British Excavations at

Jerablus Tahtani, Syria 1998

Supported by:

the British Academy, the British Museum, the Council for British Research in the Levant, the Jenny S. Gordon Memorial Foundation, the National Museums of Scotland and the University of Edinburgh

Edgar Peltenburg, University of Edinburgh, Scotland-U.K.


This is a report on results of renewed 1998 excavations at Jerablus Tahtani, following a study season in 1997. Investigations continued in the four Areas previously explored: on the south-east edge of the tell and its lower shoulder where high status Tomb 302 had been excavated (Areas I and II), on the western side of the tell inside and below the Early Bronze fort wall (Area III) and on the northern side of the tell inside the wall of the fort (Area IV). Virtually all the results belong to the 4 th and 3rd millennia BC, that is Period 1, the Late Uruk occupation, and Period 2, the Early Bronze Age.

Table of contents


Period 1 The Late Uruk occupation

Period 2 Early Bronze Age fort wall

Period 2 Stratigraphic observations outside the postern

Period 2 Early Bronze Age occupation inside the fort

               List of References


The sixth season of University of Edinburgh excavations at Jerablus Tahtani took place from 22 March to 10 June 1998. The project conducted its archaeological research with the authority of the Directorate General of Antiquities and Museums as part of the international Tishreen Dam rescue programme in the Aleppo Mohafazat. The team is most grateful to Professor Dr. Sultan Muhesen, Director General of Antiquities and Museums, Dr. Adnan Bounni, Director of Excavations and Dr. Wahid Khayatah, Director, National Museum, Aleppo for their considerable help and encouragement. Most valuable assistance was rendered by our Government Representative, Mr. Mohammed Ali, who re-joined the team for the fifth time.

The project had three main aims this season: 1) to evaluate the earlier pre-fort occupation of the tell by a sounding in Area III, 2) to investigate the Early Bronze Age fort in Area IV, and 3) to assess the relations between the fort and a monumental tomb complex by connecting Areas I and II. For accounts of previous discoveries in these Areas, see Peltenburg 1997 and Peltenburg et al. 1995, 1996, 1997.

Back to Table of Contents

Period 1: The Late Uruk occupation

Plate1: The Uruk occupation (centre) lies before the fort wall (foreground).

The stone-based mudbrick wall of the Early Bronze Age fort divides Area III into two zones (Pl. 1). Excavations in front of and below this wall in 1996 disclosed a pre-fort Early Bronze settlement oriented at an acute angle to the overlying fort. This year we were able to show that the settlement was of no great duration and that it had expanded over an open space previously reserved for concentrations of pits. Initially, at least one rectilinear building with stone-paved courtyard, B 2242, was placed over the pits, but pitting resumed here before the settlement became firmly established. There are thus two pit phases. The latest consists of large round pits lines with pinkish white silicates.

The earliest comprises seven rectangular pits in a 42 m2 area. Many cut other pits, hence there is some time depth to this phase. Pottery from these levels belongs to the early part of the EBA. The rectilinear pits were cut into external laminated deposits of ashy and clay-rich materials, some containing evidence for primary refuse, others for in situ material. As in remaining deposits here, all units contained Late Uruk pottery with virtually no local Late Chalcolithic. Associated wall fragments were badly disturbed by the rectilinear pits, but the small bun-shaped bricks with rounded corners were entirely different from overlying EBA brick types.

The first coherent walls were found 1 m. below the top of the Late Uruk deposits. Here was a complex, 2185, consisting of corners of two rooms, one with a bin and associated stone-paved courtyard (Pl. 1). Cut into the wall of the other was a pit for a secondary burial, T. 2165. Disarticulated long bones were neatly collected here, with all proximal ends oriented to the north. The two rooms were linked by an extraordinary wall distinguished by bricks of five different colours. It was only preserved for 2 courses, but because of the multicoloured character of adjacent fills, there is no reason to doubt that the polychrome brick arrangement continued upwards to stunning effect. A thick buff mud render concealed this design on the exterior face.

Artefacts from the environmentally rich associated deposits include Reserved Slip jars, bevel-rimmed bowls, numerous tiny disc beads, a fine copper awl with twisted handle and a clay sealing with part of a complex spiraliform motif. Two contrasting trends are conspicuous in the transition from Late Uruk to EBA in this exposure. Artefacts, including the ceramic repertoire and brick types, undergo significant changes. There is, however, a striking measure of continuity in the orientation, position and perhaps functions of complex 2185 and overlying B 2242. This includes the recurrence of stone pavings against north walls. About 1.2 m. thick deposit separates these two entities, presumably covering the much-discussed post-Uruk transitional period, one that in the Tishreen has different characteristics from the transition in the Euphrates Valley to the north, in the area of Arslantepe (Fragipane and Palmieri 1983; Jamieson 1993) . It will obviously be helpful to obtain independent chronometric information on the length of this transition.

Back to Table of Contents

Period 2: Early Bronze Age occupation inside the fort

With increased exposures of occupation beside the EBA fort walls it has become clear that the nature of activities vary considerably. We have not been able to explore the central area where yet further differences are to be expected. Construction materials comprise mudbrick walls with standard sizes of bricks only occasionally founded on stones. There is a discontinuity in Area IV where, above the absolute level of the eroded top of the fort wall, stones were used extensively for construction purposes. Levels with such stone walls are either very late in the EBA history of fort or post-date it.

In Area I, located on the south-east of the fort, more work was carried out on Room 1569 which we initially regarded as a possible casemate (Peltenburg et al. 19996, 8 Fig. 7). Room 1569 now seems more likely to have been reserved for storage or some industrial purpose. Silicates, probably of grain stalks, were ubiquitous and two square ‘silos’, like those at Tell ‘Atij for example, were constructed in this rectangular, doorless space (cf. Fortin 1997, 54 Fig. 2b). However, above each silo in the west wall are blocked apertures that also contain silicates. One of them leads to a silt-filled pit adjacent to a silo. A channel extends from the south wall into the southern silo and there is a platform beside the northern silo. It is not clear how these features operated in grain storage. It may well be that the silicates arrived during an abandonment phase and that Room 1569 was originally intended for other purposes.

Plate 2: The fort wall crosses diagonally (from top right to top left). Inside it, to the left, are rooms in the course of excavation.

Immediately to the south of this room, between it and the postern (see below), lay a group of rooms and courts of a more domestic nature (Pl. 2). The best preserved, Room 2024, has a bench, rectilinear hearth and pot placements like those which occur elsewhere in the fort as well as on the platform at the base of the glacis. Thus, domestic and special-purpose rooms are grouped together in the south-east.

On the west, in Area IIIB, a succession of superimposed rooms inside the fort was investigated in order to understand the nature of the fort occupation in the west.This sequence is adjacent to two previously excavated rooms against the fort wall, each of which served different functions.

Typical of the new rectilinear rooms is a liberal use of mudbricks for thickening walls and platform construction contemporary with the glacis period of the fort. Only parts of each could be revealed and both retain portions of two nested jars 0.35m deep. Below this lay Room 1980 with a brick floor and possible door to the west. A tanur was placed away from the walls in the first phase. It was subsequently sealed by a platform on which stood two contiguous, plastered bins that occupied much of the recovered space. The result of 1998 excavations inside the western part of the fort is that we now have three adjacent rooms, each serving different functions. Although small and large scale storage and household activities are evident, it remains to understand how these architectural units were integrated.

Our major effort to expose a sequence of fort occupations lies in Area IV at the north where 120 m2 is exposed behind the fort wall (Pl. 3). In previous seasons, we noted that the dead were buried between building phases here. As excavations continued, this was shown to be an activity largely confined to the last stages of occupation, in the phase characterised by stone buildings.

Plate 3. Sequences of buildings in Area IV

It was particularly noticeable in the south of Area IV where a series of burials, T. 1518, 1670, 1687 and 1850 were cut into last walls of a major, often re-built structure. The graves consist of pit and corbel-walled small tombs, most with multiple interments associated with pottery, pins and necklaces (Pl. 4).

Plate 4.
A selection of material from one of the tombs inside the fort in Area IV.

The exposure is divided into two by a persistent feature, a paved passage running parallel with and some 7.5 m. from the fort wall. Thus, in this quarter of the late fort, communication was by a peripheral ring passage that, together with the fort wall, governed spatial organisation. It is conceivable that the inside of the fort was terraced, with buildings to the south of the ring passage higher than those to the north, between the passage and the fort wall. Evidence for erosion to the north of the ring passage was widespread, but there is still no stratigraphic proof for terracing. In a final phase of its use, the eastern end of the exposed passage had been blocked with a stone upright and the area was disturbed by pits and a possible hearth.

South of the last surviving paving of the ring passage, 1092, stood a number of superimposed stone-built rectilinear structures, 6 x >5 m. Most of their interiors were destroyed by a large Islamic pit that had also truncated at least one prolific EBA grave sunk into the west side of the buildings. Hence it is difficult to establish the function of these solid structures. The penultimate building, B 1775, had an entrance in the east where there was a series of yards, some with a tanur, others with pits containing much burnt debris including a seal impression. Indicating a crowded and often refurbished zone were several walls parallel structures were less substantial. Associated occupation extends beyond the limit of excavations. A poorly provisioned burial,T. 1782, had been placed in the narrow passage that separates these structures.

As already stated, the area between the passageway and fort wall suffered erosion before B 1000 was established in this locale (Peltenburg et al. 1996,10). Recurrent Iron Age finds in animal burrows suggest that occupation of this period exists elsewhere on the mound or was truncated by Period 5 terracing. In spite of erosion, traces of what must have been at least one relatively rich pre-B 1000 burial were recovered. T. 1703 yielded copper objects that include double tubes and a crescentic axe (cf. Philip 1989). Post-abandonment disturbance, therefore, obliterated many walls and surfaces with the result that we only have very fragmentary structures and tenuous relations. It should also be noted that excavations here are at a lower absolute level than exposures to the south of the passageway (but note the possibility of terraces).

Three features characterise these fragmentary remains of what seems to be a single row of contiguous rooms and yards. Unlike the area on the other side of the passageway, stonemasonry is confined to wall foundations of one or two courses; only southern walls seem to be stone founded; and north-south walls are built at right angles to the curving fort wall with the result that available space defined by these walls is narrower in the south than the north. In general, the western part of this exposure, where perhaps as many as three rectangular structures, c. 3 x 4 m., were located, was better preserved than the eastern which may have sometimes served as a courtyard.

Three spaces were discerned. In the northwest was a series of superimposed structures with three freestanding walls. Only one contained in situ features: a hearth, a conical feature, pebble flooring, threshold to the south. Beyond the threshold, the space between the buildings and the passageway became a discard zone for ash, pottery and stones. Amongst these was a bull figurine of a type that recurs in fort contexts. To the east were remnants of three superimposed wall bases and another, wider mud brick wall with a niched platform. Although there were in situ domestic features including tanurs and plaster basins, there were also metalworking finds like a mould and crucible (Pl. 5), unfortunately in fill rather than primary context.

Back to Table of Contents

Period 2: the Early Bronze Age fort wall

In order to establish the relations between the large T. 302 and the fort in the south-east, we linked Areas I and II by a series of extensions. Our assumption that a postern gate in the fort wall may have led to the tomb via Passage 990 was largely confirmed this year. As expected, removal of a stone threshold recorded in 1996 (Peltenburg et al. 1997, 5 Fig. 4) revealed a blocked entrance below. This postern, 2225, is only 1.6 m. wide, its northern jamb plastered white like the external wall of the fort. The wall protruded above the eroded glacis which terminated in front of the north jamb of the postern where it was revetted by a stone wall. However, we have still to reach those levels of the entry and passage that will establish the chronological relationships between the fort and tomb.

The fort wall was well preserved at the north end of what is now a 48 m. long exposure. It proved to be a distinct component, 1680, and its relationship with the wall continuing to the south, 108, indicates a more complex fort history than previously thought. Wall 1680, projecting 2 m. into the northern excavation area, is a sturdy structure founded on a battered, 2.2 m. high stone foundation or bulwark and bonded with a return which has been traced for c. 5 m. A late rectilinear tower(?) projected 1.75 m. from the wall face and there were no traces of a glacis here. The section of the wall which continues to the south-east corner of the fort, 108, has a different structure comprised of distinctively alternating brown and yellow brick courses and is only founded intermittently on a stone base. Although certainty is not possible, it may have been a two-phase extension that abutted the corner of a pre-existing fort the wall of which still stands to a height of 5.7 m. If wall 108 is indeed an annex, this has the effect of prolonging the history of the fort and making T. 302 very late within that history.

Back to Table of Contents

Period 2: stratigraphic observations outside the postern

It is clear that, in its final stages, the passage leading from the postern to high status T. 302 and other areas was affected by erosion and fell into disrepair. In the penultimate stage of use, a series of terrace walls was placed parallel to the fort wall in front of the entry, so entailing what seems to be a narrowly confined dog-leg approach. This followed gullying and slopewash events that swept away passage surfaces. Prior to that, the passage with a cobbled ramp that curved towards the entrance was oriented at a slightly different angle form the latest refurbishment. There were more gravel lenses below, some within drainage channels, and at this stage the passage was flanked by mudbrick walls rather than the stone walls used later in the EBA (Pl. 5). There is, thus, a partially excavated complex sequence of pavings, refurbishings and erosion that needs to be evaluated in order to relate T. 302 with the fort.

In its final stages, at least, Passage 990 passed over the southern terrace wall, 587, that runs generally parallel to and some 15 m. outside of the fort wall. It retained a platform on which stood domestic structures. In one exposure at the north-east of Area II, the eroded glacis was seen to underlie the platform. Since T. 302 postdates the earliest terrace wall, the tomb was constructed after the glacis. We know from drainage arrangements in Area IV that the glacis was a secondary feature of the fort (Peltenburg et al. 1996, 7-9) hence the tomb was erected some time after, and probably well after, the installation of the fort.

Plate 5. Fragmentary crucible with copper still adhering to the interior. From Area IV

Back to Table of Contents

List of References

Fortin, M. 1997; Urbanisation et 'redistribution' de surplus agricoles en Mésopotamie septentrionale (3000-2500 av. J.-C.).

In: W. Aufrecht, N. Mirau & S. Gauley (Eds) Urbanism in Antiquity. From Mesopotamia to Crete. Sheffield, JSOT Supp 244, pp. 50-81.

Frangipane, M. and A. Palmieri 1983; Cultural developments at Arslantepe at the beginning of the third millennium, Origini 12: pp. 523-74.

Jamieson, A. 1993; The Euphrates Valley and Early Bronze Age Ceramic Traditions, Abr Nahrain 31: pp. 36-92.

Peltenburg, E. 1997; Excavations at Jerablus Tahtani 1992, Chronique archéologique en Syrie I, 1992: pp. 42-44.

Peltenburg, E., S. Campbell, P. Croft, D. Lunt, M. Murray & M. Watt, 1995;Jerablus Tahtani, Syria, 1992-4: Preliminary Report, Levant 27: pp. 1-28.

Peltenburg, E., D. Bolger, S. Campbell, M. Murray and R. Tipping, 1996; Jerablus Tahtani, Syria, 1995: Preliminary Report, Levant 28: pp. 1-25.

Peltenburg, E., S. Campbell, S. Carter, F. M. K. Stephen and R. Tipping, 1997; Jerablus Tahtani, Syria, 1996: Preliminary Report, Levant 29: pp. 1-18.

Philip, G. 1989; Metal Weapons of the Early and Middle Bronze Ages in Syria-Palestine (BAR Int Ser 526) Oxford.

Back to Table of Contents

Back to home