern side (fig.3, red).  All stratigraphy, gravel floor with the bases, Hellenistic domestic remains and Early Bronze walls and plaster floor, were revealed in less than half a meter of stratified soil (fig.4).

Daadli 2019 20 photo 3

Fig. 3. Schematic plan of the different layers, blue Early Islamic, yellow Hellenistic and red Early Bronze.

Daadli 2019 20 photo 4

Fig.4. 1.Northern wall of the qasir, Early Islamic, 2. Hellenistic walls, 3. Early Islamic wall, 4. Early Bronze wall, 5. Early Islamic column base.



The first phase of the palace is dated by historical sources to the reign of Muʿāwiya who served as the governor of the province before he became caliph.  The primary analyses of the finds date the hypostyle hall or the mosque to Muʿāwiya or at least before Abd al-Malik’s days.  This unit and the scale of the works on the northern part of the qasir clarify the nature of the palace that may be described more as a complex since a further unit to the north, the dar, was previously revealed (fig.5).  Although few small finds were found, they added further confirmation regarding the date of the palace complex.

This unit standing to the north emphasizes the place of the northern entrance.  Here we were facing a major problem regarding the southern entrance. T hat entrance is bordered by two square towers; although it leads to the ground surrounding the basilica, it does not face the main façade.  This mean that visitors approaching the qasir from the south were not able to enter the main hall from the main entrance that probably was situated in the northern part parallel to the apse.  Hence, when visitors entered the qasir from the relatively narrow entrance in the north they faced the main façade of the main hall or the throne hall.  When approaching the qasir from the north, visitors could also enter the mosque situated on the same entrance.  On the opposite side of the qasir the bath was situated behind the rear wall of the basilical throne hall.  It could be due to the early date of this palace that it follows few of the expected ‘blueprints’ for Umayyad palace construction.

Daadli 2019 20 photo 5

Fig.5. The palace complex

We may conclude that by the end of the third season we reached the limit of an excavation of such scale.  To further explore the northern part of the qasir, at least the area border to the north of the church, on which the dar unit was established, there will be need for serious earth removal and an overall change in the park shape and plan.  The same can be said about the area extending to the west.

Tawfiq Daʿadli
Dept. of Islamic and Middle Eastern Studies and Art History
The Hebrew University of Jerusalem



Alexandre Y.

2013 Tel Bet Yerah, the Inverted Siphon Pipeline. HA–ESI 125.

2014 Tel Bet Yerah: The Bridge to el-Ṣinnabra. HA-ESI 126.

2017 The Inverted Siphon Pipeline to al-Ṣinnabra,” Bet Yerah, Vol. III: Hellenistic Philoteria and Islamic al-Ṣinnabra. R. Greenberg, O. Tal and T. Da‘adli, eds. Jerusalem, IAA Reports 61.

Applebaum, S.

1988 The Synagogue at Beth-Yerah: Its Character and function, Synagogues in Antiquity. Kasher, et al., eds. Jerusalem. 181-84.

Da'adli T.

2017 Stratigraphy and Architecture of the Fortified palace, Bet Yerah, Vol. III: Hellenistic Philoteria and Islamic al-Ṣinnabra. R. Greenberg, O. Tal and T. Da‘adli, eds. Jerusalem, IAA Reports 61.

Delougaz, P., and R. C. Haines.

1960 A Byzantine Church at Khirbet al-Karak. Chicago, OIP 85.

Elad A.

1999 The Southern Golan in the Early Muslim Period: The Significance of Two Newly Discovered Milestones of ‘Abd al-Malik.” Der Islam 76: 33-88.

Humphreys S.

2006 Mu‘awiya ibn Abi Sufyan: from Arabia to Empire. Oxford. Yāqūt.

1990 Mu’jam al-Buldān. Ed. F. ‘Abd al-‘Azīz al-Jundī. vol. III. Beirut. Whitcomb D.

2002 Khirbet al-Karak Identified with Sinnabra.” al-‘Usur al-Wusta: The Bulletin of Middle East Medievalists 14/1:1-6.


1990 Mu’jam al-Buldān. Ed. F. ‘Abd al-‘Azīz al-Jundī. vol. III. Beirut.

Whitcomb D.

2002 Khirbet al-Karak Identified with Sinnabra.” al-‘Usur al-Wusta: The Bulletin of Middle East Medievalists 14/1:1-6.