Excavation on the "High Mound" Excavation in the Lower Town The Tombs Regional Surveys and Excavations Remote Sensing

Field School

2008 Season 2009 Season

Excavation on the "High Mound"

We began excavations on the high mound in 1989 with two 5 by 5 m soundings (Operations 1 and 2) just east of the 1970s Area IV. Our intention was to expand one of them into what would be the first in the planned series of 10 by 10 m squares across the northern end of the mound. Because of time constraints we were forced to restrict the actual area excavated in each operation in order to reach a floor level contemporary with the Area IV building (ca. 2 m below the surface of the mound). In 1991 we enlarged Operation 1 to a 10 by 10 m square (with the original sounding forming the northwestern corner of the unit), but excavated only its northern half to a floor level contemporary with the Area IV building.

In 1993 we spent the first part of the field season working on architectural and stratigraphic problems left unresolved in 1991. This involved much research, study and work-- effort comparable to obtaining several online degrees. Soon though, we were able to resolve the issues and begin a new phase of work. We then removed the walls in the western part of the unit and began a sounding that eventually reached virgin soil. Though a stratigraphic sounding was not originally part of our field strategy, our work in Operation 1 had provided a large area cleared to more than 2 m below the surface; so, we made an opportunistic decision to continue down. We saw the sounding as contributing to the long-term goals of the Tell es-Sweyhat Project since we had initiated a new phase of regional work in 1993 and the sounding provided a reliable stratigraphic and local ceramic sequence into which we could tie other sites in the area. Our sounding measured 5 by 6 m, and we cleared to virgin soil over the whole of the unit.

We began a second 10 by 10 m square (Operation 12) 6 m east of Operation 1 in 1993. Eventually, we extended this unit 6 m farther to the east in order to investigate its architectural remains.We also laid out a 10 by 10 m square (Operation 13) in line with Operations 1 and 12 on the east side of the mound in the southwest corner of Square F9. However, having cleared a substantial (1.5 m thick) wash layer that sealed the architectural remains, we decided to close the operation in order to focus our efforts elsewhere.

We returned to Operations 1 and 12 in 1995 with limited goals. In Operation 1, we wanted, first, to clear the as yet largely unexcavated parts of the upper occupation levels and, second, to expand our 1993 deep sounding to clarify the stratigraphy and dating of the lowest occupation levels. We also wanted to link Operations 1 and 12 by excavating the 6 by 10 m baulk between them. We dubbed the excavation of the baulk Operation 20. After clearing surface fill over the whole of the area, we restricted deeper digging to the central part of the operation, with the deepest excavations in a 1.5 m wide trench in the south central part, where we were 4.4 m below the surface of the mound, ca. el. 87.70.

In addition to work in Operations 1, 20 and 12, we also laid out a new 10 by 10 m excavation unit (Operation 21) in line with those operations, approximately 10 m to the east (or up the slope of the mound), with the aim of clearing what we had assumed would be Hellenistic or Roman occupation levels in preparation for later work on third millennium levels.We began digging, but stopped when we encountered remains of 3rd millennium occupations immediately below the surface. The results of excavations in Operation 20 had already convinced us of the impossibility of correlating such up-slope 3rd millennium occupation levels with down-slope levels without physically linking the units. We will continue Operation 21, but only as we proceed more systematically up the mound.

Operations 1, 20 and 12

We have tentatively divided the almost six meters of accumulated debris from virgin soil to the surface of the mound in Operations 1, 20 and 12 into six occupation phases.



The phases group together building levels or depositional events and mark what we perceive to be continuities or discontinuities in the occupational sequence. The six phases span the 3rd and early 2nd millennia, with Phase 1 attributed to the first half of the 3rd millennium and Phase 4, our main occupation phase, to the end of the 3rd millennium.

Calibrated radiocarbon dates for charcoal and carbonized grain associated with the Area IV building (Buiterenhuis 1983:132-33; Holland 1977:62-63).


Lab No. P 2338 P 2324 GrN 10349 GrN 10350
Locus Area IV, Rm.7 Area IV Bldg. Area IV F 1.15 Area IV P 1.5
Material Charcoal Carbonized Grain Charcoal Charcoal
BP Date 3730+70 3640+70 3675+40 3810+35
Intercepts 2135, 2071, 2063 BC 2011, 2009, 1977 BC 2033 BC 2272, 2258, 2204 BC
1 Sigma Calibration 2271-1982 BC 2125-1898 BC 2129-1976 BC 2288-2146 BC
Probability Estimate 2271-2263 (.03) 2125-2081 (.20) 2129-2079 (.42) 2288-2191 (.90)

Quoted as date: (Relative contribution to probabilities)



2203-2020 (.90)

2001-1982 (.07)



2044-1898 (.80)




2046-2010 (.30)

2010-1976 (.29)



2159-2146 (.10)




2 Sigma Calibration 2391-1917 BC 2194-1775 BC 2176-1927 BC 2396-2065 BC
Probability Estimate 2391-2389 (.00) 2194-2153 (.04) 2176-2167 (.01) 2396-2380 (.02)
Quoted as date: (Relative contribution to probablilities) 2334-1917 (1.00)


2149-1866 (.89)

1844-1775 (.07)

2142-1927 (.99)


2347-2135 (.98)

2071-2065 (.01)


Phase 1


We reached virgin soil at 14.56 m (el. 85.44) below the cement bench mark on the top of the mound. Phase 1 consisted of a series of ashy soil layers that accumulated 90 cm to 1.20 m deep on virgin soil.

Our study of the Operation 1 ceramics is not yet complete, but we can make a few preliminary remarks about the Phase 1 pottery. With the exception of a relatively few jars and wide-mouth pots made of coarse, low-fired, grit-tempered cooking-pot ware the vessels were generally made of a high-fired, medium-textured fabric that varied in color from buff to pink. The clay was tempered by sand, possibly a natural inclusion, supplemented by the occasional addition of chaff or grit. While some sherds were covered in a cream-colored slip, reserve-slipdecoration was absent, as were other forms of decorative surface treatment. Several sherds bore incised marks made of several strokes of a stick in the wet clay. These were generally found on jars, but a few examples occurred on bowls as well (f).


Although examples of straight-sided bowls were attested in Phase 1, the most common bowl/cup form had an incurving side and a rounded rim (a-b). A bowl with S-shaped profile was also present (c). Small jars had an everted neck and rounded rim (d). A unique base for a small jar consisted of a tripod of three small knobs (e). The standard, medium-to-large jar had a wide mouth, a short, everted neck and rim that was usually rounded and frequently thickened (g-k). Some examples had a depression on the interior just below the rim, forming a shallow ledge, perhaps to hold a lid (h); others had a distinct groove on the outside of the rim (i). The depression and groove were the result of pressure exerted by the fingertip and fingernail in finishing the jars' mouths. Larger storage jars had low necks and rolled or rounded rims (l).

The assemblage represented by our Phase 1 pottery is essentially the same as that already published from Sweyhat, Area IIA, Phases A-F (Holland 1976: 39-48). Close parallels can also be drawn with the pottery from Tell Ahmar, Area A (Jamieson in Bunnens 1990) and Hadidi, Area RII, Stratum 1, Levels 1-4 and Stratum 2, Level 1 (Dornemann 1988). More generic parallels exist with Kurban Höyük, Period V (Algaze 1990: 281-309; cf.Wilkinson 1990: 214-17). All these parallels indicate that our Phase 1 is to be dated to the first half of the third millennium, that is, to the early part of the Early Bronze Age (EBA).

Phase 2


Phase 2 included the earliest architecture on the western edge of the mound recorded to date. The excavated remains can be divided into earlier and later occupations, but those distinct chronological periods have not been assigned subphase numbers. Our earliest architecture consisted of three partially cleared buildings constructed on or associated with the latest Phase 1 soil layer: a stone construction oriented northwest to southeast in the northwest corner (hereafter referred to as the NW Building) of the excavated area; a fragmentary structure running southeast to northwest across the northeast corner of the sounding (hereafter referred to as the NE Building); and, a pit house to the south.The NW and NE Buildings stood for the duration of Phase 2, while the occupation of the pit house was short-lived.



Phase 3


Phase 3 is as yet poorly understood. In Operation 1 the occupation phase that includes floors, debris layers and pits that follow on Phase 2 remains and preceed the Phase 4 buildings.Though we have not yet connected Operations 1, 20 and 12 at the level of our Phase 3 occupation, we have uncovered substantial architectural remains in Operations 12 and 20 that are earlier than Phase 4 and that should probably be attributed to Phase 3. The most important of these is a curved structure in Operation 12, which, if round, had a diameter of more than 12m.


The extant remains of the fortress at Phase 3 of excavation.
Model of the successive building layers of the fortress before it was resurfaced by a platform.


Phase 4


Phase 4 represents the main phase of Tell es-Sweyhat's occupation and the period of its floruit. In Operation 1 it includes parts of threebuildings to the east of the Area IV building uncovered in the 1970s. We found evidence of extensive terracing operations undertaken in preparation for the construction of the buildings, and probably for the inner fortification wall and Area IV building as well. The western side of the existing mound was cut down ca. 95 cm and roughly leveled. A retaining wall running northeast-southwest was then built against the cut face of the mound, with the base of the wall at el. 87.84. The 70 cm wide wall consists of a single line of stones fifteen to twenty courses or 2.3 m high at its highest preserved point. The wall was carried up in mud brick and the exterior face of the stones and mud brick covered in a thick plaster. The area behind (or to the east of) the wall was filled in presumably with earth from the western side of the mound (see the south section). We have identified two soil layers behind the retaining wall, which presumably respresent a deliberate sorting of fills and two distinct filling episodes. The lower fill, 30-60cm thick, is a rubble with whole and fragmentary mud bricks; in Operation 20 this fill includes a large number of substantial stones. The upper fill, more than 2 m thick in places, is a relatively clean and homogeneous brown soil. The top of the brown fill presumably approximates the surface level of the north central part of the late third millennium settlement.

As can be seen in the south section (see south section), the fill behind the retaining wall goes up to and over the top of the round building or platform uncovered in Operation 12 (see above), meaning that the structure must be earlier than our Phase 4 structures.



The buildings uncovered to the west in Operations 1 and 20 abutted the stone retaining wall described above.Those structures--and presumably the inner fortification wall and the Area IV complex--would have stood on a lower level than whatever structures existed to the east. Though parts of three buildings lay within the two operations, a single large,still not completely excavated, complex occupies all but the north-northeast part. This building provides evidence for both the original construction and a subsequent rebuilding. We uncovered part of an alleyway or street (Locus 1.14), which separated the large building from the Area IV building, in the western end of the operation. In the building itself wedefined two rooms (Locus 20.8 and 20.10) east and two rooms (Locus 1.13 and1.15) west of an L-shaped courtyard (Locus 1.16).

The number of work surfaces and cooking installations, as well as grinding stones and mortars, in the original Phase 4 building suggest that large-scale food preparation was carried out in it, and we have dubbed it the "kitchen building."

The "kitchen building" was rebuilt once in the course of its existence. Though the layout of the rebuilding nearly duplicated that of the original structure, sufficient differences existed to suggest that the function of the building had changed. Most importantly, two of the ovens went out of use. In Locus 1.13, the oven in the northern end of the room was replaced by a firepit, ca. 1.4 m in diameter, located to the south near the doorway to Locus 1.16. In Locus 20.8 the oven was given up and a doorway opened from Locus 1.16 directly into Locus 20.8. The arch between Locus 20.8 and Locus 20.10 was blocked at the same time. Only minor modifications were introduced in other rooms. We found few artifacts on the floor of the rebuilt building. In addition to the "kitchen building," we excavated parts of three rooms (Locus 1.11 and Loci 1.10 and 1.12) apparently associated with two buildings to its north. We cleared the spaces only to the level of the Phase 4 rebuilding.



Our Phase 4 is contemporary with the Area IV burned building uncovered in the 1970s. The Area IV building is securely anchored in time by four radiocarbon dates run on charcol and carbonized grain associated with its original floor [see c14 table above]. The calibrated dates suggest that the Area IV building and our Phase 4 date to the end of the 3rd millennium or, more specifically, are probably later than 2150 b. c. Though our study of the pottery is not complete, we can at least note that aside from the expected parallels with the Area IV pottery, strong parallels exist with Kurban Höyük, Period III and Tell Bi'a's Akkad "silo" fill (Einweg 1993: 46-49).


Phase 5

When the Phase 4 rebuilt complex was abandoned, it must have stood open for some period of time, and the walls gradually collapsed. The central part of the northern wall of Locus 1.16 was still preserved to a height of ca. 1.5 m above floor level; the eastern wall to ca. 90 cm; and, the other walls to 40-60 cm. At least parts of the abandoned and crumblingbuilding were apparently used as a dumping ground and were still beingwalked around in as fragmentary tamped earth surfaces were uncovered inLoci 1.15 and 1.16.

Phase 5 consists of the debris that was dumped over the ruins of therebuilt "kitchen building." The debris consisted of alternating brown sandyloams and gray ashy layers, apparently representing the remains of floorand yard sweepings and refuse from fires or kilns.

Our Phase 5 debris layers yielded substantial quantities of pottery, which is still being studied, as well as bone, and the soil samples which we took for flotation proved particularly rich in botanical remains. Small finds from the debris included a fragment of a model house presumably similar to those from excavations at Tell Ali il-Haj, near Rumeileh (Masuda 1933: 153-160) and Assur (Andrae 1922: 36-38 and pls.13-17) and to one in the Aleppo Museum (Khayata 1974/75); a nearly complete figurine of a male with slightly cocked head, applied eyes, and incised beard, eye brows and hair; and, the head of a chalk figurine of a male with eyes, nose, mouth and beard standing out in relief on a flat face.


Above: This fragment of a model house was found in the Phase 5 debris layers.

Below: Figuines found in the Phase 5 debris layers.

Above: Sketch of a complete model house.


Phase 6


Phase 6 represents four successive, poorly preserved building levels in the accumulation--more than 1.2 m deep in places--from the top ofthe debris layers that sealed the "kitchen building" up to the surface ofthe mound in Operation 1.

Fragmentary architecture and pits in Operation 20 and the western portion of Operation 12 seem to be contemporary with the second and third building levels of Phase 6. A fragmentary hollow bone cylinder with incised decoration, and several broken figurines were recovered from the Phase 6 soil layers in the area.




Our excavations in Operations 1, 20 and 12 have produced a long and potentially useful stratigraphic and ceramic sequence. Whether this represents an unbroken sequence remains as yet to be determined, as does the more detailed dating of individual occupation phases. As for the former, the terracing undertaken in connection with the construction of the Phase 4 "kitchen building" resulted in the removal of a substantial accumulation of debris from the western edge of the mound. The shallowness of our Phase 3 remains in Operation 1 on the west side of the terrace wall and the substantial remains we have tentatively assigned to Phase 3 in Operations 20 and 12 on the east side of the terrace wall suggest that excavations there are likely to yield additional stratigraphic data and a more complete ceramic sequence.

As for dating, our Operations 1, 20 and 12 occupation phases are anchored at two points in the sequence: Phase 1 and Phase 4. As detailed above, comparisons of the pottery recovered from the debris layers that accumulated on virgin soil in Operation 1 with the pottery from nearby sites suggest that Phase 1 dates to the first half of the 3rd millennium.Yet the marked degree of continuity in ceramics of the early 3rd millennium (Jamieson 1993: 36) precludes our suggesting a more restricted span of time. Without radiocarbon dates we cannot determine whether the succession of soil layers we attributed to Phase 1 spans four hundred hundred or fifty years. Phase 4, on the other hand, is securely pinned by radiocarbon dates to the last 150 years of the 3rd millennium.

With those chronological "anchors" in mind, we have tenatively assigned Phases 2-3 to the mid-to late 3rd millennium. The pottery from those Phases includes "metallic ware" and appears to be contemporary with the pottery from the tombs we uncovered in the outer town (cross-ref. below here) and with the pottery from Kurban Höyük Period IV (Algaze 1990:311-68) or Titrish Höyük Mid-to-Late EBA (Matney, Algaze and Pittman n.d.). Phases 5-6 likely date to the early years of the second millennium. Fortunately, they yielded substantial quantities of pottery. We anticipate that with the pottery from our Phases 4--and recently excavated ceramic sequences from sites such as Qara Quzak (Olávarri 1992), Tell Banat (Porter1995) and Mari (Pons n. d.)-- we will eventually be able to provide asecure ceramic sequence for the late third and early second millennium, a period as yet poorly documented for northern Syria.

As regards Tell es-Sweyhat's main Phase 4 occupation, our excavations in Operations 1, 20 and 12 uncovered evidence for substantial labor investment in terracing activities in the late 3rd millennium settlement and revealed that a "kitchen building" was located to the east of a deliberately laid-out alleyway from the Area IV building that contained bulk stores of processed grain. The two complexes were almost certainly connected in terms of their function and might be thought of asancillary to whatever existed upslope in the "core" area of the settlement.


Return to The Jezireh Regional Project

All Content © 2006-2010 Michael Danti
All rights reserved.